Working Group Integrated Protection and Production in Viticulture. Proceedings of the meeting. Staufen im Breisgau (Germany) 1 4 November 2009

Transcription

1 IOBC / WPRS Working Group Integrated Protection and Production in Viticulture Proceedings of the meeting at Staufen im Breisgau (Germany) 1 4 November 2009 Editors: Agnès Calonnec, Carlo Duso, Cesare Gessler, Hanns-Heinz Kassemeyer, Michael Maixner, Denis Thiéry & Tirtza Zahavi IOBC wprs Bulletin Bulletin OILB srop Vol. 67, 2011

2 The content of the contributions is in the responsibility of the authors The IOBC/WPRS Bulletin is published by the International Organization for Biological and Integrated Control of Noxious Animals and Plants, West Palearctic Regional Section (IOBC/WPRS) Le Bulletin OILB/SROP est publié par l Organisation Internationale de Lutte Biologique et Intégrée contre les Animaux et les Plantes Nuisibles, section Regionale Ouest Paléarctique (OILB/SROP) Copyright: IOBC/WPRS 2011 The Publication Commission of the IOBC/WPRS: Dr. Ute Koch Schillerstrasse 13 D Moerlenbach (Germany) Tel , Fax u.koch_moerlenbach@t-online.de Dr. Annette Herz Julius Kühn-Institute (JKI) Federal Research Center for Cultivated Plants Institute for Biological Control Heinrichstr. 243 D Darmstadt (Germany) Tel , Fax Annette.Herz@jki.bund.de Address General Secretariat: Dr. Philippe C. Nicot INRA Unité de Pathologie Végétale Domaine St Maurice - B.P. 94 F Montfavet Cedex (France) ISBN Darmstadt (Germany), 2011

3 IOBC/WPRS Working Group Integrated Protection and Production in Viticulture Liaison officer: Sylvia Blümel, Austrian Agency of Health and Food Safety (AGES), Austria Convenor: Agnès Calonnec, INRA Bordeaux, France Deputy convenor: Andrea Lucchi, Università degli studi di Pisa, Italy Scientific committee: Sub-group Integrated pest management: Biological biotechnological control methods Host Plant interactions : Carlo Duso, University of Padua, Italy Tirtza Zahavi, Ministry of Agriculture, Israel Sub-group Biology and epidemiology of pathogens, fungal, bacterial and physiological diseases, including grapevine trunk disease Forecast modeling : Cesare Gessler, Swiss Federal Institute of Technology ETH, Zürich, Switzerland Hanns-Heinz Kassemeyer, State Institute for Viticulture and Oenology, Freiburg, Germany Sub-group Biology and population dynamics of insects and moths and modeling : Denis Thiéry, INRA Bordeaux, France Michael Maixner, Julius Kühn Institute, Bernkastel-Kues, Germany I

4 II

5 III Preface When I accepted to become the convenor of this working group at our meeting in Marsala (Italy) in October 2007, it was a great honour for me, and the pleasant occasion to expand my culture of epidemiologist to IPM. At that time, I did not realize the extent of the editing task! However the most important challenge is yours: always go further in our knowledge of the agro-ecosystem and put forward improvements to its management following the rules of IPM and of the agro-ecology. This implies having an even more integrated view of our cultural system, to better know each other and to exchange our points of view and disciplines. To reinforce this goal we tested a new structure of the meeting with an increase plenary session with one full day dedicated to IPM contributions, biological approaches and modelling (session 1 15 oral presentations). This was much appreciated from everybody and a nice way for pathologist, entomologist and generalist people to meet together. I still have in mind to increase this plenary session and who knows, maybe one day, to have only one plenary session (64% of you are in favour of it). The other sessions were more conventional for our meeting with session 2 dedicated to the Biology and epidemiology of pathogens, fungal, bacterial and physiological diseases, including grapevine trunk disease (15 oral presentations) and session 3 dedicated to the Biology and population dynamics of insects and moths and modeling (17 oral presentation). I take this opportunity to introduce you my new sub-convenors Tirtza Zahavi and Carlo Duso in charge of session 1, Hanns-Heinz Kassemeyer and Cesare Gessler in charge of session 2, Michael Maixner and Denis Thiéry for session 3 and Andrea Lucchi, our deputy convenor, who was in charge of the poster sessions. It was of great help to have them as chairmen and co-editors and I really thank them for their assistance. I shall also thank all participants for their participation to the building of this bulletin with 67% of presentations running into a paper. Finally, I would like to express gratitude to Hanns-Heinz Kassemeyer for accepting to organize the meeting in Germany despite the fact I contacted him very late. This is always a risky task to find enough financial support to get the registration fee as low as possible. Thanks to its professionalism, experience and energy, Hans and his collaborators made this meeting in Staufen a success and hosted 70 participants for 47 oral presentation and 19 posters. We had very active sub-group meetings on modelling and insect s behaviour and this encourage us to dedicate more time to discussions in the future. Finally, I am still working on improvements of our meeting and functioning of our group and you would see in the near future a web-site more especially dedicated to your IOBC group with your oral presentations, leaflets, propositions for collaborations, discussion, job or training advertisements and links to other web-sites. According to my survey, 87% of you are favourable to this web-site. I am also working on new improvements: create a price for the best student s paper, invite speaker s specialist from other discipline or working group, build a common European project? The next meeting is planned in October 2011, near Bordeaux and Denis Thiéry and I would be in charge of it. I wish you an exciting reading. Agnès Calonnec Convenor IOBC/WPRS working group Integrated protection and production in Viticulture

6 IV

7 V List of participants Bagnoli, Bruno Istituto Sperimentale per la zoologia agraria Entomologist via Lanciola, 12/A Firenze Italy bruno.bagnoli@isza.it Baus, Ottmar Geisenheim Research Center Fachgebiet Phytomedizin - Department of Phytomedicine POB 1154, Geisenheim Germany baus@fa-gm.de Birner, Erich BASF Germany Erich.Birner@basf.com Bleyer, Karl Referat Weinbau und Rebschutz Staatliche Lehr- und Versuchsanstalt für Wein- und Obstbau Traubenplatz 5, Weinsberg Germany karl.bleyer@lvwo.bwl.de Breth, Karl Bachstraße 15, Alsheim Germany weingut.breth@t-online.de Caffi, Tito Università Cattolica del Sacro Cuore Istituto di Entomologia e Patologia Vegetale Piacenza Italy tito.caffi@unicatt.it Calonnec, Agnès INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France calonnec@bordeaux.inra.fr

8 VI Cartolaro, Philippe INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France Chuche, Julien INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France Cocco, Arturo Università degli studi di Sassari Dipartimento di Protezione delle Piante Sassari Italy Delrio, Gavino Università degli studi di Sassari Dipartimento di protezione delle Piante Via E. De Nicola, Sassari Italy Denzer, Heinrich Walter Pessl Instruments GmbH Austria Di Marco, Stefano IBIMET CNR Bologna Italy Díez-Navajas, Ana M. NEIKER-Tecnalia Department of Plant Production and Protection Box 46, Vitoria-Gasteiz Spain

9 VII Eriksson, Anna University of Pisa Dept. CDSL Sect. Entomologia agraria Italy Fermaud, Marc INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France Fulchin, Emma Association in sustainable viticulture 1 cours du Général de Gaulle, Gradignan France ard-vd@enitab.fr Garganii, Elisabetta elisabetta.gargani@entecra.it Gessler, Cesare ETH Zürich Department of Agronomy Universitätstrasse 2, 8092 Zürich Switzerland cesare.gessler@agrl.ethz.ch Hein, Detlef Justus-Liebig-University Giessen Professorship for Organic Agriculture Giessen Germany Detlef.F.Hein@agrar.uni-giessen.de Herrbach, Etienne INRA UMR Santé de la Vigne et Qualité du Vin BP 20506, Colmar Cedex France Etienne.Herrbach@colmar.inra.fr Herrmann, Josef Valentin Bavarian state Institut for viticulture and Horticulture Germany josef.herrmann@lwg.bayern.de

10 VIII Hill, Georg DlR Rheinhessen-Nahe Department of Viticulture DlR R-H-N, Oppenheim Germany Hoffmann, Christoph Julius-Kühn-Institute Federal Research Institute for Cultivated Plants PlantProtection in Fruit Crops and Viticulture Brüningstr. 84, Bernkastel-Kues Germany Hommay, Gérard INRA UMR Santé de la Vigne et Qualité du Vin BP 20507, Colmar Cedex France Hummel, Hans E. Justus-Liebig-University Giessen Professorship for Organic Agriculture Giessen Germany Huth, Claudia Agricultural Service Centre Palatinate Neustadt a.d. Weinstraße Germany Ipach, Dr. Ulrike Agricultural Service Centre Palatinate Neustadt a.d. Weinstraße Germany Jermini, Mauro Agroscope Changins-Wädenswil ACW Centro di Cadenazzo 6594 Contone Switzerland

11 IX Kast, Walter K. Leiter der Abteilung Wein- und Obstbau Staatliche Lehr- und Versuchsanstalt Traubenplatz 5, Weinsberg Germany Kecskeméti, Elizabeth Germany Kehrli, Patrik Agroscope Changins-Wädenswil ACW Department of Entomology Route de Duillier CP 1012, 1261 Nyon Switzerland Knauf-Beiter, Gertrude Syngenta Crop Protection Münchwilen AG Research Biology WST , CH-4332 Stein Switzerland Kortekamp, Andreas Sachgebietsleiter Mykologie und Bakteriologie Abteilung Phytomedizin Breitenweg 71, Neustadt a.d. Weinstraße Germany Kuntzmann, Philippe Institut Français de la Vigne et du Vin Pôle Alsace, Biopôle Colmar France Latinovic, Nedeljko University of Montenegro - Biotechnical Department of Plant Protection Kralja Nikole bb, Podgorica Montenegro nlatin@cg.yu

12 X Lecomte, Pascal INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France Legler, Sara Elisabetta Università Cattolica del Sacro Cuore Istituto di Entomologia e Patologia Vegetale Piacenza Italy SaraElisabetta.Legler@unicatt.it Lentini, Andrea Università degli studi di Sassari Dipartimento di Protezione delle Piante Via E. De Nicola, Sassari Italy lentini@uniss.it Linder, Christian Agroscope Changins-Wädenswil ACW Department of Entomology Route de Duillier CP 1012, 1261 Nyon Switzerland christian.linder@acw.admin.ch Lorenzon, Mauro University of Padua Department of Environmental Agronomy and Crop Science viale Dell' Università, 16, Legraro Italy mauro.lorenzon@unipd.it Lucchi, Andrea Università degli studi di Pisa Dipartimento di Agricultural Entomology Via S. Michele 2, Pisa Italy alucchi@agr.unipi.it Maixner, Michael Julius-Kühn-Institute Federal Research Institute for Cultivated Plants PlantProtection in Fruit Crops and Viticulture Brüningstrasse 84, Bernkastel-Kues Germany Michael.Maixner@jki.bund.de

13 XI Mazzoni, Valerio IASMA Research Center SafeCrop Centre Via E. Mach 1, San Michele all'adige Italy Merdinoglu, Sabine INRA UMR Santé de la Vigne et Qualité du Vin 28, rue de Herrlisheim, Colmar France Miladinovic, Zoran University of Montenegro - Biotechnical Department of Plant Protection Kralja Nikole bb, Podgorica Montenegro zmiladinovic@t-com.me Molitor, Daniel Geisenheim Research Center Phytomedicine Von-Lade-Str. 16, 5366 Geisenheim Germany daniel.molitor@fa-gm.de Mugnai, Laura DiBA-Patologia vegetale P.,le delle Cascine 28 Italy laura.mugnai@unifi.it Naud, Oliver Cemagref UMR ITAP 361, rue J.F. Breton, BP 5095, Montpellier France olivier.naud@cemagref.fr Ortiz, Amaia NEIKER-Tecnalia Department of Plant Production and Protection Box 47, Vitoria-Gasteiz Spain aortizb@neiker.net

14 XII Peressotti, Elisa INRA UMR Santé de la Vigne et Qualité du Vin BP 20507, Colmar Cedex France Raynal, Marc ENTAV/ITV FRANCE Department of Viticulture 39 Rue michel Montaigne, Blanquefort France Reineke, Annette Geisenheim Research Center Phytomedicine Von-Lade-Str. 1, Geisenheim Germany Rossi, Vittorio Università Cattolica del Sacro Cuore Istituto di Entomologia e Patologia Vegetale Piacenza Italy vittorio.rossi@unicatt.it Savino, Francesco Cbc (Europe) Ltd Department of Biocontrol Via E. Majorana 2, Nova Milanese Italy fsavino@cbc-europe.it Schirra, Karl-Josef DLR Rheinpfalz Phytomedicine Breitenweg 71, Neustadt/Weinstraße Germany karl-josef.schirra@dlr.rlp.de Schnee, Sylvain INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France sylvain.schnee@bordeaux.inra.fr

15 XIII Schwappach, Peter Bayerische Landesanstalt für Weinbau und Gartenbau Rebschutz und Physiologie Herrnstraße 8, Veitshöchheim Germany Sentenac, Gilles Institut Français de la Vigne et du Vin Pôle Bourgogne Beaujolais Jura Savoie Beaune France Serra, Giuseppe CNR Istituto per lo Studio degli Ecosistemi Via E. De Nicola, Sassari Italy Sharon Ministry of Agriculture Kiryat Sh'mona Israel Thiéry, Denis INRA UMR1065 Santé végétale 71 avenue E. Bourlaux, Villenave d'ornon France Tirello, Paola Università degli Studi di Padova Dipartimento DAAPV Viale dell'università 16, Legnaro Italy Van Helden, Maarten INRA -ENITA UMR 1065 INRA/ENITA Santé Végétale 71 avenue E. Bourlaux, Gradignan France

16 XIV van Zeller de Macedo Basto Gonçalves, Maria Isabel Italy Veronelli, Vittorio Cbc (Europe) Ltd Department of Biocontrol Via E. Majorana 2, Nova Milanese Italy Viret, Olivier Agroscope Changins-Wädenswil ACW Département fédéral de l'économie DFE Route de Duillier CP 1012, 1261 Nyon Switzerland Wicht, Barbara Agroscope Changins-Wädenswil ACW Centro di Cadenazzo 6594 Contone Switzerland Zahavi, Tirtza Ministry of Agriculture Plant Protection Kibbutz Gshur Ramat Hagolan Israel Zeisner, Norbert Austrian Agency of Health and Food Safety Institute for Plant Health Spargelfeldstraße 19, Vienna Austria

17 XV Contents Preface... I List of participants... V Contents... XV Session 1: Integrated pest management and host-pathogen-interactions Distribution and host plant preferences of Hyalesthes obsoletus, the vector of bois noir disease, in Switzerland P. Kehrli, S. Kessler, S. Schaerer, N. Delabays Adult monitoring improves control of the flavescence dorée leafhopper Scaphoideus titanus in Gironde (France) while using less pesticide! M. van Helden, E. Fulchin, A. Verpy, F. Gil, C. Garcia Effectiveness of hot water treatments against the eggs of Scaphoideus titanus Ball C. Linder, L. Schaub, F. Klötzli-Estermann Downy, powdery mildew and botrytis risk assessment by climate parameters, pest and spores monitoring in Rioja Alavesa vineyard (Northwest of Spain) A.-M. Díez-Navajas, A. Ortiz-Barredo Biological control of Heliococcus bohemicus Sulc (Hemiptera Pseudococcidae) with the natural enemy Ericydnus sipylus (Walker) (Hymenoptera Encyrtidae) G. Sentenac, P. Kuntzmann, L. Perez, A. Gili, P. Kreiter Experiences about the effect of growth regulators on cluster structure and fruit rot diseases of the vine varieties Pinot Meunier, Pinot Blanc, Pinot Noir and Rhine-Riesling K. Bleyer, W. K. Kast Control of grey mould by application of gibberellin acid inhibitor (Regalis ) and management practices in Vitis vinifera (cv. Riesling) O. Baus, A. Reineke, B. Berkelmann-Löhnertz Variety-dependent responses of seeded vine cultivars to different growth regulators S. Böll, T. Lange, H. Hofmann, P. Schwappach Monitoring nematode populations to adapt fallow periods against Xiphinema vectors of grapevine fanleaf virus (GFLV) M. van Helden, L. Villate, C. Laveau, E. Morin, G. Darrieutort, C. van Leeuwen

18 XVI VitiMeteo a future-oriented forecasting system for viticulture G. Bleyer, H.-H. Kassemeyer, M. Breuer, R. Krause, O. Viret, P.-H. Dubuis, A.-L. Fabre, B. Bloesch, W. Siegfried, A. Naef, M. Huber The expert system OiDiag a useful tool for the precise scheduling of sprays against powdery mildew of vine (Erysiphe necator Schwein.) W. K. Kast, K. Bleyer an interactive platform for a better management of grapevine diseases and pests O. Viret, P.-H. Dubuis, A.-L. Fabre, B. Bloesch, W. Siegfried, A. Naef, M. Hubert, G. Bleyer, H.-H. Kassemeyer, M. Breuer, R. Krause Modelling and experimenting crop protection decision workflows: some lessons from GrapeMilDeWS research O. Naud, P. Cartolaro, L. Delière, B. Léger Session 2: Biology and epidemiology of pathogens, fungal, bacterial and physiological diseases, including grapevine trunk disease Impact des données météo de précision sur l évaluation locale du risque épidemique modelisé sur le vignoble. Premiers résultats de validation spatiale des orages de grêle de mai 2009 sur les vignobles de Bordeaux et Cognac M. Raynal, C. Debord, S. Guittard, M. Vergnes, J. Congnard, D. Grimal Long-term survival of Plasmopara viticola oospores T. Caffi, V. Rossi, M. Lusitani Dispersal of the sexual stage of Erysiphe necator in northern Italy V. Rossi, T. Caffi, S. E. Legler, R. Bugiani, P. Frisullo Modelling the effect of the grapevine growth and susceptibility on the dynamics of a powdery mildew epidemic A. Calonnec, S. Schnee, P. Cartolaro, M. Langlais Consideration of dynamical plant-pathogen interactions for an improved management of powdery mildew epidemics in grapevine S. Schnee, J. Jolivet, A. Calonnec Blackrot - downy mildew control in small vineyards in southern Switzerland C. Gessler, M. Jermini First study on the population genetic structure of Guignardia bidwellii M. Jermini, A. Angst, M. Raynal, C. Gessler, G. Broggini Methods for screening new Ampelomyces strains to be used as biocontrol agents against grapevine powdery mildew S. E. Legler, T. Caffi, L. Kiss, A. Pintye, V. Rossi

19 XVII Water activity at the fruit surface: a potential indicator of grape berry susceptibility to Botrytis cinerea M. Fermaud, C. Deytieux-Belleau, J. Roudet, G. Darrieutort, L. Geny Investigation of contamination pathways of Esca-associated fungal pathogens and of fungicide-based control strategies A. Kortekamp, J. Köckerling Impact of biotic and abiotic factors on the development of Esca decline disease P. Lecomte, G. Darrieutort, C. Laveau, D. Blancard, G. Louvet, J.-P. Goutouly, P. Rey, L. Guérin-Dubrana Session 3: Biology and population dynamics of insects and moths and modeling Electrospun nanofibers as novel carriers of insect pheromones: communication disruption strategy against Lobesia botrana D. F. Hein, M. Breuer, H. E. Hummel, A. Greiner, J. H. Wendorff, C. Hellmann, A. Vilcinskas, A. Kratt, H. Kleeberg, G. Leithold Control of grape berry moth larvae using parasitoids: should it be developed? D. Thiéry, L. Delbac, C. Villemant, J. Moreau Life history of Lobesia botrana on Daphne gnidium in a Natural Park of Tuscany A. Lucchi, L. Santini Current status of grapevine leafminers in north-eastern Italy C. Duso, A. Pozzebon, M. Baldessari, G. Angeli Factors affecting the post-release dispersal of Trichogramma cacoeciae Marchal in the vineyard G. Hommay, J. C. Kienlen, C. Gertz, C. Bihry, J. Pizzol Mating disruption field trials to control the vine mealybug Planococcus ficus A. Cocco, M. Coinu, A. Lentini, G. Serra, G. Delrio Biological characteristics of Heliococcus bohemicus and Ericydnus sipylus in controlled conditions A. Fleisch, L. Pérez, P. Kreiter, A. V. Odor, G. Sentenac Survey on Scaphoideus titanus egg distribution on grapevine B. Bagnoli, E. Gargani Preliminary study of the aggregative behaviour of Scaphoideus titanus larvae J. Chuche, A. Boursault, D. Thiéry Vibrational signals associated to the mating behaviour of Hyalesthes obsoletus Signoret (Hemiptera: Fulgoromorpha) V. Mazzoni, G. Anfora, C. Ioriatti, M. Virant-Doberlet, A. Lucchi

20 XVIII A beneficial species becomes a pest - the common earwig Forficula auricularia (Linnaeus 1758) C. Huth, K.-J. Schirra, A. Seitz, F. Louis Effects of some insecticides on Kampimodromus aberrans: laboratory and field studies P. Tirello, S. Vettore, A. Pozzebon, M. Lorenzon, C. Duso Development and reproduction of the predatory mites Kampimodromus aberrans, Typhlodromus pyri and Amblyseius andersoni on different food sources M. Lorenzon, A. Pozzebon, C. Duso Susceptibility and sensibility of grape cultivars to the leafhopper Empoasca vitis D. Fornasiero, A. Pozzebon, F. Pavan, C. Duso Can Harmonia axyridis affect the taste of European wines? C. Linder, F. Lorenzini, P. Kehrli Feeding behaviour of Lobesia botrana on leaves and shoots of grapevine A. Lucchi, E. Pozzolini, G. Anfora, V. Mazzoni, M. Tasin, E. Leonardelli, C. Ioriatti Attractiveness of different colours to Scaphoideus titanus Ball (Hemiptera: Cicadellidae) adults V. Mazzoni, F. Trona, C. Ioriatti, A. Lucchi, A. Eriksson, G. Anfora The impact of a number of insecticides on Empoasca vitis populations in north-eastern Italy A. Pozzebon, M. Pederiva, R. Moret, C. Duso Small insect enclosure field cages: A simple method to assess mating disruption F. Briand, C. Vergely, P.-J. Charmillot, P. Kehrli Influence of nettle control along a ditch on spatial distribution of Hyalesthes obsoletus Signoret in a neighbouring vineyard N. Mori, N. Reggiani, A. Pozzebon, C. Duso, F. Pavan Electropenetrography, a tool to investigate the feeding behaviour of sucking insects: development of this technique to Scaphoideus titanus J. Chuche, N. Sauvion, D. Thiéry Gaps in knowledge for modern integrated protection in viticulture: lessons from controlling grape berry moths Denis Thiéry Authors index

21 Session 1: Integrated pest management and host-pathogen-interactions

22

23 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp. 3-8 Distribution and host plant preferences of Hyalesthes obsoletus, the vector of bois noir disease, in Switzerland P. Kehrli 1, S. Kessler 1,2, S. Schaerer 1, N. Delabays 1 1 Station de recherche Agroscope Changins-Wädenswil ACW, CP 1012, CH-1260 Nyon, Switzerland. 2 Institute of Biology, University of Neuchâtel, CP 158, Neuchâtel, Switzerland Abstract: In Europe, the polyphagous planthopper Hyalesthes obsoletus Signoret (Hemiptera, Cixiidae) is assumed to be the most important vector of the grapevine yellows disease bois noir, which is caused by phytoplasmas of the stolbur 16Sr-XII-A group. For a better understanding of the epidemiology of bois noir in Switzerland, distribution and host plant preferences of H. obsoletus were studied in the field as well as in the laboratory. A national survey revealed that H. obsoletus is present in vineyards of southern, western and northern Switzerland; however, no specimens were caught in the east of Switzerland, where evidence for the disease is lacking. Even though field bindweed (Convolvulus arvensis L.) was much more abundant in vineyards than stinging nettle (Urtica dioica L.), H. obsoletus adults were captured almost exclusively on the latter. A second, more targeted field study confirmed H. obsoletus' clear preference for U. dioica. From June to September 2008, adults of H. obsoletus were primarily captured on U. dioica. However, few individuals were also caught on C. arvensis, hedge bindweed (Calystegia sepium L.) and four other dicotyledons. Emergence traps set up above U. dioica and C. arvensis highlighted that significantly more nymphs emerged from stinging nettle than field bindweed. Besides, a clear association between the distribution of U. dioica patches and the incidences of bois noir infected vines could be observed in a commercial vineyard. In conclusion, there is good evidence that H. obsoletus is also the most important insect vector of bois noir in Switzerland and that the insect prefers to feed and develop on U. dioica. Stinging nettle therefore plays a major role in the epidemiology of bois noir in Swiss vineyards. Key words: Viticulture, weeds, plant-vector associations, Vitis vinifera Introduction Flavescence dorée and bois noir are two important grapevine yellows diseases in Europe that can cause serious economical losses in grapevine production. Symptoms of these two yellows diseases on grapevines (Vitis vinifera L.) are both provoked by phytoplasmas. Phytoplasmas are wall-less and phloem-restricted plant pathogenic bacteria that are usually transmitted by phloem-sucking insects. Bois noir is caused by phytoplasmas belonging to the stolbur taxon (Ca. Phytoplasma solani; IRPCM, 2004) (16Sr-XII), which besides grapevines also infects a wide range of other crops and wild plant species. The polyphagous planthopper Hyalesthes obsoletus Signoret (Hemiptera: Cixiidae) is assumed to be the most important vector of bois noir in Central Europe (Maixner, 1994; Sforza et al., 1998). However, grapevine is only an erroneous food resource on which H. obsoletus nymphs can not develop. The principal difficulty for understanding the epidemiology of bois noir disease is the wide host plant range of both its phytoplasma and its vector. Even though plant preferences of H. obsoletus vary considerably among regions, field bindweed (Convolvulus arvensis L.) and stinging nettle (Urtica dioica L.) are its preferred host plant species in Central Europe (Johannesen et al., 2008). 3

24 4 Even though bois noir is present in Switzerland since more than twenty years, the interaction of the phytoplasma, its original host plants, its insect vectors and vines was only poorly understood. A research program was therefore launched in 2008 in order to gain insight into the significance and epidemiology of this grapevine yellows disease. In a national survey the distribution and abundance of H. obsoletus was examined in Swiss vineyards. Besides, H. obsoletus host plant preferences were studied in additional field trials. Material and methods National survey on the presence and abundance of H. obsoletus The presence of H. obsoletus was studied by sweep netting and sucking sampling of U. dioica and C. arvensis in all major wine regions of Switzerland. The national survey has been conducted between the end of June and July Host plant preferences of H. obsoletus Host plant preferences of H. obsoletus adults was studied with a D-vac sucking sampler in five vineyards of Western Switzerland from the end of May until the mid of September Three of the five vineyards were located in the Valais (Noës, Flanthey and Salgesch) the other two were situated in the 3-Lake area (Le Landeron and Vallamand). All five vineyards are infected by bois noir. Sucking samples were taken on altogether ten herbaceous plant species as well as grapevines. Each plant species was sampled for a period of five minutes (Table 1). To study the importance of U. dioica and C. arvensis for the development of H. obsoletus nymphs, emergence traps were set up directly above these two potential host plant species. Three traps per plant species were set up in the beginning of June 2009 in a commercial vineyard close to Geneva. Emergence traps were controlled once a week from the mid of June until the beginning of August. In the end of summer 2009, this commercial vineyard close to Geneva was heavily infected with bois noir. Thus, the distribution of infected vines and U. dioica was mapped the 25 September. Results National survey on the presence and abundance of H. obsoletus The presence of H. obsoletus could be confirmed in vineyards of Southern, Western and Northern Switzerland (Figure 1). However, no individuals were captured in the East of Switzerland, where evidences for the presence of bois noir are lacking. Even though C. arvensis was much more abundant in Swiss vineyards than U. dioica, adults of H. obsoletus were captured almost exclusively on the latter. Overall, 667 individuals were caught on U. dioica compared to 8 specimens on C. arvensis. However, the number of captured adults varied significantly between locations (e.g. between 1 and 265 individuals), without pointing to any concealed cultural or climatic factor. Host plant preferences of H. obsoletus Hyalesthes obsoletus was much more abundant on U. dioica than on C. arvensis or any other plant species (Table 1). Nonetheless, some specimens were also caught on Calystegia sepium L., Clematis vitalba L., Plantago lanceolata L., Polygonum aviculare L. and Taraxacum

25 5 officinale Weber, but never on Amaranthus retroflexus L., Ononis pusilla L., Solanum nigrum L. or V. vinifera. Overall, 96 H. obsoletus adults were captured in the three emergence traps set up above U. dioica, whereas only one adult emerged from C. arvensis. However, this individual emerged a week before the first adults were trapped above U. dioica. The emergence of H. obsoletus above U. dioica started around 20 June 2009, reached its peak at the end of June and lasted until the end of July. About 22% of all vines were infected with bois noir in the vineyard close to Geneva. Besides, the distribution of U. dioica patches and the distribution of infected vines was correlated (Figure 2). Table 1. Total number of H. obsoletus adults captured in the five study sites between May and September Location Plant species N of samples Total N of H. obsoletus Salgesch Amaranthus retroflexus 3 0 Convolvulus arvensis 8 1 Polygonum aviculare 6 0 Solanum nigrum 3 0 Taraxacum officinale 6 0 Vitis vinifera 6 0 Noës Amaranthus retroflexus 4 0 Convolvulus arvensis 7 0 Vitis vinifera 6 0 Flanthey Clematis vitalba 7 1 Convolvulus arvensis 7 0 Ononis pusilla 5 0 Vitis vinifera 6 0 Vallamand Calystegia sepium 7 11 Urtica dioica Vitis vinifera 6 0 Le Landeron Convolvulus arvensis 9 3 Plantago lanceolata 5 1 Polygonum aviculare 5 5 Solanum nigrum 6 0 Taraxacum officinale 6 4 Urtica dioica Vitis vinifera 7 0

26 6 Figure 1. Presence of H. obsoletus on U. dioica (left) and C. arvensis (right) in Swiss vineyards. Black dots ( ) represent sites were the presence of H. obsoletus was confirmed, at sites with white dots ( ) no individuals were found.

27 Figure 2. Map of the distribution of bois noir infected vines and of U. dioica in a commercial vineyard close to Geneva. Infected vines are indicated in black, U. dioica patches in grey. Discussion There is good evidence that H. obsoletus is the most important insect vector of bois noir in Switzerland. First of all, its presence is confirmed in most Swiss regions and its appearance accords with the distribution of bois noir. And second, molecular analyses conducted on more than 30 potential vector taxa revealed that only specimens of H. obsoletus carried bois noir phytoplasma (Kessler, 2009). Our field study clearly demonstrated that, with the exception of the Valais, stinging nettle is the preferred host plant of H. obsoletus in Switzerland. Not only did we capture significantly more adults on U. doica than any other potential host plant species, nymphs also developed almost exclusively on the former. Even though field bindweed is one of the most abundant accompanying plant species in Swiss vineyards, bois noir infected vines are almost exclusively found in vineyards sheltering stinging nettle (Kehrli et al., 2009). This association could even be observed at the micro-spatial scale. The distribution of U. dioica and bois noir infected vines was clearly correlated in the commercial vineyard mapped. In conclusion, there is good evidence that H. obsoletus is also the most important insect vector of bois noir in Switzerland and that the insect prefers to feed and to develop on U. dioica. Stinging nettle therefore plays a major role in the epidemiology of bois noir disease in Swiss vineyards. Acknowledgements We thank Michael Maixner and Stéphane Emery for their useful suggestions and Helen Johnston, Denis Pasquier, Martine Rhyn, Susanne Tagini, Constance Wagner, Carine Vergely, Pierre-Adrien Roux and Luc Egli for their assistance. References Johannesen, J., Lux, B., Michel, K, Seitz, A. & Maixner, M. 2008: Invasion biology and host specificity of the grapevine yellows disease vector Hyalesthes obsoletus in Europe. Entomol. Exp. Appl. 126:

28 8 Kehrli, P., Kessler, S., Schaerer, S. & Delabays, N. 2009: Die Verbreitung der Schwarzholzkrankheit und ihres Überträgers in der Schweiz. Schweiz. Z. Obst- Weinbau 145 (14): 4-6. Kessler, S. 2009: Epidémiologie du Bois noir de la vigne en Suisse et biologie de son vecteur Hyalesthes obsoletus Signoret (Hemiptera, Cixiidae). Master thesis, University of Neuchâtel, Switzerland, 118 pp. Maixner, M. 1994: Transmission of German grapevine yellows (Vergilbungskrankheit) by the planthopper Hyalesthes obsoletus (Auchenorrhyncha: Cixiidae). Vitis 33: Sforza, R., Clair, D., Daire, X., Larrue, J. & Boudon-Padieu, E. 1998: The role of Hyalesthes obsoletus (Hemiptera: Cixiidae) in the occurrence of bois noir of grapevines in France. J. of Phytopathol. 146:

29 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Adult monitoring improves control of the flavescence dorée leafhopper Scaphoideus titanus in Gironde (France) while using less pesticide! M. van Helden 1,2, E. Fulchin 2, A. Verpy 3, F. Gil 3, C. Garcia 4 1 ENITA, Université de Bordeaux, UMR INRA/ENITA Santé Végétale, ISVV, CS Gradignan, France; 2 ARD-VD, 1 Cours du Général de Gaulle, Gradignan, France; 3 GDON du Libournais, 14 rue Guadet, Saint Emilion, France; 4 Direction Régionale de l'alimentation, de l Agriculture et de la Forêt, Service Régional de l Alimentation, 51 rue Kiéser, Bordeaux Cedex, France Abstract: The flavescence dorée is caused by a phytoplasme spread by the leafhopper Scaphoideus titanus and is present in the south of France. Leafhopper control is based on up to 3 obligatory insecticide sprayings (T1; T2; T3) using neurotoxic insecticide compounds over large areas (Imposed Sprayed Area = ISA). Spraying dates are defined according to the observation of the first larvae (T0) at T1= T0+4 weeks, T2 = T0+6 weeks (T1 and T2 aim the larvae) and T3 = T0+10 weeks (aiming at adults). Such schedules and area definition are very efficient but lack scientific basis on insect dynamics and migration and encounter opposition from farmers. Over the last 5 years progress was made to reduce pesticide use while maintaining sufficient vector control. A project was launched in the Saint Emilion region after two plots were found contaminated in ha of vineyards entered the ISA. Farmers and regional Plant protection service organised into a small association to test a more integrated approach for vector control. Spraying zones were limited to a buffer zone of 2 Km around the contaminated plots. Farmers between 500m and 2 Km were allowed to reduce sprayings to only 1 single application at T0 + 5, while adult trapping was used to monitor overall results. Traps were yellow delta traps (Piège Tri- nglué) that are highly efficient. In most of this zone a single insecticide spray reduced trap catches to almost nonexistent, considered as a proof of sufficient control of the vector. In some areas trap catches were considered too important (>3 adults on a single trap in a single week) and a second spraying (T3) was imposed. This was probably due to some farmers either not spraying at all or organic farmers that only were allowed Rotenone, that is rather inefficient. This approach was repeated in Total cost of trapping and monitoring of larvae (traps, technicians, monitoring) is around 3 to 4 Euros per ha. These costs were largely compensated by the reduction in spraying (64 and 72 percent reduction in 2007 and 2008 compared to the traditional approach). Moreover, the combination of reduced spraying and vector monitoring seemed well accepted by the farmers. In addition all plots were monitored for plants showing symptoms of phytoplasm, around 1200 samples were analyzed first year. Over 80% of the samples appeared to be Bois Noir phytoplasm but some Flavescence Dorée samples were found in both 2007 (2 samples) and 2008 (6 samples). Scouting and analysis costs were around 25 Euro per ha. In other areas where spraying frequency was reduced simultaneously but without vector monitoring, a strong increase in FD symptoms was observed, together with high insect populations. This probably reveals that many farmers do not apply the single imposed spraying. The trapping seems a useful tool both to show efficient vector control and to incite farmers to respect the minimal spraying frequency. Key words: flavescence dorée, Scaphoideus titanus, vector, insecticide, trap, monitoring, imposed sprayings, scouting 9

30 10 Introduction Flavescence dorée is a grapevine yellows caused by a phytoplasme that spread by the leafhopper Scaphoideus titanus. This disease is present in the south of France, and more generally in southwest and center Europe. The only way to manage it is to control the vector by using neurotoxic insecticides and to uproot the contaminated vines. In France, when an infected plant is discovered, a large Imposed Sprayed Area (ISA) is defined, including the contaminated district and all the adjacent ones. On this ISA, the leafhopper management schedule is based on up to 3 obligatory insecticide sprayings (T1; T2; T3). Spraying dates are defined according to the observation of the first larvae (T0) at T1= T0+4 weeks, T2 = T0+6 weeks (T1 and T2 aiming at larvae) and T3 = T0+10 weeks (aiming at adults). If well respected, such insect management is very efficient. Yet the decision rules lack scientific basis on insect dynamics and migration. Moreover, this spraying schedule is not associated with scouting obligations, whereas it is necessary to proove by prospecting that there is no infected plant in the vineyard of the district. For this very few districts are allowed to go out of the ISA. Thus, the feeling of an inefficient management, the additional cost of these imposed sprayings and the reluctance of some wine-growers to use insecticides often lead to a non respect of the sprayings obligations and then to a failure of the fight. During the past few years, pest management adjustments have been allowed: the average number of imposed sprayings per hectare has decreased. Farmers and regional Plant protection service organised into small associations to test a more integrated approach for vector control. Thanks to this alternative approach, a huge progress was made over the last 3 years to reduce pesticide use while maintaining an efficient vector control. Material and methods Concept of the alternative pest management of S. titanus As a response to failures of the classical management of flavescence dorée, wine-growers decided, with Plant Protection Service agreement, to implement a new S. titanus management. The main goal is to use minimum insecticide to control the vector populations in the ISA. With this alternative management, one single spraying aiming at larvae is realised (instead of 2) and a second spraying aiming at adults may be launched only if insects are observed. So a network of yellow delta traps is set up at the rate of one trap per 30ha (based on a mobility under 500m) and weekly observed during adult flight period to check the presence or absence of the insect (Catalano, 2008). The second purpose is to scout the area in order to identify and suppress the infected plants. Such prospecting should help and speed up the decontamination of the area and eventually allow outing of the ISA. Example of implementation on the Saint Emilion vineyard In 2006, two plots were found infected by flavescence dorée. This led to an ISA of more than ha from The wine-growers reacted quickly and strongly, asking to implement a more integrated management of the issue than the classic one (Figure 1).

31 11 C L A S S I C A L T E R N A T I V E ISA: contaminated district (3 sprayings) + adjacent districts (1 spraying) T0: first larvae MAY T1: T0 + 4 weeks T1 + T2 + T3 JUNE T1-2: T0 + 5 weeks T2: T0 + 6 weeks ISA: contaminated area (1or 2 sprays) + adjacent areas (1 spray) T1-2 (+ T3 if catches) T1-2 T1 T3: T weeks JULY T3 ONLY IF CATCHES Figure 1: classic and alternative schedules for S. titanus management Instead of using administrative borderline, several sprayed areas were delimited as concentric circular zones (buffer). The center of the circles was an infected plot. The number of sprayings was defined depending on the distance to the source: it was decreasing when moving away from the contaminated patch : - plots less than 500m from the infected one (based on the hypothesis that S. titanus flight distance ability is under 500m and on the low number of infected plants) had to be sprayed 1 time at T0 + 5 and a second time if there are catches in the traps. - plots between 500m and 2km from the source were allowed to be sprayed only 1 single time at T A network of about 350 yellow traps was used to monitor adults populations of S. titanus over all the area related to the project (Figure 2). A huge scouting was also realised: ha (all the vineyard related to the project) in 2007 and about 4000ha (around the infected plots) in 2008 were monitoed by walking through. In 2007, it represented 1,350 days of work at the end of the summer. The use of a service provider was needed to provide such a quantity of work in such a short periode. 50 hikers organised in teams walked about 30 km per day and per person. Example of implementation on the Pessac-Leognan vineyard Since 2003, a part of the Pessac-Leognan vineyard has been bound to apply 2 imposed sprayings against S. titanus. In 2008, a new plot was found infected by flavescence dorée out of the ISA of that moment. This led to a new larger ISA of about 1,420ha in The winegrowers decided to implement an alternative management of the issue. 50 yellow traps were set up in the vineyard. The contaminated districts had to apply 1 single spraying on larvae, plus 1 on adults in case of trap catches. The adjacent ones only had to apply the spraying on larvae (Figure 3).

32 12 Figure 2: Map of the ISA and of the network of traps in 2008 in the St Emilion vineyard The scouting was organised on 850ha in It was realised by a team of about 10 employees provided by the chateaux during 2 weeks and formely trained to recognise the flavescence dorée symptoms. The contaminated districts were partly prospected (at 50%) whereas the adjacent ones were totally scouted, in order to allow them to leave the ISA if no infected plant was found.

33 13 1 spraying on larvae + 1 on adults if catches 1 spraying on larvae Figure 3. Imposed spraying schedules on the Pessac-Leognan vineyard in 2009 Results and discussion Results of trapping in the Saint Emilion project In most of the ISA, a single insecticide spraying reduced trap catches to almost nonexistent, considered as a proof of sufficient control of the vector. In only one small area (< 500m), trap catches were considered too high (>3 adults on a single trap in a single week) and a second spraying (T3) was imposed. This probably was due to some farmers either not spraying at all or organic farmers that are allowed only Rotenone, that is rather inefficient. Besides, there were very few catches in zones that formerly belonged to an ISA (before 2007). This seems to show that the effect of insecticides use on population levels can last several years. This can be an argument in favor of the implementation of only one imposed spraying every two years on the zones of the ISA that are far away from infected plots. In 2008, the reduction of insecticide sprayings aiming S. titanus reached 72% thanks to the alternative approach (Figure 4). Results of trapping in the Pessac-Leognan project On the contaminated districts, almost no insect was caught in the traps and the second spraying was not launched in On the contrary, some catches were observed on the adjacent districts, but there was no second imposed spraying on this zone. This was probably due to wine-growers not spraying at all in the ISA and to the proximity of untreated zones. In 2009, the reduction of insecticide sprayings aiming S. titanus reached 54 % thanks to the alternative approach (Table 1).

34 14 A) CLASSIC PEST MANAGEMENT B) ALTERNATIVE PEST MANAGEMENT sprayings sprayings Figure 4. ISA and associated number of sprayings on the St Emilion vineyard in 2008 with a) the classic approach and b) the alternative one. Table 1. Reduction of the number of sprayings tanks to the alternative schedule compared with the classic one on the Pessac-Leognan vineyard in Classic schedule Alternative schedule Léognan (650 ha, previously infected) 2 sprayings 1 spraying Martillac (500 ha, newly infected) 3 sprayings 1 spraying Cadaujac / St Médard (270 ha, adjacent districts) 1 spraying 1 spraying Total sprayed surface in ha 1420 ha Results of scouting in both projects Scouting allowed in both cases to find infected plants but in different proportions: very few contaminated vines were found in St Emilion, whereas more significant infestations were discovered in Pessac-Leognan. On the contrary, Bois Noir seemed more present in St Emilion (Table 2). In both examples, some uncontaminated or decontaminated zones were highlighted and could go out of the ISA. According to us, a monitoring should be maintained on these areas to check if there is no recontamination. But the high cost of scouting may make it hard to accept by the wine-growers. Indeed, whereas trapping is quite cheap to implement (around 3 /ha), prospecting is much more expensive if realised using a service provider (20 /ha) (Table 3).

35 15 Table 2. Results of prospecting in the St Emilion and Pessac-Leognan vineyards St Emilion in 2007 Pessac-Léognan in 2009 Prospected area ha 850 ha Number of sampled plants Positive to FD Positive to Bois Noir Negative Outing of the ISA 2 districts in district from 2010 Table 3. Annual costs and savings of the St Emilion and Pessac-Leognan projects ANNUAL COST Total budget In sprayings (hectares) ANNUAL SAVINGS In money * Saint Emilion (26 /ha) (28 /ha) Saint Emilion (17 /ha) (19 /ha) Pessac-Léognan (10 /ha) (29 /ha) * Cost of 1 insecticide evaluated around 25 /ha Profitability, interests and limits of the approach If scouting out of ISA raises funding issues, trapping and prospecting in ISA are profitable, even from the first year of implementation: the savings made thanks to the reduction of sprayings are higher than the costs of the approach. The profitability of the alternative approach is of course an important argument to convince wine-growers to implement it, but other gains can also be promoted, such as the benefits for environment and human health. Besides the insecticide reduction and the savings, this method presents a lot of interests: - consciousness raising of the issue of flavescence dorée by the farmers - improvement of their knowledge about it and about some other diseases, - a better communication with the wine-growers about flavescence dorée management, - a better acceptance and respect of the imposed sprayings, - improvement of scientific knowledge on some questions (incubation period of the phytoplasme in plants, grape variety sensitivity ) by analysing the results of trapping or scouting. Nevertheless, this approach is difficult to implement, especially because the expenses are paid collectively by a union, whereas the savings are made at the individual level. The money is often very hard to collect, because it is collected in advance at the beginning of the year with no guarantee that savings will be done later in the year. Moreover, the savings may be different between people depending on the possible reduction of insecticide use in their zone (especially if a second spraying is launched because of trap catches). The discovery of new infected vines in plots that formerly were in ISA highlighted some limits of this management. Without maintaining insects monitoring and scouting on areas after

36 16 outing the ISA, recontaminations are possible and might not be identified. This underlined the lack of scientific knowledge concerning the incubation period of the phytoplasme in plants, the grape variety sensitivity Some new contaminations or spreading are just unexplained at the moment. Finally, if this approach can stop the spreading of the disease in infected areas and help decontaminating them, it doesn t solve the problem of introduction of infected plants from nurseries. The hot water treatment is an efficient way to prevent this problem and could be imposed to the nurseries for any plant sold in this zone. Acknowledgements We thank all the wine-growers who participated to the different projects, especially those who started them. We also thank the FREDON organism of Aquitaine for its help and collaboration in these projects. References Boudon-Padieu, E. 2000: Les jaunisses à phytoplasmes de la vigne. In: Maladies à virus, bactéries et phytoplasmes de la vigne, Editions Féret, Bordeaux, France: Boudon-Padieu, E.: Grapevine phytoplasmas. In: First internet conference on phytopathogenic mollicutes, web site Catalano, M. 2008: Dispersion de la Cicadelle de la flavescence dorée. Master degree report. Milou, C. 2008: Flavescence dorée, une augmentation des périmètres de lutte. Viti, 338: 36. Sforza, R. 2008: La cicadelle vectrice du phytoplasme de la Flavescence Dorée. In: Ravageurs de la vigne (second edition), Editions Féret, Bordeaux, France: Verpy, A., Dufour, M. C., Garcia, C. and van Helden, M. 2008: La flavescence dorée dans le Libournais: une gestion efficace est possible avec 60% de traitements en moins. Phytoma, La Défense des Végétaux 614:

37 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Effectiveness of hot water treatments against the eggs of Scaphoideus titanus Ball C. Linder, L. Schaub, F. Klötzli-Estermann Research Station Agroscope Changins-Wädenswil ACW, CH-1260 Nyon, Switzerland Abstract: The leafhopper Scaphoideus titanus (Hemiptera: Cicadellidae) is the principal vector of flavescence dorée, a severe grapevine disease spreading in Europe. The project PROPSCAPH united four European research institutes in order to improve risk management schemes against the spread of flavescence dorée's vector on propagation material. Amongst others, the impact of the standard hot water treatment against the disease on the development of the S. titanus eggs was studied. In autumn 2008, one-year-old and two-year-old cuttings were collected in S. titanus infested vineyards in France, Italy and Switzerland and sent to our institute. After storage in a cold chamber until March 2009, half of the cuttings were exposed for 45 minutes to a hot water treatment at 50 C. Thereafter, treated and untreated cuttings were placed in cages and the hatching of S. titanus nymphs was assessed. At least 10-times more nymphs hatched from two-year-old cuttings than from the one-year- old cuttings. This confirmed the ability of S. titanus to lay eggs on propagation material. However, hot water treatments significantly reduced the number of hatched nymphs, e.g. the exposure to hot water killed about 80% of deposited eggs. In conclusion, hot water treatments do not only have a major impact on flavescence dorée, they also kill most of its vector's eggs. Thus, the hot water treatment of rootstocks and scions before grafting is strongly recommended in order to reduce the risk of S. titanus' propagation. Before commercialisation grafts should be hot-water treated once again, since the chemical protection of rootlings is not always 100% effective in the nursery fields. All these measures together will ensure the propagation of plant material free of flavescence dorée and they will minimise the risk of disease s and vector's spread. Introduction Flavescence dorée is a severe quarantine grapevine diseases spreading in Europe causing in the infested areas serious economic losses. This yellows disease is caused by phytoplasmas (i.e. wall-less and phloem-restricted plant pathogenic bacteria) and it is transmitted by the leafhopper Scaphoideus titanus Ball. (Hemiptera: Cicadellidae). A direct control of the disease by antibiotics is infeasible, but the exposure of phytoplasmas to heat reduces their chance of survival significantly (Caudwell, 1966). The treatment of stocks and scions in 50 C hot water for 45 minutes is therefore strongly recommended for reducing the risk of flavescence dorée's propagation (Caudwell et al., 1990; Caudwell et al., 1997; Tassarts- Subirats et al., 2003; Mannini, 2007). According to Caudwell et al. (1997) these hot water treatments (HWT) also kill the eggs of flavescence dorée's vector, but data were not presented. In the framework of the European programme EUPHRESCO, the project PROPSCAPH aimed, amongst others, to fill this gap. The project united four European institutes from four different countries (INRA France, CRA Italy, UP ZRS Slovenia and ACW Switzerland). At Agroscope Changins-Wädenswil the efficacy of HWT against S. titanus eggs was tested. 17

38 18 Material and methods In the winter , one-year-old and two-year-old cuttings were collected in S. titanus infested, but untreated vineyards in France, Italy and Switzerland (Figure 1). I. Contone Switzerland II. Bordeaux France III. Veneto - Italy 1- and 2-year-old Cabernet Franc Figure 1. Provenance, age and variety of plant material tested. Cuttings were sent to our institute, where they were stored in a cooling chamber at 4 C ± 1 C and 95% RH until March Cuttings of each lot were homogenised and assigned randomly to the two treatments (hot-water treated or untreated control). However, the number of S. titanus eggs per cutting was not assessed in order to prevent eggs of being damaged. In March 2009, cuttings assigned to the HWT were taken out of the cooling chamber. These cuttings were humidified and stored for another 24 hours at room temperature (~15 C) in an aerated chamber. Thereafter, cuttings were treated in a RECS hot water soaking tank ( according to standard procedures used to eliminate phytoplasma (Boudon- Padieu & Grenan, 2002). The immersion time was 45 minutes at a temperature of 50 C ± 0.5 C. After 24 hours of storage at room temperature, cuttings were returned to the cooling chamber. One week later, a first series of 2 to 2.5kg of treated and untreated cuttings of each lot was placed on its own in an insect proofed cage. Cages were of 50x50x50cm dimension and they contained a potted vine of 'Chasselas' and 'Gamay' originating from areas free of S. titanus. These 16 cages were then placed in a temperate greenhouse (min. 18 C max. 25 C). Over 2 months, cages were monitored on a regular basis and the number of hatchings was recorded. In mid-may, the second series of treated and untreated cuttings was set up in these cages and the number of hatchings was once again recorded over the next 2 months. To assess the efficacy of HWT, the number of hatched nymphs per kilogram cuttings was calculated.

39 19 Results and discussion Not a single S. titanus nymph hatched from the untreated one-year-old cuttings of Switzerland, whereas a few insects hatched from the two Italian lots (Table 1). However, at least 10 times more nymphs hatched from untreated two-year-old cuttings compared to untreated one-year-old cuttings. These results confirm that females of S. titanus prefer to deposit their eggs on older parts of vines. Nevertheless, S. titanus is also able to lay its eggs on one-year old shoots. This is consistent with observations of Schvester et al. (1969). Thus, the risk to spread S. titanus via propagation material is real even if the risk can be rated as very low. HWT significantly reduced the number of hatched nymphs (Table 1). Altogether, the exposure of cuttings to hot water killed around 80% of eggs deposited. Whereas HWT killed all eggs deposited on one-year-old cuttings, nymphs hatched from each two-year-old lot. A special case is the two-year-old 'Cabernet Franc' lot from Italy. On this lot, more vectors emerged from the treated cuttings than from the untreated control. This artefact can probably be explained by the heterogeneous egg-laying behaviour of S. titanus females. Overall, HWTs failed to completely prevent the hatchings of S. titanus, which is in opposition to the results obtained by Caudwell et al. (1997). Table 1. S. titanus hatchings observed on the different lots of cuttings. Country Variety Cuttings' age S. titanus / kg cuttings Untreated HWT % reduction CH Merlot I Cabernet Franc Kober 5BB F Merlot CH Merlot I Cabernet Franc Kober 5BB In conclusion, HWTs do not only have a major impact on flavescence dorée, they also kill most of its vector's eggs. If done properly, HWTs of one-year-old scions and rootstocks do not compromise the survival of grafts (Tassarts-Subirats et al., 2003; Dupraz & Schaub, 2007; Mannini, 2007). This treatment therefore ensures that the few potential eggs laid on the parent material are killed and that treated rootlings can not only be considered free of flavescence dorée but also free of S. titanus. Subsequently, grafted grapevine rootlings should either be grown in nurseries located in areas free of S. titanus or they have to be protected with insecticides in order to avoid egg laying and the transmission of flavescence dorée by the vector. However, nurseries have to be aware that these mandatory chemical treatments are not

40 20 always 100% effective against the vector and diseases' transmission (not presented data of the PROPSCAPH project). Even if the wax used to protect the grafting point and the waxing of grafts after picking probably reduce the egg-laying opportunities, HWT of the young plants before commercialisation is strongly recommended. All these measures together will reduce the risk of the propagation of plant material contaminated with either flavescence dorée or S. titanus. Acknowledgements We thank D. Thiéry and D. Papura (INRA-Bordeaux), B. Bagnoli and E. Angelini (CRA- Italy) and M. Jermini (ACW) for the provision of plant material and for their fruitful suggestions and Isabelle Lavoie-Fleury for her technical support and her heartiness. References Boudon-Padieu, E. & Grenan, S. 2002: Hot Water Treatment. International Council for the Study of Virus and Virus-like Diseases of the Grapevine, Methods. 2 pages. Caudwell, A. 1966: L'hinibition in vivo du virus de la flavescence dorée par la chaleur. Annales des Epiphytes, 17 hors-série: Caudwell, A., Larrue, J., Valat, C. & Grenan S. 1990: Les traitements à l'eau chaude des bois de vignes atteints de la Flavescence dorée. Progrès Agricole et Viticole 107: Caudwell, A., Larrue, J., Boudon-Padieux, E. & McLean, G. D. 1997: Flavescence dorée elimination from dormant wood of grapevines by hot-water treatment. Australian Journal of Grape and Vine Research 3(1): Dupraz, Ph. & Schaub, L. 2007: Lutte contre le phytoplasme de la flavescence dorée l'eau chaude a été réinventée! Revue suisse Vitic. Arboric. Hortic. 39(2): Mannini, F. 2007: Hot water treatment and field coverage of mother plant vineyards to prevent propagation material from phytoplasma infections. Bulletin of insectology 60(2): Schvester, D., Carle, P. & Moutous, G. 1969: Nouvelles données sur la transmission de la Flavescence dorée de la vigne par S. littoralis Ball. Annales de Zoologie et d'ecologie Animale 1: Tassarts-Subirats, V., Clair, D., Grenan S., Boudon-Padieux, E. & Larrue J. 2003: Hot water treatement: curing efficiency for phytoplasma infection and effect on plant multiplication material. 14 th ICVG Conference, Locorotondo, th September, 2003:

41 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Downy, powdery mildew and botrytis risk assessment by climate parameters, pest and spores monitoring in Rioja Alavesa vineyard (Northwest of Spain) A.-M. Díez-Navajas, A. Ortiz-Barredo Department of Plant Production and Protection. NEIKER-Tecnalia, Box 46, E Vitoria- Gasteiz., Spain Abstract: The integrated production Regulation for grapevine in the Basque Country was published in 2004 (BOPV 221/2004, official bulletin). A previous 3-year study was carried out to evaluate the real situation of growers to assume this regulation in their exploitations. Several deficiencies were reported, including technical and management aspects. This work is focused in Rioja Alavesa, an area with 12,869ha of vineyard in the Northwest of Spain producing 81,325,000 wine grape kilo. Integrated disease control interest is rising, although the number of fungicides applications in the last two years against downy mildew and other fungal diseases reached up to 12 treatments, being usual 5 or 6 per season. These data prompted us to implement the strategy of fungal disease control. We installed 12 monitoring weather stations in different trials, according to different environment and geographical characteristics. The stations were equipped with sensors of temperature, relative humidity, leaf wetness, soil temperature, radiation sensors, rain gauge, and an anemometer. Registered data can be discharged and recuperated from internet, at the same time that the risk assessment for downy and powdery mildew and botrytis diseases. In parallel, one sticky glass trap to capture spores by impact were installed near each station and two more traps in each field. Pheromone traps to capture Lobesia botrana were also distributed in the fields close to the spore traps, as the abundance of this insect gives an idea of the possible future botrytis damages and spread. After these assays we would be able to validate the stations for each studied zone and assure that the risk assessment made by climate parameters follows a real model, taking into account the disease incidence in the field as well as the spore and vector populations. Key words: Downy mildew, powdery mildew, botrytis, disease risk assessment, agrometeorological station, Lobesia botrana, Empoasca vitis Introduction Rioja Alavesa is situated in the South of Álava, one of the three provinces setting up Basque Country in the Northwest of Spain. Rioja Alavesa is a wine producing region with 12,869ha of vineyard and harvesting over 81,325,000 wine grape kilo. The Regulation for grapevine integrated production in the Basque Country, published in 2004 in the BOPV (Boletín Oficial del País Vasco) Regional bulletin (BOPV number 221, 2004), contributed with several guidelines to the producers interested in integrated production. These guidelines involved the registration of parcels for phytochemical management, equipments and staff, preparation and handling of soil, origin, age, cultivar and rootstock, pollination and environmental protection, fertilization, irrigation and harvest. The rules included the possession of a notebook where all the activities during the corresponding season in reference to the grapevine culture must be noted. The interest of this notebook was based in the traceability of the crop, interesting for wineries and also to receive 21

42 22 incentives from governments and public entities, more and more interested in reducing the number of phytochemical applications. We wanted to know the real situation of grapevine growers in Alava and evaluate if published Regulation for grapevine integrated production could be applied in the exploitations. To discern it, a study involving 14 wine farms with a total of 375 parcels (170ha) and 4 wineries with 750ha, was taken during 3 years in Alava. The guidelines in the rules shown in Table 1 were evaluated. After the study, several deviations from the rules and deficiencies were found in the exploitations (Table 1). In 2008, the Regulation was updated (BOPV number 43) in base to some difficulties for growers. Table. 1. Deviation from the rules and cause of this deviation observed after a 3 year study. Item in the Regulation Yes Deviation Register of parcels X Difficulty Exploitation: Installation, equipments and staff Soil: preparation and handling Vegetal material: origin, age and planting Pollination and environmental protection X No X X X Why? - Deficient personal protection: safety glasses, suit/overall, gloves - No applicator licence: 15% Fertilization X - Less than recommended: 56% - Higher: 12% Irrigation X Not usual in Rioja Alavesa Phytochemical management: application and residues Harvest X - Residues in samples: 15% - Treatment machinery no suitable: 45% - No authorized products: 40% Notebook X - No: 10% - Deficient: 40% X Farmers had to solve deviations and deficiencies from the rules, as phytochemical product dose, personal protection deficiencies or machinery checking. Our aim was to help them to reduce the number and timing of applications by studying the pest and pathogen dynamics. For this, we installed weather stations to assess downy and powdery mildew and botrytis risk by climate parameters, spore traps to compare the presence or absence with the risk models offered by the stations, and traps to monitor populations of insects present in the vineyard.

43 23 In this report, we show preliminary results about dynamics of insects populations, disease assessment based on spore data and its relationship with risk models offered by the weather stations in Rioja Alavesa vineyard. Material and methods Weather stations 12 weather stations imetos (Pessl Instruments GmbH) were installed in 12 different and geographical emplacements to monitor environmental parameters corresponding to 12 parcels. The stations had several devices to measure its correspondent parameter: an anemometer for wind speed (m/sec), air and soil temperature sensors ( C), a relative humidity sensor (%), a rain gauge for precipitation (mm), a solar radiation sensor (W/m 2 ), a leaf wetness sensor (min), and tensiometers for soil matricial potential (cbar). Climate data download and disease risk models were given by the MetwinII software of the imetos station system. This software offers the risk models for downy and powdery mildew and botrytis in grapevine. For the downy mildew risk model, it uses the climatical parameters of precipitation, leaf wetness, air temperature and relative humidity. It is possible to obtain information about primary (sexual) and secondary (asexual) infections of the cycle. Air temperature and leaf wetness data are used for the powdery mildew and botrytis risk models. Powdery mildew risk model is based on the secondary cycle, so the model shows the date for possible infections by ascospores. Traps for insect capture Two kinds of sticky traps were installed to capture insects: (1) chromotrophic traps, yellow and blue to capture Tetranychus urticae, Panonychus ulmi, Lobesia botrana, Eupocelia ambiguella, Sparganothis pilleriana and Empoasca vitis, and (2) pheromone traps (Trècè, Kenogard) to capture Lobesia botrana. Chromotrophic traps were replaced once a week and insects over them identified and counted. Pheromone capsules to capture L. botrana males were replaced once a month, following manufacturer s instructions, and once a week the number of male captured was reported. The total counting of insects for each parcel at the same date was calculated adding up the number recorded in the three devices installed in the same parcel. Traps for spores To capture airborne spores of downy and powdery mildews and botrytis, sticky glass traps were installed in supports for them to be oriented towards the four cardinal points (Iturritxa and Ganley, 2007). At least 3 supports were put in each parcel, at 65 and 100cm from soil. Traps were replaced twice a week. Microscopy to identify and count the number of spores of interest After acidic lactofucsine staining (0.1% lactofucsine in lactic acid) traps were observed under visible light at microscope. Spores for each species were identified and the number reported. Phenological stages Phenological stages during the season were reported at the same day when other data were taken. We followed the growth stage keys described by Eichhorn and Lorenz (EPPO, 1984). In the text we use the abbreviation GSK for growth stage key.

44 24 Results and discussion Spore detection and disease risk assessment In this study we only show the results obtained in four stations, in the parcels named Matacaballos, las Muñecas, Durana and Villabuena. We found a lag of several days from the peak in spore counting respect to disease risk for the three studied pathogens (Figure 1). A MATACABALLOS LAS MUÑECAS DURANA VILLABUENA

45 25 B MATACABALLOS LAS MUÑECAS DURANA VILLABUENA

46 26 C MATACABALLOS LAS MUÑECAS DURANA VILLABUENA Figure 1. Risk models for downy mildew (A), powdery mildew (B) and botrytis (C) obtained from the weather stations in the parcels of Matacaballos, Las Muñecas, Durana and Villabuena. Arrows indicate a peak detection of spores at that moment along the vegetative period. Graphics of downy mildew (A) correspond to asexual cycle.

47 27 We evaluated the presence of downy mildew sporangia (asexual spores). The weather station software uses the recorded data of precipitation, leaf wetness, air temperature and relative humidity to calculate a risk model. In the four parcels, it was observed a lag of several days from the first peak of spores up to the infection risk (Figure 1A), the detection of spores occurring earlier than the calculated risk. The software determined the powdery mildew and botrytis risk infection models using air temperature and leaf wetness. The spore peak for powdery mildew (Figure 1B), the same as for downy mildew, was earlier than slope risk advice started up. For botrytis (Figure 1C), only in Las Muñecas, spore peaks appear at the same or almost at the same time than the slope risk model. In Matacaballos and Durana risk models were also delayed respect to the spore peaks. In any case, botrytis risk did not exceeded 40%. In Villabuena, the model showed the same shape as for the other disease risk models. In a near future, we would like to compare and assess the risk models offered by the weather stations installed in our parcels, with other models calibrated and validated on agrometeorological data collections (Orlandini et al., 2008) for sexual (Rossi et al., 2007) and asexual cycles, always compared to the presence or absence of spores until validation in our area. Disease spread depends on the weather conditions, and their detection is related to the favourable conditions of dispersion. Spore presence may assure information obtained from agrometeorological stations about the date in which disease risk was detected, validating obtained information from stations for the area where they were installed, and giving the possibility to use this information in an agricultural warning service. So advices informing about disease risks might be broadcasted, in order to apply phytochemicals treatments only when necessary, contributing in this way to reduce the phytochemical inputs into the environment. Pest monitoring To compare the number of males of Lobesia botrana captured in each parcel, we elaborated graphics with total number of specimens from each parcel (Figure 2). Shown data corresponding to grapevine moth were reported in the same parcels as the weather station data. The shape in the graphics was quite similar in Durana and Las Muñecas, having the first peak in the phenological moment from inflorescence swelling and flowers closely pressed together (GSK 16) to inflorescence fully developed and separated flowers (GSK 17), and the second one in the beginning of ripening (GSK 32). Curve of Durana presents an intermediate cycle in full flowering (GSK 23). Respect to Matacaballos and Villabuena the first peak was between fully developed inflorescence (GSK 17) and full flowering (GSK 23), and the second in the beginning of ripening (GSK 35). So a lag in the flight of grapevine moth is evident, probably due to the air temperature, as Durana and Las Muñecas are situated in warmer areas. Respect to other pests, practically only Empoasca vitis specimens were observed and counted over the coloured traps. Otherwise, the population captured over the yellow traps was larger than over the blue ones, so results are shown from yellow traps and for green leafhopper, in the parcels Páganos and Rodezno. We show the difference of E. vitis populations between these two parcels in the figure 3. Along the vegetative period the population in Rodezno is larger than in Páganos. Three cycles appear: (1) during cracking, this is when first indications of growth appeared (GSK 2), (2) in the period from bloom (GSK 23) to berry touch (GSK 33), and (3) the last one overlapped with the end of the second one, from berry touch (GSK 33) to veraison (GSK 36). We think

48 28 the difference between the numbers of specimen trapped in all the parcels might be due to the differences in relative humidity and air temperature in each one. For example, in Rodezno, where population is higher, relative humidity is up than in Páganos and just the opposite for air temperature Lobesia botrana DURANA LAS MUÑECAS MATACABALLOS VILLABUENA number /04/ /04/ /05/ /05/ /05/ /06/ /06/ /07/ /07/ /08/2009 date Figure 2. L. botrana distribution in Durana, Las Muñecas, Matacaballos and Villabuena. Empoasca vitis number Rodezno Páganos 17/04/ /05/ /05/ /05/ /06/ /06/ /07/ /07/ /08/ /08/2009 date Figure 3. Green leafhopper specimens recorded in two different parcels: Rodezno and Páganos.

49 29 We compiled shown results during one season, so successive studies must be done to confirm the population dynamics of green leafhopper and grapevine moth, as other authors made during successive growing seasons (Decante and van Helden, 2008; Sciarretta et al., 2008). We will continue the study of pest capture to be able to predict the spatial and temporal distribution of them in our parcels. References Decante, D. & van Helden, M. 2008: Spatial and temporal distribution of Empoasca vitis within a vineyard. Agricultural and Forest Entomology 10: EPPO. 1984: Crop growth stage keys. EPPO Bull. 14(2): Iturritxa, E. & Ganley, R Dispersión por vía aérea de Diplodia pinea en tres localidades de la cornisa cantábrica. Bol. San.Veg. Plagas. ISSN , Vol. 33, Nº 3, 2007: Orlandini, S., Massetti, L., Dalla Marta, A. 2008: An agrometeorological approach for the simulation of Plasmopara viticola. Computers and electronics in agriculture 64: Rossi, V., Caffi, T., Giosuè, S., Bugiana, R. 2008: A mechanistic model simulating primary infections of downy mildew in grapevine. Ecological modelling 212: Sciarretta, A., Zinni, A., Mazzocchetti, A., Trematerra, P. 2008: Spatial analysis of Lobesia botrana (Lepidoptera: Tortricidae) male population in a mediterranean agricultural landscape in central Italy. Enviromental entomology 37(2):

50 30

51 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Biological control of Heliococcus bohemicus Sulc. (Hemiptera: Pseudococcidae) with the natural enemy Ericydnus sipylus (Walker) (Hymenoptera: Encyrtidae) G. Sentenac 1, P. Kuntzmann 2, L. Perez 3, A. Gili 3, P. Kreiter 3 1 Institut Français de la Vigne et du Vin, Pôle Bourgogne Beaujolais Jura Savoie, Beaune, France. 2 Institut Français de la Vigne et du Vin, Pôle Alsace, Biopôle, Colmar, France. 3 Institut National de la Recherche Agronomique, Unité Expérimentale de Lutte Biologique, Valbonne, France. Abstract: The control of mealybugs by growers currently relies on the use of neurotoxic insecticides. The possible use of the predator Chrysoperla lucasina (Lacroix) is the unique alternative method. We propose to evaluate the biological activity in situ of the bred Ericydnus sipylus (Walker), major parasitoid of Heliococcus bohemicus Sulc. The releases of parasitoids were carried out in 2007 and During the 2007 season, the first three releases of E. sipylus did not appear to have any effect on the population of mealybugs. However the fourth and fifth release made later in the season did show significant differences between the experimental plots. The biological control showed positive results in 2008 and 2009, as the number of H. bohemicus in the release blocks was significantly lower than those of the control blocks. A reduction of 50 to 60% of the number of H. bohemicus was found for the 2007 winter generation and from 75 to 85% for the 2008 summer generation. Despite of stopping the release of E. sipylus rates of 80% were found in the 2008 winter generation, and 80% in early 2009 summer generation. If the difference of the population of Pseudococcidae between the experimental plots was due to the introduction of the parasitoid E. sipylus, the lower level of the population found in summer 2008 and 2009 in the control blocks was probably due to: a) the presence of other parasitoids like Anagyrus szodensis or predators like Exochomus quadripustulatus (Linné), Hippodamia variegata Goeze, Harmonia axyridis (Pallas), Coccinella septempunctata Linné, Nephus quadrimaculatus (Herbst) and Chrysoperla carnea (Stephens) b) the dispersion and activity of E. sipylus beyond the plots where it was introduced. Key words: biological control, mealybugs, Pseudococcidae, Heliococcus bohemicus, Ericydnus sipylus Résumé: Les stratégies mises à la disposition des viticulteurs pour lutter contre les cochenilles farineuses, reposent sur l emploi d insecticides neurotoxiques. La seule méthode alternative évaluée repose sur l utilisation du prédateur généraliste Chrysoperla lucasina (Lacroix). Nous nous proposons, en conditions de plein champ, d évaluer l activité biologique de l Encyrtidae Ericydnus sipylus (Walker), parasitoïde majeur d Heliococcus bohemicus Sulc dont l élevage est maîtrisé. Les apports de parasitoïdes sont effectués en 2007 et Si les trois premiers lâchers réalisés en 2007 semblent ne pas avoir eu d effet sur les populations de cochenilles, les quatrièmes et cinquièmes lâchers engendrent des différences entre modalités statistiquement significatives. Les efficacités sont modestes de l ordre de 35 et 45%. Le contrôle biologique est plus soutenu au cours des campagnes 2008 et 2009, la modalité «lâcher» est alors significativement inférieure à la modalité «témoin». Une réduction de 50 à 60 % des populations est obtenue sur les femelles et larves s de la génération 2007, elle atteint des valeurs comprises entre 75 et 85% sur les jeunes larves de la génération L efficacité augmente encore en 2009, année où les lâchers ont été interrompus: 80% sur les femelles et larves de la génération 2008, 80 à 90% sur les jeunes larves de la génération

52 32 Si les différences de densités d H. bohemicus entre modalités sont dues à l efficacité du parasitoïde introduit, le faible niveau des populations rencontrées au cours de l été 2008 et durant l année 2009 dans la modalité «témoin» est certainement imputable pour une part à la présence des prédateurs suivants: Exochomus quadripustulatus (Linné), Hippodamia variegata Goeze, Harmonia axyridis (Pallas), Coccinella septempunctata Linné, Nephus quadrimaculatus (Herbst) et Chrysoperla carnea (Stephens), pour une autre part à la dispersion et à l activité d Ericydnus sipylus bien au-delà des zones où il a été introduit. Mots clés: lutte biologique, cochenilles farineuses, Pseudococcidae, Heliococcus bohemicus, Encyrtidae, Ericydnus sipylus Introduction La volonté de réduire de moitié l usage des produits phytopharmaceutiques rend prioritaire la nécessité de disposer de méthodes alternatives comme la lutte biologique, entre autre. Le cortège des parasitoïdes d Heliococcus bohemicus Sulc est à présent mieux connu, Ericydnus sipylus (Walker), Anagyrus szodensis Erdös et Leptomastidea bifasciata Mayr sont les auxiliaires qui assurent la plus grande part de la régulation. Son élevage étant maîtrisé, l évaluation d une lutte biologique par augmentation d E. sipylus a été envisagée. Matériel et méthodes Le dispositif expérimental retenu est un essai en blocs éclatés à 6 répétitions implanté sur le site de Strohbach à proximité de Colmar (Haut-Rhin-France). Avant d identifier les parcelles élémentaires, une estimation du niveau des populations est réalisée le 6 et le 14 septembre 2006 sur différentes zones du site expérimental d une superficie de 0.45 hectare. L écartement est de 1.60 mètres, l espacement de 1.35 mètres. Douze parcelles élémentaires de cinq ceps de Riesling chacune (1 rang x 5 ceps) sont sélectionnées parmi les 17 qui ont fait l objet d une évaluation des populations d H. bohemicus le 21 septembre 2006, observations portant sur 10 feuilles par cep. Les blocs sont formés, a posteriori, de parcelles ayant des niveaux d occupation équivalents. La validation de ce dispositif repose donc sur l'analyse statistique des résultats, nombre de cochenilles par lot de 10 feuilles, issus de l observation réalisée le 21 septembre Les lâchers programmés d E. sipylus adultes, réalisés en 2007 et 2008, sont au nombre de deux sur la génération n-1 et de quatre (toutes les 5 semaines dès présence d au moins 30% de larves dont la taille est à 1mm) sur la génération n d H. bohemicus. Les parasitoïdes ( + ) sont conditionnés à raison de 10 à 50 individus par tube en plastique (bouchon aéré) et expédiés par transporteur en moins de 24 heures. La densité des lâchers sera liée au niveau de production de l élevage, probablement aux alentours de plusieurs dizaines d E. sipylus par cep. En début de campagne (génération n-1), de début avril (gonflement du bourgeon - 02) à mi-mai (8 à 9 feuilles étalées- 16) les observations portent sur le bois d'un an ainsi que sur l'ensemble de la végétation (feuilles, rameaux) de chaque cep. Par la suite (génération n), les notations ne concernent que 10 feuilles par cep, 5 sur chaque face, de rang 1 à 5. La variable observée est le nombre de cochenilles H. bohemicus par cep ou par lot de 10 feuilles. Leur taille, lors des différents dénombrements, est évaluée au moyen d un papier millimétré transparent.

53 33 Les résultats des différents comptages, exprimés en nombre de cochenilles par cep ou lot de 10 feuilles, sont soumis à une analyse statistique du type comparaison des moyennes, méthodes des couples, test unilatéral, α = 5%. Afin de déterminer le niveau de parasitisme à un instant donné et d apprécier indirectement la dispersion d E. sipylus, des H. bohemicus sont prélevés, hors parcelles élémentaires et élevés à 24 C, dans des boîtes de 60mm de diamètre, à raison de cinq larves âgées ou femelles par boîte d élevage qui comprend un amas de coton humidifié et un disque foliaire (Ø = 26mm) de marronnier. Le coton est ré-humidifié deux fois par semaine, le disque foliaire est changé selon le même rythme, ceci afin d augmenter la durée de vie des hôtes. Deux observations par semaine sont effectuées au minimum, chaque momie observée est placée individuellement dans un tube jusqu à l émergence du parasitoïde qui est identifié. Résultats et discussion Validation du dispositif Le 21 septembre 2006 avant traitement, un suivi des populations de cochenilles a été réalisé sur 17 parcelles élémentaires (PE) de 5 ceps chacune. Nous pouvons à partir de ces dernières, constituer les 6 répétitions des deux modalités prochainement à l'étude, les blocs étant formés, a posteriori, de parcelles élémentaires présentant des densités d H. bohemicus équivalentes. La variable observée est le nombre de cochenilles, H. bohemicus, par lot de 10 feuilles (voir Tableau 1). t0.95 = 2.015, l hypothèse H 0 égalité des moyennes est acceptée. L analyse statistique (comparaison des moyennes, méthodes des couples) ne met pas en évidence de différence significative, le dispositif expérimental en blocs éclatés est validé. Tableau 1. Nombre d H. bohemicus par lot de 10 feuilles le 21 septembre 2006 Bloc 1 mod. lâcher 2 témoin différences BI BII BIII BIV BV BVI PE non retenues PE non retenues PE non retenue moyenne SCE t obs

54 34 Évaluation de l efficacité de la lutte biologique par augmentation En 2007 et 2008, les deux lâchers programmés sur la génération n-1 ont été effectués respectivement les 12 et 25 avril, les 23 avril et 06 mai. Les lâchers ciblant la génération n sont mensuels, de juin à septembre. A chaque dépôt, 20 à 60 E. sipylus sont apportés par cep, ainsi 7200 parasitoïdes ont été introduits en 2007, 7050 en La mortalité «postréception» évaluée par le dénombrement des cadavres récents de parasitoïdes retrouvés dans les tubes de chasse deux à trois heures après la pose, est très faible, inférieure à 5%. Ces données sont reprises dans le tableau 2. Le conditionnement (boîte isotherme en polystyrène, bloc réfrigérant, microgouttes de miel) a permis une excellente survie des agents de lutte biologique. Tableau 2. Caractéristiques des lâchers réalisés en 2007 et 2008 date d expédition (INRA Antibes) date de réception (ITV Colmar) stade phénologique nombre d E. sipylus/tub e nombre d E. sipylus/cep % mortalité «postréception» lâcher 1/6 11 avril avril 2007 éclatement du bourg % 06 lâcher 2/6 24 avril avril à 6 fles. étalées % lâcher 3/6 19 juin juin 2007 ferm. grappe % lâcher 4/6 24 juillet juillet 2007 début véraison % lâcher 5/6 28 août août % véraison 37/ % lâcher 6/6 24 sept sept post-récolte 20 ou % lâcher 1/6 22 avril avril 2008 pointe verte et % lâcher 2/6 05 mai mai à 3 fles. étalées % lâcher 3/6 24 juin juin 2008 baies taille de pois % 31 lâcher 4/6 29 juillet juillet ère baie vérée % lâcher 5/6 26 août août % véraison % lâcher 6/6 23 sept sept pré-récolte % Sachant qu E. sipylus obtient un nombre optimal de descendants quand il parasite des H. bohemicus de classe 2: 1mm taille < 2 mm ou classe 3: 2mm taille < 3mm voire classe 4: 3mm taille, (Pérez, 2006), les lâchers ont été programmés en fonction de la dynamique et de la structure de la population d H. bohemicus. Ces derniers ont été effectués dans des conditions idéales au regard de la répartition des effectifs dans les classes (voir tableau 3), les lâchers «estivaux» débutent dès que 30% des larves ont atteint la classe 2.

55 35 Tableau 3. Structure de la population d H. bohemicus au moment des lâchers lâcher n classe 1 classe 2 classe 3 classe 4 0 % 10% 80% 10% lâcher n 2 0% 0.2% 23.2% 76.6% lâcher n % 65.3% 0.3% 0% lâcher n 4 5.5% 72.5% 21.8% 0.2% lâcher n 5 9.7% 76% 14% 0.3% lâcher n 6 0% 38.9% 60.9% 0.2% lâcher n % 37.7% 53.1% 9.1% lâcher n 2 0% 3.6% 32.8% 63.6% lâcher n 3 70% 30% 0% 0% lâcher n % 69.1% 2.7% 0% lâcher n 5 0% 64% 36% 0% lâcher n 6 0% 34.8% 58.4% 6.7% Les momies (stade nymphal de l auxiliaire) sont obtenues de 11 à 73 jours après l acte de parasitisme (Sentenac et al., 2005). Deux mois après le premier lâcher, le suivi de la dynamique des populations devrait permettre de mettre en évidence l efficacité de la lutte biologique par augmentation. La reprise d activité des H. bohemicus, en tout début du mois d avril, est très précoce lors de la première année d expérimentation. Le 10 avril la densité dépasse les 50 cochenilles par cep, mais les nombres de cochenilles rencontrées dans les modalités témoins et lâchers ne sont pas différentes statistiquement. Il en est de même le 20 avril, huit jours après le premier lâcher, ainsi que lors des comptages portant sur la génération 2007 réalisés le 31 mai, le 20 juin, le 12 et le 25 juillet (voir tableau 4 et figure 1). Les trois premiers lâchers, deux sur la génération 2006 et un sur la génération 2007, n ont pas eu d effet sur la dynamique des populations d H. bohemicus. Il faut attendre le 29 août, un mois après le quatrième lâcher, et le 25 septembre, un mois après le cinquième lâcher de la campagne pour mettre en évidence des différences statistiquement significatives entre le nombre de H. bohemicus des modalités témoins et lâchers. Les efficacités rencontrées à ces deux dates sont modestes de 35 à 45%. La situation évolue en 2008 (voir tableau 4 et Figure 1). Le nombre de H. bohemicus est significativement inférieur dans la modalité «lâcher» par rapport à la modalité témoin durant toute la seconde année d étude c est à dire de la première notation du 16 avril 2008 à la dernière notation du 15 octobre Les efficacités obtenues sont comprises entre 50 et 60% sur la génération 2007, 75 et 87% sur la génération 2008, elles sont donc bien supérieures à celles rencontrées en En 2009, en l absence de lâchers, les différences entre les deux modalités sont encore statistiquement significatives. Les efficacités obtenues sont supérieures à celles des années précédentes: 80% sur la génération 2008, de 80 à 90% sur le début de la génération 2009 en mai et juin.

56 36 Tableau 4. Nombre d H. bohemicus par cep ou par lot de 10 feuilles Génération 2006 Différence Date Lâcher Témoin Efficacité de Stade phénologique moyennes 04 avril 2007, gonflement des bourgeons n.s. 10 avril 2007, bourgeons dans le coton n.s. 12 avril 2007, éclatement du bourgeon 1 er lâcher d'e.sipylus 20 avril 2007, 3-4 feuilles étalées % n.s. 25 avril 2007, 5-6 feuilles étalées- 2 ième lâcher d'e. sipylus Génération mai 2007, nouaison, grain de plomb % n.s. 20 juin 2007,fermeture de la grappe % n.s. 20 juin 2007, fermeture de la grappe 3 ième lâcher d'e. sipylus 12 juillet 2007, fermeture de la grappe % n.s. 25 juillet 2007, début véraison % n.s. 25 juillet 2007, début véraison 4 ième lâcher d'e. sipylus 29 août 2007, 100 % véraison maturité % n.s. 29 août 2007, 100 % véraison maturité 5 ième lâcher d'e. sipylus 25 septembre 2007, post récolte % s. 25 septembre 2007, post récolte 6 ième lâcher d'e. sipylus 10 octobre 2007, début chute des feuilles % n.s. Génération avril 2008, bourgeon dans le coton (03) % s. 23 avril 2008, pointe verte (05) %- s. 23 avril 2008, pointe verte (05) 1 er lâcher d'e. sipylus 6 mai 2008, 2-3 feuilles étalées (09) % n.s. 6 mai, 2-3 feuilles étalées (09) 2 ième lâcher d'e. sipylus 21 mai 2008, 8-9 feuilles étalées (16) % n.s.

57 37 Génération 2008 Différence Date Lâcher Témoin Efficacité de Stade phénologique moyennes 4 juin 2008, tout début floraison (19) % s. 25 juin 2008, baies à taille de pois (31) % s. 25 juin baies à taille de pois (31) 3 ième lâcher d'e. sipylus 16 juillet, fermeture de la grappe % s. 30 juillet 2008, toute 1 ère baie vérée % s. 30 juillet 2008, toute 1 ère baie vérée 4 ième lâcher d'e. sipylus 27 août 2008, véraison % s. 27 août 2008, véraison 5 ième lâcher d'e. sipylus 16 septembre 2008, 100% véraison/maturité % s. 24 septembre2008, maturité pré-récolte % s. 24 sept. 2008, maturité pré-récolte 6 ième lâcher d'e. sipylus 1 er octobre 2008, post-récolte % s. 15 octobre 2008, chute des feuilles % s. G avril 2009, 2 à 3 feuilles étalées % s Génération mai 2009, début floraison % s 16 juin % n.s. 24 juin 2009, baie taille pois % s 09 juillet 2009, fermeture de la grappe % s 30 juillet 2009, début véraison % s 27 août 2009, fin véraison % n.s. 09 septembre 2009, maturité % s 22 septembre 2009, maturité % s Activité et dispersion de l agent de lutte biologique Ericydnus sipylus Des prélèvements destructifs ont été réalisés sous les écorces pour quantifier les larves âgées et les jeunes femelles en mars-avril, et les femelles en tout début de période de parturition en mai. Ils permettent, suite à l élevage des H. bohemicus, de déterminer le niveau de parasitisme à un instant donné et d apprécier indirectement la dispersion d E. sipylus. Les ceps faisant partie du dispositif expérimental ne font pas l objet de prélèvement. Dans la zone «lâcher» nous avons prélevé en avril 2008 sur les ceps adjacents aux ceps inoculés puis sur les ceps situés à moins de 10 mètres d un cep inoculé. Dans la zone témoin, les prélèvements sont faits sur les ceps situés aux alentours de 40 à 60 mètres du cep inoculé le plus proche. La figure 2 reprend les résultats obtenus. En avril 2008, 102 jeunes femelles ou larves âgées ont été prélevées dans chaque zone. Dans la zone «lâcher» le taux de parasitisme, à mettre à l actif d E. sipylus, est de 12%, il est de 8% dans la zone «témoin» où A. szodensis est le parasitoïde majeur. En mai 2008, 30% des

58 38 femelles âgées sont parasitées dans la zone «lâcher», le parasitoïde majeur est E. sipylus, son activité est ainsi prouvée. Seulement 10% des femelles prélevées à la même date dans la zone «témoin» sont parasitées, A. szodensis et E. sipylus y sont actifs à part égale. En 2009 des difficultés ont été rencontrées pour prélever 100 H. bohemicus, le temps de récolte a donc été limité à 4 ou 5 heures dans la zone «lâcher», à 2 heures dans la zone «témoin». En mars les taux de parasitisme sont très voisins 42.5% et 38.2%, E. sipylus est toujours le parasitoïde majeur dans la zone «lâcher», son activité est plus discrète dans la zone «témoin» où le parasitoïde majeur est à nouveau A. szodensis. En mai 2009 la récolte de 100 H. bohemicus n est possible que dans la zone «témoin», le niveau des populations étant trop bas dans la zone «lâcher». Les taux de parasitisme sont voisins 25.8% et 23.5%, respectivement dans la zone «témoin» et la zone «lâcher». E. sipylus est le parasitoïde majeur dans les deux zones, sa dispersion au sein du site expérimental est donc effective. N om b re d e P seu dococcid aepar c ep o u p ar lot d e 1 0 fe uilles E ssa i d e lutte biologique contre H. bo hem icus a u m o yen d 'E. sipylus S troh bach : L â cher T ém oin L âchers â d 'E. sipylus génération 2006 génération 2007 génération 2007 génération génération 2008 génération /1/07 31/1/07 2/3/07 1/4/07 1/5/07 31/5/07 30/6/07 30/7/07 29/8/07 28/9/07 28/10/07 27/11/07 27/12/07 26/1/08 25/2/08 26/3/08 25/4/08 25/5/08 24/6/08 24/7/08 23/8/08 22/9/08 22/10/08 21/11/08 21/12/08 20/1/09 19/2/09 21/3/09 20/4/09 20/5/09 19/6/09 19/7/09 18/8/09 17/9/09 17/10/09 T em ps Figure 1. Essai de lutte biologique contre H. bohemius au moyen d E. sipylus à Strohbach: 2007, 2008, 2009.

59 39 T a ux d e p a rasitism e et a ctivité d es p a rasitoïdes P rélèvements d estructifs A c t iv ité d es p arasito ïd es % 7.8 zone"témoin" (n = 102) autres Anagyrus szodensis Ericydnus sipylus taux de parasitisme zone "Lâcher" (n = 102) zone"témoin" (n = 99) 29.7 zone "Lâcher" (n = 185) 38.2 zone"témoin" (n = 68) 42.5 zone "Lâcher" (n = 40) zone"témoin" zone "Lâcher" (n = 90) (n = 17) % T a ux d e p arasitism e avr-08 mai-08 mars-09 mai-09 Figure 2. Taux de parasitisme et activité des parasitoïdes après prélèvements destructifs Décroissance des populations dans la modalité témoin Dans la modalité témoin, le pic des populations en septembre 2006 était de 72 H. bohemicus/ lot de 10 feuilles, en septembre 2007, après une campagne de lâcher, ce pic est de 73 H. bohemicus/lot de 10 feuilles, en septembre 2008 il n est plus que de 24 H. bohemicus/lot de 10 feuilles, pour atteindre en septembre 2009 la valeur de 3 H. bohemicus/lot de 10 feuilles. L activité du parasitoïde A. szodensis renforcée par celle d E. sipylus après dispersion à partir de mai 2008 peut en partie expliquer la baisse des populations rencontrée dans la zone «Témoin». La présence dès mai 2008 d insectes coccidiphages et leur maintien, pour certains, sur le site expérimental en 2009, ont indéniablement participé à l érosion de la population d H. bohemicus. Citons les coccinelles Exochomus quadripustulatus (Linné), Hippodamia variegata Goeze, Harmonia axyridis (Pallas), Coccinella septempunctata Linné, Nephus quadrimaculatus (Herbst) et la chrysope Chrysoperla carnea (Stephens). N. quadrimaculatus a été retrouvé fréquemment en 2009, au stade adulte et au stade larvaire. Pour indication 18 larves de N. quadrimaculatus ont été prélevées lors de la collecte des 90 femelles d H. bohemicus sous les écorces le 27 mai 2009, soit un rapport de 1 pour 5! Conclusions Si les trois premiers lâchers réalisés en 2007 semblent ne pas avoir eu d effet sur les populations de cochenilles, les quatrièmes et cinquièmes lâchers engendrent des différences entre modalités statistiquement significatives. Les efficacités de parasitisme sont modestes de l ordre de 35 et 45%. Le contrôle biologique est encore plus important au cours des campagnes 2008 et 2009, la modalité «lâcher» est alors significativement inférieure à la

60 40 modalité «témoin». Une réduction de 50 à 60% des populations est obtenue sur les femelles et larves de la génération 2007, elle atteint des valeurs comprises entre 75 et 85% sur les jeunes larves de la génération L efficacité de parasitisme augmente encore en 2009, année où les lâchers ont été interrompus: 80% sur les femelles et larves de la génération 2008, 80 à 90% sur les jeunes larves de la génération Si les différences de densités d H. bohemicus entre modalités sont dues à l efficacité du parasitoïde introduit, le faible niveau des populations rencontrées au cours de l été 2008 et durant l année 2009 dans la modalité «témoin» est certainement imputable pour une part à la présence des prédateurs suivants: Exochomus quadripustulatus (Linné), Hippodamia variegata Goeze, Harmonia axyridis (Pallas), Coccinella septempunctata Linné, Nephus quadrimaculatus (Herbst) et Chrysoperla carnea (Stephens), pour une autre part à la dispersion et à l activité d Ericydnus sipylus bien au-delà des zones où il a été introduit. Remerciements Cette étude a reçu un soutien financier de la part du Conseil Régional de Bourgogne et de FranceAgriMer. La réalisation de ce projet a été possible grâce à la collaboration d Isabelle Pavoine*, Christina Otel *, Damien Leplus*, Thuy Pham**, Valentin Gertz*, Pierre Feuillet* et Olivier Delloye*** *: CDD Institut Français de la Vigne et du Vin, unité expérimentale de Beaune **: CDD, Station Régionale Bourgogne Centre Est ***: stagiaire, Licence des Sciences de la Vigne, Institut Jules Guyot, Université de Bourgogne Références Pérez, L. 2006: Étude des potentialités biologiques d Ericydnus sipylus (Walker), parasitoïde de la cochenille de la vigne Heliococcus bohemicus (Sulc). Mémoire de fin d études pour l obtention du Master II Professionnel, mention Systèmes Écologiques Spécialité Gestion Intégrée des Agrosystèmes et des Forêts. 31 pp. Sentenac, G., Pham, T., Souvignet, J., Dufaÿ, J., 2005: Éléments de biologie des parasitoïdes de la cochenille farineuse Heliococcus bohemicus (Sulc). Compte rendu d activité technique, document interne, 8 pp.

61 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Experiences about the effect of growth regulators on cluster structure and fruit rot diseases of the vine varieties Pinot Meunier, Pinot Blanc, Pinot Noir and Rhine-Riesling K. Bleyer and W. K. Kast State Institute for Viticulture, Oenology and Fruit Technology, D Weinsberg, Germany Abstract: In field trials, effects of the growth regulators Gibb 3 and Regalis on cluster structure of grapes and the reduction of botrytis disease and sour rot disease were evaluated. Gibb 3 can only be used on Pinot-varieties. Regalis also can be used on Rhine-Riesling. The growth regulators Gibb 3 and Regalis were very effective in loosening the cluster structure at Pinot-varieties. Reduction of botrytis disease was compared to fungicide treatments. In most cases, the yield of grapes decreased without an increase of the sugar content. Results of Regalis on Rhine-Riesling varied greatly without relevant reasons. No differences were found when Regalis was applied with different amount of water and at different development stages. Gibb 3 proved to have a better effect if applied with greater amounts of water. The quality of red wines increases because of higher anthocyan- and phenolcontents. White wines produced from grapes treated with Regalis didn t show better tasting results in spite of lower yield. Key words: growth regulator, cluster structure, Botrytis disease, sour-rot Introduction In Southern Germany climate changes and higher temperatures cause an advance of shooting, flowering and veraison of about eighteen days (Kast and Rupp 2009). The vigor of the vines increased and the grapes became more and more compact. The grapes are much more affected by diseases like botrytis-, penicillium- and sour-rot when this occurs. Cluster density has become one of the biggest problems of German viticulture in all Pinot-varieties and Rhine- Riesling, which cover 40% of German vineyard area. To solve these problems there are some options in viticulture like cutting the clusters in the middle, defoliation of the leaf zone, or spraying pesticides that work against Botrytis disease. Another possibility is to spray growth regulators like Gibb 3 (Gibberellin acid GA 3) or Regalis (Prohexidione-Calcium), a gibberillic acid-inhibitor, to get a looser cluster structure and avoid the problems caused by the pressing of berries against each other. Material and methods Field Trials Seven field trials with 4 replications were carried out: In 2003 and 2004 Gibb 3 at Pinot Meunier and Pinot Noir; in 2006 Gibb 3 and Regalis at Pinot Meunier, Regalis at Rhine- Riesling; in 2007 and 2008 Gibb 3 and Regalis at Pinot Blanc and Regalis at Rhine- Riesling and in 2009 Gibb 3 and Regalis at Pinot Meunier and Pinot noir, Regalis at Rhine- Riesling. 41

62 42 Application The applications were made with a tunnel-sprayer. Growth regulators were usually applied in the blossom stage (BBCH 65) with 400 liters of water per hectare to the cluster zone. Applications were carried out in the early morning or late evening for a longer wetness period. In 2008 and 2009 results of the applications at different stages of blossom (BBCH 63, 65 and 68) and different amounts of water (200l/ha, 400l/ha and 800l/ha) were compared. Ratings Ratings of botrytis disease and sour-rot disease were made with the official EPPO-guidelines (incidence % and severity %). Compactness of the bunches were described with a Bunch density index in five categories: 1 = Very loose (open) cluster structure, berries don't touch each other. It's possible to bend the rachis over 90, 2 = Loose cluster structure, berries touch each other a little bit. It's possible to bend the rachis from 45 to 90, 3 = Robust cluster structure, berries are still flexible. It's possible to bend the rachis from10 to 45, 4 = Compact cluster structure, berries are unmovable but do not compress. It's possible to bend the rachis till 10, 5 = Very compact cluster structure, berries are unmovable and are deformed from the neighbour berries. It isn't possible to bend the rachis. Tasting scores 2007 and 2008: wines of the trials were made by taking the grapes of the four replications together. Afterwards the must was separated into two replications. The treatments in Pinot blanc were Gibb 3 and Regalis ; in Pinot noir Gibb 3 and Regalis in Rhine-Riesling. The wines were tasted by ranking and describing the attributes of fruit flower, "phenolicstructure and body" by a group of more than 26 experienced tasters. Mean ranks or attributes of the taters were used for calculation between two or three wines. Results Gibb 3 Applications with Gibb 3 always produced looser clusters in Pinot varieties, which resulted in the redution of severity of the botrytis disease and sour rot. Levels of the sour rot disease in the trials of pinot varities were low. Although it reduced the yield sometimes by about 25% no increase of sugar content was observed (Figure 1-3; Pic. 2). Regalis Like Gibb 3, applications with Regalis always caused looser clusters in Pinot varieties. The clusters had more maiden berries (Pic. 3). As a result botrytis disease severity and sour rot severity became lower (Figures 1-2). Sour rot disease in our trials in pinot varities was low. Yield was sometimes reduced by more than 40% without an increasement of the sugar content (Figure 3).

63 bunch density-index 3 2 3,0 2,3 1,9 1 0 untreated Gibb 3 Regalis Figure 1. Bunch density-index (LSD 5%: 0.37), Pinot Meunier, Gundelsheim ,0 Incidence 4 2 1,9 1,1 0 untreated Gibb 3 Regalis Figure 2. Botrytis disease severity (LSD 5%: 1.4), Pinot Blanc, Wildeck Castle kg/a must density Oe untreated Gibb 3 Regalis 59 Figure 3. Yield and sugar content (Yield: LSD 5%: 17.0), Pinot Blanc, Wildeck Castle 2008

64 44 Applying Regalis at Rhine Riesling did not produce uniform results. In most of the trials there were looser clusters and less botrytis disease, but sometimes there was no effect. If there was an effect, the yield was reduced by about 25% and the bunch-density index was lower. There was also a very good effect against sour-rot disease (Figure 4) severity 6 5, ,7 0 untreated Regalis Figure 4. Sour-rot disease severity (LSD 5%: 1.8), Rhine-Riesling, Weinsberg 2008 Picture 1. Compact grapes with botrytis disease, Pinot Blanc, Wildeck Castle Picture 2. Grapes applied by Gibb 3 at blossoming, Pinot Blanc, Wildeck Castle 2008.

65 45 Picture 3. Grapes applied with Regalis at blossoming, Pinot Blanc, Wildeck Castle Loose clusters with maiden berries. Effect of water volume and stage by application of Gibb 3 and Regalis When applying Gibb 3, the amount of water had an effect on the bunch-density index (Figure 5). Application of the same amount of Gibb 3 in 800l water into the cluster zone had significantly better results than 400l and 200l water had an even worse effect. The effect of development stages, on the other hand, had no significant effect. Spraying Regalis in different stages or with different water amounts didn t show any difference. The bunch-density index ranged from 1.2 to 1.3. The same effect was found in 2008 after applying Regalis at Rhine-Riesling ,0 1,5 1,2 1,8 1,6 1,4 1,3 1,2 1,3 1,2 1,2 0 untreated Gibb L Gibb L Gibb L Gibb L prebl. Gibb L postbl. Regalis 400 L Regalis 800 L Regalis 200 L Regalis 400 L prepl. Regalis 400 L postbl. Figure 5. Bunch density-index (LSD 5%: 0.2), different times of application and water amounts, Pinot Meunier, Gundelsheim 2009.

66 46 Wine tasting When tasting the wines of Pinot Blanc vintage 2007 and 2008 treated with Gibb 3 or Regalis, the tasters didn t notice a great difference. The Wines of the variants with Regalis had the same quality than the untreatet variant.the variants applied with Gibb 3 tended to be better than the untreatet. In mean ranking, they showed a small advatage and were described as more of a fruitflower with more body. Phenol content in wines variants treated with Gibb 3 were higher. Regalis showed diffferent results. (Table 3) The Pinot Noir wines that had been treated with Gibb 3 were always found to taste better. In the mean ranking they showed advantages and were also described as having more fruitflower, more body and better phenol-structure. The wines of the Pinot varieties had more phenols and colours. The wines of treated Rhine Riesling did not show higher quality scores than the untreated variants and were never described with more fruitflower and body (Table 1 and 2). Table 1. Mean ranking of wines tasted by tasters in 2008 and 2009 (significant differences are marked with different letters) Pinot Noir 2007 Control 2.5a Gibb 3 1.4b Pinot Noir 2008 Control 2.5a Gibb 3 1.5b Rhine-Riesling 2007 Control 2.4a Regalis 2.4a Fungicide 2.7a Rhine-Riesling 2008 Control 2.6a Regalis 2.6a Fungicide 2.3a Table 2. Phenols and Sum of colours in Pinot Noir treated by Gibb 3. Pinot Noir 2007 Sum of Phenols mg/l Sum of Colours nm/1cm Control Gibb Pinot Noir 2008 Sum of Phenols mg/l Sum of Colours nm/1cm Control Gibb ,35

67 47 Table 3. Phenols in Pinot Blanc treated by Gibb 3 and Regalis. Pinot Blanc Sum of Phenols mg/l Sum of Phenols mg/l Control Gibb Regalis Discussion Gibb 3 and Regalis are very effective in loosening the cluster structure of Pinot varieties. The yield after applying Gibb 3 is reduced about %. The effect of Regalis on Pinot varieties is sometimes so strong that we did not get much yield. Sometimes there is a reduction of up to 40%, this may not be tolerable for many vine growers. In Rhine-Riesling, the loosening effect of Regalis varied extremely. Sometimes the clusters would become very loose, though at other times there was no effect. The yield was reduced between 0-25%. Despite this, there are many advantages to be gained after applying growth regulators. The effect of botrytis and sour-rot disease is very serious. Healthy grapes are very important for making good wines. In red wines, the colour found in the skin of the berries would be destroyed by botrytis disease and if there is sour-rot disease in the grapes, they cannot be used for it would be a lot of work to sort them. In our experiments there were no differences in the bunch-density index and botrytis-disease at the different times of application. The producer recommends that Gibb 3 should be applied at full bloom. The effect of the amount of water when using the same amount of Gibb 3 (150g/ha) seemed to be greater. This observation was made in terraced culture. The effect was found in rows near the retaining walls, where it is warmer and blossoming begins some days earlier than in the other rows. In these vineyards, we always sprayed with the spray hose so that the flower clusters would get a lot of water. There was no great effect using Regalis with different water amounts and at different stages on the bunch-density index and in Botrytis incidence. For that, it would be necessary to apply growth regulators in both full bloom and pre blossom. When applying Gibb 3, it is necessary to use a lot of water. The quality of red wine gets significantly better after applying Gibb 3, while with Pinot Blanc, there is a trend for better wines when the vines are applied with Gibb 3. This fact should be object of further research. The quality of Rhine-Riesling and Pinot Blanc wines were not found to be better than those of the control wines. The only effect of those variants was the reduction of botrytis disease and sour-rot disease and this result is very important. There are both advantages and disadvantages to be found in both of these growth regulators. It is very important to weight the loss of yield and the healthiness of the grapes, especially if grapes without botrytis disease and sour-rot disease are necessary in the making good wines. References Kast, W. K. and. Rupp, D. 2009: Effects of climatic change on phenology and ripening conditions of grapevine. Mitteilungen Klosterneuburg 59: 3-7.

68 48

69 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Control of grey mould by application of gibberellin acid inhibitor (Regalis ) and management practices in Vitis vinifera (cv. Riesling) O. Baus, A. Reineke, and B. Berkelmann-Löhnertz Geisenheim Research Center, Department of Phytomedicine, D Geisenheim, Germany Abstract: Grapevine varieties with tight clusters are extremely susceptible to infection by grey mould (Botrytis cinerea). Severe pathogen attack leads to reduced yield and affects wine quality. A couple of management practices can reduce tightness of berry clusters, among them moderate defoliation and the use of plant growth regulators. Prohexadione-Calcium (trademark Regalis ) is a gibberellin acid biosynthesis inhibitor with low toxicity and persistence in the plant. Over a period of three consecutive years the effect of prohexadione-ca on grape cluster architecture was tested in cv. Riesling and combined with anti-botrytis fungicide applications and leaf removal measures. While prohexadione-ca treatments significantly increased the flexibility of grape clusters in 2007, no significant effects were evident in 2008 and However, even a minor reduction of index values of may have an influence on the tightness of clusters and thus on the susceptibility to B. cinerea infections. In all three years severity of B. cinerea was significantly reduced in Riesling grapes when the following measures were combined: application of Regalis, fungicide treatments and moderate defoliation. Therefore, future strategies to control grey mould should aim at a combination of different management strategies. Key words: Botrytis cinerea, Prohexadione-Calcium, integrated control Introduction Various fungal and bacterial organisms can infect grapes and result in bunch rots, which significantly reduce grape yield and wine quality. The most common of these is botrytis bunch rot caused by Botrytis cinerea. The development of bunch rot is multifaceted resulting on the interaction of a range of variables. Among them, injury to the berries is the primary factor influencing bunch rot development. Berry wounds arise from birds and insects feeding, hail, and infection by the powdery mildew fungus. In addition, tight-clustered varieties are particularly susceptible to bunch rot, as berries may experience fruit abscission or splitting from growth pressure. Infection of grapes by grey mould can result in premature berry dropping, leading to reduced yield. Mycelium can block filter membranes at filtration processes and wines can easily taste mouldy and oxidise. Various fungicides are actually registered in Germany for the chemical control of B. cinerea including active ingredients such as boscalid, fludioxonil, cyprodinil, fenhexamid and pyrimethanil. In addition, for control of botrytis bunch rot a couple of management strategies are available to the grower, among them various cultural control options such as canopy management practices (shoot thinning and leaf removal before fruit set). Another option in botrytis bunch rot control is the use of plant growth regulators in order to reduce tightness of berry clusters. These plant growth regulators are available in practical viticulture for a few years now (Vaquero-Fernandez, 2009). Such foliarly applied plant growth regulator is prohexadione-calcium (trademark Regalis ), which inhibits the synthesis of gibberellin, a naturally occurring plant hormone. Prohexadione-calcium shows a low toxicity and persistence in the plant. It is used in diverse crops and ornamental plants, mainly in order to control vegetative plant growth. For grapevine, effects of prohexadione-calcium on grape yield 49

70 50 components, wine composition and sensory characteristics have been previously shown (Vaquero-Fernandez, 2009; Lo Giudice, 2004). Here, we report the effect of the application of prohexadione-calcium on grape cluster architecture in cv. Riesling on reducing susceptibility of grape clusters towards B. cinerea infections. Effects of this plant growth regulator were tested over a period of three consecutive years, both alone as well as in combination with anti-botrytis fungicide applications and leaf removal measures. Material and methods Chemical and cultural measures According to EPPO-Guideline PP 1/17(2) field trials were conducted in experimental vineyards (cv. Riesling) of the Geisenheim Research Center in randomised blocks based on four replicates with 15 plants each. Various agents were applied using a pneumatic application gear (Schachtner). The plant growth regulator prohexadione-calcium (trademark Regalis, BASF) was applied at a dosage of 1.8kg/ha at growth stage BBCH 65 to the clusters only. Defoliation measures in the grape zone were performed in a moderate way by applying a defoliator (Binger Seilzug Defoliator EB 490) both sides of the rank just before the application of Regalis. Anti-Botrytis fungicides were applied at growth stage BBCH 77 (Cantus, a.i.: boscalid) and BBCH 81 (Switch, a.i.: fludioxonil and cyprodinil). Four replicates of an untreated control were included in the randomised trial design. Assessments Bunch architecture was measured by assessing a flexibility-index (Ipach et al. 2005) at growth stage BBCH 79. The assessment scale is based on five categories from 1 = very loose-berried clusters to 5 = very tight clusters (the main axis of bunches cannot be bended). The infestation level of B. cinerea was assessed twice (14 days before harvest and just before harvest). For each cluster disease severity was classified in nine categories (0; 1; 3; 6; 11; 20; 35; 60; 100%) and 100 clusters per replication were assessed. Data were tested for statistic significance using Tukey HSD test for pairwise comparisons of the treatments and a Dunnett s test to compare the mean of each treatment to the mean of the untreated control. Results and discussion Flexibility index of grape clusters Application of Regalis influenced grape cluster architecture at different levels. While prohexadione-ca treatments significantly increased the flexibility of grape clusters in 2007, no significant effects were evident in 2008 and 2009 (Figure 1). However, observations in the vineyard indicated that even a minor reduction of index values of may have an influence on the tightness of clusters and thus on the susceptibility of berries to B. cinerea infections. Disease severity of Botrytis bunch rot on grape clusters In all three years, severity of B. cinerea was significantly reduced in Riesling grapes when the following treatments were combined: fungicide treatments, application of Regalis, and moderate defoliation (Figure 2). Nevertheless, the application of anti-botrytis fungicides alone or in combination with Regalis or leaf removal was an effective control measure as well. In all three years, two consecutive applications with anti-botrytis fungicides have been very effective against grey mould.

71 51 5 4, ,5 Flexibility index 3 2,5 2 1,5 1 0, ,5 4 3,5 a c bc bc ab untreated Regalis Regalis defoliation Regalis defoliation fungicide defoliation a defoliation fungicide 2008 Flexibility index 3 2,5 2 1,5 1 0, ,5 a a a a a a a untreated Regalis Regalis defoliation Regalis fungicide Regalis defoliation fungicide defoliation defoliation fungicide ,5 Flexibility index 3 2,5 2 1,5 1 0,5 0 a a a a a a a untreated Regalis Regalis defoliation Regalis fungicide Regalis defoliation fungicide defoliation defoliation fungicide Figure 1. Flexibility index of grape clusters in 2007, 2008 and 2009 (cv. Riesling) in the untreated controls and for the different treatments. Values with the same letters are not significantly different at P < 0.05 after both Tukey and Dunnett s test. Bars represent standard errors of the mean.

72 52 Botrytis disease severity (%) on grapes A a BC de BC bc BC bcd untreated fungicide Regalis Regalis defoliation C e Regalis defoliation fungicide ABC b BC cde defoliation defoliation fungicide Botrytis disease severity (%) on grapes a ab ab ab b b ab ab untreated fungicide Regalis Regalis defoliation Regalis fungicide Regalis defoliation fungicide defoliation defoliation fungicide Botrytis disease severity (%) on grapes A a ABC ab ABC ab ABC ab C b C b AB ab AB b untreated fungicide Regalis Regalis defoliation Regalis fungicide Regalis defoliation fungicide defoliation defoliation fungicide Figure 2. Severity of B. cinerea on grapes (cv. Riesling) after different management practices in 2007, 2008 and Values with the same letters at each assessment date are not significantly different at P < 0.05 after both Tukey and Dunnett s test. Bars represent standard errors of the mean.

73 53 In contrast to this, all treatments with bioregulator application and defoliation measures alone, respectively, or their combination gave unsufficient results. This was most evident in These studies indicate, that the grower has different options in order to adapt the intensity of applied grey mould control measures to individual viticultural demands: grape variety, specific oenological production ambitions and the state of economical affairs, respectively. Acknowledgements We like to thank Isidoro Giampapa, Doreen Richter and Greta Brachmann for technical assistance in the vineyard. References Ipach, R., Huber, B., Hofmann, H. & Baus, O. 2005: Examination of the effectivity of growth regulators to loosen bunch structure by defining a flexibility index. Unpublished German guideline and proposal for an EPPO-Guideline. Lo Giudice, D., Wolf, T. K. & Zoecklein, B. W. 2004: Effects of prohexadione-calcium on grape yield components and fruit and wine composition. Am. J. Enol. Vit. 55: Vaquero-Fernandez, L., Fernandez-Zurbano, P., Sanz-Asensio, J., Lopez-Alonso, M. & Martinez-Soria, M. T. 2009: Treatment of grapevines with Prohexadione Calcium as a growth regulator. The influence on production, winemaking and sensory characteristics of wines. J. Int. Sci. Vigne Vin 43:

74 54

75 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Variety-dependent responses of seeded vine cultivars to different growth regulators S. Böll 1, T. Lange 2, H. Hofmann 1, P. Schwappach 1 1 Bavarian State Institute for Viticulture and Horticulture, D Veitshöchheim, Germany; 2 Institut für Pflanzenbiologie, Technische Universität Braunschweig, D Braunschweig, Germany Abstract: GA-levels and turnover rates during early berry development were much higher and longer lasting in developing berries of Pinot gris compared to Sylvaner and were negatively related to their sensitivity to exogenous applied GA 3. Anthers and pollen are known to be major sources of endogenous GAs during prebloom, and as such a higher pollen tube abundance could be responsible for higher endogenous GA levels during early berry development. Pinot gris and Sylvaner showed the expected differences in pollen tube abundance that corresponded to the observed endogenous GA levels. Furthermore, pollen tube analyses of a broad spectrum of additional cultivars with known GAsensitivity support the hypothesis, displaying a broad range of pollen tube abundance that correlates negatively with the known sensitivity of the cultivars to GA 3 applications. The observed variety-specific pollen tube abundance seems to lead to differential developmental patterns: postbloom, Pinot gris had no unfertilized ovaries while in Sylvaner up to 35% of the ovaries were unfertilized. Furthermore, there were variety-specific differences in the degeneration of ovules: Pinot gris had on average 3-4 fertilized ovules per ovary while Sylvaner only had 1-2 fertilized ovules per ovary. However, during further berry development endosperm development was aborted in about half of the ovules in Pinot gris, while in Sylvaner all fertilized ovules developed into seeds. Also in shattered berries the percentage of unfertilized ovaries was much higher in Sylvaner than in Pinot gris. Altogether, variety-dependent effects were much stronger than treatment effects. The results imply that a simple microscopic pollen tube analysis of varieties with hitherto unknown gibberellin sensitivity might predict the sensitivity to exogenous GA 3 -treatment including possible adverse side effects. Key words: Vitis, endogenous gibberellin status, pollen tube abundance gibberellin, prohexadione-ca, variety-specific fertilization rates Introduction Berry shattering during early berry development is a natural process that depends in its extent on genetic and physiological conditions of the cultivar as well as on the prevailing climate conditions (Müller-Thurgau 1886, Koblet 1966, Schneider & Staudt 1978). However, with an increase of favourable weather conditions during the past decade Botrytis and other bunch rots have become a severe problem due to high fruit set with tight and compact clusters. In grapevine, application of gibberellic acid (GA 3 ) during full bloom usually induces reduced fruit set in seeded varieties (Alleweldt 1967, Hofmann 2004, Weaver 1972), leading to reduced infestation rates with bunch rots, such as Botrytis cinerea and sour rot (Hill et al. 2003, Petgen 2004, Roschatt et al. 2003). However, different seeded vine varieties react quite differently to GA 3 -applications at full bloom: sensitive varieties, such as Sylvaner, respond with a much higher degree of cluster loosening compared to less sensitive ones, such as Pinot gris (Böll et al. 2009, Hofmann 2004, Julliard & Balthazard 1965). Furthermore, sensitive 55

76 56 varieties suffer from severe side effects in the following year showing reduced fertility in cluster numbers and weights as well as late bud break (Alleweldt 1967, Hill et al. 2003, Julliard & Balthazard 1965, Weaver 1960). The growth regulator prohexadione-ca, a gibberellin biosynthesis inhibitor, also leads to a reduced fruit set in some varieties. In this study, two cultivars with known differential sensitivity towards gibberellin as well as prohexadione-ca were treated with these growth regulators over three years. Histological and hormone-physiological analyses of developing and shattered berries during early development were carried out to determine the underlying causes of variety-specific differences in response to these growth regulators. Material and methods For the 3-year field study Sylvaner and Pinot gris, grown in Franconian vineyards of the Bavarian State Institute for Viticulture and Horticulture, were chosen as typical representatives of a very sensitive and a rather insensitive cultivar towards gibberellin applications, while in contrast, Sylvaner was known to respond less towards prohexadione-ca applications than Pinot gris. Accordingly, the following practice-oriented treatment regimes were used over the years: during full bloom Sylvaner was treated with 28g/ha Gibb3 (7ppm GA 3 ), Pinot gris with 80g/ha Gibb3 (20ppm GA 3 ) into the cluster zone. For the prohexadione- Ca treatment 1,8kg/ha Regalis were applied to Sylvaner and 1,35kg/ha Regalis to Pinot gris. At harvest total cluster weight/ vine was determined of 20 vines for each treatment group of Sylvaner and Pinot gris, Gibberellin extraction and GC-MS analysis In 2006, berries of the control group of Sylvaner and Pinot gris were collected and immediately frozen for quantitative determination of endogenous GAs. They were taken at four developmental stages (prebloom, BBCH 61, postbloom, BBCH 68, berries groat-sized, BBCH 73, berries begin to touch, BBCH 77) from 7-10 clusters to investigate GA-levels of the two cultivars over developmental time. Freeze dried plant material (0.01g and 0.1g dry weight, respectively) was pulverized under liquid nitrogen and spiked with 17,17-d2-GA standards (1ng each; purchased from Prof. L. Mander, Australian National University, Canberra, Australia). Samples were extracted, purified, derivatized, and analyzed by combined GC-MS using selected ion monitoring as described elsewhere (Lange et al. 2005). Object of analysis were GA 20, GA 1 and GA 8, three successive GAs of the 13-hydroxylation gibberellin biosynthesis pathway, with GA 20 being the inactive precursor and GA 8 the catabolic successor of the active plant hormone GA 1. Pollen tube preparations In 2006 and 2007, at full bloom (BBCH 65) ovaries of several clusters of Pinot gris and Sylvaner, in 2008, ovaries of 40 clusters of these cultivars and 9 additional vine varieties were collected in the field and fixed in FAA solution (50% ethanol:acetic acid:formalin = 8:1:1 v/v/v). In the laboratory 20 ovaries of each variety were divided longitudionally in half, rinsed with tap water and stained with aniline blue for 24h (2.3g K 3 PO 4 in 100ml aqua dest g aniline blue). Thereafter, the specimens were rinsed twice with tap water, mounted on microslides in glycerine and slightly squeezed. Pollen tubes in the styles of half ovaries were counted under a fluorescent microscope with blue light excitation.

77 57 Histologigal sections of ovaries of developing and shattered berries Developing and shattered berries of all treatment groups of Sylvaner and Pinot gris were collected during shatter periods in 2007 and Due to a technical defect of a freezer half of the specimens of 2008 were lost, so that a pooled sample of 2007 and 2008 specimens was analysed. For each treatment group 20 ovaries were embedded in paraffin and longitudinal serial sections were made at 6µm. Sections were stained with astra blue (0.2%) and safranin red (0.1%). 4,5 4 Sylvaner kg / vine 3,5 3 2,5 2 1,5 1 0,5 0 control Regalis Gibb control Regalis Gibb control Regalis Gibb ,5 4 Pinot gris kg / vine 3,5 3 2,5 2 1,5 1 0,5 0 control Regalis Gibb control Regalis Gibb control Regalis Gibb Figure1. Total cluster weight/vine at harvest of the treatment groups of Sylvaner and Pinot gris,

78 58 Seed number determination At the beginning of ripening (BBCH 81) for each treatment of both cultivars, berries of 20 clusters were sorted according to size. The number of seeds was determined in 10 berries of each of the bigger size classes (Ø 11, 13, 15mm). Results and discussion In spite of the low application rate of Gibb3 and due to the high application rate of Regalis in Sylvaner, both treatment groups showed reduced crops. The effects were a combination of increased looseness of the clusters as well as side effects after the first year of treatment i.e. reduced number of clusters per vine. As expected, Pinot gris hardly responded to the Gibb3 treatment and only showed a minor crop reduction in the Regalis group. These and other differential reactions of Sylvaner and Pinot gris to GA 3 applications (Böll et al. 2009) correspond to the endogenous GA-levels determined in these cultivars (Figure 2): GA-levels and turnover rates were much higher and longer lasting in developing berries of Pinot gris compared to Sylvaner. This indicates that exogenous GA-applications might only have a minor impact on the hormone balance of cultivars with high endogenous GA-levels like Pinot gris while the greatest impact is observed in cultivars with low endogenous GA concentrations. ng GA/ g dry berry weight Pinot gris > 500 > 500 Syl - GA 20 Syl - GA 1 Syl - GA 8 PG - GA 20 PG - GA 1 PG - GA 8 0 BBCH 61 BBCH 68 BBCH 73 BBCH 77 developmental stage Figure 2. Endogenous gibberellin concentrations of the successively metabolized gibberellic acids GA 20, GA 1, GA 8 of the gibberellin biosynthesis pathway during early berry development in Sylvaner (Syl) and Pinot gris (PG) Anthers and pollen are known to be major sources of endogenous GAs during prebloom (Pharis & King 1985, Singh et al. 2002), and as such a higher pollen tube abundance could be

79 59 responsible for higher endogenous GA levels during early berry development. Pinot gris and Sylvaner showed marked differences in their pollen tube abundance in the styles during bloom with Pinot gris being much more potent (Figure 3). The pollen tube analyses of a broad spectrum of additional cultivars with known GA-sensitivity support this hypothesis, displaying a broad range of pollen tube abundance (Figure 3) that correlates negatively with their known sensitivity towards GA 3 applications (Böll et al. 2010). Variety-specific sensitivities of cultivars towards prohexadione-ca are not as well known, but there seems to be a converse tendency that more responsive cultivars (Pinot cultivars) have a higher pollen tube abundance than less responsive ones (e.g. Sylvaner, Mueller-Thurgau). More data are needed to see if this holds true. Postbloom development patterns of the ovaries of Sylvaner and Pinot gris also showed variety-specific differences that are probably related to the differences in pollen tube abundance: while in Pinot gris ovaries were always fertilized in developing berries, in Sylvaner 10% to 35% of the developing berries, depending on the treatment (Table 1), were completely unfertilized. Furthermore, in Pinot gris, 3-4 ovules per ovary were fertilized, while in Sylvaner only 1-2 ovules per ovary were fertilized (Tab.1). However, half of the fertilized ovules were aborted during further development in Pinot gris, while in Sylvaner all fertilized ovules developed into seeds leading in both varieties on average to two seeds per berry (Table 2). Table 1. Percentage of fertilized ovules/ ovary during postbloom development in Sylvaner and Pinot gris Sylvaner Pinot gris BBCH 69 control Regalis Gibb3 control Regalis Gibb3 % fertilized ovules /ovary mean + SD median % unfertilized ovaries Table 2. Mean number of seeds per berry at the beginning of ripening. control Regalis Gibb3 Pinot gris Sylvaner In shatter berries of both cultivars some unfertilized ovaries were found. However, as in developing berries, the percentage was much higher in Sylvaner (60% in all treatments) than in Pinot gris (33% to 48% depending on the treatment). Altogether, in developing as well as shatter berries variety-dependent effects were much stronger than treatment effects.

80 60 A A B Kerner Traminer Sylvaner B B/C Riesling C M. Thurgau Portugieser C C Pinot meunier C Dornfelder Pinot blanc C C Pinot gris Pinot noir Figure 3. Pollen tube abundance in styles of different cultivars with A - high, B - moderate, C - low GA-sensitivity The results of this study indicate that differing sensitivities of different cultivars towards GA 3 applications (and probably also towards prohexdione-ca applications) are most likely dependent on the variety-specific endogenous GA concentrations that in turn, seem to lead to variety-specific pollen tube abundances, fertilization rates and ovule degeneration patterns.

81 61 Our results imply that a simple microscopic analysis of the pollen tube abundance of varieties with hitherto unknown gibberellin sensitivity might well indicate the endogenous GA-status of a given variety which in turn is expected to predict the sensitivity to exogenous GA 3 -treatment including possible adverse side effects. Acknowledgements We would like to thank A. Liebrandt, M. Adelhardt and A. Baumann for their valuable technical assistance. This work was supported by the Forschungsring Deutscher Weinbau (grant ) References Alleweldt, G. 1967: Physiologie der Rebe. Forschungsergebnisse der Jahre Vitis 6: Böll, S., Hofmann, H. & Schwappach, P. 2009: Einsatz der Wachstumsregulatoren Gibb3 und Regalis warum Sorten unterschiedlich reagieren. Rebe & Wein 5/2009: Böll, S., Lange, T., Hofmann, H. & Schwappach, P. 2010: Gibberellin-sensitivity corresponds to pollen tube abundance in seeded vine varieties. Mitteilungen Klosterneuburg (in press). Hill, G., Hill, M., & Butterfass, J. 2003: Gibberelline kleiner, weniger, besser? Das Deutsche Weinmagazin 19/2003: Hofmann, H. 2004: Traubendesign zur Fäulnisprävention. Rebe & Wein 5/2004: Julliard, B. & Balthazard, J. 1965: Effets physiologiques de l acide gibbérellique sur quelques variétés de vigne (Vitis vinifera L.). Ann. Amélior. Plantes 15: Koblet, W. 1966: Fruchtansatz bei Reben in Abhängigkeit von Triebbehandlung und Klimafaktoren. Weinwissenschaft 21: , Lange, T., Kappler, J., Fischer, A., Frisse, A., Padeffke, T., Schmittke, S., Pimenta Lange, M. J. 2005: Gibberellin biosynthesis in developing pumpkin seedlings. Plant Physiol. 139: Müller-Thurgau, H. 1883: Über das Abfallen der Rebenblüten und die Entstehung kernloser Traubenbeeren. Weinbau 9: 87-89; Petgen, M. 2004: Was bringen Gibberelline? Das Deutsche Weinmagazin 3/2004: Pharis, R. P. & King, R. W. 1985: Gibberellins and reproductive development in seed plants. Ann. Rev. Physiol. 36: Roschatt, C., Haas, E. & Pedri, U. 2003: Der Einsatz von Gibberellinen im Weinbau gegen Essigfäule. Schweiz. Z. Obstbau Weinbau 4/2003: Singh, D. P., Jermakov, A. M. & Swain, S. M. 2002: Gibberellins are required for seed development and pollen tube growth in Arabidosis. The Plant Cell 14: Schneider, W. & Staudt, G. 197: Zur Abhängigkeit des Verrieselns von Umwelt und Genom bei Vitis vinifera. Vitis 17: Weaver, R. J. 1972: Plant growth substances in agriculture. W. H. Freeman and Company, San Francisco.

82 62

83 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Monitoring nematode populations to adapt fallow periods against Xiphinema vectors of grapevine fanleaf virus (GFLV) M. van Helden 1,4, L. Villate 2, C. Laveau 1, E. Morin 1, G. Darrieutort 4, C. van Leeuwen 3 1 ENITA, Université de Bordeaux, UMR INRA/ENITA Santé Végétale, ISVV. CS 40201, Gradignan CEDEX, France. 2 UMR INRA-Agrocampus Ouest (Rennes) BiO3P, Domaine de la Motte, BP 35327, Le Rheu cedex, France. 3 ENITA, Université de Bordeaux, UMR EGFV, ISVV. CS 40201, Gradignan CEDEX, France. 4 ARD-VD, 1 Cours du Général de Gaulle, Gradignan, France. m-vanhelden@enitab.fr Abstract: The nematode Xiphinema index is, economically, the major virus vector in viticulture, transmitting specifically the grapevine fanleaf virus (GFLV), the most severe grapevine virus disease worldwide. A second virus (ArMV) and its vector (X. diversicaudatum) do exist in the Bordeaux region but seem of lower importance. Plots that have become completely infected, no longer providing sufficient yield or quality, have to be uprooted to be replaced by new healthy plants. Increased knowledge of the spatial distribution of this nematode, both horizontally and vertically, and of correlative GFLV plant infections, is essential for efficient control of the disease. Vertical distribution data showed that the highest numbers of individuals occurred at 40 to 110cm depth, corresponding to the layers where the highest densities of fine roots were observed. Horizontal distribution revealed a significant aggregative pattern with patches of 6 to 8m diameter, together with a significant neighbourhood structure of nematode densities, thus identifying the relevant sampling scale to describe the nematode distribution. Nematode patches often, but not always, correlate significantly with those of GFLV-infected grapevine plants. The nematodes are able to survive in the soil even without food for several years. Since nematodes live deep in the soil, they cannot be attained by nematicides (that are, anyway, progressively forbidden). Because of this lack of efficient control measures of the vector, a fallow period of 7 years is advised to farmers but rarely observed for economic reasons. In some cases, high re-infestation levels have occurred with short fallow periods but this does not always occur. Sampling of numerous plots in the Bordeaux region, uprooted because of virus infestations, showed that nematode numbers are unpredictable. No clear relation was found between the percentage of each virus and the population of its vector. X. diversicaudatum was rarely present in significant amounts, even though infestation percentage of plants with ArMV could be high. X. index was sometimes present in high numbers (up to 300 per litre of soil) in some samples but mostly numbers were quite low or nematodes even not detected. These results suggest that the fallow period can be adapted following nematode sampling. In the case of very low nematode numbers it could be reduced to 1.5 to 3 years whereas longer periods need to be observed when high numbers are found. Nematode sampling grid should be rather dense, and sampling should ideally be done twice, just after uprooting and before replanting. The use of nematicidal antagonistic plants against X. index is currently being investigated in our lab. Key words: Xiphinema, GFLV, ArMV, viticulture Introduction Two nematodes of the genus Xiphinema, X. index and X. diversicaudatum, are major virus vectors in European viticulture transmitting respectively the grapevine fanleaf virus (GFLV) and the Arabis mosaic virus (ArMV) (Demangeat 2007). Virus infestation affects yield 63

84 64 quality or quantity, and plot renewal (uprooting, fallow period, replanting) is the only solution when a plot is fully infested. In vine plots Xiphinema nematodes are mainly concentrated in the deeper soil layers (Esmenjaud et al. 1992, Villate 2008) where the density of small roots provides food resources for them. Nematode spatial distribution on existing plots generally shows a clearly aggregated distribution with nematode patches of several meters in diameter and large gaps with very low nematode numbers (Villate 2008). Nematodes are capable of surviving in the soil for several years, experiments (Demangeat et al. 2005) have described survival for at least 4 years in laboratory conditions, without any food (root residues) in the stored soil. When vines are uprooted the roots remnants in the deeper soil layers will persist for several years, and these can provide nutritional resources for nematodes Therefore herbicide devitalisation of the vine plants is strongly recommended to kill off all root residues. Systemic herbicides such as glyphosate or trichlopyr (depending on authorization) can be applied just after final harvest and plants are uprooted in the next spring (Magnien 1998). Subsequent fallow periods of up to 7 years are recommended to farmers. Such periods are rarely applied. Using shorter (2.5 years) fallow periods, sometimes but not always causes rapid re-contamination of the plots. As part of a larger project (Villate 2008) we studied the X. index and X. diversicaudatum nematode populations in many vineyard plots, all uprooted in the year of sampling. All plots were located in the Bordeaux wine production region. This paper summarizes overall results of such samplings showing the variability of nematode species composition and population level in vineyard plots. Material and methods Nematode sampling and extraction Nematode sampling was performed by digging 1 to 1.5 meter deep trenches using an excavating machine. Then samples were collected using a hand spade on 3 to 5 points of the sides of the trench in the undisturbed areas, most often at 0.8 to 1 meter deep where most small vine roots were (initially) located, under the compaction area. Total sample size was 5 litres. Samples were stored at room temperature (20 C) when extraction was done less than 1 week after sampling or at 4 C when processed later. A total of 47 plots were sampled. The number of samples on each plot was very variable (2 to 50, see table 1) according to plot size and sampling objectives. Nematodes were extracted as described by (Villate 2008) from 2 litre sub-samples. Results and discussion Detailed results are presented in table 1 and summary in table 2. Since the number of samples per plot is very variable we will not focus on individual plots but we will try to extract a general tendency out of these data. As can be observed in table 2 on a total of 10 farms, 47 plots and 334 samples X. index is more present than X. diversicaudatum on all possible criteria. X. index was found on 32 plots against 21 for X. diversicaudatum. However, it is likely that the real values are a little higher because on some plots we have not found the nematodes due to insufficient sample effort. The percentage of positive samples on positive plots is 53% (X. index) or 43% (X. diversi-

85 65 caudatum). Therefore multiple samples are clearly necessary to get a reasonably good image of the population. Data from Villate (2008) show spatial structure with 6 to 8 meter infestation patches. To get a good image of the spatial distribution and the population levels (as used in our research program on some plots) very intensive sampling grids are necessary (every 5 meters), but these seem unrealistic as a diagnostic tool for the farmer. Table 1. Sampling results of 47 uprooted plots on 10 farms in Bordeaux for Xiphinema index and X. diversicaudatum nematodes in a 2 litre soil sample. Freq. shows de fraction of samples containing the nematode species. Max shows the highest value in these samples. Nr of X. index X. diversic. Nr of X. index X. diversic. Farm Plot Farm Plot samples Freq. Max Freq. Max samples Freq. Max Freq. Max % 7 0% % 68 0% % 4 50% % 22 0% 0 A G % 3 100% % % 12 B C D E F % 1 60% % 276 0% % 0 0% % 9 0% % 28 20% % 1 0% % 171 0% % 73 0% % 0 0% % % % 0 50% % 6 25% % 3 0% % 7 0% % 0 50% % 3 67% % 0 0% % 0 0% % 37 40% % 36 67% % 1 20% % 6 0% % 0 0% % 3 14% % 0 50% % 7 33% % 6 0% % 0 25% % 0 0% 0 J % 6 33% % 1 0% % 2 0% % 0 0% % 12 67% % 0 0% % 5 50% % 0 0% % 6 67% % 5 0% 0 Mean (334) 49% 8,1 10% 0, % 0 0% % 0 0% 0 H I For simple diagnostics sampling should always be done on multiple points, even if we just want information on presence or absence of the species. Taking about 10 samples per Ha seems an acceptable compromise for such qualitative diagnostics and combining the numbers found in such samples and the frequency of occurrence gives at least some indication on the population levels.

86 66 Table 2. Summary of overall data. All plots X. index X. diversicaudatum Number of plots % of plots 68% 45% Number of samples % of positive samples 49% 10% Mean nem. / sample 8,1 0,8 'Positive' plots only % of positive samples 58% 43% Mean nem. / sample 23,5 2,6 The dominance of X. index in our samples clearly shows that this species is the major problem in the Bordeaux area. X. diversicaudatum is mostly found together with X. index (17 plots harbouring both species, only 4 plots showed X. diversicaudatum only). Therefore management should focus on X. index, also because the GFLV virus is causing much more quantitative and qualitative damage. When comparing multiple plots on the same farm (and on similar soil characteristics) the population levels are very different (see for example farms B and I, table 1). No explanation for these differences has been found. On these ancient plots (most have been cultivated for at least 3 to 4 centuries) the presence or absence of a nematode species could be linked to historical contamination events far beyond the last cropping cycle, but this does not explain the population levels. More data on population dynamics of the nematodes and the ecological interactions between plant and nematode, and on the soil community are necessary. Some other Xiphinema species were observed occasionally (X. pachtaicum, X. rivesi) but these are not known as virus vectors in France. Data on the presence of the two viruses prior to uprooting were not always available. Most plots were uprooted for loss of yield quality or quantity attributed to virus infection but no clear data on the percentage of each of the two viruses present during uprooting, the plant material (especially rootstock), the management applied during the last plot renewal, and the state of the plots at uprooting such as the number of missing plants and the plant (root) vigour. The data strongly suggest that plot fallow following uprooting can be adapted according to nematode population numbers, for this a diagnostic sampling at uprooting is recommended, if possible repeated before replanting. The use of adapted management practices such as the use of suppressive plant species ( nematicidal plants) is currently under study in our laboratory. This approach is complementary to the development of nematode or virus resistant rootstocks (Ollat et al. 2005) Acknowledgements We thank the following persons for help in sampling and extraction: Bernard Chauvin, Bram Hanse, Florent Delamarre, Vincent Fiolleau and Olivier Auroy; our colleagues Nathalie Ollat, Gérard Demangeat, Daniel Esmenjaud and Olivier Plantard for their input in this project; and all the wine growers for their help and financial support.

87 67 References Demangeat, G., Voisin, R., Minot, J.-C., Bosselut, N. & Fuchs, M. 2005: Survival of Xiphinema index in vineyard soil and retention of Grapevine fanleaf virus over extended time in the absence of host plants. Phytopathology 95: Demangeat, G. 2007: Transmission des Népovirus par les nématodes Longidoridae. Virologie 11(4): Esmenjaud, D., Walter, B., Valentin, G., Guo, Z. T. & Cluzeau, D. 1992: Vertical distribution and infectious potential of Xiphinema index in field affected by grapevine fanleaf virus in vineyards in the Champagne region of France. Agronomie 12: Magnien, C. 1998: Lutte contre le court-noué. La dévitalisation des ceps avant arrachage : une mesure préventive de grand intérêt. Phytoma - La Défense des Végétaux 510: Ollat, N., Demengeat, G., Esmenjaud, D., van Helden, M. & Bouquet, A. 2005: Les perspectives françaises en matière de sélection de nouveaux porte-greffes." Progrès Agricole et Viticole 122(7): Villate, L. 2008: Origine, variabilité et gestion des populations de Xiphinema index, le nématode vecteur du Grapevine fanleaf virus. (PhD thesis Rennes University Agro-Campus Ouest presented December 19 th 2009.

88 68

89 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp VitiMeteo a future-oriented forecasting system for viticulture G. Bleyer 1, H.-H. Kassemeyer 1, M. Breuer 1, R. Krause 2, O. Viret 3, P.-H. Dubuis 3, A.-L. Fabre 3, B. Bloesch 3, W. Siegfried 4, A. Naef 4, M. Huber 4 1 Staatliches Weinbauinstitut Freiburg, D Freiburg, Germany 2 GEOsens Ingenieurpartnerschaft, D Ebringen Germany, 3 Agroscope RAC Changins, Route de Duillier, CP 254 CH-1260 Nyon, Switzerland, 4 Agroscope FAW Wädenswil, Postfach 185, CH-8820 Wädenswil, Switzerland. Abstract: Disease management has been greatly improved with the new technologies of the Internet. Current information and forecasting systems have not only a positive ecological, but an economical one for wine growers as well. Now, it is possible to access the VitiMeteo forecasting system free twice a day on the Internet ( The first VitiMeteo module created was VitiMeteo Plasmopara. The core of this system is found in the database Agrometeo, where all the data from different weather stations are stored. The purpose of Agrometeo is to integrate weather data into various other software models. The next component is VitiMeteo Growth which was programmed in cooperation with Hans-Reiner Schultz from the Geisenheim Research Center. More software followed, such as VitiMeteo Insects, VitiMeteo Oidium and VitiMeteo Data Graph. VitiMeteo Insects is a programme that simulates the development of insects and other pests. Walter Kast, from the State Institute for Viticulture, Oenology and Fruit Technology Weinsberg, was able to create an algorithm based on OiDiag-2.2, which calculates which dates the spraying against powdery mildew (Erysiphe necator) should begin and the time increments between each spray. Weather data are controlled and presented with VitiMeteo Data Graph. The VitiMeteo-system is a cooperative project between the State Institute of Viticulture and Enology, Freiburg (Germany), the Swiss Research Station ACW, Changins-Wädenswil (Switzerland), and the company GEOsens, Ebringen (Germany). The company Meteoblue (Basel, Switzerland) has provided weather forecasts, which have been integrated with our software modules since This is the first time that an actual forecast of biological processes e.g. incubation period or growth was possible. The requirements of the VitiMeteo forecasting system can be entered quickly, making it an extremely flexible system. Key words: forecasting, grapevine, fungal diseases, downy mildew, powdery mildew, grape berry moth, pests, Internet access Introduction The use of Internet is important to the progress for disease management in viticulture. Up-todate information and decision support based on forecasting systems have an obvious economical and ecological advantage for advisers and winegrowers. With the forecasting system VitiMeteo, actual data referring to infection risk and to other important information for plant protection, is created twice a day on the Internet for free ( The system gives information about downy mildew, powdery mildew, grape berry moth and weather data. The system was elaborated in cooperation with the Swiss Research Station ACW, Changins-Wädenswil and the company GEOsens (Germany). One objective was to create modern and flexible tools for research and further development of models. Another very important purpose was the application of VitiMeteo in practical viticulture. In this paper, the current stage of this system will be presented. 69

90 70 Material and methods Forecasting system VitiMeteo Figure 1 shows the structure of the forecasting system VitiMeteo. Basically, the system consists of data sources (weather data), a database, the expert models (software) and the presentations on the Internet. The data flow is organized in the following steps: Weather data are stored in a database. The expert models receive the parameters from the database. They allow calculations by adjusting the parameters to optimize the models as well as the publication of practice-oriented results via Internet. The heart of the system is the database Agrometeo, where weather data from the different weather stations are stored. VitiMeteo Plasmopara was the first module of the VitiMeteo system; it calculates the most important stages of the life cycle of downy mildew (Plasmopara viticola) (Müller and Sleumer, 1934; Bläser, 1978; Bläser et al., 1979; Viret and Bloesch, 2002; Hill, 1989; Hill, 1997, Loskill, 2004). The second software component was the VitiMeteo Growth, a growth model, developed by Schultz from the Geisenheim Research Center (Schultz, 1992). It was integrated in the forecasting model for downy mildew. Other expert software components followed, such as VitiMeteo Insects which simulates development of insects and other pests. VitiMeteo Data Graph is an extensive software program for presenting, administrating, and controlling weather data. VitiMeteo Oidiag was developed with the algorithms of Walter Kast (Kast and Bleyer, 2009) from the State Institute for Viticulture, Oenology and Fruit Technology Weinsberg. Weather forecasts have been provided by the company Meteoblue from Basel in Switzerland since 2009 and have also been integrated in all expert software modules. Source of Data Database Expert software Models New! Presentation Internet New! VM Plasmopara New! VM Downy mildew New! Weather stations VM Growth New! VM Downy mildew New! Lufft Campell Adcon Hoffmann VM Oidiag New! VM Insects New! VM Powdery mildew New! VM Grape Berry Moth New! Weather forecast Meteoblue New! AGROMETEO 4 New: real forecast VM DataGraph New! VM Weather data New! VM Edit VM = VitiMeteo Scab Scab Figure 1. Structure of the VitiMeteo forecasting system in 2009: Overview shows how weather data are acquired, processed and subsequently offered to the user as calculated model.

91 71 Validation of VitiMeteo Plasmopara Results of the first model of downy mildew VitiMeteo Plasmopara were compared with the established models (Bleyer and Huber, 1996) which were implemented in the electronic wheather stations BIOMAT (Berghof company; Ehingen, Germany) and in the HP 100 (Lufft company, Felbach, Germany). After comparing the two programs for five years, we found that the results of both models were similar. Next, we validated results of VitiMeteo Plasmopara with field observations and experiments. These observations were: i) symptoms of primary infections in untreated plots since the year 2001, ii) amount of sporulation under field conditions in 2008, iii) marked oil spots in the varieties Pinot Noir and Müller-Thurgau evaluated for 35 consecutive days in June and July of 2008 in Freiburg on two different sites, iv) the spread of downy mildew on leaves observed between 2005 and 2009 in untreated control plots. In 2005, the calculation of the incubation period was investigated with two experiments in May and June with three varieties respectively in vineyards. Growth model in VitiMeteo Plasmopara Data about the duration of biological activity of fungicides against downy mildew are essential for determining the interval between treatments. Huber et al. (2002, 2003) conducted extensive studies on protective action fungicides in commercial vineyards. These studies have demonstrated that the duration of fungicide efficiency is mainly restricted by the growth of new, unprotected leaves. For this reason, a method was needed to measure the growth of the grapevine, since this data allows us to determine the effective period of a fungicide in the vineyard. Up to this time, the increase of growth was a mere observed value. Schultz at the research station in Geisenheim already described a growth model in the 1990s for the variety Riesling cv. (Schultz, 1992.). The number of leaves and leaves size per main shoot can be simulated according to weather date with this model. A growth model was developed in 1999 for the important varieties Müller-Thurgau and Pinot Noir (Schulz, 2003). The growth models have been tested at the State Institute for Viticulture in Freiburg since 2000 on the varieties Müller-Thurgau and Pinot Noir and in 2007 on Riesling. Validation of VitiMeteo insects The first practical application is the determination of the beginning of the flight of the European grapevine moth (Lobesia botrana) and the European grape berry moth (Eupoecilia ambiguella). The mating disruption method is currently used in 54% of the vine growing area of Baden-Württemberg in order to control the moths. The dispensers with the specific pheromone have to be placed in the vineyard before the start of the flight period of the moths in order to achieve efficient control. Since the year 2008, the beginning of the flight period has been calculated with VitiMeteo insects on approximately 35 locations in Baden- Württemberg. On selected locations the forecasted the start of the flight period was compared with actual start of the flight period. Results and discussion VitiMeteo Plasmopara The model correctly calculated the primary infections in most cases in the years 2001 to In 2005, the calculations were frequently applicable in western Switzerland, while in Baden- Württemberg and eastern Switzerland the calculated primary infections were too late. Entering primary infections manually in the system, however, could solve this problem. In winter 2005, an additional algorithm for the calculation of primary infections was

92 72 programmed. Primary infections have been more accurately calculated since 2006 with this new algorithm (Siegfried et al., 2007). However, the calculated date of the first primary infections under variable climatic conditions and locations was in some cases earlier compared to the date of the observed oil spots at the end of the incubation period. Anyhow, in practice it is better for the advisors and growers to calculate an infection too early than too late. They get early information of a possible primary infection and so they have the possibility to interpret the calculated data with their own experience. The comparison between calculated and observed sporulations was sufficient. Figure 2 shows a comparison between spread of downy mildew in untreated control plots and the values calculated by VitiMeteo Plasmopara. The variants were artificially inoculated on 13 th of May 2004 in order to obtain a homogenous infection pressure. On the 2 nd of June, no additional infection to the inoculated leaves could be observed. On the 14 th of June, however, infestation escalated from 0 to 30%, which can be attributed to the rainfalls from the 2 nd to 5 th of June (Gobbin et al., 2007). The model permanently showed during this phase infection conditions. The rainfall can explain the next increase from 30% to 80% on the 11 th and 12 th of June. The model calculated permanent infections also for those days. The rating results show that the model predicts reality accurately. Generally VitiMeteo Plasmopara calculated the epidemiological important infections in the years 2005 to 2009 very well. Figure 3 presents some of the results of the experiments with the calculation of the incubation period. These results indicate that symptoms are expressed after relative long incubation periods (>10 days) and in physiological young leaves already from 70% incubation period. After relative short incubation periods (< 7 days) and physiological old leaves symptoms are only visible after approximately 90% incubation period These results show that VitiMeteo Plasmopara calculates the progression of the incubation time very accurately (Bleyer et al., 2008). VitiMeteo Growth The results of the field experiments about the duration of biological activity of fungicides indicate, that even with high infection pressure treatment intervals of 2 to 3 leaves or 300 to 400 cm 2 can be tolerated without risk (Bleyer et al., 2003). Figure 4 shows one result of the comparison between model and reality. In some years the growth model slightly underestimates the number of leaves of the variety Müller-Thurgau and slightly overerestimates Pinot noir. However, the growth model reflects the real growth in adequate way (Schulz, 2003). Therefore, it is a valuable tool for grapevine protection (Huber et al., 2002, 2003). VitiMeteo insects In the vine growing areas in the Southwest of Germany, the flight of Eupoecilia ambiguella usually begins when the temperature reaches Kd In some extraordinary cases and vineyards, the insects hatch earlier at a temperature of Kd 900. In 2009, the prediction was quite accurate in most of the locations (Table 1). Only in two cases, the date was too early or too late. Anyhow, for practical use in viticulture the accuracy of this prediction method is sufficient, to hang up the pheromone dispensers.

93 73 biology Grade hours at lw (h) biology Grade hours at lw (h) Spores(s/xm 2 x10 3 ) Event LW Rainfall Leaf wetness (LW) Soil infection density of sporangia Course of infection Average temperature Grade hours at leaf wetness Sporulation Dying-off of spores Soil infection (%) Relative humidity Oospores ready to germinate Infection Incubation Frequency of leaf infection % Figure 2. Validation of VitiMeteo Plasmopara : Comparison of ratings with results of simulation calculations in Leaves % visible non visible 75% / 9 days 80% /10 days 90% /11 days 100% /12 days 100% /13 days 100% /14 days 18 May 19 May 20 May 21 May 22 May 23 May Oil spots well visible from 21 May Incubation times/ days Figure 3. Plasmopara viticola; Validation of the incubation period with VitiMeteo Plasmopara ; Incubation period and symptoms (oil spots) n = 100; Freiburg, Schlierberg, Pinot Noir ; artificial infection on the 9 th of May 2005.

94 Number of Leaves Model Number of Leaves Rating 24 July 17 July 10 July 03 July 26 June 19 June 12 June 05 June 29 May 22 May 15 May 08 May 01 May Figure 4. Validation of the growth model. Graphic with results of the simulation. Freiburg, variety Müller-Thurgau, 2005 Table 1. Comparison between forecasted and actual start of the flight period of the grape berry moth in 2009 Forecasting of the start of the flight period; grape berry moth Actual start of the flight period; grape berry and grapevine moth location early date (Eupoecilia ambiguella) Kd 900 average date (Eupoecilia ambiguella) Kd 1082 (Eupoecilia ambiguella) (Lobesia botrana) Blankenhornsberg 15 Apr 25 Apr 27 Apr 27 Apr Eichstetten 11 Apr 19 Apr 26 Apr 26 Apr Königschaffhausen 10 Apr 18 Apr 21 Apr 21 Apr Niederrotweil 17 Apr 26 Apr 24 Apr 24 Apr Ortenberg 19 Apr 28 Apr 27 Apr 04 May Station Apr 22 Apr 21 Apr 20 Apr Wagenstadt 22 Apr 02 May 24 Apr 24 Apr VitiMeteo The feedback from plant protection service and growers was very positive. According to the opinion of the users, one of the most important advantages is the actuality and the clear presentation of information, besides the free access on the Internet twice a day ( In regular training events, users are informedand so their requirements can be considered.

95 75 The parameterization of the models proved to be a tremendous advantage in the development of the models. One of the main progresses of the VitiMeteo-system is the integration of the five-day weather forecast in all expert software modules. For example Figure 5 demonstrates the weather data, the downy mildew risk, vine growth and the forecasted weather data, infection risk and vine growth for 5 days (behind the grey area) ahead. A detailed overview for downy mildew is plotted in Figure 6. The graph shows calculated biological process as soil infection, incubation period in past, in present and in future (5 days). This integration is considered as a milestone in the development of the VitiMeteo -system. In this way the first time a real forecast of biological processes e.g. incubation period, vine growth, the start of the period of the grape berry moth has been possible. 05 May: 14mm precipitation 02 May: 16 C average daily temperature 06 May: medium red = medium infection Forecasting: 11 May: light red = weak infection Infection risk 03 May: green = no infection leaves Leaf area (cm 2 ) Rel. humidity 07 May 4 leaves are unfolded and 165cm² leaf surface/primary shoot are grown Forecasting: 14 May 6 Blätter leaves are unfolded 446cm² leaf surface/ primary shoot are grown Figure 5. Downy mildew Risk graph; weather data, infection risik and vine growth (daily) from 01 to 09 May; forecasted weather data, infectionrisik and vinegrowth for 5 days (grey area) from 09 to 14 May The VitiMeteo system offers the following advantages: User s requirements of the systems, for example scientists, from plant protection service and growers, may be considered by appropriate communication. Other models e.g. diseases or irrigation management as well as new results, for example with respect to downy mildew of grapevine, which is based on meteorological data may be integrated. Weather stations of different manufacturers may be used. Use of weather data obtained by different means is possible.

96 76 The group of users can be extended by different distribution mechanisms (Internet, Intranet, LAN, WAN, etc.). Scientists can calculate What if scenarios to optimise and validate the models. Grade h leaf wetness Biology beginn of incubation time 06 May forecasted end of incubation time 13 May Spores event Soil infection Rel. humidity Temperature Forcasted sporulation Leaf wetness Relative humidity Leaf wetness (LW) Soil infection density of sporangia Course of infection Weather forecast Average temperature Grade hours at leaf wetness Sporulation Dying-off of spores Soil infection (%) precipitation Oospores ready to germinate Infection Incubation Germination grade days Figure 6. Downy mildew: Detailed overview, biology, hourly weather data from 28 Apr to 09 May; forecasted weather data, incubation time, possible sporulations and infections for a 5 days (grey area) from 09 to 14 May The two collaborating research institutions Swiss Research Station ACW, Changins- Wädenswil and the State Institute of Viticulture and Enology, Freiburg in Germany intend to develop the existing models and create new ones in the future. The development is not dependent on commercial success because the research institutes own the software modules of the VitiMeteo system. The Research institutes task the company GEOsens with programming, in the first line for them but the Software modules may be purchased. The majority of the earnings will be used for further development of the VitiMeteo project while the other part will be used by GEOsens for distribution and support. The forecasting system VitiMeteo system is extremely sustainable since results of research and requirements of growers may be integrated quite quickly. References Bläser, M. 1978: Untersuchung zur Epidemiologie des Falschen Mehltaus an Weinreben, Plasmopara viticola (Berk. & Curt, ex de Bary) Berl. & de Toni. Dissertation. Universität Bonn.

97 Bläser, M. und H. C. Weltzien 1979: Epidemiologische Studien an Plasmopara viticola zur Verbesserung der Spritzterminbestimmung. Z. Pflanzenkrank. Pflanzenschutz 86(8): Bleyer, G. und B. Huber 1996: Bekämpfung der Peronospora nach dem Freiburger Prognosemodell. Deutsches Weinbau-Jahrbuch 47: Bleyer, G., B. Huber, V. Steinmetz, and H.-H. Kassemeyer 2003: Growth-models, a tool to define spray intervals against downy mildew (Plasmopara viticola). IOBC/WPRS Bull. 26(8): Bleyer, G., Kassemeyer, H.-H., Krause, R., Viret, O. & Siegfried, W. 2008: "Vitimeteo- Plasmopara"-Prognosemodell zur Bekämpfung von Plasmopara viticola (Rebenperonospora) im Weinbau. Gesunde Pflanzen 60: Gobbin, D., G. Bleyer, S. Keil, H.-H. Kassemeyer and C. Gessler 2007: Evidence for sporangial dispersal leading to a single infection event and a sudden high incidence of grapevine downy mildew. Plant Pathology 56: Kast, W. K. and Bleyer, K. 2009: Oidium-Bekämpfung mit EDV-Unterstützung. Der Deutsche Weinbau 6: Hill, G. K. 1989: Effect of temperature on sporulation efficiency of oilspots caused by Plasmopara viticola (Berk. & Curt. ex de Bary) Berl. & de Toni in vineyards. Wein- Wissenschaft 44: Hill, G. K. 1997: Peronospora: Dem Rätsel der Primärinfektion auf der Spur. Deutsches Weinbau- Jahrbuch 48: Huber, B. et al. 2002: Studies on the effective period of protective fungicides against downy mildew (P. viticola). In: Proceedings of the 4th International Workshop on Powdery and Downy Mildew in Grapevine, Napa, California: Huber, B., G. Bleyer and M. Gesiot 2003: Neue Entwicklungen bei der Bekämpfung des Falschen Rebenmehltaus. Schweizer. Z. Obst-Weinbau 9: Loskill, B. 2004: Untersuchungen zur Epidemiologie von Plasmopara viticola (Berk. & Curt.) Berl. & De Toni auf der Basis serologischer und genetischer Studien. Dissertation, Universität Göttingen. Müller, K., and M. Sleumer 1934: Biologische Untersuchungen über die Peronosporakrankheit des Weinstockes mit besonderer Berücksichtigung ihrer Bekämpfung nach Inkubationszeitmethode. In: Landwirtschaftliches Jahrbuch 79: Schultz, H. R. 1992: An empirical model for the simulation of leaf appearance and leaf development of primary shoots of several grapevine (Vitis vinifera L.) canopy-systems. Scientia Hortic. 52: Schultz, H. R. 2003: Wachstumsmodell der Rebe: Jetzt wächst die Rebe am Bildschirm. Das Deutsche Weinmagazin 10: Siegfried, W., O. Viret and M. Sacchelli 2007: Die Kaskaden des Falschen Rebenmehltaus. Schweiz. Z. Obst-Weinbau Nr. 8/07. Viret, O., and B. Bloesch 2002: Observation on germination of oospores and primary infection of Plasmopara viticola (Berk & Curt.) Berl. & De Toni under field conditions in Switzerland. In: Proceedings of the 4th International Workshop on Powdery and Downy Mildew in Grapevine, Napa, California:

98 78

99 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp The expert system OiDiag-2.2 a useful tool for the precise scheduling of sprays against powdery mildew of vine (Erysiphe necator Schwein.) W. K. Kast and K. Bleyer State Institute for Viticulture, Oenology and Fruit Technology, D Weinsberg, Germany Abstract: OiDiag-2.2 is a system of two tools. The first one helps the vine growers to find the date for the start of the sprays against powdery mildew (Erysiphe necator). The second tool calculates an index value considering the climatical and ontogenic inputs for the risk of powdery mildew attack. Vine growers should get a table of relevant fungicides with values for the maximal time lag after the spray, which is completely covered by these fungicides considering these index values. Key words: Powdery mildew, Erysiphe necator, risk-model, ontogenetic susceptibility, climatical effects Introduction Powdery mildew caused by Erysiphe necator in spite of the development of highly effective fungicides in southern Germany is an upcoming problem, because of favourable climatical change in this region (Rupp and Kast, 2009). In order to get the best results in the control of powdery mildew, winegrowers need to know a suitable date for the first application of fungicides and the maximal permissible time lag for the subsequence sprays. OiDiag-2.2 is a system of two separate tools that allows determining the date and the time lag for these applications. OiDiag-2.2-system was developed to improve the OiDiag-1.0:-tool (Kast, 1997). It combines the knowledge about ontogenetic resistance of grapes (Stark-Urnau and Kast, 1999), influence of climatically conditions and long-term observations of powdery mildew development. Material and methods Data analyses for the calculation of the date of the first application of fungicides referred to records of temperature of the wine region Wuerttemberg over 53 years and records of the powdery mildew disease severity (Kast et al., 2004). Records of disease severity were performed by different individuals using a six class system: 0 = no powdery mildew, 1 = only a few leaves with late season powdery mildew, 2 = late season symptoms in a few vineyards, 3 = late season symptoms in most vineyards on leaves and lateral grapes; 4 = a few vineyards had powdery mildew on grapes, 5 = more than 5% of the vineyards had damage on grapes. Using this data, correlations and linear regressions were calculated. Based on the results of the regression analysis a function was developed for a practical or relevant phenological range. The index values for ontogenetic resistance took into account the findings of Stark-Urnau and Kast (1999) and Gadoury et al. (2003). In a first step, a fixed, weather-independent function based on the number of days after the three leaves stage was calculated for the vine regions Wuerttemberg and Baden. A new version OiDiag-2.3 will be integrated into the Vitimeteo-system. This system calculates the number of leaves based on the weather-data 79

100 80 according to the model of Schultz (1992). Therefore, in future the ontogenetic function will fit more appropriately to other regions. The climatical index values were calculated by considering temperature, humidity, rainfall and leaf wetness. For the index temperature, the findings of Delp (1954) and Fessler and Kassemeyer (1995) were considered. For the index humidity, the findings of Delp (1954) were neglected. The humidity function of OiDiag-1.0 (Kast, 1997) was reused because this relation was mainly confirmed by Carol and Wilcox (2003). Infections under completely dry conditions (days completely with RH <65) are of minor relevance for the epidemic. It was assumed that the factors humidity and temperature both should be close to the optimum for achieving appropriate conditions for the fungus. Therefore, the indices of humidity and temperature were combined by multiplying both factors. No relevant published data were found for the influence of rainfall and leaf wetness. The findings of Gadoury and Pearson (1990) about ascospore release show, that low rain volume in a special period even may favor the disease. As a result, a rainfall part index was developed as a combination of leaf wetness duration and amount of rainfall. The rainfall index was integrated into in the formula as a reduction of the temperature-humidity combination. Results ad discussion Date of the first spray The severity of powdery mildew disease was correlated significantly 1) to the mean of lowest temperatures of the two preceding winters (december-february) (r=0.45**) and 2) to the disease incidence of the preceding year (r=0.35**). Based on these results, a time lag for the first spray was calculated starting at the development stage of three leaves unfolded (BBCH13). Date of the first spray = Date of three-leaves-stage + (1,3xT-5xD + 12) days T= mean of the absolute value of the minimum temperatures of the two preceding winters (assumption: they are always negative and lower than -5 C) D= disease severity of the vine-site in the last year (6-class index) In cases of extreme disease severity in the preceding year and of warm winter conditions, the first spray is recommended at three leaves stage (T = -10 C, disease record D=5). If no disease was reported and low temperatures were measured during the preceding winters the start of the sprays is recommended five weeks after three leaves stage, which is normally after fruit set. Counting of spores (Falacy et al., 2007) could be an alternative to the use of this function but is much more time consuming and expensive than the use of weather data. Maximal time lag for the subsequent sprays, the OiDiag-2.2 Index As base for the maximal time lag for the following spray indexes were calculated. These index-values integrate the calculations for ontogenetic resistance, temperature, humidity, rainfall and leaf wetness. Ontogenetic index [On]: The ontogenetic resistance of grapes was factored according to Stark-Urnau and Kast (1999). Index values were higher for the time span 10 days before and after the blossoming period and decreased after this period of extreme susceptibility (Figure

101 81 1). A function based on number of days past three leaves stage was used, which proved to be suitable for the wine regions Wuerttemberg and Baden. ON = (D³ x (SQRT(D) x 176.6)-( LN(D) x 114,9*)-(D x 14.0)-66.0))/100) D= number of days since 3 leaves stage Figure 1. Ontogenetic index used for OiDiag-2.1 and 2.2 in 2009, based on number of days past three leaf stage.(x=days past three leaf stage) These calculations will be replaced by the number of leaves of the temperature dependent Geisenheim model (Schultz 1993) which is already in use for Vitimeteo (OiDiag-2.3.) The Vitimeteo-system used the Geisenheim model of Schulz (1992) for the development of a number of leaves. A similar function depending on the number of leaves offered the chance of a similar function with broader adaptation to different climatic situations. It is intended to allow experienced people to set their own ON-values in a table. Climatical index [CL]. The calculations of these indices were based on the input of temperature (daily means), hours with humidity between 65% and 80% and hours with humidity >80% per day, duration of leaf wetness (hours per day) and the amount of rainfall (mm per day). The climatical index was calculated from three indexes temperature (T), humidity (H) and rainfall and leaf wetness (R). T= temperature part index H= humidity part index R= rainfall and leaf wetness part index The calculated value was limited to a minimum of zero and a maximum of 1.1.

102 82 Temperature index (T): For the calculation of the temperature index, daily mean values were used. By use of multiple regressions a function was developed, that best fitted to the data of Delp (1954) and Fessler and Kassemeyer (1995). T= (((0.11xTM)-(0.0025xTMxTM))-0.6) x1.63 TM= daily mean of temperature Humidity index [H]. For the calculation of the humidity index, the function developed for OiDiag-1.0 was used. OiDiag-1.0 weighted hours with humidity level between 65% and 80% to 0.7 and more than 80% to 1.0. H = (H80+(H65x0.7))/24 H65 = hours with humidity >65 and <80 H80 = hours with humidity >80 For values below 65% RH the contribution to the OiDiag part index humidity is 0. This is used in spite of the results of Delp (1954) and Caroll and Wilcox (2003), which demonstrates the fact, that powdery mildew could infect vines even under lower humidity. The low contributions are ignored, because this spreads the index values and gives a better differentiated index.. The error effects of extremely high humidity near to 100% are corrected later on by the rainfall and leave wetness part index. Rainfall and leaf wetness index [R]. Low rainfall events could provoke ascospore release (Gadoury and Pearson, 1990) and therefore could favour the epidemic. Rainfall events with less than 2.5mm per day were neglected in OiDiag and events of >2.5mm/day were considered to be negativ. >10m/day were weighted 3-fold because we assume a strong diminishing of the powdery mildew epidemic. Leaf wetness hours are assumed to be negative for the epidemic as it may favour mildew antagonists and hyperparasites. The part index is calculated by weighting the hours of leaf wetness with the amount of rainfall: if PREC <2.5 then R1 = 0, otherwise R1= LWET x 1,5 if PREC >10 then R = LWET x 3, otherwise R = R1: PREC = rainfall/day [l/m²] LWET = Leaf-wetness [hours/day] OiDiag-2.2 index-values Index for a specific day. The index of a specific day was calculated by multiplication the ontogenic index (On) to the global climatic index. Indeed, each factor can limit the upcoming of infections and disease could propagate best if ontogenetic resistance is low and if optimal climatic conditions exist. Id=Cl x On Cl=climatic index On=ontogenic index Sliding means. Sliding means of the daily index values for the last seven days were calculated as final results. A period of seven days seems to be the most relevant result or practical application, since this is the shortest spray interval. In a period of seven days, an infection of powdery mildew may develop new spores already under optimal conditions. Therefore, good

103 83 climatic conditions for the fungus over this 7-day period define the maximal risk in field. These final values are given out as % values for better practical handling. Mean ((Id (day-6), Id (day-5)), Id (day-4), Id (day-3), Id (day-2), Id (day-1), Id (actual day) ))x100) Table 1: Maximal time lag in days for fungicides considering OiDiag Index-values I. Last sprayed fungicide OiDiag Indices Range >80 Wettable Sulfur Topas Systane 20 EW Prosper Discus Stroby Vento Power Universalis Cabrio Top Talendo Vivando Flint Collis Application of index values under consideration of the active period of fungicides The OiDiag index values were collocated into five classes. Fungicides were ranged for their preventive period for each of these five classes and results were summarized in a table that allows winegrowers to take a reading for the next spray. All information available, own results and published results were used to establish a draft and the results were finally discussed with members of the fungicide-manufacturers and trade organizations. Best information were given by the remarks about the fungicides of competitive organizations. The time lags ranged from 6 to 23 days depending on fungicide and the index value. The range of fungicides is not fixed over time. New fungicides will be developed and the effect of some fungicides will decrease by the development of resistance or a shifting. Application Test in wine estates in the Württemberg wine region with 43 different vine fields used the information of OiDiag for planning their sprays in Powdery mildew disease severity

104 84 was generally low in In 34 of these fields the first spray and the time lags were according to OiDiag or even earlier or shorter. In none of these vineyards, powdery mildew could be found. The required intervals were not fulfilled in 7 cases. In 4 of these, disease symptoms on a low level were found (maximal disease severity 2.4%). One estate with 4 vineyards started too late and had powdery mildew disease (severity 0.5%) in a high susceptible variety Trollinger. However the results in 2009 are less relevant because of the low disease pressure. An excel sheet with the complete OiDiag-2.1 calculations can be freely downloaded from or from References Caroll, J. E. & Wilcox, W. F. 2003: Effects of humidity on the development of grapevine powdery mildew. Phytopathology 93(9): Delp, C. J. 1954: Effect of temperature and humidity on grape powdery mildew fungus. Phytopathology 44: Falacy, J. S., Grove, G. G., Mahaffee, W. F., Galloway, H., Glawe, D. A., Larsen, R. C., Vandemark, G. J. 2007: Detection of Erysiphe necator in air samples using the polymerase chain reaction and species-specific primers. Phytopathology 97(10): Fessler, C. & Kassemeyer, H.-H. 1995: The influence of temperature during the development of conidia on the germination of Uncinula necator. Vitis 34(1): Gadoury, D. M. & Pearson, R. C. 1990: Ascocarp dehiscence and ascospore discharge in Uncinula necator. Phytopathology 80: Gadoury, D. M., Seem, R. C., Ficke, A., & Wilcox, W. F. 2003: Ontogenetic resistance to powdery mildew in grape berries. Phytopathology 93(5): Kast, W. K. 1997: A step by step risk analysis (SRA) used for planning sprays against powdery mildew (OiDiag-system). Vitic. Enol. Sci 52: Kast,.W. K., Rupp, D., Schiefer, H.-C., & Tränkle, L. 2004: Statistische Beziehungen zwischen Witterungsdaten und dem Auftreten von Krankheiten und Schädlingen im Weinbaugebiet Württemberg/Deutschland. Mitt. Mittsc. Klosterneuburg 54(4): Rupp, D. & Kast, W. K. 2009: Auswirkungen der klimatischen Veränderungen auf Phänologie der Rebe und die Bedingungen während der Traubenreife. Mitt. Klosterneuburg 59(4) (in press). Schulz, H. R. 1992: An empirical model for the simulation of leaf appearance and leaf area development of primary shoots of several grapevine, Vitis vinifera L., canopy systems. Scientia Horticulturae 52: Stark-Urnau, M. & Kast, W. K. 1999: Development of ontogenetic resistance of powdery mildew in fruit of differently susceptible grapevines (cvs Trollinger and Lemberger). Mitt. Klosterneuburg 49(5):

105 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp an interactive platform for a better management of grapevine diseases and pests O. Viret 1, P.-H. Dubuis 1, A.-L. Fabre 1, B. Bloesch 1, W. Siegfried 1, A. Naef 1, M. Hubert 1, G. Bleyer 2, H.-H. Kassemeyer 2, M. Breuer 2, R. Krause 3 1 Agroscope Changins-Wädenswil ACW, CH-1260 Nyon; 2 Staatliches Weinbauinstitut Freiburg, Freiburg, Germany; 3 GEOsens Ingenieurpartnerschaft, Ebringen, Germany Abstract: To control the main fungal diseases in accordance with the epidemiological development of the fungi in viticulture, the use of forecasting systems is one of the major progresses. Agrometeo is an interactive platform ( and a general tool for agriculture, including actual and historical weather data, modules for field crops, grapevine and fruit orchards. The grapevine module contains forecasting for downy mildew and grape berry moths; leaf area adapted spraying, growth development model, pesticides index, and descriptions of the main diseases and pests. The forecasting modules use data from a weather station network covering the whole viticulture area of Switzerland. Grapevine downy mildew forecasting is done with Vitimeteo, a new expert-system for the forecast of grapevine downy mildew, designed by the research institute of Freiburg (Germany), Agroscope ACW and the company GEOsens. The software generates graphics and tables freely available for the growers on the Internet. Leaf area adapted spraying was developed in analogy to the tree row volume concept on fruit orchards. The experiments show a reduction of about 20 to 30% of the use of pesticides by calculating the precise dose needed for a given leaf area. The module in allows a ready to use solution for the calculation of the amount of plant protection products based on the width and height of the leaf canopy and the row distance, depending on the registered amount indicated on the package. The number of Agrometeo users is in expansion confirming the grower's interest. Key words: grapevine, fungal diseases, pests, forecasting, agro-meteorology, Internet access, cropadapted dosage of pesticides Introduction To control fungal diseases in vineyards planted with Vitis vinifera varieties, the regular use of fungicides against downy and powdery mildews and Botrytis grey mould is necessary to ensure yield and quality of the grapes. Growers spray fungicides from budburst until beginning of ripening at regular intervals, generally independently of the disease pressure, by considering weather forecast to define the spray intervals. Using this strategy under Swiss climatic conditions, over ten fungicides can be applied. Insecticides and acaricides are nearly absent in the vineyards, mites (Tetranychus urticae and Panonychus ulmi) being controlled by predatory mites (e.g. Typhlodromus pyri, Amblyseius andersoni) and grape berry moths (Lobesia botrana and Eupoecilia ambiguella) by sexual confusion or Bt-based products. In organic vineyards, only copper, sulphur and few partially effective natural compounds are allowed which have to be applied more frequently than synthetic fungicides. Elicitors to reinforce natural defence mechanisms of vine are still in evaluation but do not reach the performance of chemical active ingredients. The only possible progress for a restricted use of 85

106 86 fungicides is an accurate prediction of infections and simulation of the epidemiological development of the fungi by models. The most interesting alternative to pesticides remains the breeding and planting of resistant varieties. This second point is in contradiction with current European regulations, prohibiting so-called hybrids, the only known possibility to introduce resistance genes from other Vitis species in V. vinifera. Growers planting resistant varieties can only produce table wines and are most of the time excluded from the "appellation d'origine contrôlée, AOC". In Switzerland, the Agrometeo platform project started in 2000 with the aim to centralise weather data from microclimatic measurement units in the field. Two years later, the first data were available on the Internet ( and the new expert-system Vitimeteo for the forecast of grapevine downy mildew was elaborated in collaboration with the Grapevine Research Institute of Freiburg (Germany) and the company GEOsens. Based on historical climatic and epidemiological data and recent results on the biology of Plasmopara viticola, the first tests could be performed in This paper presents the current stage of the platform Agrometeo, the grapevine downy mildew forecasting tool Vitimeteo with corresponding field epidemiological observations, the use of the forecast for the growers, and the leaf area adapted dosage of fungicides to spray at the right moment the right dosage of plant protection products. Material and methods The setup of a national network of weather data at the microclimatic level started in 2000 and ended in 2002 with the Internet display of the values. During this time, different weather stations available on the market were tested for their reliability, precision and compatibility under field conditions, using weather data from MeteoSwiss as reference. One year later, the first data transfer into existing models for apple scab (Rimpro, Welte) and fire blight (Maryblight) predictions were performed. In 2003, all data from field measurements of weather stations from ACW and extension services in the different Swiss cantons were daily transferred by GSM to central servers in Changins for the French and Italian parts and in Wädenswil for the German part of Switzerland. Weather station net 150 weather stations (Campbell CR10X, Campbell CR 1000, Lufft HP-100, Lufft Opus) are measuring temperature ( C), relative humidity (%), leaf wetness duration (h) and rainfall (mm) under field conditions, covering the whole country. Ten minutes data are sent via GSM, two times a day (4-6 a.m, 4-6 p.m). Forecast of downy mildew The model Vitimeteo (Bleyer et al., 2008a, 2009; Viret et al., 2005) developed by the company Geosens (Germany) in collaboration with experts from Agroscope ACW (Switzerland) and the Grapevine Research Institute of Freiburg (Germany) is used since 2005, simulating the main development steps of the epidemiology of Plasmopara viticola (Bleyer et al., 2008b; Viret et al., 2007). All parameters included in the model can be adjusted after experimental values by the experts. Results are presented as summarized tables and graphs for each region with the possibility to access to detailed tables containing all data from the first of January. The predicted downy mildew risk for the next five days appears greyish on the tables, based on five days weather forecast from Meteoblue (Basel, Switzerland) for temperature, rain and relative humidity.

107 87 Growth model The growth model of Schultz (1992) is implemented and gives information on the number of developed main leaves and the leaf surface per shoot. This parameter can be of importance for the evaluation of the newly unfolded and unprotected leaves after the last treatment. Leaf area adapted dosage of fungicides The leaf area adapted dosage of fungicides developed by Siegfried et al. (2007) allows a rapid calculation of the dose of plant protection products to apply, using row distance and measured high and width of the leaf canopy. An easy to use calculation is available for the growers on Internet access All data and results of calculation are actualised twice a day (9 am and 7 pm in Changins, 10 am and 3 pm in Wädenswil) and freely available on the Internet for the growers under Field experiments On farm field experiments have been conducted on standard vineyard of 6500 to 7000 plants per ha ( x 0.8m) at different places in the French part of Switzerland and in the experimental plots of Agroscope-ACW in Changins (VD), Leytron (VS) and Wädenswil (ZH). The experiments were conduced with the sensitive varieties cv. Chasselas, Müller- Thurgau, or Pinot noir. The control strategy against downy mildew was to wait for the first primary infection calculated by Vitimeteo and to place a contact fungicide (active ingredient folpet) at 80-90% of the incubation time or a penetrating fungicide few days after the beginning of the first secondary infection (Viret et al., 2001). Spray intervals were determined using Vitimeteo, by considering duration of efficacy of 10 days for contact fungicides and of 12 days for penetrating and systemic active ingredients. In all plots, an unsprayed part of at least 200m 2 was used to follow the epidemic of downy mildew. Results were obtained by regular scoring of downy and powdery mildews in the control plots, counting four replicates of 100 leaves and 50 bunches, compared with the sprayed part after Vitimeteo. The diseased leaf surface was visually estimated (0, 1=0-2.5%, 2=2.5-10%, 3=10-25%, 4=25-50%, 5=>50%) and disease frequency and severity was calculated. Experiments were specifically designed to evaluate the efficacy of spraying schedules used by the growers (different active ingredients) conventionally dosed, compared to the adapted dosage according to Siegfried et al. (2007). The fungicides used were contact active ingredients (folpet, sulphur, copper), penetrating a.i. (strobilurines, cymoxanil, triazols, amide carbamates) or systemic a.i. (Alfosetyl, phenylamides) applied at the recommended concentration, adapted to the leaf area (Siegfried et al., 2007) or to the growth stage (Viret and Siegfried, 2009). Downy mildew laboratory under natural conditions An observation laboratory under field conditions has been build up for the validation of Vitimeteo. Control sensitive cv. Pinot noir and Gamay vines are planted with a stock of downy mildew infected leaves underneath, placed every year in autumn. To follow oospores maturation, leaf pieces of approx. 5mm 2 are selected under the binocular (to ensure the presence of oospores) and placed in Falcon tubes of 50ml deep in the soil. In spring, at regular intervals, the leaf pieces are transferred in Petri dishes at 100% rh and the emergence of primary sporangia counted every day under the binocular. Oospores are considered mature, when germination occurs within 24 h. From that date, trap-plants (cuttings with 6-7 unfolded leaves cv. Chasselas grown during winter in greenhouse) are placed always before rainfalls over a stock of infected leaves and placed in the greenhouse after the rain, to check for oil spots (100% rh, darkness, 20 C).

108 88 Results and discussion Effective epidemics and Vitimeteo calculation Validation of the model Vitimeteo by comparing data from field observations and calculated infections shows a good correspondence (Table 1). Under the climatic conditions of Changins, oospores reach maturity between April 30 (2007) and May 11 (2006), when vine have two to four unfolded leaves (Table 1). Maturity of the oospores is not a limiting factor for the beginning of the epidemics under the given Swiss climatic conditions. Quantitatively, the maximum number of sporulating oospores after that date occurs in the same period of time and no correlation could be found between the amount of germinated oospores and the severity of the infection during summer. The total amount of rain from January to May had no influence on the germination rate of the oospores or on severity of the epidemics. In Vitimeteo, the date of maturation of the oospores can be indicated manually or is calculated using the daily temperature sum of 160 C (>8 C) from the first of January. Table 1 shows, that this date can be delayed (in 2005 and 2009) compare to the observed dates of germination. To be more in accordance with the effective germination, the limit value for the temperature sum >8 C was adjusted in Vitimeteo at 140 C for the French part of Switzerland. Calculation of the primary infection can only be confirmed by the appearance of oil spots on trap-plants or on vine. The data obtained during the last 5 years with trap-plants indicate that the predicted infections are generally before the occurrence of downy mildew in the vineyard beneath. This can be explained by the vulnerability of the cuttings used as trap-plants, and the physical vicinity of the healthy leaves to the infected leaf stock. The calculated date of the first primary infections by Vitimeteo under variable climatic conditions is generally earlier as the date related to the observed oil spots at the end of the incubation time. For the rest of the growing season, the number of infections indicated by Vitimeteo is in accordance with the disease development in the field reflected by the severity. On-Farm experiments with Agrometeo and Vitimeteo Generally, the use of Vitimeteo for the management of diseases and pests is evaluated positively by the extension services and by the vine-growers. Under the Swiss conditions with locally high downy mildew pressure, Vitimeteo is a precious tool for a more precise control of the disease. Growers using it are spraying more in accordance with the epidemic and, in dry years, they can objectively delay the first spray and decide to enlarge spray intervals, reducing the number of sprays. Globally considered, the reduction of the number of sprays can be of two to three, when disease pressure is weak (Table 2). Generally the total number of infections calculated by the model Vitimeteo before and after flowering is proportional to the disease severity and related to the number of spray applications by the grower (Table 2). Vitimeteo is every year more used by the growers, as shown by the statistics of requests on the internet site. Between 2006 and 2009, the mean yearly increase of requests reaches 15%, with over 100'000 in One of the most important points is the quality of measurements in the field and the reliability of the data which have to be checked before being integrated in any models. In the Agrometeo database, this check is automatically performed when the values are communicated by the weather stations and the stations are regularly surveyed in the field by specialists.

109 89 Table 1. Biological and epidemiological characteristics of downy mildew on cv. Chasselas, Changins (VD), , under natural conditions. Oospores maturation, trap plants, number of infections, growth stage of the vine, and rainfalls. Downy mildew epidemic: 0, absence; (+), few oil spots; +, weak presence; ++, strong epidemic; +++, very strong epidemic, locally with economical losses. Date of observed oospores maturation (germination within 24h) Growth stage BBCH Date of the maximal number of oospores sporulating after the date of oospore maturity Growth stage BBCH Mean number of germinated oospores per count May 11 May 30 April 5 May 4 May May May April May May Sum of rain (January to May) 339 mm 478 mm 348 mm 369 mm 190 mm Temperature sum (>8 C) = 160 Growth stage BBCH 8 May May April May 52 8 May 51 Date of first primary infection after Vitimeteo 14 May 18 May 30 April 16 May 27 May Date of first oil spots observed in vineyard 06 June 22 May 29 May 26 May 25 May Downy mildew observed on trap-plants never 12 May 12 June 26 May 26 May Beginning of exponential development of epidemic end July end May mid June end May end June Total number of secondary cycles after Vitimeteo before bloom (observe disease 1 (+) 4 (+) 15 (+) 17 (++) 1 (0) severity) Total number of secondary cycles after Vitimeteo during bloom (observed 2 (0) 5 (+) 4 (+) 3 (+) 9 (0) disease severity) Total number of secondary cycles after Vitimeteo after bloom (observed disease 17 (+) 34 (+) 37 (+++) 24 (++) 20 (+) severity) Sum of rain (May to August) 256 mm 319 mm 536 mm 338 mm 215 mm 53

110 90 Table 2. Downy mildew on cv. Chasselas at Perroy (VD). Rainfall sum (mm), date of the first primary infection calculated by Vitimeteo, appearance of the first oil spots, sum of monthly infections, total number of sprays to control downy and powdery mildew from May to August Date of primary infection indicated by Vitimeteo 16 May 16 May 27 April 16 May 15 May First oil spots 9 June 29 May 6 June 29 May 5 June First spray application 24 May 19 May 1 May 23 May 5 May Beginning of bloom 13 June 14 June 29 May 14 June 9 June Number of sprays Rain (mm) May June July August Sum Number of downy mildew infections indicated by Vitimeteo May June July August Sum Table 3. Downy mildew on leaves and bunches (frequency and % intensity in parentheses) observed at the end of August (average of 4x 100 leaves and 4x 50 bunches per plot) at Perroy (VD) cv. Chasselas, on unsprayed control plots, plots sprayed with dosage of fungicides adapted to the leaf area, and plots sprayed with standard dosage. The adapted dosage is a sigmoid adaptation of the amount active ingredient in function of the leaf area of the vine after Siegfried et al. (2007). The adaptation rate is the difference between the leaf area adapted dosage and the standard dosage expressed in %. Unsprayed control Leaf area adapted dosage according to Siegfried et al. (2007) Standard dosage Leaves Bunches Leaves Bunches Leaves Bunches Number sprays Adaptation rate (40.9) 7.3 (0.5) 8 (0.7) 0 (0) 7.3 (0.5) 0 (0) (61.5) 42 (9.6) 2.6 (0.1) 31.9 (7.1) 0.6 (0.01) (32.7) (10.1) 4.2 (0.3) 18 (6.7) 3.3 (0.1) 9 22 (94.1) (44.6) (48) 51.3 (10.2) 27.3 (7.6) 49.7 (9.9) 16.5 (4) (44.7) (38.4) 52.7 (6) 7.3 (1.3) 0.7 (0.1) 1.7 (0.5) 1.3 (0.1)

111 91 Leaf area-adapted dosage of fungicides The disease pressure of downy mildew at Perroy, a place located on the lake of Geneva in a relative humid area is presented in Table 3. Every year, except in 2005, the unsprayed control is nearly destroyed by the pathogen and no yield can be harvested. The leaf area adapted spraying follows the growth curve of the vines. This approach allows a reduction of the amount of fungicides against both, downy and powdery mildews of 15 to 26% compared with the linear growth stage adapted dosage (Table 3). Yearly variations depend on the climatic conditions and on the spraying schedule, especially before and during bloom. The years with very high downy mildew pressure (2006, 07, 08), both dosages present diseased leaves and bunches, but no significant differences could be found. In these years, no economical losses could be found, except in the unsprayed control. The adaptation of the dosage to the leaf surface is easy to use in Agrometeo, but can only be performed in trellised vineyards with properly calibrated spraying equipments. References Bleyer, G., Kassemeyer, H.-H., Krause, R., Viret, O. & Siegfried, W a: "Vitimeteo- Plamopara" Prognosemodell zur Bekämpfung von Plasmopara viticola (Rebenperonospora) im Weinbau. Gesunde Pflanzen 60: Bleyer, G., Kassemeyer, H.-H., Viret, O., Siegfried, W. & Krause, R b: Vitimeteo- Plasmopara a modern tool for integrated fungicide strategies. IOBC/WPRS Bull. 36: Bleyer, G., Kassemeyer, H.-H., Viret, O., Siegfried, W. & Krause, W.R. 2009: "VitiMeteo": innovatives Prognosesystem. Der deutsche Weinbau 13: Viret, O., Bloesch, B., Taillens, J., Siegfried, W., & Dupuis, D. 2001: Prévision et gestion des infections du mildiou de la vigne (Plasmopara viticola) à l'aide d'une station d'avertissement. Revue suisse Vitic. Arboric. Hortic. 33(2): I-XII. Viret, O., Bloesch, B., Fabre, A.-L., Siegfried, W., Bleyer, G., Huber, B., Kassemeyer, H.-H. & Steinmetz, V. 2005: Vitimeteo: un nouveau modèle de prévision pour le mildiou de la vigne ( Revue suisse Vitic. Arboric. Hortic. 37(1): Viret, O., Bloesch, B., Fabre, A-L., Siegfried, W. 2007: Prévision du mildiou sur et gestion de la lutte. Revue suisse Vitic. Arboric. Hortic. 39(1): Viret, O. and Siegfried, W. 2009: Application des produits antiparasitaires. In: Le guide Viti Revue suisse Vitic. Arboric. Hortic. 41(1): Siegfried, W., Viret, O., Huber, B. & Wohlhauser, R. 2007: Dosage of crop protection products adapted to leaf area index in viticulture. Crop Protection 26(2): Schultz, H. R. 1992: An empirical model for the simulation of leaf appearance and leaf development of primary shoots of several grapevine (Vitis vinifera L.) canopy-systems. Scientia Hortic. 52:

112 92

113 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Modelling and experimenting crop protection decision workflows: some lessons from GrapeMilDeWS research O. Naud 1, P. Cartolaro 2, L. Delière 2, B. Léger 3 1 Cemagref UMR ITAP, BP 5095, Montpellier Cedex 5 - France; 2 INRA-Bordeaux UMR INRA-ENITA 1065 Santé Végétale ISVV, BP81, Villenave d Ornon France; 3 Arvalis - Institut du végétal, station de la minière, Guyancourt - France. Abstract: The political roadmap about pesticides in France is to reduce quantities by 2, if possible, before Research is needed to design and evaluate new pest management solutions. A decision workflow system, name GrapemilDeWS was designed, at the plot scale, to handle grapevine powdery and downy mildews. GrapeMilDeWS stipulates throughout the season if and when fungicide sprayings should take place. GrapeMilDeWS has been experimented on a network of plots in different French wine regions. We give some results about the number of sprayings generated by GrapeMilDeWS and the crop protection performance obtained. We discuss methodological aspects such as partnership and data required to test and check such decision system. Key words: Plasmopara viticola, Erysiphe necator, decision, workflow Introduction The French political roadmap on pesticides ( Grenelle de l environnement, Ecophyto 2018 ) is to divide sprayed quantities by 2, if possible and before This emphasizes need for research to design and evaluate new pest management solutions. When it comes to decide about when to protect grapevine against powdery and downy mildews, the framework of Protection raisonnée constitutes the current basis for French advisers on how to avoid unnecessary treatments. Protection raisonnée is related to Integrated Crop Protection in the sense of using indicators and statements often called rules about how to use these indicators. The current status about Protection raisonnée of grapevine against mildews is the following: the information provided to growers is mostly bioclimatic at a micro-region scale; it is mostly used, in France, to reduce spraying in the beginning of the season. Where and when can we do more with these information and indicators, can we combine risk management and proven tactics to design environmentally efficient crop protection decision systems? This paper contributes to answer these questions and is also a call for building international research collaboration on operational decision systems and models for grapevine. The authors of this paper study reasoning methods about when to spray and when not to spray. More precisely, we design decision systems, which, throughout the season, output decisions based on the current epidemic and bio-climatic inputs. On the track to reduce pesticides use in viticulture, this work is very complementary with work on dose management undertaken by e.g. IFV (Davy, 2007) and Agroscope Changins (Siegfried, 2007), and also on optimisation of the spraying process itself (Gil, 2007). 93

114 94 Previous work GrapeMilDeWS stands for Grape Mildews Decision Workflow System. GrapeMilDeWS manages the decision making about control of both powdery and downy mildews. It was originally designed and tested in 2005 on 4 experimental plots in the Bordeaux region (Léger, 2008). The original guidelines consisted in 7 stages defined on a phenological timeline, 1 decision array per stage, and a set of indicators, including 3 field observations per year. Then, a model of the decision system was made in the Statecharts language. The Statecharts language (Harel, 1987) describes dynamic processes on the basis of state machines, which react to events, as described in Figure 1. The Statecharts language possesses mechanisms such as hierarchy and concurrency for combining state-machines, which ease the building of models. GrapeMilDeWS was elicited with careful interviews of each expert of the phytopathologists team, and the interviews were made on the basis of successive versions of the statechart diagram (Leger, 2009). The resulting Statechart model was introduced to OILB s grapevine group in 2007 (Léger, 2007). It was successfully checked against actual behavior of experts during the experiments of 2005 & 2006 (Léger, 2008). The model has thus been shown to be consistent with the experts how-to decide knowledge on both a declarative and a behavioural basis. state Event [guard] / effect other state e.g. healthy contact [sensitive] / infection infected Figure 1: States and events in Statecharts graphical language. Experimenting GrapeMilDeWS in different wine regions Basics of GrapeMilDeWS: a quick reminder GrapeMilDeWS is precisely described in the Statecharts language in (Leger, in press). The 7 decision stages are decomposed as follows. Each stage leads to at most one treatment against each of the two diseases. Whenever it is appropriate in a stage, if both diseases have to be treated, the spraying combines products against both diseases in order to limit the workload, consistently with growers practice. There are three stages before flowering, one from bud break to 5 leaves unfolded (stage 0), a second one starting after a field survey to be performed between 5 and 7 leaves unfolded (stage 1), and the last starting two weeks after, also initiated by a field survey (stage 2). At flowering, the decision stage 3 consists in a unique and mandatory spraying against both diseases. Stage 4 follows and manages fruit set. There is no field survey at this stage, the treatments are decided according to the epidemics levels estimated before flowering, and according to the local downy mildew risk. A third field survey has to be done before stage 5. It allows deciding if it is necessary to protect fruit and leaves before grape closure. The last stage is a mandatory copper treatment at ripening.

115 95 GrapeMilDeWS experiment protocol Since 2008, experiments on GrapeMilDeWS rely on a protocol with a decision procedure written in quasi natural language that was derived from the Statechart model. This protocol gives detailed instructions on how to decide at each stage, and when to move from one stage to the other. It describes the sampling methods for getting epidemic level in the field: number and choosing of vinestocks, leaves, and grapes. It specifies how to record the local risk indicators for the control of downy mildew: bioclimatic info and rain forecasts. It stipulates that the decision path that leads to spraying is part of experimental data and is to be recorded. The protocol naturally includes instructions for the evaluation of protection performance. Thanks to this protocol, it became possible to extend the experimental network. The extent of this network as of 2009 is described below. After the 2008 campaign, due to a few cases mentioned in the results section, the first stages (0, 1 & 2) of GrapeMilDeWS were slightly transformed in order to handle more efficiently high and early downy mildew pressure. The network of experimental plots (since) 2005 / 2006 (since) 2007 (since) plots 17 plots 14 plots Figure 2: Network of plots were GrapeMilDeWS is tested in France. Since the beginning, GrapeMilDeWS is experimented on real size plots. The sprayings are performed with the usual spraying equipment of the farm. The experimental network has grown and diversified. Whereas in 2005, 2006 and 2007, the plots experimented were on research estates, since 2008, the network includes vineyards run by extension services (ES) as well as professional growers. In the later case, the field surveys are done by researchers or ES

116 96 staff. As can be seen on Figure 2, the network spans over a number of wine producing regions: Bordeaux, Languedoc-Roussillon, Provence, Côtes du Rhône, Burgundy, Jura, Champagne and Cognac. The total number of plots has reached 39 in Results Spraying intensity As the goal of GrapeMilDeWS is to provide a satisfactory level of crop protection with a few sprayings, we give in table 1 a brief synthetis of results on how many treatments were done on testing plots from 2005 to The conventional number of sprayings was obtained thanks to surveys. Table 1. Number of spraying with GrapeMilDeWS compared to conventional practice. Year Nb of sprayings Downy mildew Nb of sprayings Powdery mildew (Number of plots) Median (min max) Median (min max) Location GrapeMilDeWS Conventional GrapeMilDeWS Conventional INRA plots 4 6 (4-12) 2 (2-4) 5 (3-11) Bordeaux INRA plots 4 7 (5-8) 2 (2-3) 6 (3-10) Bordeaux 2007 network 6 plots 6 (5-6) 10 (7-15) 2 (2-3) 6 (4-9) Bordeaux 2008 network >12 plots Bordeaux (network > 10 plots) Languedoc Roussillon 6 (5-6) 4 (3-5) 10 (7-14) (2-4) 6 (3-13) plots Bordeaux + Cognac 12 plots Languedoc Roussillon 5 (4-7) 4 (3-6) - 2 (2-4) 2 (2-5) - 7 plots Beaujolais - Bourgogne - Champagne - Jura 4 (2-6) - 2 (2-3) - Agronomical performances On Figure 3, the analysis is related to the hypothesis that severity of diseases on bunches should be kept below 5% for satisfactory protection, since it has been shown for powdery mildew (Calonnec, 2004) that limited contamination do not cause loss of wine quality. The level of severity has to be considered together with the yield, which is given here relatively to a yield objective (e.g. a quota). Most of the plots where GrapeMilDeWS has been tested are

117 97 meant to produce wine under designation of origin label. In parts 2 and 4 on Figure 3, yield is under the objective. These situations can result from disease level, or from other problems such as physiological trouble (flower abortion, millerandage) or hydric stresses. Whatever the cause, the grower s income is then below its expected maximum. Yet as far as GrapeMilDeWS is concerned, only cases situated in part 4, i.e insufficient yield and severity above 5%, are considered as possibly problematic Disease severity on bunches (Downy and Powdery Mildews) (%) Aquitaine 2008 Languedoc Roussillon Deviation from yield objective (%) Figure 3. Disease severity on bunches and deviation from yield objective The analysis of these cases in 2008 led to modifications of stages 0 and 1. Before 2009, a field survey had to be done before the first treatment, would it be within stage 0 or within stage 1. However, in the Bordeaux region, the BBCH phenological stage 15 (5 leaves unfolded) occurs beginning of May. Bank holidays occurred close to week-ends and make it difficult to schedule human resources for a field survey before a possibly contaminating rainfall. Therefore it may be difficult to schedule a spraying when downy mildew epidemics start fast and remain invisible because of the delay between contamination and expression of symptoms which was what happened in 2008, near Bordeaux. When regional risk is estimated high, it is then important that GrapeMilDeWS gives the possibility to schedule a first treatment against downy mildew during stage 0, without doing a preliminary field survey, and before the field survey that precedes stage 1. The protocole was modified from 2009 on, so as to account for this conclusion.

118 98 Discussion: What does the decision workflow approach change for? Experimenting decision strategies Decision workflow versus pre-scheduled spraying A decision workflow acting on a pathosystem produces a decision that changes according to inputs (variables, events) and workflow state, state which is the consequence of a sequence of past events. This is very different from a pre-schedule spraying, as happens for example in phytosanitary products homologation tests. With the decision workflow, the epidemiological facts that are accumulated during experiments are attached to the decision path, which includes timed traces of indicators, estimated phenology, and the decisions for scheduling treatments. The notion of decision path for experiments is illustrated in Figure 4. A B B A B B A B B A C A C A C C C C Year Y 1, plot P 1 Year Y 2, plot P 1 Year Y 2, plot P 2 Figure 4. Decision path for different test cases (a case is a year and a plot). Recording decision paths along with data from the field and the weather from a wide range of cases across regions and test years, allows us to systematically explore the decision workflow and link its behaviour to agronomical results. For instance, we can check which decision paths have been taken so far, and which have not. We thus identify which parts of the decision system have been tested and according to which conditions. For example, in Figure 4, one can identify that the path A. C has not been tested. A decision system is analogous to a computer program in this regard. Computer scientists and engineers are well aware that systematic testing is a tedious but necessary activity for the sake of users safety (e.g. embedded electronics of cars) or users activity (e.g. personal computer programs). Model-checking are formal methods for performing such systematic verification procedures, and assess behavioural properties of systems (Müller-Olm, 1999; ten Teije, 2006). Interdisciplinary research Choosing operational concerns as a subject of study such as how to decide treatments according to epidemics, wheather forecasts, and resources available required and fostered interdisciplinary research. The research about GrapeMilDeWS is done within projects which blend phytopathology, agronomy, automation, computer science, economics and management. Because it is very difficult to simulate downy and powdery epidemics at the plot scale, the most convenient way to validate the technical performance of a decision system is to test it, in many different situations and for several years. This means that the research on such decision systems requires the creation of experiment networks with extension services and growers. This is what we did. As a result, the transfer to development and the accumulation of knowledge for further research are happening simultaneously.

119 99 GrapeMilDeWS has been designed to be generic and functional in a variety of situations and regions. What will be made available to the growers in different regions may be a set of local adaptations, or even completely new designs that will take into account facts demonstrated by GrapeMilDeWS. For example, it has been made clear that the lower susceptibility of grapes to powdery mildew disease after pea-size can be used to skip treatments from the beginning of berry touch in case of low epidemics. Handling risk with low fungicide input is a new territory for grapevine growers. We believe that the job of a farmer, who is not a phytopathologist and has many things to do to run his estate, is not to explore this territory on its own. It should fall on the side of research and development to design and test prescriptive solutions that are «safe routes» in the search space of crop protection decision and make it possible to reach the production target with little fungicide input. This applied and interdisciplinary research needs to be conducted in close relation with the professional community. Conclusion We have learnt from GrapeMilDeWS experiment how the pathosystem behaves under low input in different regions and we have improved the design of GrapeMilDeWS. We have run large field experiments without usual blocks and repetitions, still the accumulation of cases allows to consolidate scientifically sound knowledge. What extension workers have been learning is that some level of disease can be accepted. They have seen a number of cases from 2007 to 2009 which showed that it is sometimes possible to limit powdery mildew treatments to two sprayings, even on sensitive cultivars (like Chardonnay, Gamay,,). GrapeMilDeWS has been designed and experimented at the plot scale. At this scale, extension workers have experienced that low input strategies are technically feasible, and have handled the decision workflow, followed a specific observation protocol and interpreted the decision variables. What growers participating to this research have been learning is that (i) some disease can be tolerated without economical loss and going out of business and that fungicides may be reduced significantly (repeatedly >40% less treatments). They have seen scientifically sound research work made in large fields (fields in the network are >0.5ha) and gained confidence in the possibility to control disease risk. Further experiments will be conducted at the plot scale. Besides, future research should investigate several points such as decision at the farm scale, optimisation of sampling in different plots of an estate, enhanced modelling formalisms to account for anticipation and revision of anticipated decisions to follow evolution of bioclimatic forecasts, model-checking of decision systems. Acknowledgements This work has been carried out with the financial support of the French National Research Agency (ANR) under the Programme Agriculture et Développement Durable, project ANR-05-PADD-001, Vin et Environnement and is now supported by A2PV program of French Agricultural Ministry, project SyDéRéT.

120 100 References Calonnec, A., Cartolaro, P., Poupot, C., Dubourdieu, D., Darriet, P. 2004: Effects of Uncinula necator on the yield and quality of grapes (Vitis vinifera) and wine. Plant Pathology 53(4): Darriet, P., Pons, M., Henry, R., Dumont, O., Findeling, V., Cartolaro, P., Calonnec, A., Dubourdieu, D. 2002: Impact odorants contributing to the fungus type aroma from grape berries contaminated by powdery mildew (Uncinula necator) ; incidence of enzymatic activities to the yeast Saccharomyces cerevisiae. Journal of Agricultural and Food Chemistry 50: Davy, A. 2007: Le programme Optidose: optimisation agronomique et environnementale de la pulvérisation. EUROVITI, novembre 2007, pp Gil, Y., Sinfort, C., Brunet, Y., Polveche, V., Bonicelli, B. 2007: Atmospheric loss of pesticides above an artificial vineyard during air assisted spraying. Atmospheric Environment 41: Harel, D. 1987: Statecharts: A visual formulation for complex systems. Science of Computer Programming 8(3): Léger, B., Naud, O., Bellon-Maurel, V., Clerjeau, M., Delière, L., Cartolaro, P., Delbac, L. 2010: GrapeMilDeWS: a formally designed integrated pest management decision process against grapevine powdery and downy mildews. In: Decision Support Systems in Agriculture, Food and the Environment: Trends, Applications and Advances, eds. B. Manos, K. Paparrizos, N. Matsatsinis and J. Papathanasiou, IGI Global. Léger, B., Naud, O. 2009: Experimenting Statecharts for Multiple Experts Knowledge Elicitation in Agriculture. Expert Systems With Applications 36(8): Léger, B. 2008: Recueil et Formalisation de procédés experts pour conduire une protection intégrée du vignoble. PhD Thesis with core contents in english, Supagro Montpellier (France). Léger, B., Cartolaro, P., Delière, L., Delbac, L., Clerjeau, M., Naud, O. 2008: An expert based crop protection decision strategy against grapevine's powdery and downy mildews epidemics: Part 1) formalization. IOBC/WPRS Bull. 36: Müller-Olm, M., Schmidt, D. A., Steffen, B. 1999: Model checking: a tutorial introduction. Proc. 6th Static Analysis Symposium, SAS 99, Venice (Italy), G. File and A. Cortesi, eds., Springer LNCS 1694, pp Siegfried, W., Viret, O., Huber, B., Wohlhauser, R. 2007: Dosage of plant protection products adapted to leaf area index in viticulture. Crop Protection 26(2): ten Teije, A., Marcos, M., Balser, M., van Croonenborg, J., Duelli, C., van Harmelen, F., Lucas, P., Miksch, S., Reif, W., Rosenbrand, K., Seyfang, A. 2006: Improving medical protocols by formal methods. Artificial Intelligence in Medicine 36(3):

121 Session 2: Biology and epidemiology of pathogens, fungal, bacterial and physiological diseases, including grapevine trunk disease

122

123 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Impact des données météo de précision sur l évaluation locale du risque épidemique modelisé sur le vignoble. Premiers résultats de validation spatiale des orages de grêle de mai 2009 sur les vignobles de Bordeaux et Cognac M. Raynal 1, C. Debord 1, S. Guittard 1, M. Vergnes 1, J. Congnard 2, D. Grimal 2 1 Institut Français de la Vigne et du vin (IFV), Blanquefort, France; 2 Météo France, Mérignac, France Résumé: Les orages de grêle très marqués de ce printemps 2009, nous ont permis de tester sur le terrain, les outils de validation spatiale des données météo pluviométrique à maille fine (1x1km). Ces premiers tests, effectués sur la grêle à défaut de pouvoir les cibler sur des foyers de développement maladie, révèlent la pertinence de ce travail d enquête dans la mesure où ils peuvent porter sur des phénomènes de type accidentel, ou fortement localisés. La plate forme de saisie Web, incitant les techniciens et viticulteurs, à partager leurs observations nous a également donné satisfaction car elle a fait l objet d un accueil favorable auprès de la profession. Du point de vue du zonage de la grêle, on retient que l information radar permet de définir le caractère grêlés ou indemne d une parcelle avec un taux réussite proche de 75%. Le radar ne permet cependant pas de cerner l approche quantitative des dommages subis par la vigne. Ce travail d expérimentation participative en réseau sera prochainement appliqué aux thématiques épidémiologiques liées à la prévision des risques d épidémie. Key words: agro-météorologie, pluie, grêle, radar pluviométrique, validation spatiale, zonage, modélisation Introduction Depuis 2007 un partenariat expérimental est élaboré entre les équipes de Météo France et de l Institut Français de la Vigne et du Vin (IFV) de Bordeaux pour évaluer et apprécier l impact des données météorologiques à maille fine sur la qualité de l appréciation des risques d épidémie. Dans le cadre de ce partenariat, l IFV teste les données: - Antilope, donnée pluviométrique obtenue par la fusion des informations sur la localisation des zones de pluies par le radar et la quantification des hauteurs de pluie par le réseau de stations déployées au sol; cette donnée est délivrée à la maille 1km - Safran, donnée de température modélisée par Météo France à la résolution de 6 x 8km. Ces données servent ainsi d intrant au calcul des modèles Potentiels Systèmes mildiou, oïdium et black rot de S. Strizyk (société SESMA). L évaluation des risques de développement épidémique concernant ces différentes maladies est ainsi progressivement établie à partir de 6000 et 7000 points de calculs respectivement pour les Ha et Ha des vignobles de Bordeaux et Cognac, en lieu et place des 45 et 38 stations du réseau Démeter jusqu à présent exploitées à cette fin. Quel que soit le système de mesure des données Météo France ou Démeter utilisées, les résultats tant météorologiques que portant sur l évaluation des risques, sont représentés à l aide du Système d Information Géographique (SIG) Epicure. L expérience de la campagne 103

124 nous a montré l aptitude du modèle a fortement réagir sur des évènements climatiques très localisés comme le montre la figure 1. MAX MIN 60% 3 km 2% 45 Stations Déméter 6000 points ANTILOPE FTA Démeter FTA Antilope Pluviométrie du 24 mai 2007 FTA (%) mildiou le 4 juillet 2007 Zoom FTA mildiou Figure 1: représentation spatiale de l impact de l orage du 24 mai 2007 sur l évaluation de la Fréquence Théorique d Attaque (FTA) de mildiou simulée par le modèle Potentiel Système début juillet 2007 sur les vignobles de Gironde et Dordogne; zoom sur la commune de Fougeyrolles (24). Le 24 mai 2007, un orage avec formation de grêle provoque, sur un foyer de 2 à 3km de diamètre, une précipitation estimée à 107 mm d eau près de la commune de Fougeyrolles (24), qui n est pas enregistrée par le réseau de stations Démeter dont la maille, de l ordre de 10 à 20km entre chaque poste, ne permet pas la détection. Quelques semaines plus tard, le modèle indique sur ce foyer une FTA de 60% alors qu elle n atteint que 2% à quelques kilomètres de distance. De même que pour les stations météo, le réseau d une cinquantaine de témoins non traités déployé sur les vignobles de Gironde et Dordogne, ne permet pas de matérialiser la réalité du phénomène décrit par le modèle. Un dispositif ponctuel d observation de terrain, basé sur un protocole d enquête permettant d apprécier le zonage d attaque a ainsi été élaboré pour vérifier la validation spatiale fine de tels foyers infectieux. Un tel phénomène aussi marqué ne s étant pas reproduit lors des deux millésimes passés, la validité de ce dispositif est testée sur les épisodes de grêle observés au cours du mois de mai Matériel et méthodes La donnée Radar Le RADAR de Bordeaux-Mérignac est un RADAR Mélodi en bande S, mis en service en Ce RADAR fonctionne tous les jours 24h/24h et 7j/7j. Il explore l horizon sur 3 sites différents selon des angles de 0,8; 1,2 et 1,8 degrés par rapport à l horizontale. Il accomplit un tour complet en 1 min environ et délivre une information utile dans un rayon de 80 à 100km. Météo France étudie la concordance entre l écho radar et les cellules pluvieuses. La donnée de quantification des pluies, Antilope, est ainsi établie par calibration à partir des relevés de pluies relevées par le réseau de station au sol déployé par Météo France. Contrairement aux données Stations qui représentent la pluviométrie en un point précis, les données Antilope représentent la pluviométrie moyenne sur un pixel d environ 1km 2.

125 105 Un travail de seuillage de l écho radar a récemment été établi par Météo France pour traiter le signal grêle. Deux niveaux d écho ont ainsi été établis pour définir des zones de grêle: - 0: absence de grêle - 1: grêle possible - 2: grêle probable La donnée biologique relevée sur le terrain Pour établir une correspondance entre le signal radar et les dégâts engendrés par la grêle sur le vignoble, nous avons du élaborer une échelle de notation des dégâts sur la végétation. Echelle de notation des impacts de grêle La notation des dégâts de grêle sur la végétation porte principalement sur l observation des stigmates sur les rameaux de la vigne. Les six niveaux suivants sont retenus: - 0: absence d impacts - 1: quelques impacts - 2: impacts réguliers - 3: nombreux impacts - 4: très nombreux impacts, nombreux rameaux endommagés - 5: très nombreux rameaux endommagés Munis de cette échelle, les techniciens de l institut Français de la Vigne et du Vin (IFV) ont sillonné le vignoble pour réaliser les observations. Dispositifs d observation Relevés effectués par les techniciens de l IFV. Quatre zones d étude ont été définies ; trois portent sur le vignoble de Bordeaux, sur les secteurs d appellation du Médoc, des Graves et 1eres cotes de Bordeaux, du Saint-Emilionnais et du sud-ouest de l entre deux mers. La quatrième zone porte sur la partie Sud du vignoble de Cognac. Les observations sont réalisées sur quatre journées, les 27, 28, et 29 mai sur le vignoble bordelais et le 15 juin sur celui de Cognac. Les tracés de relevés sont globalement prédéfinis pour cerner les secteurs établis par le radar les plus intéressants. Les relevés sont effectués à intervalles réguliers, de l ordre de quelques centaines de mètres, voire du kilomètre, sur les parcelles rencontrées au gré des déplacements. Les relevés sont effectués à l aide d un mobile mapper qui permet de géo-référencer l observation à l aide d un GPS différentiel avec une précision de l ordre du mètre. 232 points d information ont ainsi été relevés en quatre jours de travail, avec respectivement 157 et 57 points relevés sur les vignobles de Bordeaux et de Cognac. Relevés effectués par les observateurs sur l interface WEB Parallèlement, l IFV a développé une interface de saisie sur le web pour des observateurs distants. Dès sa finalisation, l adresse du site (http/ a été largement diffusée à tous nos partenaires techniciens ou viticulteurs pour tenter de multiplier les observations relevées sur le terrain. L interface de saisie utilise la technologie GoogleMAP pour géo-référencer l observation: une fois connecté sur le site, l observateur introduit le nom de sa commune pour cibler la fenêtre à proximité de son parcellaire. L image satellite lui permet de zoomer sur son parcellaire et de faire glisser le pointeur sur sa parcelle observée. Les coordonnées du point sont automatiquement délivrées. L observateur renseigne ensuite la date du phénomène observé et le niveau de dégâts établi selon l échelle précédemment défini en cliquant sur la

126 106 note retenue comme illustré dans la figure 2. L identification de l observateur et son adresse mail sont requises pour pouvoir, le cas échéant, vérifier la véracité des relevés effectués. L envoi des observations se fait par simple clic. Il renseigne directement par mail l opérateur IFV de l arrivée d une nouvelle observation et stocke l information dans la donnée base de données EPICURE constituée par l IFV. Cette base de données est interfacée avec un SIG qui renvoie directement à l observateur la carte de son observation également par l ensemble des observations déjà effectuées. Ce système permet à la fois de restituer à l observateur le visu de son travail, resitué dans le contexte des observations environnantes, tout en garantissant la confidentialité de l observateur assurée par l IFV. Figure 2: illustration de l interface de saisie développée sur le Web et du formalisme de retour d information offerte à l observateur 126 points d informations ont ainsi été saisis par les observateurs sur le site Web entre les mois de juin et juillet Méthode d analyse des données Analyse statistique globale L analyse des correspondances entre seuils de détection par le radar et impacts de grêle au vignoble est basée sur la répartition des proportions d individus dans des tableaux de contingence: pour un seuil de détection fixé, la parcelle observée, tout comme le pixel radar concerné, est déclaré positif ou négatif. Les deux seuils de détections peuvent ainsi indépendamment varier sur les deux échelles propres de l écho radar analysé ou des impacts de grêle observés (Tableau 1). Tableau 1. Formalisme du tableau de contingence Radar Absence de Présence de observation grêle * grêle * Absence d impacts * A B Présence d impacts * C D (*) au niveau de seuil fixé. L ensemble des couples pertinents possibles est ainsi analysé. Pour chacun d eux, la proportion de vrai positifs, vrai négatifs, faux positifs et faux négatifs détermine la performance du couple retenu, à partir des ratios suivants.

127 107 Non détection = C / (C+ D) Performance alerte = D / (B+C+D) Fausse détection = B / (B+D) Coïncidence = (A+D) / (A+B+C+D) La fausse détection et la coïncidence sont les deux indicateurs les plus pertinents retenus et présentés pour la suite de l étude. Ces tableaux de contingence servent usuellement de base aux analyses par la méthode ROC (Receiver Operating Characteristics) qui pourrait ici avantageusement compléter l analyse du signal radar donnant les meilleures performances. Le caractère discret de la donnée fournie par Météo France rend cette analyse inopportune dans notre cas d étude. Analyse géostatistique descriptive Le SIG traite l ensemble des données sur le plan de leur distribution spatiale et permet d affiner et de compléter l analyse statistique globale par des considérations d ordre qualitatives. Résultats et discussion Le tableau 2 suivant résume la performance des correspondances pour chaque couple (observation, radar) testé. La nomenclature retenue est exprimée par O pour observation et R pour Radar, accompagné du niveau de seuil retenu dans les deux échelles pour déclarer l individu positif ou négatif. Tableau 2: Résumé de l analyse de la performance des couples (écho radar/dégâts observés) les plus pertinents. jeu de Taux données O1/R1 O2/R1 O1/R2 O2/R2 analysé fausse alerte complet coincidence complet fausse alerte IFV coincidence IFV fausse alerte Web coincidence Web L ensemble de couples possibles a été testé. Seuls sont présentés dans le tableau 2 les couples les plus pertinents. Les chiffres en gras figurent les niveaux les plus performants, soit les plus faibles pour la fausse détection et les plus forts pour la coïncidence. Analyse sur l ensemble du jeu de données Considérons tout d abord l ensemble du jeu de données complet. Au seuil de grêle dit «possible» (R1) La performance optimale de détection par le radar correspond très nettement au niveau 1 d observation sur le terrain, soit «quelques impacts». Il est globalement, sur l ensemble du jeu de données analysé, proche des 75% (73%), et le taux de fausses alertes, de l ordre de 22%, reste inférieur à 25% soit moins d une chance sur quatre.

128 108 Au seuil de grêle dit «probable» (R2) Le moindre niveau de fausses alertes est là aussi nettement favorable au niveau correspondant aux quelques impacts observés sur les parcelles. Il est globalement de 15% sur l ensemble du jeu de données. La coïncidence optimale est obtenue pour le niveau d impacts réguliers observés sur le vignoble. La performance (63%) baisse légèrement par rapport au seuil précédent, mais reste d un niveau très correct. Analyse de l origine des données selon le dispositif d observation L ensemble du jeu de données d observations relevées sur le terrain est divisé entre celles faites par les techniciens de l IFV, et celles issues de l interface Web. La première source est réputée à priori fiable, parce que réalisée par un petit nombre d observateurs et que leur origine en est ainsi maitrisée; Au contraire, la seconde est saisie par un grand nombre d observateurs différents; nombre d entre eux ne sont d ailleurs pas directement connus par nos équipes, le bouche à oreille ayant fait son office lors de l opération de diffusion de la plate forme Web. L analyse du tableau 2 montre clairement que les niveaux de performance entre les deux sources de données sont du même ordre de grandeur et tout à fait comparables entre eux, ce qui tend à prouver l intérêt que peut revêtir cette méthode d observation participative en réseau. Analyse géostatistique descriptive La figure 3 regroupe les illustrations des passages de grêle et des impacts observés à l issue des quatre épisodes de grêle des 9, 11, 13 et 25 mai La carte de gauche montre l ensemble des observations issues du radar au seuil retenu comme «probable», et relevées sur le terrain pour l ensemble du vignoble. Les histogrammes verts correspondent aux relevés effectués par les techniciens de l IFV, ceux en bleu ceux recueillis par les autres observateurs sur le Web. Figure 3: Illustration des zones de grêle «probable» (rouge) et «possible» (orange), du positionnement et de l intensité des dégâts observés par l IFV (histogrammes verts) ou recueillies sur la plate forme Web (histogrammes bleus) Ce graphique appelle deux commentaires; d une part il montre très nettement le contrôle réalisé perpendiculairement aux trois couloirs de grêle orientés du Sud-Ouest vers le Nord-Est. Dans le sud du Cognaçais. Cette observation transversale permet de bien délimiter la largeur de ces couloirs et d en vérifier la pertinence qui, conformément aux analyses issues des tableaux de contingence, est globalement bonne.

129 109 Cette carte matérialise par ailleurs, en ce qui concerne les observations recueillies sur le Web (histogrammes bleus), qu elles concernent quasiment toutes les zones de grêle «probables»: très peu d observateurs, non concernés par cette suite d accidents climatiques, se sont manifestés pour indiquer l absence de dégâts sur leurs parcelles. Ce type d information «négative» revêt pourtant autant de signification qu une alerte «positive» dans le dispositif de validation, surtout lorsqu elles sont recueillies à proximité des zones atteintes. Cette observation montre la nécessité d insister sur ce point lors de la diffusion d un tel message. A défaut, l observation recueillie ne pourra être considérée comme représentative de l ensemble du vignoble, et elle aura nettement tendance à surévaluer l importance du phénomène considéré à cette échelle. Les deux cartes de droite illustrent un zoom effectué sur un secteur particulier du vignoble. Elles montrent, pour des relevés entourés d un cercle jaune, l absence de relation nette entre la matérialisation de l écho radar et les dégâts relevés sur le terrain: d une part des impacts peu nombreux sont parfois relevés sur certaines parcelles nettement situées en zone rouge; d autre part, nous avons pu observer des dégâts forts sur des zones «oranges» supposées moins grêlées, voire sur des secteurs «blancs» matérialisant des secteurs qui devraient être indemnes de grêle. Cette observation nous laisse penser que, si le radar est performant pour localiser les zones de grêle et ainsi délivrer une information qualitative, il semble par contre inopérant pour quantifier l ampleur des phénomènes, ou tout au moins des dégâts qu ils entrainent. Enfin nous avons pu observer, sur le secteur du Blayais fortement atteint, qui fut l objet d un nombre important de relevés, la puissance de l outil participatif développé sur le Web qui permet, grâce à la participation assidue des acteurs locaux, de délivrer un niveau de validation spatiale très fin, dont la résolution peut être bien supérieure à la maille minimale de 700 m proposée par l analyse de l écho radar. Conclusion et perspectives Les orages de grêle très marqués de ce printemps 2009, nous ont permis de tester sur le terrain, les outils de validation spatiale dont la nécessité s était faite sentir dès la première année d exploitation des données météo à maille fine en La forte pression mildiou, relativement uniforme évaluée depuis ce millésime ne nous avait guère laissé cette opportunité. Ces premiers tests sur la grêle révèlent la pertinence de ce travail d enquête dans la mesure où il peut porter sur des phénomènes de type accidentel, ou tout au moins fortement localisés. La plate forme de saisie Web, incitant les techniciens partenaires et professionnels, à partager leurs observations nous donne également satisfaction. Elle a fait l objet d un accueil favorable auprès de la profession: bien qu elle fut élaborée et diffusée près de 2 à 3 semaines après les premiers épisodes de grêle, et qu elle concerna un évènement au caractère douloureux pour certains professionnels. Malgré ce, le nombre assez conséquent des données recueillies, et leur qualité comparable à celle relevées par nos équipes a permis de convenablement conduire cette étude de zonage. Du point de vue du zonage de la grêle, on retient que l information radar permet de définir le caractère grêlés ou indemne d une parcelle avec un taux réussite proche de 75%. Le radar ne permet cependant pas de cerner l approche quantitative des dommages subis par la vigne. Nous envisageons ainsi de poursuivre le développement de cet outil d expérimentation participative en appliquant ce type d interface de saisie au domaine de l épidémiologie: cette

130 110 plate forme nous semble être un outil prometteur pour l animation de nos réseaux, particulièrement adapté à la déclaration des premiers foyers, de mildiou par exemple. La mobilisation d un nombre important de relevés pourrait ainsi permettre de rapidement bien localiser les foyers primaires et étudier en grandeur nature au vignoble, les conditions météorologiques propices à leur installation. Remerciements Nous tenons à remercier les équipes IFV ainsi que toutes les structures et partenaires associés, viticulteurs ou techniciens qui participent à ce projet et ont rendu cette étude possible. Bibliographie Bois, B. 2006: Spatialisation des précipitations, Investigations sur l utilisation des lames d eau HYDRAM issues de l imagerie radar comme outil de spatialisation des champs de pluie en Gironde. Etude préliminaire. Météo France, ENITA Bordeaux, INRA, 30 pp. COLLECTIF MÉTÉO FRANCE. 2005: La mesure RADAR: principe et description. Documentation technique. Météo France, 10 pp. Griaud, K. 2007: Les données pluviométriques RADAR Antilope: évaluation et impact sur l exploitation du modèle Potentiel Système mildiou dans le vignoble bordelais, mémoire de master viticulture-œnologie de l université Victor Ségalen Bordeaux 2: 39. Fernandez, N. 2008: L'utilisation des relevés climatiques Radar ouvre la voie à la modélisation du mildiou à l'échelle de l'exploitation agricole. Mémoire de fins d études ENSAT: 80.

131 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Long-term survival of Plasmopara viticola oospores T. Caffi, V. Rossi, M. Lusitani Istituto di Entomologia e Patologia Vegetale, Università Cattolica del Sacro Cuore, Piacenza, Italy. tito.caffi@unicatt.it Abstract: A research was carried out in order to verify the long-term viability of Plasmopara viticola oospores. It is commonly assumed that oospores may survive across grapevines growing seasons, but no experimental evidence of this assumption has been provided to date. Grapevine leaves were collected from untreated plants in Emilia-Romagna in fall 2004 to Fragments of these leaves were kept in soil under natural conditions. In spring 2005 to 2009, part of the leaf fragments collected in the previous autumn were used to evaluate oospore germination using the floating disk bioassay; in spring 2009, the bioassay was also performed using the leaf residue collected in fall 2005 to Results demonstrated that P. viticola oospores remained viable for the whole period considered in this work, i.e., about 65 months (from fall 2004 to spring 2009). Key words: grapevine downy mildew, oospores germination, floating disk assessment Introduction The heterothallic oomycete Plasmopara viticola is the causal agent of the downy mildew, a major disease in all grape-growing areas throughout the world (Lafon & Clerjeau, 1988). It causes both direct and indirect yield losses which include rotting of inflorescences, clusters and shoots, reduction of photosynthetic activity of the affected leaves and premature defoliation of vines. P. viticola is dimorphic for its reproductive forms, with both sexual and asexual spores. Oospores represent the sexual stage of the pathogen; they are formed in infected tissues at the end of the season and constitute the unique source of inoculum for infections in the following spring (Branas, 1974). Oospores reach morphological maturity in the middle of November (Vercesi et al., 1999) but they are not yet able to germinate (Zachos, 1959) because of dormancy (Galet, 1977). Dormant spores fail to germinate even under favourable conditions because of internal factors which are broken by ageing (Oliver & Schweizer, 1999). When dormancy has broken, oospores start germination under favourable environmental conditions and complete it by producing the sporangia. Several studies were carried out over decades about oospores morphology (Vercesi et al., 1999) and physiology during their overwintering period (Burruano et al., 1987 and 1989; Rossi & Caffi, 2007). While the dynamics of oospores maturation and germination was essentially described, there is still a lack of knowledge about the long-term survival of oospores. It was assumed that oospores may survive across grapevines growing seasons (Zachos, 1959; Kennelly et al., 2006) but no experimental evidence of this assumption has been provided to date. The aim of the present work was to verify the long-term capability of oospores to germinate. 111

132 112 Material and methods Grapevine leaves showing typical mosaic symptoms were collected in untreated plots of commercial vineyards through Emilia-Romagna region during five consecutive falls, from 2004 to 2008; presence of oospores in the infected leaf tissue was confirmed by microscopic observations. Leaves were dried, crumbled, put into grid plastic boxes on a layer of sand, and kept in soil under natural conditions at the University campus of Piacenza (northern Italy). In the following spring (2005 to 2009), part of the leaf fragments were removed and used to evaluate oospore germination using the floating disk bioassay (Hill, 1998). Between late March and early July, samples of 10g of the leaf residue were collected from the box once per week, uniformly arranged on the base of a glass jar (12cm in diameter) covered with a layer of non woven tissue (mod. Agricova, Maniver Srl, Comignago, Novara, Italy), and flooded with 150ml of distilled water. Leaf disks (20mm in diameter) were excised from potted plants of Barbera grown in a greenhouse isolated from any external source of P. viticola inoculum; 10 disks were floated, abaxial side down, per glass jar. Jars were then incubated under natural light at 20±2 C for 24 hours. Afterwards, leaf disks were collected and placed in Petri plates covered with wet blotting-paper, with the abaxial surface exposed, and incubated again. This procedure was repeated for 10 days after the leaf litter fragments had been flooded with water. All the leaf disks were inspected daily for 14 days, to note the appearance of disease symptoms (i.e., sporangia elongating from leaf stomata). Thus, the evidence of oospores germination was obtained indirectly by the infections caused on green leaf disks. This method allows detecting the production of a low number of zoospores deriving from the oospores germination process (Pertot & Zulini, 2003). Boxes containing the leaf residue were kept in soil again until spring In spring 2009, the bioassay was used to determine oospore germination in all the samples collected during the 5-year period, from 2004 to 2008 (Table 1). Table 1. Experimental design: x shows the time when the germination ability of the Plasmopara viticola oospores was assessed using the floating disk bioassay. Oospore collection Oospore germination (spring) (fall) x x 2005 x x 2006 x x 2007 x x 2008 x Results and discussion The leaf residues collected in fall 2004 to 2008 (with exception for 2005) contained viable oospores in the following spring, because they released zoospores in water that caused infection in the green leaf disks during the bioassay. All these leaf residues (with exception for those collected in 2005) still contain viable oospores in 2009 (Table 2). In 2005 very few

133 113 oospores were formed in the mosaic affected leaves and no infection was observed in the leaf disks, neither in 2006 nor in Results from this work clearly demonstrated that P. viticola resting spores are capable of surviving for long time in soil; the longest period considered in this work, a period of about 65 months (from fall 2004 to spring 2009), did not exhaust all oospores as a fraction still germination and successfully infect leaf disks. Contribution of the long-term survival of P. viticola oospores to the development of downy mildew epidemics is difficult to estimate. Quantitative information on the survival rate of oospores will help to this task. Quantitative data on germination dynamics of the oospores were collected during the present work and will be published elsewhere. Table 2. Germination of Plasmopara viticola oospores assessed using the floating disk bioassay in the spring following their formation and in spring Oospore Oospore germination (spring) collection (fall) (next year) (2009) 2004 Yes Yes 2005 No No 2006 Yes Yes 2007 Yes Yes 2008 Yes Yes References Branas, J. 1974: Viticulture. Déhan ed., Montpellier, France. Burruano, S., Di Graziano, M., Faretra, F., Nalli, R., Pennisi, A. 1989: Indagini sulla germinazione delle oospore di Plasmopara viticola in aree geografiche differenti. Phytopathol. Med. 28: Burruano, S., Strazzeri, S., Laviola, C. 1987: Influenza dell'acqua sulla germinazione delle oospore di Plasmopara viticola. Phytopathol. Med. 26: Galet, P. 1977: Les Maladies et les Parasites de la Vigne. Anonymous, Tome I: Hill, G. K. 1998: Studies on the germination of Plasmopara viticola oospores with a floating disc test. IOBC/WPRS Bull. 21(2): 1-2. Lafon, R. & Clerjeau, M. 1988: Downy mildew. In: Compendium of Grape Diseases, ed. Pearson and Goheen, APS Press, St. Paul, Minnesota, USA: Kennelly, M. M., Gadoury, D. M., See, R. C., Wilcox, W., Magarey, P. 2006: Recent investigations of the biology of Plasmopara viticola: consideration for forecasting and management of grapevine downy mildew. Proceeding of the 5 th International Workshop on Grapevine Downy and Powdery Mildew, eds. Pertot, Gessler, Gadoury, Gubler, Kassemeyer and Magarey: Oliver, R. & Schweizer, M. 1999: Molecular fungal biology. Cambridge, UK: Cambridge University Press. Pertot, I. & Zulini L. 2003: Studies on Plasmopara viticola oospore germination in Trentino, Italy. IOBC/WPRS Bull. 26(8):

134 114 Rossi, V. & Caffi, T. 2007: Effect of water on germination of Plasmopara viticola oospores. Plant Path. 56: Vercesi, A., Tornaghi, R., Sant, S., Burruano, S., Faoro, F. 1999: A cytological and ultrastructural study on the maturation and germination of oospores of Plasmopara viticola from overwintering vine leaves. Mycol. Res. 103: Zachos, D. G. 1959: Recherches sur la biologie et l épidémiologie du mildiou de la vigne en Grèce. Bases de prévision et d avertissements. Annals Institute de phytopathologie, Benaki 2:

135 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Dispersal of the sexual stage of Erysiphe necator in northern Italy V. Rossi 1, T. Caffi 1, S. E. Legler 1, R. Bugiani 2, P. Frisullo 3 1 Istituto di Entomologia e Patologia Vegetale, Università Cattolica del Sacro Cuore, Piacenza, Italy; 2 Plant Protection Service, Emilia-Romagna Region, Bologna, Italy; 3 PRIME Department, Università di Foggia, Foggia Abstract: Either by themselves or in combination with mycelium in the dormant buds, ascospores produced in chasmothecia (the sexual fruiting bodies) of Erysiphe necator are an important source of primary inoculum for grapevine powdery mildew disease. In northern Italy, E. necator overwinters mainly as the sexual stage (i.e., mycelia are not usually present in dormant buds), but no data are available on the abundance of chasmothecia in the vineyards. Therefore, the dispersal of chasmothecia was studied in commercial vineyards of northern Italy, from 2005 to 2007 (16 combinations of 10 vineyards X 3 years); the vines were not sprayed with fungicide during this study. The dispersed chasmothecia were collected on filter papers in funnels that were placed on the trunk of affected grapevines from mid-august to the end of leaf fall. Each filter paper was examined microscopically, and the chasmothecia were counted. The first chasmothecia of the season were dispersed between late August and early October; cumulative numbers of the chasmothecia dispersed subsequently increased (with different dynamics depending on the vineyard and year) but dispersal stopped at the end of leaf fall. Over all three years and 16 vineyards, chasmothecia averaged 3/cm 2 of trap surface, with a maximum of 15/cm 2. The numbers of mature chasmothecia that dispersed roughly depended on the powdery mildew severity on leaves: high numbers of chasmothecia were associated with disease severity 80%. According to estimates based on the current data, vine bark in vineyards trained with a Guyot, Geneva double curtain, or spurred cordon pruning system could contain till 18, 19, or 44 million chasmothecia/ha, respectively. Key words: grape powdery mildew, dynamics of the dispersal, chasmothecia Introduction Erysiphe necator (syn. Uncinula necator) (Schw.) Burr., the causal agent of grape powdery mildew, is a heterotallic fungus, i.e., the fruiting bodies (chasmothecia, formerly cleistothecia) form when two opposite mating types encounter and their antheridium and ascogonium combine. This process is time dependent: the longer the time period, the higher the probability that the two mating types will be in close proximity so that mating can occur (Schnathorst, 1965; Gadoury and Pearson, 1988). Environmental factors also determine the rate of increase of both mating types and, therefore, directly control the probability that they will be sufficiently close to mate. Once initiated, the chasmothecia reach maturity in three stages: initially they are white, then yellow, then brown, and finally they become black at maturity. These events are affected by temperature and host resistance (Gee et al., 2000). During maturation, chasmothecia are fixed to the powdery mildew colony by anchorage hyphae; when chasmothecia are morphologically mature, characteristic appendages appear, the anchorage breaks down, and the fruiting bodies are dispersed by rain splashes (Yossifovitch, 1957; Gadoury and Pearson, 1988; Cortesi et al. 1995). The rains deposits the chasmothecia onto the bark of the vine trunk, onto the soil, or onto leaf litter on the soil surface; however, survival of these fruiting 115

136 116 bodies is consistently higher in the cracks of the vine trunk than on the other substrates (Gadoury and Pearson, 1988). Some chasmothecia can release ascospores in autumn (Gee et al., 2000; Caffi et al., 2008) but most overwinter and repeatedly release ascospores in the following spring. Despite an initial controversy about the importance of the ascospores produced in chasmothecia as a consistent source of primary inoculum for powdery mildew disease on grapevine, ascospores are now considered an important form of primary inoculum, either alone or in combination with the mycelium in the dormant buds (Pearson and Gadoury, 1987; Gadoury and Pearson, 1988; Cortesi et al., 1997; Jailloux et al., 1998; Füzi, 1999). Control of powdery mildew is traditionally based on the management of secondary infections. For this purpose, some mathematical models have been developed for simulating epidemics caused by asexual (secondary) infections (Sall, 1980; Chellemi and Marois, 1991; Kast, 1995; Gubler et al., 1999) and/or for evaluating the risk of powdery mildew infection and thereby for the scheduling of fungicide applications. However, recent findings on the role of ascosporic infections in triggering powdery mildew epidemics indicate the need for a new, early-season strategy for controlling the disease. The aim of this work was to acquire data about the abundance of chasmothecia in North Italian vineyards where flag shoots are usually absent in the spring (indicating the absence of mycelium in dormant buds) and where E. necator overwinters only as the sexual stage. Collection of these data is the first step for developing a decision support system for controlling powdery mildew epidemics by reducing the incidence of primary infections caused by ascospores. Materials and methods Vineyards Ten commercial vineyards were sampled in Northern Italy in 2005 to 2007 (16 combinations of 10 vineyards over a period of 3 years); some vineyards were on the plain while others were located on the Apennine hilly area. Vineyards were representative of the vine-growing area in terms of soil type, vine variety, trellis system, and cropping regime (Table 1). As is typical for this region, flag shoots were not observed in any of 16 vineyards in any of three years. In each vineyard, a plot of at least 10 rows of vines, 100m long, was not sprayed against powdery mildew for the entire season. From mid-august to complete leaf fall, severity of the powdery mildew epidemics (percentage of leaf area affected) was periodically assessed on a random sample of 100 leaves. Trapping of chasmothecia The chasmothecia that dispersed from bunches and leaves to the vine bark were monitored using plastic trap funnels (13cm diameter) as described by Gadoury and Pearson (1988). In mid-august of each year, five funnels were attached to each of three randomly selected vines per vineyard (15 funnels per vineyard) as follows: three funnels were secured to the trunk (at 1/3, 2/3 and 3/3 of the trunk height), and two funnels were secured to the permanent cordon or, for the Guyot pruned vineyards, to the fruiting arm (at 1/3 and 2/3 of the cordon or arm length). The funnels were oriented so that the wide part was facing up. A disk of filter paper (15cm diameter) was fixed in the mouth of each funnel by pushpins so that any material falling into the funnel top remained on the filter. In mid-august, E. necator colonies were present on leaves and/or on bunches but no chasmothecia had formed; microscopic observations on sample leaves and berries collected at the time of funnel installation

137 117 confirmed the absence of visible fruiting bodies. Filters were replaced every 2 weeks until leaf fall was complete. Each filter was examined using a stereomicroscope (20 x magnification), and the number of chasmothecia was counted. The number of chasmothecia trapped per cm 2 of trap surface was then calculated by dividing total counts by the total area of the funnel top opening. Numbers of chasmothecia trapped over time were also expressed as a cumulative proportion of the total chasmothecia trapped in each vineyard. Data analysis An analysis of variance was performed to determine whether the vineyard, the position of the funnel on the vine, and their interaction influenced the number of chasmothecia trapped; before analysis, the data were transformed (ln(x+1)) to make variances homogeneous. A hierarchical cluster analysis was carried out to classify vineyards based on powdery mildew severity and the total number of chasmothecia trapped during the trapping period; this analysis used the Centroid method for grouping vineyards and the Euclidean quadratic distance method for measuring similarity. All the analyses were performed with SPSS statistical package (ver.13.0, SPSS Inc., Chicago, IL, USA). Table 1. Characteristics of the vineyards. Year Vineyard Location Altitude (m a.s.l.) Vine variety Trellis system CàBosco 2 Pinot nero GDC a 2 Canneto Pavese 230 Barbera Guyot (one arm) 3 Castel'Arquato 192 Cabernet sauvignon Guyot (one arm) 4 Correggio 31 Sangiovese Pergola romagnola 5 Tebano 177 Pinot nero Spurred cordon CàBosco 2 Pinot nero GDC 7 Castel San Pietro 60 Sangiovese Spurred cordon 8 Scipione 250 Sangiovese Guyot (one arm) 9 Tebano 177 Pinot nero Spurred cordon CàBosco 2 Pinot nero GDC 11 Castel San Pietro 60 Sangiovese Spurred cordon 12 Marzeno 370 Pinot nero Guyot (two arms) 13 Modigliana 373 Chardonnay Spurred cordon 14 Modigliana 373 Sangiovese GDC 15 Scipione 250 Sangiovese Guyot (one arm) 16 Tebano 177 Pinot nero Spurred cordon a Geneva double curtain Results and discussion The first chasmothecia of the season were trapped between late August (in two vineyards in 2007) and early October (in two vineyards in 2005); no fruiting bodies were trapped at Cà

138 118 Bosco (vineyard 10) or Marzena (vineyard 12) in 2007 when powdery mildew was absent or present only as sporadic small colonies. In Cà Bosco in 2005 and 2006 the presenced of powdery mildew corresponded to the trapping of chasmothecia (Table 2). After the first trapping of the season, the proportion of chasmothecia trapped progressively increased and stopped with complete leaf fall (Figure 1), which occurred between late October and late November depending on the year (Table 2). In 2007, the trapping of chasmothecia progressed slowly until the end of September (Figure 1). The number of chasmothecia trapped significantly differed (P<0.001) among the vineyards: from none (vineyard 10 and 12), and very few chasmothecia (0.08/cm 2 of trapping surface) in 15 to a high of 15.4 chasmothecia trapped per cm 2 in vineyard 16 (Table 2). Table 2. Dates of first and last trapping of E. necator chasmothecia, and total number of chasmothecia trapped. Year Vineyard Disease severity a Date of first trapping Date of last trapping Chasmothecia (n/cm 2 ) b September 07 November October 07 November September 07 November October 07 November September 07 November September 31 October September 31 October September 31 October September 31 October No disease September 20 November Sporadic August 20 November September 20 November August 06 November September 20 November a Leaf area affected by powdery mildew (%) b Total number of chasmothecia trapped per cm 2 of trap surface The cluster analysis distinguished three clusters of vineyards (Figure 2). Cluster I consisted of the three vineyards with no or very low disease severity and no or few chasmothecia. In cluster II, the vineyards had 60-70% powdery mildew severity and < 0.5 chasmothecia/cm 2. In cluster III, the vineyards had 80% disease severity and, with the exception of vineyards 1 and 8, > 0.5 chasmothecia/cm 2.

139 119 Position of the trapping funnels on the vine (Figure 3) did not significantly influence the number of chasmothecia trapped (P=0.23 for the main factor and P=0.96 for the interaction position x vineyard). Proportion of seasonal chasmothecia Aug Sep Oct Nov Aug Se Oct Nov Figure 1: Seasonal dynamics of the chasmothecia trapped in 2005 to 2007; data are expressed as the cumulative proportion of the total cleistothecia trapped over time. Each value is the mean (± SE) of the vineyards shown in Table 1. Chasmothecia / cm I Powdery mildew severity (%) Figure 2. Relationship between powdery mildew severity and number of the chasmothecia trapped per cm 2 of trap surface in the 16 vineyards; ellipses delineate three clusters (I to III) while white boxes show the centroid of each cluster. II III 15.4

140 120 Position on V IV III II I Chasmothecia / cm 2 Figure 3: Relationship between the position of the trapping funnels on the vine (I to V) and number of the chasmothecia trapped per cm 2 of trap surface; I, II, and III = 1/3, 2/3, and 3/3 of the trunk height, respectively; IV and V = 1/3 and 2/3 of the cordon or fruiting arm length, respectively. Values are the means (+ SE) of 16 vineyards. Conclusion In conclusion, the present work provides data about the abundance of chasmothecia in some vineyards of northern Italy where E. necator overwinters only as the sexual stage only (i.e., flag shoots indicating the presence of overwintering mycelium were absent). Within each growing season, chasmothecia were first dispersed between late August and early October; the cumulative numbers of dispersed chasmothecia subsequently increased at different rates (depending on the year and vineyard), and dispersal ended at leaf fall. Air temperature and rain pattern may be the main factors affecting the dynamics of chasmothecia dispersal; temperature influences the formation and maturation of the fruiting bodies on grape leaves (Gadoury and Pearson, 1988; Rossi et al., 2009) while rain splashes physically transport the mature chasmothecia from leaves to bark (Cortesi et al., 1995). The numbers of mature chasmothecia that dispersed roughly depends on the severity of the powdery mildew epidemics on leaves: high numbers of dispersed chasmothecia are associated with disease severity 80%. However, absence of rainfall or the prevalence of only one powdery mildew mating type could account for the trapping of few chasmothecia in some severely affected vineyards (Cortesi et al., 1995). Based on the data collected in this work, very high numbers of chasmothecia can be produced in severely affected vineyards. A vineyard trained with a Guyot pruning system has 3,000 to 4,000 plants/ha, each plant having approximately 850 to 1,500cm 2 of wood bark; considering the average number of chasmothecia found in this work, i.e., 2.97/cm 2, a primary inoculum of 7.5 to 18 million of chasmothecia per hectare of vineyard can be calculated. For a GDC trained vineyard (about 1,500 plants/ha with 2,900 to 4,300cm 2 of bark/plant), the estimated inoculum increases to million of chasmothecia/ha, while for a spurred cordon (3,000 to 4,000 plants/ha with 2,500 to 3,700cm 2 of bark/plant), the estimate is million chasmothecia per hectare. In one hectare of vineyard 16, which contained 3,400 plants with an average of 3,100cm 2 of bark/plant, more than 162 million chasmothecia were potentially present on the plants after leaf fall of 2007.

141 121 Because ascospores produced within chasmothecia are the sole form of inoculum for spring infection by E. necator in the considered vineyards as well as in many other grapegrowing areas of the world (Pearson and Gadoury, 1987; Cortesi et al., 1994), disease management actions aimed at reducing the primary inoculum should have a key role in controlling powdery mildew. These actions include an efficient control of the disease after grape harvest to prevent chasmothecia from forming and treatments that destroy chasmothecia in those vineyards where severe epidemics have developed from late summer to leaf fall. The formation and maturation of chasmothecia can be prevented by fungicides or by biological control agents like the hyperparasitic fungus Ampelomyces (Falk et al., 1995a,b), while eradication requires the use of fungicides such as lime sulphur (Gadoury et al., 1994). Studies are necessary to define the activity of both fungicides and biological control agents to prevent chasmothecia from forming and to eradiate chasmothecia once they have formed. References Analytis, S., 1977: Über die Relation zwischen biologischer Entwicklung und Temperatur bei phytopathogenen Pilzen. Phytopath. Zeit. 90: Caffi, T., Cavagna, S., and Rossi, V. 2008: Ascospore maturation and discharge in Erysiphe necator. J. Plant Pathol. 90 (supplement 2): Cortesi, P., Gadoury, D. M., Seem, R. C., and Pearson, R. C Distribution and retention of chasmothecia of Uncinula necator on the bark of grapevines. Plant Dis. 79: Cortesi, P., Bisiach, M., Ricciolini, M., and Gadoury, D. M. 1997: Chasmothecia of Uncinula necator An additional source of inoculum in Italian vineyards. Plant Dis. 81: Falk, S. P., Gadoury, D. M., Cortesi, P:, Pearson, R. C., and Seem, R. C. 1995a: Parasitism of Uncinula necator cleistothecia by the mycoparasite Ampelomyces quisqualis. Phytopathology 85: Falk, S. P., Gadoury, D. M., Pearson, R. C., and Seem, R. C. 1995b: Partial control of grape powdery mildew by the mycoparasite Ampelomyces quisqualis. Plant Dis. 79: Füzi, I. 1999: The epidemic role of chasmothecia of grapevine powdery mildew at Szekszard vine-growing region. Növényvédelem 35: Gadoury, D. M., and Pearson, R. C. 1988: Initiation, development, dispersal, and survival of chasmothecia of Uncinula necator in New York vineyards. Phytopathology 78: Gadoury, D. M., Pearson, R. C., Riegel, D. G., Seem, R. C., and Becker, C. M. 1994: Reduction of powdery mildew and other diseases by over-the-trellis applications of lime sulfur to dormant grapevines. Plant Dis. 78: Gee, L. M., Stummer, B. E., Gadoury, D. M., Biggins, L. T., and Scott, E. S. 2000: Maturation of cleistothecia of Uncinula necator (powdery mildew) and release of ascospores in southern Australia. Australian Journal of Grape and Wine Research 6: Jailloux, F., Thind, T., and Clerjeau, M. 1998: Release, germination, and pathogenicity of ascospores of Uncinula necator under controlled conditions. Can. J. Bot. 76: Pearson, R. C., and Gadoury, D. M. 1987: Chasmothecia, the source of primary inoculum for grape powdery mildew in New York. Phytopathology 77: Schnathorst, W. C. 1965: Environmental relationships in the powdery mildews. Ann. Rev. Phytopathol. 3: Yossifovitch, M. 1923: Contribution à l étude de l oidium de la vigne et son traitement. Thèse Doc. Univ. Toulouse, 172 pp.

142 122

143 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Modelling the effect of the grapevine growth and susceptibility on the dynamics of a powdery mildew epidemic A. Calonnec 1, S. Schnee 1, P. Cartolaro 1, M. Langlais 2 1 INRA-Bordeaux, UMR INRA-ENITA 1065 Santé Végétale BP81, Villenave d Ornon, France. 2 UMR CNRS 5251, IMB & INRIA Futurs, Anubis, case 26, Université Victor Segalen Bordeaux2, Bordeaux, France. calonnec@bordeaux.inra.fr Abstract: Simulations are performed to explore the sensitivity of epidemics of powdery mildew of grapevine to variation of parameters related to the pathogen, the plant growth or the crop management. For early inoculation, the three parameters linked to the sporulation (δ), the dispersion process (cid) and the plant vigour (vig) are the most discriminating. The parameter of ontogenic resistance (τ) is less discriminating, and the height of shoot topping and the distance between buds are not discriminating. We also examined the relationship between vine vigour and disease variables at key periods in the epidemic process. An increase of the parameter of vigour from 0.2 to 1 amounted to a higher number of leaves at flowering (Nflo) and a higher rate of leaves emergence (RLE). The RLE was correlated with the rate of diseased leaves emergence and with the severity of the disease at shoot topping, whereas Nflo was correlated to the diseased leaves area at day 240. The percentage of young leaves during first sporulation event (s1s2) was correlated to the rate of shoot development (RDS). These two variables from host allowed to discriminate the years. The duration and dynamic of infectious tissue can considerably vary function on the development of secondary shoots therefore on the climatic conditions and vigour. [1 line free] Key words: model host-pathogen, sensitivity analyses, plant growth and disease epidemics11 pt] [2 lines free] Introduction The grape-powdery mildew pathosystem is characterised by a polycyclic pathogen capable of explosive multiplication, a host population with a high degree of spatial structure at the field level and with a complex architecture at the individual plant level exhibiting rapid changes over time. As well as environmental differences, the high degree of human interference during vine development and the wide diversity of cropping systems enhance variability from one crop to another. Furthermore, because of the tight relationship between powdery mildew and its host (Doster & Schnathorst, 1985, Gadoury et al., 2003) and of the spatial location of primary infections on the vine stock, we hypothesized that the dynamical changes in crop structure should be considered as key factors for explaining variability in the severity of epidemic behaviour. Indeed, by modifying the movement of inoculum, or by altering the susceptibility of the leaf population, natural and management-induced changes in crop growth and crop architecture may significantly affect the course of the epidemics (Calonnec et al., 2009). For a better understanding of these host/pathogen dynamical interactions and of the capacity of host development to modify disease progress, we developed an epidemiological simulation model coupling vine growth with the dispersal and disease dynamics of the airborne plant pathogen Erysiphe necator (Calonnec et al., 2008). The simulation model is a complex discrete deterministic model which incorporates explicitly the dynamics of host 123

144 124 growth (distance between organs and their susceptibility) and the development and dispersion of the pathogen. The development of the spatial arrangement of host organs within the vine stock is captured within a 3D architectural model. It allowed simulating the spatio-temporal dynamics of host growth and epidemic development beginning from a range of climatic conditions, production systems and initial conditions for the density and location of the pathogen. Particularly, the model takes into account shoot topping which has for effect to enhance the development of secondary shoots then the emergence of new susceptible leaves during the epidemic process. Input variables are environmental (temperature, wind speed and direction) or related to the pathogen (location and onset of primary infection). Input parameters characterise the crop system (number of buds, distance between buds, shoot topping, vigour), and conditions of growth for the vine and the pathogen. Output describes, at each time step, number, age and pattern of the healthy and infected organs, infected and infectious leaf area and aerial density of spores released. In the following, we 1) explore the sensitivity of the epidemic to variation of parameters of pathogen, plant growth or crop management and 2) more precisely we examine the relationship between host and disease variables at key periods in the epidemic process for different conditions of vine vigour. Material and methods Simulations to explore the sensitivity of the model To preliminary explore the sensitivity of the model, simulations were performed by examining the effect on epidemic development of variations of parameters linked either to the pathogen (sporulation), to the dispersion, to the plant development (vigour, leaf susceptibility) or to plant management (height of shoot topping, distance between buds). A total of 972 simulations were performed combining the variation of 6 parameters at two to three levels for inoculations at two phenological stages (Table 1). Table 1. Parameters and their levels of variation in the simulations of sensitivity analysis. Parameter function Name a Level Effect Decay of spores dispersed with the distance cid higher level = lower dispersion Amount of spores produced δ higher level = higher sporulation Decay of infection with leaves age τ higher level = lower infection Vigour vig higher level = higher secondary leaves Distance between buds (cm) d-buds Height of shoot topping (cm) st-height Phenological stage at inoculation 1 leaf 4 leaves a cid is introduced in the equation of the quantity of spores Q c cached by a leaf of surface S, function upon its d distance to the source: Qc QrS e cid..δ is a scaling parameter in the equation describing the quantity of spores Q δ. s produced by a colony of size S during a given period: Q = βe. τ is the decay rate of leaf susceptibility in the equation of spores capable of infecting a leaf of age, A, where I 0 is the maximum infection rate at optimum τ * A temperature, and F(T n ) is the Gamma function with change in temperature T. I = I0 F( T n ) e. For more details about the role of each parameters cf. Calonnec et al., 2008.

145 125 Simulations to explore the effects of crop growth on the disease In order to identify favourable or unfavourable effects of crop growth, on the dynamics of the pathogen, we simulate epidemics using different environmental data and vine growth parameters that reflect: 3 contrasting seasons: 2003 characterized by an early bud break (day 104) and an early flowering (day 152), 1998 a late bud break (day 114), late flowering (day 159), and 2004, later bud break (day 118) and later flowering (day 163) with an increased development rate (Figure 1). For simulations, the day of bud break and the day of flowering are achieved when the accumulated sum of the mean daily temperature above 10 C reaches 90 and 380 respectively starting from day 1 (1 rst of January). Shoot topping was simulated 10 days after flowering. 7 levels of vine vigour: these levels result in an increased number and development of secondary shoots (Figure 1), especially after shoot topping. Total leaf area (cm²) Total leaf area (cm²) Evolution of leaf area for the highest level of vigour (1) Calendar Day Temporal leaf surface evolution function of the vigour for 2004 vig. 0.2 vig. 0.3 vig. 0.5 vig. 0.6 vig. 0.8 vig. 0.9 vig Calendar Day Number of leaves Number of leaves Evolution of the number of leaves for the highest level of vigour (1) Calendar Day Temporal leaf number evolution function of the vigour for 2004 vig. 0.2 vig. 0.3 vig. 0.5 vig. 0.6 vig. 0.8 vig. 0.9 vig Calendar Day Figure 1: Comparison of the total leaf area and of the number of leaves per vine for simulations varying for the climatic conditions or for the vigour of the vine. Data analyses To explore the sensitivity of the model, a principal component analysis was performed. Correlation between variables from host development and variables characterising the level of disease early in the season or later in the season are analysed. These variables are respectively: - the number of secondary leaves at flowering -F2 flo, at day 183 -F2j 183 (10 days after shoot topping) or at the end of the season -F2 201

146 126 - the number of diseased primary leaves at flowering -F1D flo - the total number of diseased leaves at flowering -FD flo - the number of diseased secondary leaves at shoot topping -F2Dst - the number of diseased secondary leaves at day 201 -F2D the diseased surface area at day 201 -SD 201. Then, individuals corresponding to simulations from each parameter at its different level of variation are positioned on the graph of principal. To further explore the relationships between host development and disease variables, another principal component analysis was performed. Correlation between key variables from host development and disease were examined: the vine leaf age structure at the first sporulating event (proportion of young susceptible leaves) (s1s2), the rate of shoot growth (RSD), the rate of leaves emergence (RLE), the number of leaves at flowering time (Nflo), the rate of diseased leaves emergence (RDLE), the severity at shoot topping (Sev st ) and at the end of the epidemic (Sev 240 ), the number of infected leaves at flowering (NIflo), the diseased leaf surface at shoot topping (SDst), the diseased leaf surface at day 240. Individuals corresponding to simulations from each climatic scenario and at the different level of vigour are positioned on the graph of principal. Results and discussion Sensitivity of the model When considering the set of simulations for early inoculation, the three parameters linked to the pathogen (δ), the dispersion process (cid) and the plant growth (vigour) are the most discriminating (the means from each parameter level differ significantly). The three parameters have a similar effect on the disease at shoot topping with an enhanced effect for vigour at day 201 (Figure 2). Lower level of cid and higher level of δ, result in an increase level of disease (higher number of diseased leaves at flowering, higher number of secondary diseased leaves at day 201 and at shoot topping which is correlated to the diseased surface area at day 201) (Figure 2). The average number of diseased leaves at shoot topping is on average 1.5 higher for level 3 of δ or vig compare to level 1. The difference is enhanced for vig at day 201 (x 4.8) (Figure 3). For late inoculation, the level of disease is very low and differences between parameters levels become significant only late in the epidemic (day 201). The parameter τ (ontogenic resistance) is less discriminated, with an average number of diseased leaves increasing of 1.3 at st and 1.5 at day 201 between level 1 and 3 (Figure 3). Differences are more important for late inoculation or lower level of sporulation. The last two parameters tested: the height of shoot topping and the distance between buds are not discriminating. For the height of shoot topping this is probably due to the fact that it has mostly an effect on the number of primary diseased leaves. The non-effect of the distance between buds (between 10 and 30 cm) has to be connected to the dispersion process but should be experimentally tested. This preliminary sensitivity analysis needs further explorations for other climatic scenarios. We also need to assess the range of sporulation of different isolates. Indeed an intermediate level of vigour (2) can give the same level of disease than the higher level (3) for higher levels of sporulation (Figure 4). These results could explain observations of epidemics artificially inoculated and not controlled with the same efficiency depending on the isolate. Simulations show as well that ontogenic resistance became an interesting variable when combined with low level of sporulation. It could be important to acquire more data about ontogenic resistance especially for partially resistant varieties.

147 127 F2j181 F2flo F2j201 d = 2 Cid d = 2 δ F2D201 SD201 F2Dst Fdflo F1Dflo P3 P6 d = 2 D-buds d = 2 vig d = 2 τ P1 P7 P8 Figure 2: Correlation graph between variables according to a principal component analysis based on: the number of secondary leaves at flowering (F2flo), at day 183 (F2j183) or at the day 201(F2 201 ), the number of diseased primary leaves at flowering (F1Dflo), the total number of diseased leaves at flowering (FDflo), the number of diseased secondary leaves at shoot topping (F2Dst), the number of diseased secondary leaves at day 201 (F2D 201 ), the diseased surface area at day 201 (SD 201 ). Individuals from the same cid, δ, D-buds, vig, τ parameters level are joined by the same colour on separate graph of principal. Number of diseased leaves at shoot topping Number of diseased leaves at day τ early inoculation late inoculation τ early inoculation late inoculation Parameter level Number of diseased leaves at shoot topping Number of diseased leaves at day vig early inoculation late inoculation early inoculation late inoculation Parameter level Number of diseased leaves at shoot topping Number of diseased leaves at day δ early inoculation late inoculation δ early inoculation late inoculation Parameter level Figure 3: Range of variation of disease at shoot topping or at day 201, under the three levels of parameters τ, vig and δ for early and late inoculations.

148 128 Number of diseased leaves at day spo level 1 spo level 2 spo level vigour Figure 4: Number of diseased leaves at day 201 for three levels of sporulation and three levels of vigour (cid level 3) Effects of crop growth on the disease From the simulations, an increase of the parameter of vigour from 0.2 to 1 amounted to a higher number of leaves at flowering (Nflo) and a higher rate of leaves emergence (RLE). The RLE was correlated with the rate of diseased leaves emergence (RDLE, R=0.95) and with the severity of the disease at shoot topping (Sev st, R=0.75), whereas Nflo was correlated to the diseased leaves area at shoot topping (SD st, R=0.78) or at day 240 (SD 240, R=0.88). The percentage of young leaves during first sporulation event (s1s2) was correlated to the rate of shoot development (RDS) and contributed more on second axis (Figure 5). These two variables from host allowed to discriminate the years. Under climatic conditions of 2004, characterised by a higher RSD, RLE and s1s2, disease increase early in the epidemic development (flowering (Niflo), shoot topping (Sev st )) and whereas in 2003, characterized by lower RSD, RLE and s1s2 but by a higher increase of secondary leaves number after shoot topping, severity and disease surface is higher at later stage (between shoot topping and day 240). The duration and dynamic of infectious (sporulating) tissue can considerably vary function on the development of secondary shoots therefore on the climatic conditions and vigour (Figure 6). The model strengthens experimental results observed about the effect of the rate of leaf emergence and of the number of leaves at flowering on the severity of the disease. However, the model underlines variation of the dynamics between years with possible variations on the damage. Experiments are undergone to further explore the relationship between vine growth and disease development, 1) to demonstrate if disease development is only controlled by leaf number or also by variation in leaves susceptibility and 2) to test the crop management that could better control disease level. Acknowledgements We thank the Agence Nationale de la Recherche, programme SYSTERRA ANR-08-STRA-04 for funding this work. [1 line free]

149 129 s1s2 RSD RDLE RLE Sevst NIflo SDst Nflo E04 E E03 SD240 Sev240 Figure 5: Correlation graph between variables according to a principal component analysis based on: the average percentage of <10 days leaves during first sporulation event (s1s2), the rate of shoot development (RSD), the rate of leaves emergence (RLE), the number of leaves at flowering (Nflo), the rate of diseased leaves emergence (RDLE), the severity at shoot topping (Sevst) or at day 240 (Sev240) and the number of infected leaves at flowering (NIflo). Individuals from the same vigour level (0.2 to 1) or from the same year (E98-E03-E04) are joined by the same colour on the graph of principal. Number of secondary leaves vig 0.2 vig 0.5 vig Flo St Calendar day Number of secondary leaves vig 0.2 vig 0.5 vig St Flo Calendar day Number of secondary leaves vig 0.2 vig 0.5 Vig 1 Flo 2003 St Calendar day Infectious leaves area (cm²) vig 0.2 vig 0.5 vig vig 0.2 vig vig vig 0.5 Flo St Calendar day Infectious leaves area (cm²) vig Calendar day Figure 6: Simulation of the dynamic of the secondary leaves or of the infectious surface function on the vigour and on the climatic conditions. Flo St Infectious leaves area (cm²) Vig 1 Flo St Calendar day

150 130 References Doster, M. A., Schnathorst, W. C. 1985: Effects of leaf maturity and cultivar resistance on development of the powdery mildew fungus on grapevines. Phytopathology 75: Gadoury, D., Seem, R., Ficke, A. & Wilcox, W. 2003: Ontogenic resistance to powdery mildew in grape berries. Phytopathology 93: Calonnec, A., Cartolaro, P., & Chadoeuf, J. 2009: Highlighting features of spatiotemporal spread of powdery mildew epidemics in the vineyard using statistical modeling on field experimental data. Phytopathology 99: Calonnec, A., Cartolaro, P., Naulin, J. M., Bailey, D., & Langlais, M. 2008: A host-pathogen simulation model: powdery mildew of grapevine. Plant Pathology 57:

151 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Consideration of dynamical plant-pathogen interactions for an improved management of powdery mildew epidemics in grapevine S. Schnee, J. Jolivet, A. Calonnec INRA-Bordeaux, UMR INRA-ENITA 1065 Santé Végétale, BP 81, Villenave d Ornon, France Abstract: The current control of grape powdery mildew requires news strategies, able to limit the pathogen development and requesting a reduced number of chemical sprayings. This study proposes to exploit host plasticity in order to investigate the effect of a plant growth changing on either the epidemic process, or organs susceptibility. An experimental plot combining two cultivars, three rootstocks and two opposite crop management for creating a vigour differential was surveyed regularly during the vegetative growth and different variables were measured to characterise host growth and pathogen. Shoots coming from vinestocks localized in the different vigour area were sampled for organ susceptibility assessment. Two groups of vigour were obtained, not based on the established treatment but on the individual growth potential, that differ mainly by the number of secondary leaves. The vigorous group presents the higher level of disease, increasing with the appearance of secondary leaves. A low vigour modifies the susceptibility of the tissues, by an earlier appearance of ontogenic resistance phenomenon. The relevance of vigour control and the concomitant physiologist mechanisms are discussed as integrative strategy in the grape powdery mildew management. Key words: powdery mildew, grapevine, susceptibility, vigour, epidemic Introduction Powdery mildew remains one of the major fungal diseases of grapevine, which requires repeated chemical applications to preserve the yield and the quality of the harvest. Erysiphe necator, the causal agent of powdery mildew, is a biotrophic pathogen, displaying an asexual polycyclic reproduction capable of exponential multiplication in favourable environmental conditions. The earlier discrete symptoms on leaves and the inconstant berries damages require preventive chemical sprays in classical disease control strategy without a real assessment of the infection level. Therefore, to be in accordance with sustainable agriculture, it becomes urgent to investigate more integrated strategy that would put the pathogen in a defavourable environment to its development in order to be able to control epidemics with a reduced number of applications. The occurrence of ontogenic resistance phenomenon in the cultivated grapevine (Ficke et al., 2002) as well as the spatial location of primary infections on the vine stock, allows hypothesizing that dynamical changes in crop growth and architecture could be considered as key factors for explaining variability in the severity of epidemic behaviour. Then, Human interferences during the wine development through the range of crop management may be exploited to maintain epidemics at a level easy to control with fewer treatments. Management or soil-induced changes in crop growth and architecture are known to modulate significantly the course of epidemics (Calonnec et al., 2009). To investigate the effect of host growth in the epidemic process, variations of the vine development is explored. The vigour is defined as an integrative variable that reflects the metabolic activity of organs in growth and is assessed by quantifying number and rate of plant organs development. A modification of vine development may affect crop structural 131

152 132 characteristics as the leaves areas, the rate of shoots growth, the foliage density or the landscape of susceptible tissue. Variations of vine development can be induced through several factors: i) environment (climate, soil), ii) genetics (variety and root-stock planted), iii) cultivation factors (fertilization, irrigation, cover crop, pruning, shoot topping). The combination of all these factors should allow generating a plant growth differential that could modulate the pathogen installation and further the dynamic of epidemics. In order to assess if the plant could be considered as a key element of the protection system, combined measures of host and epidemic development are performed, as well as the measure of plant organ susceptibility. Material and methods Two plots were selected on the experimental site of INRA (Domaine du Grand Parc, Latresne) with the following characteristics. The P19 plot presents various levels of vigour by a combination of rootstocks (SO4, 110R and Riparia) and crop management (chemical weed control versus perennial cover crop). The plot is designed as 6 rows planted alternatively with two susceptible varieties Merlot and Cabernet-Sauvignon in The plot is shared in 8 blocks of 30 vines across the rows (6 rows x 5 vines). Each block is constituted by 6 sub-units combined each rootstock with each cultivar, randomly distributed. The first 4 blocks are conducted with perennial cover crop whereas the 4 others with weed chemical control. The P5, planted with the variety Merlot, displays two visually distinct area of vigour. The dynamic of development of powdery mildew in relation to the plant growth was performed on the P19, whereas the susceptibility of shoots in regards to the plant development was studied on both plots. Impact of host growth on a powdery mildew epidemic On the P19, inside each sub-unit, one vinestock was selected based on its number of buds (7 for the Merlot and 8 for Cabernet-Sauvignon) and shoot s configuration. One shoot of this vinestock was inoculated at the stage 2 to 4 leaves according to Calonnec et al. (2009). Several measurements were regularly made to characterize the vegetative growth of the vinestocks. Two to three times per week, new emerged leaves were marked by colour markers and size of shoots was measured. The different measured variables (number of leaves at flowering (NLf), number of leaves at berry pea size stage (NLps) shoot length at flowering (SLF)) and calculated variables (rate of shoot development (RSD), rate of leaf appearance (RLA)) described the dynamic of leaf appearance (Table 1). In a point of view of disease survey, the percentage of diseased foliar surface was estimated weekly on all leaves of the selected shoot and two of its neighbours. The measured variables (number of diseased leaves at flowering (NDLf), number of diseased leaves at berry pea size stage (NDLps), conditional severity on diseased leaves at berry pea size stage (Scps)) and calculated variable (rate of diseased leaves appearance (RDLA)) described the dynamic of diseased leaf appearance (Table 1). Susceptibility of shoots in relation to the vine vigour For the shoots susceptibility assessments, 12 shoots coming from vinestocks localized in the different vigour area were cut early in the morning and immediately brought back to the laboratory. Before sampling, leaf petioles were marked with a colour code to specify their respective position on the shoot. In axenic conditions, leaves were disinfected (in a sterile water bath containing 65% calcium hypochlorite) and rinsed in sterile water. Three foliar

153 133 discs (Ø 22mm) were cut in each leaf to be distributed for 3 different Petri dishes corresponding to three pathogenicity tests. Leaves from a same foliar stage were distributed in Petri dishes (6 discs from 6 shoots per dish), containing an agar medium (20g l -1 ) supplemented with benzimidazole (30mg l -1 ). Petri dishes were placed in a settling tower and were artificially inoculated by blowing conidia from a 14 days infected leaf, according to Willocquet et al. (1996). The infection capacity was assessed 72 hpi by counting conidia development stage, after removal of fungal structures by a scotch application and cotton blue staining procedure. After inoculation of foliar disc by deposit of few conidia by a needle, the colony growth was measured 4, 7, 10 and 13 dpi. Sporulation assessment was performed by the measure of the quantity of generated conidia 13 dpi, by using a particle counter (Beckman Coulter) that records size cells included between 18 and 35 µm. Only sporulation data will be described in the results part. Data were submitted to statistical analysis (PCA and clustering) with software Prism and R. Table 1. Variables of plant development and disease measured and calculated at different plant growth stage, used for further analysis in this study. Variables Abbr. Units Plant development Number of leaves at flowering NLf - measured Number of leaves at berry pea size stage NLps - measured Rate of leaf appearance RLA number day -1 calculated Shoot length at flowering SLf cm measured Rate of shoot development RSD cm day -1 calculated Disease assessment Number of diseased leaves at flowering NDLf - measured Number of diseased leaves at berry pea size stage NDLps - measured Conditional severity on leaves at berry pea size stage Scps % diseased leaf -1 measured Rate of diseased leaves appearance RDLA number day -1 calculated Results Impact of host growth on a powdery mildew epidemic For Merlot (and Cabernet-Sauvignon, data not shown) the plant growth is characterized by an approximately linear development of primary leaves and a linear increase of leaves appearance on secondary shoots with a strong variations between individuals at the end of the monitoring (Figure 1). The most vigorous vines presented three times more secondary leaves than the lowest vigorous. The number of diseased leaves follows an exponential curve, with an increase starting at the flowering stage. The last notation (68 days after contamination) presents an important variability of powdery mildew incidence on leaves. The following variables of vigour are well correlated together according to a principal component analysis (Figure 2): the shoot growth variables (shoot length at flowering stage and rate of shoot development, R² = 0.96), the rate of leaf appearance and the leaf number at different phenological stages (at flowering stage and berry pea size stage, R² = 0.88). Individuals from the weed controlled blocks are characterized in average by a global increase of shoot development (SLf and RSD). However individuals of the two blocks overlap and

154 134 indicate that the experimental design is not able to control sufficiently a vigour differential. This fact could be explained by the high intra-group variability observed especially at the end of the monitoring (Figure 1). Nbr of primary leaves / 3 shoots Flo Days after inoculation Nbr of secondary leaves / 3 shoots Flo Days after inoculation Nbr of diseased leaves / 3 shoots Flo Days after inoculation Figure 1. Distribution of the number of primary leaves, secondary leaves and diseased leaves observed on the 3 surveyed shoots on each selected vinestock function of the day after inoculation for cultivar Merlot. SLf RSD SLf RSD WC RLA NLps NLf CC RLA NLf NLps Figure 2. Correlation graph between vigour variables according to a principal component analysis based on: the shoot length at flowering stage (SLf), the ratio of shoot development (RSD), the rate of leave appearance (RLA), the leaf number at stage berry pea size (NLps) and the leaf number at flowering stage (NLf). Individuals from the same treatment are displayed by the same colour (wc: weed control area, cc: cover crop area). A PCA preceded by a hierarchical ascendant clustering analysis was performed to group individuals the most similar according to the growth variables previously used (Figure 3). Two new clusters of individuals were based on the most relevant variables, namely the rate of leave appearance and the number of secondary leaves (Figure 3). These results were consequently used to investigate the distribution of vinestocks in relation to the disease variables. The correlation graph displays a correlation between the conditional severity at the

155 135 stage berry pea size and the number of diseased leaves at the stage flowering (R² = 0.66), and between the number of diseased leaves at the end of the survey and the rate of diseased leaves appearance (R² = 0.71) (Figure 4). The PCA graph displays a distinction between the two groups of vigour previously identified. Height SLf RSD RLA NLps NLf WC WC WC CC WC CC WC WC WC CC CC CC CC WC CC CC CC CC CC CC CC CC WC WC CC WC WC WC WC CC WC WC WC WC Figure 3. Hierarchical ascendant clustering analysis and principal component analysis based on the new vigour groups of the cluster analysis. Individuals from the same group are displayed by the same colour (1: high vigour individuals, 2: low vigour individuals). Scps Scps NDLF NDLF 2 NDLps RDLA NDLps RDLA 1 Figure 4. Correlation graph between disease variables according to a principal component analysis based on: the conditional severity at the stage berry pea size (Scps), the number of diseased leaves at the flowering stage (NDLf), the number of diseased leaves at the stage berry pea size (NDLps) and the rate of diseased leaves appearance (RDLA).

156 136 Susceptibility of shoots in relation to the vine vigour The effect of wine development on tissue susceptibility was measured by comparing the intensity of sporulation per infected leaf surface (Figure 5a). On the P19, the sporulation from leaves of shoots taken from the two treatments (weed control versus cover crop) follows a comparable trend. If the sporulation amplitude and the range of diseased leaf age are identical, the maximum of sporulation in low vigour group is one day delayed in regards to the high vigour group. Concerning the P5, only the maximal amplitude of sporulation is similar between the two groups of vigour (Figure 5b), however, the range of age for sporulating leaves is significantly lower in low vigour group (less 4 to 12 days old) than in high vigour group (from less than 4 to 18 days old). The age of maximal sporulation is also significantly earlier in the low vigour group (from less than leaf of 6-7 days old versus leaves from 8 to 10 days old in the high vigour group). The amplitude is more intense on the P19 than in the P5 (7.5x10 4 versus 4x10 4 conidia produced per cm² of leaf disk). Number of spores/cm² of leaf disk A 19H 19L Leaf age (days) Number of spores/cm² of leaf disk B 5H 5L Leaf age (days) Figure 5. Sporulation level in relation to the leaf age on the surveyed shoot (A: on the P19, B: on the P5, h: high vigor level, l: low vigor level). The mean of 6 repetitions at each time point for the two distinct vigour level was adjusted by the software Prism. Discussion The field experiment allowed to distinct two groups of vigour which differ mainly by the number of leaves developed on secondary shoots. At flowering and berry pea size stage, the number of leaves is correlated to the total number of diseased leaves at the end of monitoring 65 days after inoculation (respectively R²=0.89 and R²=0.84). The emergence of secondary leaves enhance the amount of susceptibility tissues for pathogen development, explaining the rapid increase of disease level after the flowering stage and maintaining a high level of epidemic inside the plot, as observed by Valdes et al. (2007). A higher level of vigour generates a greater amount of biomass, including a more intensive growth of secondary shoots, after primary shoot topping. The production of these new susceptible leaves explain the general high disease severity of the most vigorous vines, which accentuates with the increase of the vegetative host growth. In opposition, a low vigour level reduces the shoot apical growth, which sometimes cannot arrive to a sufficient height to be cut during the topping practice, and then limits the development of secondary shoots. For these vines early contaminated, the epidemic behaviour seems to be only dependent on the amount of tissue available.

157 137 Therefore the control of the vigour could modulate the severity of the epidemic in case of low pathogen pressure in vineyard, closely linked of the date of the primary contamination. A concrete and sustainable solution from these first results could be based on a precocious chemical intervention coupled to a limited foliage density by a reduction of secondary shoots, to prevent powdery mildew epidemic extension. The assessment of tissue susceptibility indicates that a low vigour could modify the mechanisms of ontogenic resistance, by an earlier appearance of resistance at fixed foliar age. On the plot with the lowest difference of vigour (P19), sporulation curves are similar independently of the vigour level and the area under the curve that represents the global dispersal inoculum are identical with as consequence to maintain a high level of pathogen pressure. On the contrary on the P5, the plot with significant visually difference of vigour, a low vigour presents a narrow area under the curve of sporulation and limits the production of inoculum. The management of the global quantity of susceptible tissues (young leaves from the primary and then secondary shoots) constitutes an important factor to reduce as possible the favourable substratum for pathogen colonization. Mechanisms of ontogenic resistance establishment remain obscure and appear specific to the considered tissue in a same variety (Gee and Gadoury, 2008). One explanatory physiological factor could be linked to the cell osmotic pressure, directly regulated by the carbohydrate metabolism. Activation of photosynthesis process and assimilation metabolism generates sugar accumulation in plant cell content, which the consequent osmotic value can promote resistance to fungal penetration (Schnathorst, 1959). The ability of leaf physiological maturity to trigger a henceforth efficient mobilization of secondary metabolism that includes the active defense reactions against pathogens attack is also closely linked to the energy provided by primary metabolic pathways (Bolton, 2009). Therefore factors limiting plant growth may modulate the balance of the plant primary metabolism that can modify the fitness of the plant-pathogen interaction. Plant growth management either quantitatively or qualitatively in sustainable agricultural strategies needs further explanations and could be an interesting prophylactic control towards powdery mildew, providing no effect on crop agronomical objectives. Acknowledgements This project was supported by the Agence Nationale de la Recherche, programme SYSTERRA ANR-08-STRA-04. We thank Romain Cargnelutti and Valérie Mayet for their active participation to the experimentation. References Bolton, M. D. 2009: Primary metabolism and plant defense fuel for the fire. Molecular Plant-Microbe Interactions 22(5): Calonnec, A., Cartolaro, P., and Chadoeuf, J. 2009: Highlighting features of spatiotemporal spread of powdery mildew epidemics in the vineyard using statistical modeling on field experimental data. Phytopathology 99: Ficke, A., Gadoury, D. M., and Seem, R. C. 2002: Ontogenic resistance and plant disease management: A case study of grape powdery mildew. Phytopathology 92:

158 138 Gee, T. G., Gadoury, D. M. 2008: Ontogenic resistance to Uncinula necator varies by genotype and tissue type in a diverse collection of Vitis spp. Plant Disease 92(7): Schnathorst, W. C. 1959: Resistance in lettuce to powdery mildew related to osmotic value. Phytopathology 49: Valdes, H. 2007: Relations entre états de croissance de la vigne et maladies cryptogamiques sous différentes modalités d entretien du sol en région méditerranéenne. Montpellier: Ecole de Montpellier SupAgro, Thesis. Willocquet, L., Colombet, D., Rougier, M., Fargues, J., Clerjeau, M. 1996: Effects of radiation, especially ultraviolet B, on conidial germination and mycelial growth of grape powdery mildew. European Journal of Plant Pathology 102:

159 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Blackrot - downy mildew control in small vineyards in southern Switzerland C. Gessler 1, M. Jermini 2 1 Plant pathology institute of Integrative Biology, Swiss Federal Institute of Technology, 8092 ETH-Zürich, Switzerland. 2 Agroscope ACW Changins, Centro di ricerca Cadenazzo, 6594 Contone, Switzerland Abstract: Blackrot caused by Guignardia bidwelli is causing damages in recuperated vineyards in southern Switzerland planted with the cultivar Isabella. The source of inoculum is attributed to abandoned vineyards. Heavy and prolonged rains may also favor infection by Plasmopara viticola on Isabella especially if a heavy load of primary inoculum can be expected. A 6 year experiment was made to test and confirm minimal necessary fungicide applications and timing to control Blackrot and downy mildew. Infection conditions during the period between flowering and up to six week after, have led to heavy losses reaching 100% without adequate control. A combination of the two pathogens was registered, with P. viticola bunch symptoms without visible sporulation. We recommend therefore using fungicides or fungicide combinations with an effect on both pathogens. Mostly two well timed sprays with an appropriate fungicide active against both pathogens are sufficient to control fully the diseases on bunches and avoid loss. Under particular rainy situation a third applications gave 100% control. Timing should be before rain events leading to prolonged leaf wetness during the time period between flowering and six week afterwards. Winegrowers in the area are not professional winegrowers or farmers, having little specific knowledge, therefore they will apply fungicides only at particular date (Saturday) following the instruction of the manufactures. Applications are made by knapsack sprayers and vineyards are usually terraced irregularly often as pergola. Therefore we developed an easy applicable scheme to determine application timing and an instruction system to determine fungicide concentration so to avoid over and under dosaging. Here we present the most relevant points we teach to our winegrowers. Key words: Guignardia bidwelli, Plasmopara viticola, Maggia, viticulture Introduction Grape production area in the canton Ticino Southern Switzerland probably dates back to the Romans. It reached its maximum extension in the middle of the 19 th century declining with the appearance of the phylloxera and downy mildew. It was still occupying about ha early in the 20 century mostly with hybrids. From 1906 onwards Merlot gained popularity in flat or gentle sloped areas mostly in the regions confining to Italy. However most terraced and step vineyards present in the upper valleys were abandoned so that the vine growing area reach a low of 700ha around the 1970ties. The abandoning was mostly due to the change from a rural and rather poor community to a service community. This transition extended the most southern part, over the years In the side valleys it was fast and the profound changes happened in less than 20 years 1950 to Grape growing was revived with state aid to the Merlot planting reaching currently about 1000ha. Testimony of the old vineyards in the northern valleys still can be found, consisting of small terraces with stone walls and the unique stone slabs ( carasch ) support of the horizontal wood structure of the pergola. Long abandoned sites are overgrown with forest trees. Close to the villages, however, since the turn 139

160 140 of the millennium, efforts of various environmental and landscape protection associations together with the local communities, strive to revitalise the vineyards and grape growing tradition. Walls and pergolas are rebuild, often a range of cultivars are planted, however all hybrids with resistance against downy mildew (Isabella, Katowba, Chamboursin, Leon millot, Cabernet, Regent). The traditional cultivar used for pergola was Isabella. Isabella is considered resistant to diseases and can be grown on its own roots, being extremely vigorous and supporting high yields with a single plant covering often 5 to 10m 2 with 1-1.5kg/m 2. Up to the 50ties, 2-3 Bordeaux mixtures were applied to control the sole relevant disease downy mildew caused by P. viticola favoured by the high and frequent rain late spring early summer and able to produce losses on Isabella. In late winter, leaf and grass residues were racketed and burned. During the growing season grass was collected underneath the pergolas. With their availability modern fungicides substituted the Bordeaux mixture. Grass (hew) was not used anymore and currently grass is mulched, often with a line trimmer. Many vineyards as said above were abandoned or at least the amount of input drastically reduced with only sporadic, if not erratic, interventions. Bunch damage such as bunches with sparse berries, black-brown shrivelled berries, green small berries (Gessler et al. 2006) are frequent in such vineyards in years with wet springs such as 2006/2007, however can be present also in well kept vineyards. We confirmed the presence of P. viticola, however, since about 10 years we attribute relevant proportion of the damage to Guignardia bidwellii (Gessler et al. 2006, 2008) causal agent of the black rot. Damages are variable over the years. In experimental plots located in the Valle Maggia we demonstrated that damage caused by both diseases can be fully avoided with two to three correctly timed applications of fungicides with eradicative action against both agents (Gessler et al. 2008). However, as the vineyards of the type described above are all maintained by non professionals, all actions are executed on free days (mostly Saturdays) and knowledge on diseases and epidemiology/biology are minimal if not absent. In this article we describe our approach to communicate the basic rules to be observed for a successful control of the diseases under the described conditions. Material and methods We assembled the various topics and available information necessary to successfully control diseases in the local vineyards. 1- Solid knowledge on the biology and conditions of infection of P. viticola and G. bidwellii: available from literature (Ferrin & Ramsdell 1977, Hoffman & Wilcox 2002, Hoffman et al. 2002, Hoffman et al. 2004). 2- Application strategy: data on successful control using a locally adapted strategy were acquired in the years (Gessler et al. 2008). The strategy recommends that fungicides effective against black rot be applied during and up to 4 weeks after flowering, if weather forecasts predict rains, which are likely to give leaf wetness periods greater than 12 h. If such a period is not covered by an application within the subsequent days, an application should be made at the first possibility, e.g. when the grapes bunches are dry and using a curative product. 3- Equipment: vineyards holder use a knapsack blower or sprayer for the application. Applications are executed on Saturdays exceptionally on other days 4- Knowledge on fungicides and applications (e.g. quantities, concentrations).

161 141 Results and discussion We identified, contrary to dealing with professional vine growers, a lack in knowledge in point 4. Recommendation of a fungicide to be used was usually acquired directly or indirectly in a garden shop, often solely on the label information. The information of quantity to be used was acquired from the label on the product package (example Mapro l/ha, 0.1%) and was for most ambiguous. Even if pamphlets distributed by the agrochemical companies clearly indicate that the quantity of fungicide should be related to the plot surface, some vineyard holders used the indication % product, clearly leading to an application quantity well below the indication. On the other hand, if the information quantity per surface was used, as did some holders, early season application clearly led to overdoses as vegetation surface was still only a fraction compared to plot surface and in some cases plot size was much larger than maximal vine surface. A general problem was that the water quantity applied in relation to vine surface or plot surface was unknown, or not considered: the water amount sprayed per leaf surface was determined by the walking speed or/and until visible wetting. Based on these constraints, we organized two participatory conferences. The topics treated were: 1) Diagnostics: Which diseases? What do I see without control or with a wrong product/wrong dosages/wrong application time? 2) Product characteristics (protective or/and curative action): How to calculate the water applied in relation to foliage surface? How to calculate quantity of product for each filling of the knapsack sprayer? 3) Duration of the protection: The duration of the protection effect of a fungicide (ten days after application and 48 hours prior if a curative fungicide was used) 4) Infection conditions: The weather data (measured rain periods always prolonged as leaf wetness period for all situations where rain was measured after dark until next morning 7 am) 5) Correct application dates: Combing the above information the dates on which the first and the following application should be executed were shown (with and without the constraint to limit applications to Saturday). 6) Validation: Data were compared to the application dates in the experimental plot and to the dates used by the participants. Participants were then requested to report on the quality of the harvest. In 2008, much of the vineyard surface was destroyed in July by an exceptional heavy hail storm, which also conditioned to harvest In 2009, in the untreated plot of the our experimental vineyard no black rot appeared, however we registered early defoliation and ca 50% berries loss (shriveled reddish, some green) due to P. viticola. Two treatments were sufficient to fully control P. viticola. In 2008, most of the participants of the participatory conferences followed the advice on products and application quantity, part also on the timing. Not following the timing strategy was mostly due to: 1. pre blooming applications out of the fear that the diseases could appear earlier or wrong diagnostic confusing symptoms of the Grape Erineum Mite (Colomerus vitis) with downy mildew; 2. application later than the 4-6 week after bloom because of damage to the foliage by P. viticola; 3. personal reason (forgotten, absent, other priorities). In all cases where an application was made during bloom and a second in the three following weeks, the participants did not report any harvest problems (we interpreted this information as no bunch damage). In cases where no treatments during the indicated period were made, we registered heavy damage by P. viticola. Currently however due to the low number of cases in each category and just one year of data, we can give no realistic estimate on the success of the action.

162 142 References Ferrin, D. M., & Ramsdell, D. C. 1977: Ascospore dispersal and infection of grapes by Guignardia bidwellii, the causal agent of grape black rot disease. Phytopathology 67: Gessler, C., Foiada, F., Jermini, M. & Pertot, I. 2008: Control of Blackrot (Guignardia bidwellii) on the hybrid vitis cultivar Isabella. IOBC/WPRS Bull. 36: Gessler, C., Blaise, P. & Jermini, M. 2006: Black rot on the hybrid vitis cultivar Isabella. OILB/WPRS Bull. 29(11): Hoffman, J. L. E. & Wilcox, W. F. 2002: Utilizing epidemiological investigations to optimize management of grape black rot. Phytopathology 92: Hoffman, J. L. E., Wilcox, W. F., Gadoury, D. A; & Seem, R. C. 2002: Influence of grape berry age on susceptibility to Guignardia bidwellii and its incubation period length. Phytopathology 92: Hoffman, J. L. E., Wilcox, W. F., Gadoury, D. M., Seem, R. C. & Riegel, D. G. 2004: Integrated control of grape black rot: Influence of host phenology, inoculum availability, sanitation, and spray timing. Phytopathology 94: Jermini, M. & Gessler, C. 1996: Epidemiology and control of grape black rot in southern Switzerland. Plant disease 80: Spotts, R. A. 1977: Effect of leaf wetness duration and temperature on the infectivity of Guignardia bidwellii on grape leaves. Phytopathology 67:

163 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp First study on the population genetic structure of Guignardia bidwellii M. Jermini 1, A. Angst 2, M. Raynal 3, C. Gessler 2, G. Broggini 2 1 Research station Agroscope Changins-Wädenswil ACW Centre of Cadenazzo, 6594 Contone, Switzerland. 2 Plant Pathology, Institute of Integrative Biology, Swiss Federal Institute of Technology, 8092, Zürich, Switzerland 3 Institut Français de la Vigne et du Vin, Blanquefort, France Abstract: The ascomycete Guignardia bidwellii is the casual agent of the black rot of grape. Native to North America introduced to Europe in the early 19th century and today regularly observed in France, Italy, Germany and southern Switzerland. The black rot is not generally considered as an important disease but it is able to induce heavy crop losses particularly in regions characterised by weather conditions favourable to Plasmopara viticola. Present studies focused on the genetic structure of the pathogen population, to get a better understanding of the epidemiology and life cycle of the disease. We successfully extracted and amplified DNA from mummified berries and developed six SSR markers for genotyping of G. bidwellii isolates. The SSR markers were used to study the pathogen populations of two Merlot vineyards in France and a mixed variety vineyard (Merlot, Regent, Isabella, Solaris, Bianca and Chambourcin) in southern Switzerland. All three populations were related and shared common SSR alleles. Gene diversity was similar for the two French populations, and the Swiss population was nearly clonal in a genotype frequently observed in France. Only a few private genotypes were found, therefore the populations differed mainly in genotype frequency. Based on our data we challenge the importance of sexual reproduction for the propagation of the fungus. The low genotype and allele diversity observed could be explained either by a generally low genotype polymorphism in Europe or by a low SSR marker polymorphism. More research is therefore needed to assess SSR marker polymorphism on an American population, to develop additional SSR markers for a higher genotype resolution and to understand an eventually difference in the genotype frequency between leaf and berry infections. Key words: SSR markers, genotype, black rot Introduction The black rot of grape, caused by the ascomycete Guignardia bidwellii, was introduced into Europe in 1885 and today it is widespread in France (Gironde, Charentes, Loire, Midi- Pyrennees, Savoy), Italy (Liguria, Tuscany, Friuli) and, since 2002, in some parts of Germany (Maixner and Holz, 2003). In Switzerland, the disease was detected for the first time in 1988 in Ticino (southern Switzerland) (Pezet and Jermini, 1989) and in 1992 in the vineyards on the lake of Geneva (Siegfried et al., 1993). On the contrary, to downy (Plasmopara viticola) and powdery (Uncinula necator) mildews, also introduced in Europe during the 19 th century, the black rot has an endemic character, which limits its impact on economics. The black rot is, however, able to induce heavy yield losses in regions characterized by rainy and prolonged wetness season (Ferrin and Ramsdell, 1977; Gessler et al., 2006; Hoffman and Wilcox, 2004; Ramsdell and Milholland, 1998). The disease cycle, epidemiology of the pathogen, the correlation between berry age, leaf wetness duration, temperature and humidity on incubation and latency periods have been described (Cartolaro et al., 1993; Hoffman et al., 2002; Spotts, 1977; Weber, 1987) as well as the role of rainfall on the release and dispersal of ascospore and conidia (Ferrin and Ramsdell, 143

164 , 1978; Spotts, 1980). Leaf infections had no clear correlation with disease on bunches and secondary infections seemed to play no major role in the disease progress on bunches (Jermini and Gessler, 1996). Hoffman and Wilcox (2004) have shown that conidia are also present on overwintered mummies and they have given some evidence for their involvement in primary inoculums. So far, no population study was performed on G. bidwellii to clarify the role of conidia and ascospores in the primary infection on the bunch infection. The aim of this study was to develop SSR marker for G. bidwellii and to investigate the genetic structure of the population in some countries. Material and method Samples Mummified berries were collected in three untreated plots from Blanquefort (Blq), Salleboeuf (Sb), both in France and planted with Merlot, and Cugnasco (Ti) in Switzerland. The plots were constituted from two rows of 25 plants/row in Blq, one row of 35 plants in Sb and 8 rows of 80 plants/row in Ti. This last plot corresponded to the REPCO plot planted with eight different cultivars: Merlot, Isabella, Solaris, Bianca, Müller-Thurgau, Gamaret and Chambourcin (Matasci et al., 2007). Up to 8 mummies per cluster of each infected plant were sampled for a total of 76 berries in Blq, 112 in Sb and 192 in Ti on the varieties Merlot, Isabella, Solaris, Regent, Bianca and Chambourcin. In addition, four in vitro mycelia culture were used for SSR development and diversity control purposes. One of them, provided by Berndt Loskill (BBA Bernkastel-Kues/Mosel), was isolated in Geissenheim (DE). The other three provided by Olivier Viret (ACW), were collected in the region of Lake Geneva (CH) isolated from Vitis vinifera, Vitis lambrusca and Parthenocissus and were used for the SSR development and the polymorphism evaluation. DNA extraction DNA extraction was based on the protocol for plant tissues of Aldrich and Cullis (1993). SSR isolation and development SSR were isolated according to the FIASCO protocol of Zane et al. (2002) with slight adjustments: four separate reactions for restriction enzyme adapter sets were made and the unspecific amplification was performed using the adapter top oligo without any selective base. The primers were designed on the clone insert sequences containing single sequence repeats using the freely accessible SSR primer development tool of the Brainsick Public Databases ( Two sets of primer were developed to amplify six different SSR loci. The first set consisted in a normal primer pair in which the forward primer has been labelled with a fluorescent dye for detection on an ABI 3130xl Sequencer. For the second set of three primer pairs was labelled as proposed by Schuelke (2000). Population genetics Gene diversity was calculated for all three populations according to Nei (1987) implemented in the Excel Microsatellite Toolkit ( of the Animal Genomics Laboratory (UCD). FST values were calculated for pairwise population comparison including all samples and after clonal correction using Arlequin (

165 145 Results and discussion DNA extraction, SSR isolation and development The existing protocol for CTAB DNA extraction from plant tissue (Aldrich and Cullis, 1993) suitable to extract fungal DNA from mummies of grape berries was used. Lyophilisation of the mummified berries is of particular importance, although they seem to be dry their moisture otherwise makes grinding impossible. Amount and purity of the extracted DNA fulfilled the qualification for successful PCR amplification and microsatellite analysis. Six microsatellites were developed and they showed little variation in allele length, with allele number varying from one to five (table 1). Msp2_2 was the only monomorphic marker with one allele at 141 bp, followed by Msp9_2. Table 1. Allele sizes in bp identified per each SSR locus across all investigated populations. A2 Mse10 Mse18 Msp2_2 Msp9_2 Mse10_2n Mse18 and Msp2_2 were not considered for the population study. The first because it showed weak amplification in all samples and the second was discarded due to missing variability. In vitro cultured mycelia were used as control for all markers in order to evaluate polymorphism. But due to lack of amplification for some loci, the genotypes of these were incomplete and are not shown. For all further considerations, only results from mummies are taken into account. Population genetics Out of the 380 mummies collected from three different vineyards, 300 samples showed amplification of all SSR markers allowing determining the genotype. For 93% of these only one allele per loci was found, whereas for 7% two different alleles at least one locus were observed. For 75 samples only incomplete datasets with up to 4 missing markers could be collected. For the residual 5 samples none allele was detected at all. Therefore, 280 samples were considered to be infected from a single G. bidwelli isolate, based on the 4 SSR markers. Thirteen unique combinations of allele size of the four informative SSR markers were observed in the three vineyards and defined as genotypes A to M (Table 2). The analysis of the genotypes identified on berries coming from the same inflorescence from the three vineyards showed a maximum of 5 genotypes on up to 8 berries per inflorescence (data not shown). Eight genotypes were found for the two French locations with a slightly higher frequency of the genotypes B, G and M in Sb (Table 3). The genotype variability was drastically reduced for Ti, where the population is nearly clonal with a dominant genotype B, which is also the most frequent genotype in Blq and Sb that represent the 98% of the population (Table 3).

166 146 Table 2. Genotypes found on berries with one allele per loci, together with their short notation. Genotype A2 Mse10 Msp9_2 Mse10_2n Short notation A B C D E F G H I J K L M Clustering of the same genotype on neighbouring plants was observed along rows of the vineyards, in Sb, as well as across rows in Blq (data not shown). The vineyards Sb and Blq have similar gene diversities and high frequent alleles are shared, so they are relatively closely related. On the contrary the clonal population of Ti shows low gene diversity. The alleles observed beneath the ones of genotype B are alleles with high frequency in both French vineyards. Genotypes unique to a single location were detected (genotypes A, H, I, J in Blq, genotypes C, D, E and L in Sb and genotype F in Ti) but were generally rare in the analyzed populations, except for genotype A in Blq (19.6% of isolates) (Table 3). No host specificity was observed in Ti, where mummies from 6 different grape varieties were collected. The samples collected on Merlot, Isabella, Solaris, Regent, Bianca and Chambourcin show a clonal population with the exception of genotypes M, and F, the first recorded twice only on variety Isabella and the second once only on Chambourcin. Both genotypes were also recorded in the two other Merlot vineyards (Table 3). The in vitro grown isolates were isolated from three different species (Vitis vinifera, Vitis lambrusca and Parthenocissus) and even if only partial genotypes were obtained, these isolates did not show any private allele suggesting no broader host specificity. This hypothesis must be confirmed by mean of supplementary markers. At present time there is no information of the genetic structure of the Guignardia bidwelli population in its origin country. We can anticipate that two founder effects shaped the European G. bidwelli populations: the first one occurred when G. bidwelli genotypes were introduced from the United States in France at the end of the 19 th century and is supported just by historical and not yet by experimental data; the second one occurred when at least two genotypes (B and M) found in France were introduced in Ticino.

167 147 Table 3. Genotype distribution in the two French, Blanquefort (Blq) and Salleboeuf (Sb) locations and in Cugnasco (Ti) found on single infected berries. Genotype Blq Sb Ti Total A 9 9 B C 1 1 D 2 2 E 3 3 F 1 1 G H 1 1 I 1 1 J 3 3 K L 1 1 M Total The clonality observed in the Ti population and the fact that French genotypes are found also in Ticino supports the theory that the pathogen is not forced to undergo sexual reproduction for overwintering. Conidia may therefore play a major role in the overwintering and dissemination of the black rot, as hypothesized by Hoffmann and Wilcox (2004). These first findings should be validated with new analyses. The next steps are 1) the analysis of the genotypes found in the same vineyards over at least two years 2) the development of additional SSR markers to verify the low variability observed in the present work and consequently to confirm the low importance of sexual reproduction in Europe; 3) the investigation of the degree of SSR polymorphism in the American populations; 4) to analyse the differences in the genotype frequencies between leaves and berries and their relation, because the present study has considered only the berries. Acknowledgements We thank Dr. Bernd Loskill and Dr. Olivier Viret for the supplying of the black rot in vitro cultures. References Aldrich, J. & Cullis, C. A. 1993: CTAB DNA extraction from plant tissues. Plant Molecular Biology Reporter 11(2): Cartolaro, P., Pacreau S., Clerjeau M. & Maurin, G. 1993: Sensibilité des grappes au black rot: effet de l interaction température-durée d humectation sur la gravité de l infection en condition contrôlées. In Proceedings of the IOBC Working Group «Integrated control in viticulture, ed. B. Dubos.

168 148 Ferrin, D. M. & Ramsdell, D. C. 1977: Ascospore Dispersal and Infection of Grapes by Guignardia bidwellii, the Causal Agent of Grape Black Rot Disease. Phytopathology 67: Ferrin, D. M. & Ramsdell, D.C. 1978: Influence of conidia dispersal and environment on infection of grape by Guignardia bidwellii. Phytopathology 68: Gessler, C., Blaise, Ph. & Jermini, M. 2006: Black rot on the hybrid vitis cultivar Isabella. IOBC/WPRS Bull. 29(11): Hoffman, L. E., Wilcox, W. F., Gadoury, D. A. & Seem, R. C. 2002: Influence of grape berry age on susceptibility to Guignardia bidwelli and its incubation period length. Phytopathology 92: Hoffman, L. E., Wilcox, W. F., Gadoury, D. M., Seem, R. C. & Riegel, D. G. 2004: Integrated Control of Grape Black Rot: Influence of Host Phenology, Inoculum Availability, Sanitation, and Spray Timing. Phytopathology 94(6): Jermini, M., Gessler, C. 1996: Epidemiology and control of grape black rot in Southern Switzerland. Plant Disease 80(3): Maixner, M. & Holz, B. 2003: Risiken durch invasive gebietsfremde Arten für den Weinbau. Angewandte Wissenschaft, BMVEL. 498: Matasci, C., Jermini, M., Gobbin, D., Gusberti, M., Rosselli, N. & Gessler, C. 2007: Plasmopara viticola disease severity in a mixed cultivar trial. European meeting of the IOBC/WPRS working group Integrated Protection in Viticulture, Marsala (Italy) October 25-27, Abstract Book: 78. Pezet, R. & Jermini, M., 1989: Le black rot de la vigne; symptômes, épidémiologie et lutte, Revue suisse Vitic. Arboric. Hortic. 21(1): Nei, M. 1987: Molecular Evolutionary Genetics. Columbia University Press,New York, NY, USA. Ramsdell, D. C. & Milholland, R. D. 1998: Black Rot. Compendium of Grape Diseases: Schuelke, M. 2000: An economic method for the fluorescent labeling of PCR fragments. Nature Biotechnology 18: Siegfried, W., Pezet, R. & Jermini, M. 1993: Merkblatt 613: Black Rot oder Schwarzfäule. Eidgenössisches Volkswirtschafts Departement EVD, Forschungsanstalt Agroscope Changin-Wädenswil, Spotts, R. A. 1977: Effect of leaf wetness duration and temperature on the infectivity of Guignardia bidwellii on grape leaves. Phytopathology 67: Spotts, R. A. 1980: Infection of grape by Guignardia bidwellii fyctors affecting lesion development, conidia dispersal and conidia population on leaves, Phytopathology 70: Weber, M. 1987: Le Black rot: éléments de biologie et moyens de lutte chimique. Phytoma 376: Zane, L., Bargelloni, L. & Patarnello, T. 2002: Strategies for microsatellite isolation: a review. Molecular Ecology 11(1): 1-16.

169 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Methods for screening new Ampelomyces strains to be used as biocontrol agents against grapevine powdery mildew S. E. Legler 1, T. Caffi 1, L. Kiss 2, A. Pintye 2, V. Rossi 1 1 Istituto di Entomologia e Patologia Vegetale, Università Cattolica del Sacro Cuore, Via E. Parmense 84, Piacenza, Italy; 2 Plant Protection Institute, Hungarian Academy of Sciences, P.O. Box 102, H-1525 Budapest, Hungary. Abstract: Grapevine powdery mildew is a key disease all across Europe and is currently controlled almost exclusively using fungicides. To develop a new biofungicide product against this disease, caused by Erysiphe necator, a large culture collection of Ampelomyces strains collected worldwide from different powdery mildew species was screened to assess for the first time the level of variability of different characteristics of these biocontrol agents known to occur as natural mycoparasites of powdery mildews in the field. This preliminary screening revealed a considerable diversity in many characteristics of different Ampelomyces strains including both culture patterns and mycoparasitic activities against the asexual and sexual stages of E. necator. The screening protocol developed in this work could be used in further studies on Ampelomyces in order to develop new effective commercial biofungicide products against powdery mildew infections of grapevine and other crops. Key words: grapevine powdery mildew, Ampelomyces sp., biological control Introduction Grapevine powdery mildew, caused by Erysiphe necator Schwein. (syn. Uncinula necator Schw. Burr.), is one of the major problems in grapevine production worldwide. During the past decades, its control required one of the largest amounts of fungicides used in the European plant protection practice (Hewitt 1998). In the meantime, there has been a growing interest in the biological control of agricultural pests, mainly because of public concerns on the human health and environmental effects of conventional chemical pesticides. Recently, the European Community has issued a new policy on the Sustainable Use of Pesticides that requires the implementation of Integrated Pest Management (IPM) principles and will become mandatory in According to the IPM, pesticides should be used only when there are no other ways of controlling and limiting pest damages in agricultural production (Boller et al., 2004). It is therefore necessary to study new alternatives to control grapevine powdery mildew or, at least, to reduce the amount of fungicides used to control this disease. One of these alternatives could be the development and application of highly efficient biocontrol agents (BCAs) against E. necator. Pycnidial fungi belonging to the genus Ampelomyces, known to act as natural intracellular mycoparasites of a large number of powdery mildew species worldwide (Kiss et al. 2004; Kiss 2008), have long been studied as BCAs of E. necator (Falk et al. 1995a,b; Hofstein et al. 1996; Angeli et al. 2009) and other, economically important powdery mildews (e.g., Jarvis & Slingsby 1977; Sundheim 1982; Sztejnberg et al. 1989; Philipp et al. 1990; Lee et al. 2004). An Ampelomyces strain has been commercialized in the 1990s as the active compound of a biofungicide product registered in the USA as AQ10 Biofungicide to be used in the control of grapevine powdery mildew and a few other powdery mildew diseases of different crops (Whipps and Lumsden, 2001). Later on, AQ10 Biofungicide has also been 149

170 150 registered in the EU. More recently, another Ampelomyces strain has been developed as Q-fect WP biofungicide in Korea (Lee et al., 2004) and another one has been started to be produced in India as Stanes Bio-Dewcon (S. Rarmarethinam, pers. comm). The development of all these biofungicide products, similar to other previous studies on the biocontrol potential of Ampelomyces mycoparasites, was based on the use of only one or a few strains of Ampelomyces (Falk et al. 1995a,b; Sztejnberg et al. 1989; Lee et al. 2004). Nevertheless, considerable genetic differences were found among Ampelomyces strains (Kiss & Nakasone 1998; Sullivan & White 2000; Szentiványi et al. 2005; Nischwitz et al. 2005; Liang et al. 2007; Angeli et al. 2009) which might indicate that various Ampelomyces strains differ from each other in their mycoparasitic activities and other characteristics that determine their biocontrol abilities. Thus, a comparison of the in vitro and in planta characteristics of a large number of Ampelomyces strains might reveal important differences in the biocontrol potential of various strains and might lead to the selection of new strains to be developed as commercial products. The main goals of this work were to (i) develop a protocol for an efficient screening of Ampelomyces strains based on their potential use as BCAs against powdery mildew on grapevine, and (ii) determine wether there is a considerable variation in the in vitro and the in planta characteristics in a large culture collection of Ampelomyces strains which are important for their use in biocontrol. Material and methods Mycelium growth and sporulation rate More than 150 Ampelomyces strains belonging to a culture collection of >700 strains of Ampelomyces collected worldwide from many different powdery mildew species and preserved at the Plant Protection Institute of the Hungarian Academy of Sciences, Budapest, Hungary, were initially subcultured on Czapek-Dox agar supplemented with 2% malt (MCzDA) and checked visually for their growth and sporulation abilities. Based on this preliminary comparison, 30 strains were selected for more detailed studies. These were grown in 6cm diameter Petri plates on MCzDA. After two weeks of incubation at 20 o C, mycelial disks of 1cm diameter were cut from these plates and placed in the middle of other plates containing MCzDA. The newly inoculated plates were kept at different temperature regimes for 1 month; three replicate plates were used for each temperature regime. Pictures of the colonies were periodically taken and electronic images were evaluated to quantify mycelial growth and the areas of sporulating mycelia using a computerised image-analyser. To determine conidial production, mycelial disks of 1cm diameter were cut from 2-week old colonies, put in 1ml of sterile water, shaken for 1 min and the concentrations of the obtained suspensions were measured using a Bürker haemocytometer. Germination of conidia Conidial suspensions prepared as described above were distributed in Petri plates containing 1.5% water agar (WA) and incubated at different temperatures for 24 hours. Percentages of germinated conidia were then determined under a light microscope. Mycoparasitic activity against the asexual stage of grapevine powdery mildew Detached grapevine leaves with sporulating powdery mildew colonies were inoculated with conidial suspensions of Ampelomyces strains and incubated in moist chambers to maintain leaf turgidity. After 10 days, the leaves were examined under a stereomicroscope to check the presence of intracellular pycnidia in powdery mildew conidiophores.

171 151 Mycoparasitic activity against the sexual stage of powdery mildew Potted grapevines infected with powdery mildew were sprayed with conidial suspensions of each Ampelomyces strain tested. After 12 days of incubation in cabinets with controlled environment, the leaves were detached and chasmothecia for each stage of maturation were collected under the stereomicroscope and checked under the light microscope in order to determine their parasitism by Ampelomyces. Results and discussion Development of a screening protocol Initially, culture characteristics of more than 150 Ampelomyces strains were compared on MCzDA. Many strains did not sporulate at all, i.e. did not produce pycnidia containing conidia in culture. Those developing pycnidia and, thus, exhibiting sporulating areas of their colonies on MCzDA were selected for further studies. Conidial production was determined in these strains as described above. Their conidial suspensions were used to test the production and spread of intracellular pycnidia in the asexual and sexual stages of grapevine powdery mildew, i.e. in the conidiophores and immature chasmothecia of E. necator. During these comparative studies, a comprehensive screening protocol was developed (Figure 1) which can serve as a basis for a more detailed comparison of those in vitro and in planta characteristics of various Ampelomyces strains that are important for their potential use as BCAs of E. necator. Figure 1. Screening protocol for the comparison of those in vitro and in planta characteristics of various Ampelomyces strains that are important for their potential use as BCAs of E. necator.

172 152 Variability of Ampelomyces strains All the comparisons of the growth and sporulation and conidial germination rates on MCzDA showed a considerable variability of the strains based on these characteristics. Similarly, the values determined during the studies of mycoparasitic activities of the different strains in the asexual and sexual stages of grapevine powdery mildew demonstrated the existence of important differences among them. (Figure 2). Colony growth n=30 Sporulating area 21 Production of pycnidia 17 Production of conidia 16 Conidial germination 11 Spread in pm Standardized value Figure 2. Box and Whisker plot showing variability among Ampelomyces strains (number of strains is n = 30 to 6) for colony growth (mm 2 /day), sporulating area (mm 2 /day), production of pycnidia (number/cm 2 ) and conidia (number/ml), conidial germination (%) and spread ability in powdery mildew (pm) colonies (mm 2 ); values were standardized as follows: (value mean)/ mean. Boxes show the 25 th and 75 th percentiles, the whiskers extend to the extremes while circles are outliers; the dotted line is the median. This was expected in the light of the genetic variability described in Ampelomyces (e.g., Sullivan & White 2000; Szentiványi et al. 2005; Liang et al. 2007; Angeli et al. 2009) and it is rather surprising that previous studies on the biocontrol capacities of various Ampelomyces strains did not focus on the diversity of these potential BCAs. In conclusion, this work provided a first insight into the level of variability of different characteristics of Ampelomyces strains isolated from many different powdery mildew fungi worldwide. We have shown that there is indeed a considerable diversity in many characteristics of different Ampelomyces strains including both culture patterns and mycoparasitic activities against the asexual and sexual stages of E. necator. A more detailed description of these characteristics will be provided elsewhere. The screening protocol developed in this work could be used in further studies on Ampelomyces in order to develop new effective commercial biofungicide products against powdery mildew infections of grapevine and other crops.

173 153 Acknowledgements This work was carried out within the EU FP7 project no (acronym: BCA_grape ), aimed at developing new biocontrol agents for powdery mildew on grapevines. An Hungarian- Italian bilateral project (IT-33/07) has also supported a part of the work. References Angeli, D., Pellegrini, E. & Pertot, I. 2009: Occurrence of Erysiphe necator chasmothecia and their natural parasitism by Ampelomyces quisqualis. Phytopathology 99: Boller, E. F., Avilla, J., Joerg, E., Malavolta, C., Wijnands F. G. & Esbjerg, P. 2004: Integrated Production Principles and Technical Guidelines, 3 rd Edition, IOBC/WPRS Bull. 27(2): Falk, S. P., Gadoury, D. M., Cortesi, P., Pearson, P. C. & Seem, R. C. 1995a: Parasitism of Uncinula necator cleistothecia by the mycoparasite Ampelomyces quisqualis. Phytopathology 85: Falk, S. P., Gadoury, D. M., Pearson, P. C. & Seem, R. C. 1995b: Partial control of grape powdery mildew by the mycoparasite Ampelomyces quisqualis. Plant Disease 79: Hewitt, H. G. 1998: Fungicides in crop protection, CABI Publishing; Wallingford, UK. Hofstein, R., Daoust, R. A. & Aeschlimann, J. P. 1996: Constraints to the development of biofungicides: the example of AQ-10, a new product for controlling powdery mildews. Entomophaga 41: Jarvis, W. R. & Slingsby, K. 1977: The control of powdery mildew of greenhouse cucumber by water sprays and Ampelomyces quisqualis. Plant Disease Reporter 61: Kiss, L., Russell, J. C., Szentiványi, O., Xu, X. & Jeffries, P. 2004: Biology and biocontrol potential of Ampelomyces mycoparasites, natural antagonists of powdery mildew fungi. Biocontrol Science and Technology 14: Kiss, L. & Nakasone, K. K. 1998: Ribosomal DNA internal transcribed spacer sequences do not support the species status of Ampelomyces quisqualis, hyperparasite of powdery mildew. Current Genetics 33: Kiss, L. 2008: Intracellular mycoparasites in action: interactions between powdery mildew fungi and Ampelomyces. In: Stress in Yeasts and Filamentous Fungi, eds. Avery, Stratford and van West: Academic Press, Elsevier. London, UK. Lee, S. Y., Lee, S. B. & Kim, C. H. 2004: Biological control of powdery mildew by Q-fect WP (Ampelomyces quisqualis 94013) in various crops. IOBC/WPRS Bull. 27(8): Liang, C., Yang, J., Kovács, G. M., Szentiványi, O., Li, B., Xu, X. X. & Kiss, L. 2007: Genetic diversity of Ampelomyces mycoparasites isolated from different powdery mildew fungi in China inferred from analyses of rdna ITS sequences. Fungal Diversity 24: Nischwitz, C., Newcombe, G. & Anderson, C. L. 2005: Host specialization of the mycoparasite Eudarluca caricis and its evolutionary relationship to Ampelomyces. Mycological Research 109: Philipp, W. D., Beuther, E., Hermann, D., Klinkert, F., Oberwalder, C., Schmidtke, M. & Straub, B. 1990: Zur Formulierung des Mehltauhyperparasiten Ampelomyces quisqualis Ces. Zeitschrift für Pflanzenkrankheiten und Pflanzenschutz 97: Sullivan, R. F. & White, J. F. Jr. 2000: Phoma glomerata as a mycoparasite of powdery mildew. Applied and Environmental Microbiology 66: Sundheim, L. 1982: Control of cucumber powdery mildew by the hyperparasite Ampelomyces quisqualis and fungicides. Plant Pathology 31:

174 154 Szentiványi, O., Kiss, L., Russell, J. C., Kovács, G. M., Varga, K., Jankovics, T., Lesemann, S., Xu, X. M. & Jeffries, P. 2005: Ampelomyces mycoparasites from apple powdery mildew identified as a distinct group based on single-stranded conformation polymorphism analysis of the rdna ITS region. Mycological Research 109: Sztejnberg, A., Galper, S., Mazar, S. & Lisker, N. 1989: Ampelomyces quisqualis for biological and integrated control of powdery mildews in Israel. Journal of Phytopathology 124: Whipps, J. M. & Lumsden, R. D. 2001: Commercial use of fungi as plant disease biological control agents: status and prospects. In: Fungi as Biocontrol Agents: Progress, Problems and Potential, eds. Butt, Jackson and Magan, CABI Publishing, Wallingford, UK: 9-22.

175 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Water activity at the fruit surface: a potential indicator of grape berry susceptibility to Botrytis cinerea M. Fermaud 1, C. Deytieux-Belleau 2, J. Roudet 1, G. Darrieutort 1, L. Geny 2 1 INRA, UMR Santé Végétale, ISVV Bordeaux-Aquitaine, Villenave d Ornon, France, 2 Université de Bordeaux, ISVV Bordeaux-Aquitaine, UMR1219 œnologie, Faculté d oenologie-inra, F Villenave d Ornon, France. fermaud@bordeaux.inra.fr Abstract: Water activity (Aw) is a physical property of a humid solid, which is of prime importance in food microbiology. This parameter is an assessment of available water, unbound to nutrient molecules, which allows damaging bacteria, yeasts and fungi to grow in a food product. As most moulds which do not grow below an Aw level of ca. 0.8, Botrytis cinerea was unable to grow on artificial media at an Aw of 0.93 and below. In 2008, in an experimental vineyard (Vitis vinifera L. cv. Merlot noir) near Bordeaux, Botrytis bunch rot development was assessed and the evolution of water activity was investigated at the surface of healthy grape berries during fruit development. Our results showed that Aw decreased steadily from the berry herbaceous stage to maturity and was correlated negatively and significantly (P = 0.01) with disease incidence. Furthermore, an experiment under controlled conditions showed the significant effect of relative humidity (RH) on the water activity level at the berry surface. When healthy berries at two different development stages ( herbaceous and beginning of colour change ) remained at 80% RH for 6 days, the Aw values decreased significantly (P = 0.05). On the basis of the close relationship between water activity and Botrytis bunch rot development, the potential of Aw for use as a new disease indicator is discussed as well as the influence of climatic conditions on water activity measured at the grape berry surface. Key words: grey mold, integrated control, ontogenic resistance, water availability Introduction Grey mould or Botrytis bunch rot, caused by the phytopathogenic fungus Botrytis cinerea Pers.:Fr., is one of the main aerial diseases in grapevine (Vitis vinifera L.) because it can reduce drastically both yield and wine quality. Epidemic progress on fruit is driven by various factors including notably i) the environmental conditions, particularly climate and microclimate (Fermaud et al., 2001); ii) physiology and vegetative growth of the grapevine host plant (Valdés-Gómez et al., 2008); iii) the biochemical and mechanical properties of the berry skin (Deytieux-Belleau et al., 2009). The mechanical properties may be affected by the presence of micro-cracks or pores (Mlikota-Gabler et al., 2003) as well as macroscopic injuries, such as those caused by insects allowing the fungus to penetrate into the host. Among the berry skin features, water activity (Aw) at berry surface is a critical factor affecting both in vitro growth and metabolism of fungi, yeasts and bacteria (Rousseau and Donèche, 2001). For B. cinerea, spore germination and fungal growth are induced in vitro by a high free water level associated with a high Aw value (Lahladi et al., 2007). More recently, the relationship between Aw and susceptibility of the grape berry to B. cinerea has been established using the cultivar Sauvignon blanc (Deytieux-Belleau et al., 2009). In the present study, similar data have been further investigated to confirm the results on the black susceptible cultivar Merlot noir. 155

176 156 The first objective of this work was to examine the temporal variations in Aw measured at the surface on healthy berries according to the fruit developmental stage. The second goal was to establish a relationship between the Aw evolution and the disease development in the vineyard. Lastly, the potential effect of ambient relative humidity (RH) on Aw at different developmental stages of the grape berry was also investigated. Material and methods Experimental vineyard, disease and maturity assessments In 2008, grape berries (cv. Merlot noir) were collected from an INRA experimental vineyard near Bordeaux (France) from the end of July (herbaceous stage) to over-maturity at the beginning of October. The vineyard was planted in 1991 at the rate of ca vines per ha with a traditional training system. The experimental vine plot was not treated with anti- Botrytis fungicides during the growing season. Disease development was assessed regularly by visual assessments of the percentage of the cluster surface showing typical rot symptoms with B. cinerea sporulation. At each assessment date, all the clusters from 7 blocks, each consisting of 5 adjacent vines, were sampled and rated visually for disease incidence, i.e. a total of approximately 280 clusters. Water activity (Aw) and maturity monitoring For maturity assessments, three samples of 100 cut fresh berries were randomly selected at each date and processed independently according to Deytieux-Belleau et al. (2009). As for Aw measurements, every week, a random sample of 10 berries, all being apparently healthy, was used to measure Aw according to Deytieux-Belleau et al Table 1. Experimental conditions to test the effect of relative humidity on water activity Experiment Berry stage Sampling date and first Aw assessment DAA (days after anthesis) Berry no. Incubation period (day) at 80% or 100% R.H. 1 Herbaceous 07/18/ Beginning of 08/05/ veraison 3 Overmaturity 10/17/ In order to test whether Aw and relative humidity (RH) may interact at the berry surface, 8 fresh and healthy grape berries (pedicel attached) were randomly sampled at three development stages (Table 1). Beforehand, the pedicel was surrounded with paraffin to avoid water exchanges and the Aw-meter was calibrated using a solution of K 2 SO 4 (Aw = at 25 C). To measure Aw at the fruit surface, each berry was placed in the chamber of a GBX Aw-meter (model FA-st/1 ) for ca. 1 hour at room temperature (between 25 and 29 C). The berries were then placed under controlled conditions for 3 to 6 days, using climatic chambers (EX-111; TABAI ESPEC Corp., Japan), at a constant temperature of 20 C in the darkness. At

177 157 each berry stage, half of the berries (n = 4) were maintained at 100% RH whereas the other 4 berries were placed at 80% RH. After the RH-treatment, Aw at the berry surface was measured again as described above. The experimental conditions are summarized in Table 1. Data were analysed by ANOVA, followed by the test of Newman-Keuls for mean comparison, and based on two main factors: the phenological stage and the RH treatment. Results and discussion Climatic conditions and grey mould development In 2008, the climatic conditions were conducive to Botrytis bunch rot as shown by frequent precipitations (Figure 1). Under our experimental conditions on cv. Merlot noir and without any anti-botrytis fungicide, the epidemic progress curve was characterized by an incidence of Botrytis bunch rot that increased regularly from mid-veraison onwards (Figure 2). At the end of September, i.e. just after maturity, the disease incidence reached ca. 100% of rotted clusters. 3d_rainfall (mm) 3d _ mean temp ( C) 3d_RH<40% duration (hrs) Jun 16-Jun 1-Jul 16-Jul 31-Jul 15-Aug 30-Aug 14-Sep 29-Sep 14-Oct Figure 1. Climatic conditions in 2008 corresponding to the experimental vineyard (Villenave d Ornon, France). Means are calculated from three consecutive days for air temperature ( C), rainfall intensity (mm) and duration of relative humidity RH<40% (no. hrs per day) Grape berry maturity evolution As shown in Figure 3, grape berry maturity increased regularly during the season. At maturity, the maturity index reached ca. 50 which was calculated as the following ratio: sugar content (g/l) divided by total acidity (g H 2 SO 4 /l).

178 Grey mould Incidence (%) mid-veraison Maturity Aug 20-Aug 30-Aug 9-Sep 19-Sep 29-Sep Figure 2. Epidemic progress curve of Botrytis bunch rot in 2008 (cv. Merlot noir) on the experimental vineyard (Villenave d Ornon, France). Incidence is the percentage of diseased clusters assessed from mid-veraison midv to the maturity M stage. Sugar content / total acidity H y = x x R 2 = /25/08 (206) Day of the Year in 2008 (DOY) 09/29/08 (272) Figure 3. Evolution of grape berry maturity (cv. Merlot noir) shown using the maturity index Sugar / Total acidity. Sugar content is expressed in g/l and total acidity in g H 2 SO 4 /L. Grape berry stages: H : herbaceous immature berries, M : maturity. M Evolution of water activity at the berry surface Water activity at the berry skin surface decreased progressively during grape berry development from ca to a minimum of ca on 22 September (Figure 4). As a general trend, Aw decreased linearly up to maturity (112 DAA). However, in over-mature grape berries, Aw increased at the end of the season (from 22 September onwards).

179 159 Water activity Aw jul 08 y = x R 2 = Oct DAA Figure 4. Variations in water activity at the berry surface in 2008 on cv. Merlot noir. (± standard deviation: error bars). Relationship between disease incidence and water activity Our results showed that Aw was correlated negatively with disease incidence. The linear relationship was highly significant (P < 0.01, Pearson s R = 0.96; ddf = 6) (Figure 5). Botrytis Bunch rot incidence (%) 100 y = x R 2 = Water activity Aw Figure 5. Significant correlation and linear relationship between Aw and incidence of botrytis bunch rot Effect of relative humidity on water activity at the berry surface At the surface of grape berries, cv. Merlot noir, the effect of RH on Aw, is showed in Table 2. The treatment consisting of a period of 3 to 6 days at 100% RH did not affect significantly Aw at the berry surface (P = 0.57). On the other hand, the treatment of the berries at 80 % RH resulted in a significant (P < 0.001) decrease in Aw at both herbaceous and beginning of colour change berry stages. However, at the over-maturity stage, no effect was detected as shown by the second ANOVA. This resulted from a significant interaction (P = ) between the two main factors tested.

180 160 Table 2. Effect of relative humidity on water activity, assessed at the surface of healthy grape berries (cv. Merlot noir), according to the berry developmental stage. Berry stage ANOVA 1 ANOVA 2 Control a HR 100% b Control a HR 80% b Herbaceous A A a c Beginning of colour change A A b d Over-maturity B B e e a Aw assessed at the beginning of the experiment b Aw assessed at the end of the treatment period by HR Conclusion Water activity Aw at the berry surface is considered as a possible major determinant governing grey mould development in the vineyard. These results on the black cultivar Merlot noir in 2008 strengthen those obtained in 2006 on the susceptible white cultivar Sauvignon (Deytieux-Belleau et al., 2009). Therefore, besides fruit maturity, Aw might be used to estimate the evolution of the intrinsic berry susceptibility to infection by B. cinerea. The temporal pattern was a decrease in water activity until berry maturity which may be related to an increasing amount of solutes or exudates (exosmosis) at the berry surface. Aw decreased presumably because, at the fruit surface, solutes bind with free water reducing, then, available water. In this context, we have established a clear linear regression between water activity and disease incidence in the vineyard. These findings indicated that the measurement of water activity could constitute a good indicator to characterize grape susceptibility from bunch closure to fruit maturity. However, because several environmental factors may also affect the development of B. cinerea, extrapolation of these results to various other natural situations; i.e. other grapevine growing regions, cultivars and/or years, should be made with caution. Acknowledgements This study was supported by a research grant from the Conseil Interprofessionnel du Vin de Bordeaux (CIVB). The authors are also indebted to P. Cartolaro, F. Daguisé and J.M. Brustis (UMRSV, Bordeaux) and A. L Hyvernay for their technical participation to this work. References Deytieux-Belleau, C., Geny, L., Roudet, J., Mayet, V., Donèche, B., & Fermaud, M. 2009: Grape berry skin features related to ontogenic resistance to Botrytis cinerea. Eur. J. Plant Pathol. 125: (DOI /s ). Fermaud, M., Limiñana, J. M., Froidefond, G. & Pieri, P. 2001: Grape cluster microclimate and architecture affect severity of Botrytis rot of ripening berries. IOBC/WPRS Bull. 24(7): 7-10.

181 Lahlali, R., Serrhini, M. N., Friel, D. & Jijakli, M. H. 2007: Predictive modeling of temperature and water activity (solutes) on the in vitro radial growth of Botrytis cinerea Pers. International Journal of Food Microbiology 114: 1-9. Mlikota-Gabler, F., Smilanick, J. L., Mansour, M., Ramming, D. W. & Mackey, B. E. 2003: Correlation of morphological, anatomical, and chemical features of grape berries with resistance to Botrytis cinerea. Phytopathology 93: Rousseau, S. & Donèche, B. 2001: Effects of water activity (Aw) on the growth of some epiphytic micro-organims isolated from grape berry. Vitis 40: Valdés-Gómez, H., Fermaud, M., Roudet, J., Calonnec, A. & Gary, C. 2008: Grey mould incidence is reduced on grapevines with lower vegetative and reproductive growth. Crop Protection 27:

182 162

183 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Investigation of contamination pathways of Esca-associated fungal pathogens and of fungicide-based control strategies A. Kortekamp, J. Köckerling DLR Rheinpfalz, Phytomedicine, Breitenweg 71, D Neustadt/Weinstrasse, Germany Abstract: Even though Esca may be as old as viticulture itself, new attention has been directed to this disease in recent years due to the dramatic increase in economic importance. A sudden wilting of Esca-affected grapevines, followed by the death of the entire plant, occurs in many grapevine-growing areas of the world. Esca is a complex disease that comprises several symptoms caused by a set of fungal pathogens. Unfortunately, infected plants can often not be cured, since foliar or wood treatments using fungicides do not lead to an effective control of the respective pathogens in the infested wood. Given the large number of wounds made to propagation material during the various nursery stages and also made as a result of repeated pruning, hygiene and wound protection is of particular importance. Therefore, the effect of various chemical and biological treatments on Escaassociated fungi, such as Phaeoacremonium aleophilum (Pal), Phaeomoniella chlamydospora (Pch), and Fomitiporia mediterranea (Fmed), and on the disease progress of artificially infected plants was investigated in vitro and in field and greenhouse experiments. For this purpose, several putative contamination pathways and different test systems have been evaluated to verify the ports of entry for fungal spores and the efficiency of fungicides to control the disease. In the year 2009, vineyards of the DLR research station and of local grape growers were inspected and the impact of Esca in these vineyards was evaluated. As published by other authors, the number of affected vines increased with their age. Furthermore, there are no clear differences in susceptibility and none of the varieties planted in the Palatinate showed any suitable resistance. Plant material of diseased vines was collected and tested for the presence of Esca-inducing pathogens. Mainly Phaeomoniella chlamydospora (Pch), Botryosphaeria species, Cylindrocarpon sp., Trichoderma sp. and mould fungi such as Alternaria sp. and Penicillium sp. were frequently found. All type of wounds created artificially represented suitable ports of entry for all Esca-associated pathogens. Especially, infections with Cylindrocarpon destructans significantly reduced viability of the scions and had severe effects on plant growth as tested with infected cuttings. Wound infections caused by polluted water that had been used to soak propagation material led to a reduced growth of cions, especially when contaminated with Cylindrocarpon spores. Growth tests in vitro showed that most fungicides tested exhibited antifungal capacities. The Trichoderma-based product Trichostar was an effective biological agent in the field, preventing infection of wounds when applied one day before inoculation, whereas most fungicides were not able to sufficiently prevent colonisation of wounds by fungi. Key words: Fomitiporia mediterranea, grapevine, Palatinate, Phaeoacremonium aleophilum, Phaeomoniella chlamydospora, Vitis vinifera Introduction In recent years it has become evident that grapevine plants suffer increasingly from Esca. Esca is a complex disease that comprises several symptoms caused by a set of fungal pathogens (Mugnai et al., 1999). Unfortunately, infected plants can often not be cured, since foliar or wood treatments using fungicides do not lead to an effective control of the respective 163

184 164 pathogens in the infested wood. Furthermore, the term Esca seem to be frequently misused and several putatively diverse diseases are combined to an Esca disease-complex (Surico, 2009). Since different pathogens may be involved in Esca or Esca-associated diseases, it seems to be unlikely that just one fungicide, disinfectant or grafting clay used for wound protection or any single control strategy will play the dominant role in disease prevention. The ports of entry may depend on the pathogen that can infect a grapevine plant at the very first steps of propagation or in the field when the plant reaches a critical age or status. However, given the large number of wounds made to propagation material during the various nursery stages and also made as a result of repeated pruning, hygiene and wound protection seem to be of particular importance. Even though much work was done to identify causal organisms, to validate their occurrence, frequency, dispersal, and spreading in the plant (especially in the wood), and to protect plants from Esca, there are still many open questions. Therefore, the aim of the presented work was (1) to evaluate disease frequency in vineyards located in the Palatinate, (2) to evaluate differences in susceptibility towards Esca in several grapevine varieties, (3) to identify fungal organisms isolated from diseased wood in infected plants, (4) to identify putative contamination pathways, and (5) to test the efficiency of fungicides in vitro and in the field. Material and methods More than 100 vineyards in the Palatinate with an average area of 15-20a were inspected to evaluate the infestation with Esca. For this purpose, plants with symptoms typical for Esca were counted and expressed as the percentage of diseased or dead vines, respectively. Plant material of diseased vines was collected and tested for the presence of Escainducing pathogens. Parts of the shoots were decorticated, surface sterilised, placed on agar plates containing 2% malt extract, and incubated at 21 C in a climate chamber for several days. Fungal species were characterised by their colony features and spore types. A total of 120 samples were investigated. In order to investigate putative contamination pathways, shoots of infested mother vines showing symptoms of Esca were placed in a water bath and incubated over night. Aliquots of the bath water were removed, placed on agar plates and incubated as described above. In addition, bath water was contaminated with fungal spores using spore suspensions ( spores/ml) of Phaeomoniella chlamydospora (Pch) and Cylindrocarpon destructans. Noninfected shoots were incubated in the contaminated water bath over night, potted in Perlite, and incubated in the greenhouse at 25 C. Growth and root production was evaluated 20 weeks post inoculation. The effect of Esca pathogens on growth and root production was also investigated by dipping freshly created cuttings in spore suspensions of Pal, Pch, and Cylindrocarpon. These cuttings were potted into Perlite and incubated in the greenhouse for 6 weeks. Evaluation of root growth was done after removing all cuttings 6 weeks post inoculation. Summer pruning was simulated using greenhouse plants. Branches were removed from the main shoot and the freshly created wounds were inoculated with agar plugs (diameter 5mm) containing mycelium of Pal (Phaeoacremonium aleophilum), Pch, and Fmed (Fomitiporia mediterranea). Inoculation spots were sealed with Parafilm and entire plants were incubated at 25 C for 24 weeks in the greenhouse. Then, inoculated parts of the plant were removed and placed on agar plates as described above after an evaluation of the vascular browning. Growth of the respective pathogen was analysed by the outgrowing mycelium.

185 165 Number of affected vinyards Number of affected vineyards Diseased vines with symptoms (in %) Dead vines (in %) Figure 1. Evaluation of Esca in Palatinate vineyards. Vineyards of the DRL research station and of grape growers had been investigated. Field experiments were conducted in 2007, 2008, and 2009 to evaluate the efficiency of some commercially available fungicides and a biological control agent (Trichoderma harzianum, Trichostar, Gerlach GmbH, Hannover, Germany). Pruning wounds were treated with the respective product and 24 hours later artificially inoculated with Esca pathogens (Pal, Pch). Treated and inoculated wounds were protected from rain or desiccation for one week using 2.0ml Eppendorf-tubes placed at the tops of the shoots. Inoculated shoots were harvested 7 months post inoculation and analysed as described above. The efficiency of commercially available fungicides to control spore germination and mycelial growth was also tested in vitro on agar plates containing malt extract amended with a selected fungicide. Each plate was inoculated at the centre with a fresh culture of the respective fungus and incubated at 21 C. Measurements of the colony diameter were taken 18 days after inoculation. Germination of spores on supplemented media was also determined on

186 166 fungicide-containing malt extract agar. Spores were harvested from control plates without additives. The concentration of spore suspension was adjusted to provide 10 5 spores per plate. Germinated spores were counted on randomly selected agar plugs that had been removed from test plates after incubation for 48 hours at 21 C. Survival after exposure was determined as a percentage of spores germinated on control plates. Results and discussion Assessment of disease in vineyards located in the Palatinate In the year 2009, vineyards of the DLR research station and of local grape growers were inspected and the impact of Esca in these vineyards was evaluated. For this purpose, the percentage of those grapes exhibiting typical symptoms such as tiger-striped leaves, black measles or sudden death (apoplexy) was determined. The results revealed that all vineyards investigated have diseased plants (Figure 1). In most cases, the percentage of diseased plants is less than 5% (Figure 1a). The percentage of dead vines that died during the last season or the last 2 to 3 years due to Esca was also about 5% in most cases (Figure 1b). However, some vineyards exhibit up to 50% or even more affected plants and in some plantations a majority of vines (up to 90%) were dead due to Esca. In accordance to the literature (Romanazzi et al., 2009), the number of affected vines increased with their age. Especially those vineyards planted in the 70s or 80s exhibit high levels of affected or even dead plants. Thus, the number of vines that had to be removed and replanted or that represent a permanent source of inoculum increased during time. Since grape growers intend to use plantations over a period of 35 or 40 years, the high number of diseased plants leads to a reduction of yield and quality and makes also a replacement of affected plants necessary. Both, reduced crop and increased costs are in contrast profitable to viticulture. Susceptibility of grapevine varieties towards Esca Very little is known about the different behaviour of cultivated grape varieties towards Esca and therefore information in literature leading to a recommendation to use suitable and less susceptible varieties is still missing. Since infection studies take a long time beginning from primary infection(s) to visible symptoms or the death of the entire plant, the susceptibility of varieties grown in the Palatinate was only estimated. For this purpose, the number of affected vineyards planted with a distinct variety (expressed as the percentage of indication) was compared with the total area occupied by this specific variety. As expected, there are no clear differences in susceptibility and none of the varieties showed any suitable resistance with two exceptions: The Pinots (Pinot blanc, Pinot gris, Pinot noir) seem to be less susceptible compared to other varieties. This seems also to be the case for Regent, a relatively new bred variety with partial resistance against downy and powdery mildew. However, Regent is quite recently used as a new variety to produce red wines and thus may not have reached its critical age to express typical Esca symptoms. In contrast to the situation described for Regent and the Pinots, there is increasing evidence that some varieties such as Riesling seem to be more susceptible towards Esca. Isolation of fungal organism from diseased wood of infected plants Since Esca is a disease caused by several distinct fungal pathogens, the expression of symptoms, especially in wood, may depend on the fungal community found in the affected plant. Mainly Phaeomoniella chlamydospora (Pch), Phaeoacremonium aleophilum (Pal) and Fomitiporia mediterranea (Fmed) are considered as being the most important and most

187 167 destructive pathogens in grape wood leading to vascular symptoms (tracheomycosis) and white rot in wood (Fischer & Kassemeyer, 2003). However, many other fungi were isolated (Figure 2). Among these, especially Botryosphaeria species, Cylindrocarpon sp., Trichoderma sp. and mould fungi such as Alternaria sp. and Penicillium sp. were frequently found. 16 Relative frequency (in %) Alternaria Aspergillus Fusarium Penicillium Botrytis Cladosporium Phomopsis Trichoderma Cylindrocarpon Botryosphaeria Pal Pch Fmed Other Figure 2. Relative frequency of fungi putatively pathogenic to grapevine found in Escadiseased vines in the Palatinate ( ). Even though Pal, Pch, and Fmed are important pathogens and in most cases responsible for the disease outbreak (Valtaud et al. 2009), the role of other putatively pathogenic fungi is not clear. Some of them are quite common and may represent secondary pathogens only capable to live and to propagate in weakened tissues, whereas others are well known for their ability to infect vines. Putative contamination pathways For both types of fungi, common species and pathogens, it remains nebulous how they are able to enter a plant and the wood of a trunk. Botryosphaeria sp. was frequently found in one or two year old wood and is thus able to infect freshly created wounds and to survive in such young plant tissues. This pathogen enters the plant obviously via wounds created either in summer (removal of unwanted shoots) or during winter (pruning). On the other side, Cylindrocarpon sp. (mainly C. destructans) infects the roots and seems to migrate into the rootstock or even the trunk up to the point of grafting. Trichoderma species are mainly known for their antagonistic capacities and antimicrobial features. Therefore, they had been often tested for their ability to suppress pathogenic bacteria and fungi on different hosts. Even though nothing is known about the antimicrobial effect of Trichoderma species in grape wood, they seem not to be able to inhibit infections with other fungi or to sufficiently delay their proliferation. However, first tests with T. harzianum leading to preliminary results revealed that commercial products used for biological control may have a place in control strategies in the near future. First results indicated that T. harzianum was able to inhibit fungal infections by Esca pathogens in 80% of all inoculations. It is assumed that wounds created during propagation and pruning are the most important sites where an infection takes place.

188 168 Therefore, it was tested if Esca-associated pathogens are able to enter a plant at distinct steps during propagation and grafting and if infected shoots may contaminate healthy plant material. For this purpose, shoots of infected mother vines that exhibited typical Esca symptoms were placed over night in a soaking bath to provoke infections in noncontaminated shoots. Fortunately, Esca-associated pathogens were all absent in the soaking bath and incubated samples. Thus, shoots of infected mother vines seem not to increase the risk of contamination, even though such infested material should not be used for propagation. A second experiment was conducted to evaluate the probability of infection by spores that had been added to the bath water. An inoculation with the pathogen Pch did not affect growth and root production. However, infections with Cylindrocarpon significantly reduced viability of the scions (of about 50%). Since Cylindrocarpon showed such severe effects on plant growth, the influence of Esca pathogens and Cylindrocarpon on growth and root production was determined in another experiment using freshly created cuttings that were dipped into spore suspensions of Pal, Pch, and Cylindrocarpon. The development of inoculated cuttings was investigated 6 weeks post inoculation. All pathogens were successfully and frequently reisolated from infection sites and reduced root production. Interestingly, infections with Cylindrocarpon resulted in a die back of all plants 20 weeks post inoculation. Since wounds created during the year may increase the probability of an infection with pathogens leading to Esca, summer pruning was simulated by creating pruning wounds on greenhouse plants. The freshly created wounds were either inoculated with mycelium (agar plugs) or spores (spore suspension), sealed with Parafilm, and incubated in the greenhouse for 24 weeks. Both, growth of the pathogens and discolouration of the vascular tissue were investigated. The two pathogens Pal and Pch were successfully isolated from inoculated wounds in most cases (Pal 49%, Pch 79%). Furthermore, both pathogens were able to spread within the inoculated shoot and grew up to 2cm upwards and downwards from the inoculation spot. However, it was not possible to reisolate Fmed from young inoculated plant tissues and also controls were not infected with pathogens. Efficiency of fungicides in vitro and in the field In order to test the effect of fungicides on mycelial growth and spore germination, an in vitro test using agar plates was performed. Most fungicides (12 of 15 tested) currently in the market were effective and controlled both spore germination and mycelial growth significantly (degree of efficiency 75 to 100%). However, all these fungicides showing great promise for the control of Esca-inducing pathogens such as Pal and Pch were not effective in the field, except for the Trichoderma product and some test compounds that are currently under investigation. Time will tell if just one of them has the capability to sufficiently protect vines from Esca. Acknowledgements We would like to thank Adelinde Andrae, Doris Dersch-Fischer, and Rainer Rueff for excellent technical assistance. References Fischer, M. & Kassemeyer, H.-H. 2003: Fungi associated with Esca disease of grapevine in Germany. Vitis 42:

189 Mugnai, L., Graniti, A. & Surico, G. 1999: Esca (black measles) and brown woodstreaking:two old and elusive diseases of grapevine. Plant Dis. 83: Romanazzi, G., Murolo, S., Pizzichini, L. & Nardi, S. 2009: Esca in young and mature vineyards, and molecular diagnosis of the associated fungi. Eur. J. Plant Pathol. 125: Surico, G. 2009: Towards a redefinition of the diseases within the esca complex of grapevine. Phytopathol. Mediterr. 48: Valtaud, C., Larignon, P., Roblin, G. & Fleurat-Lessard, P. 2009: Developmental and ultrastructural features of Phaeomoniella chlamydospora and Phaeoacremonium aleophilum in relation to xylem degradation in esca disease of the grapevine. J. Plant Pathol. 91:

190 170

191 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Impact of biotic and abiotic factors on the development of Esca decline disease P. Lecomte 1, G. Darrieutort 1, C. Laveau 1, D. Blancard 1, G. Louvet 1, J.-P. Goutouly 2, P. Rey 1, L. Guérin-Dubrana 1 1 UMR Santé Végétale INRA-ENITAB, Institut des Sciences de la Vigne et du Vin, Centre INRA Bordeaux Aquitaine, BP 81, Villenave d'ornon cedex, France. 2 UMR 1287, Ecophysiologie et Génomique Fonctionnelle de la Vigne, Institut des Sciences de la Vigne et du Vin, Villenave d'ornon cedex, France. lecomte@bordeaux.inra.fr Abstract: The aim of researches performed by our group is to determine which internal and external factors have a key influence on grapevine wood diseases, especially on Esca. The relationship between Esca leaf symptoms, the size of inner necrosis, the fungal endophytes associated with the disease and the whole microflora that can be detected from the grapevine wood, is discussed. Recent studies on leaf symptom development showed that summer temperature conditions likely play a major role in the expression of Esca symptoms. Cultural factors, like plant vigor or type of soils, also have presumably a significant influence. Several evidences suggest that other abiotic factors linked to the training systems (e.g. forms with very short cordons) or to pruning decisions (e.g. winter hand pruning characteristic of the modern grape-growing training systems) have also to be taken into account, alone or combined with other factors. Whole data indicate that Esca is a multi-factor disease and that many complex pathological scenarios could explain the grapevine trunk disease expression. A control strategy, mainly based on cultural measures to avoid a too early development of inner necrosis, is proposed. Key words: aetiology, cultural preventive measures, integrated control, prophylaxis Introduction Esca is a worldwide syndrome associated with the development of fungal endophytes into the grapevine wood. This major trunk decline disease is found in all European vine-growing areas and lead to entire or partial decline of vines. In France, the National Grapevine Wood Disease Survey has recently reported that the percent of no-productive vines affected by vine trunk diseases (including Eutypa dieback) is about 10% (Grosman, 2008). In the last decade, as the incidence of Esca or Bot cankers seemed to increase regularly in some regions and to appear earlier and more severely in some vineyards, it was even assumed that these grapevine trunk syndromes were emerging diseases. This context has encouraged more investigations. Based on studies in progress or on their own experience, authors of the current paper briefly review the main biotic or abiotic factors that may have an influence on grapevine trunk disease development, particularly on Esca. The final objective is to suggest cultural methods for the preventive control of these insidious diseases. Biotic factors Leaf symptoms of Esca are known to be highly variable (Surico et al., 2006), both in their incidence and in their shape (Lecomte et al., 2006). They concern leaves with various and evolutionary discolorations, interveinal necrosis and wilting of branches and include those 171

192 172 attributed to Black Dead arm (Lecomte et al., 2006, 2008a). The most dramatic symptom is the sudden wilting of entire vines, also called apoplexy. Mild symptoms appear sporadically in most of the vineyards (Mugnai et al., 1999). Leaf symptoms of Esca are generally associated with the presence of wood necrosis and decays into arms and trunks. In French vineyards, the mean percentage of vines showing Esca leaf symptoms is around 4% (Kobès et al., 2006). However, the mean percentage of vines with inner necrosis may reach 100% in some vineyards or mother vines (Dumot et al., 2007; Lecomte et al., 2008; Liminana et al., 2009). This difference between the high level of vines affected by inner necrosis development and those of vines exhibiting leaf symptoms is still a matter of questioning. To better understand the relationship between inner necrosis and leaf expression, various studies have been carried out with vines collected in Bordeaux area. Preliminary results (Guérin-Dubrana et al., 2008; Lecomte et al., 2008b) pointed out that the type of necrosis and/or their size, assessed by image analysis, is correlated with the leaf damages. However, as reported by Calzarano and Di Marco in Italy (2007), the necrosis size is likely not always sufficient to explain the variability of leaf expression of the Esca disease. Many other factors like plant defence efficacy (Gaudillère, 2003; Goutouly, 2007) or the microbial activity are also assumed to play a significant role in the leaf symptom occurrence. Isolations of fungi carried out from Esca-affected trunks showed that necrosis were usually colonised by the same fungal pathogens that can act as wood-degrading agents (Fisher and Kassemeyer, 2003; Larignon & Dubos, 1997; Péros et al., 2008). The aggressiveness of these pathogenic fungi was recently investigated (Laveau et al., 2009) based on pathogenic tests performed into the wood of young cuttings. Results confirmed that all the fungi studied can be associated with necrosis and two of them, Phaeomoniella chlamydospora and Neofusiccocum parvum developed significant canker lesions and inner necrosis. Nevertheless, only one fungus, Eutypa lata, up to now, is capable to provide necrosis and more or less regularly leaf symptoms using this bioassay (Péros & Berger, 1994). Several studies indicate that toxins are likely involved (Sparapano et al., 2000). However, no complete demonstration of the Koch s postulate have been obtained after inoculation of plants by the fungi putatively involved (Sparapano et al., 2001), indicating that conditions inducing the foliar disease expression in the summer remain unclear. Gubler et al. (2005) reported that more than 1300 different species of fungi or bacteria were found to colonize grapevine wounds in spring. There is still a lack of information concerning the role and the activity of the whole microflora colonizing or living into the grapevine wood, either necrotic or apparently healthy. Preliminary results of isolations on culture media done from 2006 (data not shown) in different Bordeaux vineyards also revealed that wood inhabiting parasites can be very diverse, including bacteria and yeast. Therefore the whole microflora that can be identified from the grapevine wood is probably not yet determined. To know more, especially about the inoculum pressure in the apparently healthy wood, we are presently characterizing the trunk-colonizing microflora by cultivable and molecular methods, i.e. the single-strand conformation polymorphism (SSCP) fingerprinting technique and microbial DNA-sequencing. The main advantages of SSCP are that it can be used to detect rapid changes in microbial communities in the absence of prior knowledge about their composition and it avoids the biases introduced by culture-based methods. Abiotic factors In France, local surveys showed that the disease expression can reach percentages higher than 20% or much more in some vineyards (Lecomte et al., 2005; 2008b). Foliar expression, depend on the years, the vineyard areas, the vine-age and the varietal susceptibility factors already well-known (Dubos, 2002; Fussler et al., 2008). Various factors may explain the

193 173 variability of Esca incidence and its sporadic character (Mugnai et al. 1999). Among them, climate is often the first abiotic factor cited to explain variable disease incidences between years or between areas. Surico et al. (2000) did not detect in Italy (Florence and Siena) any weather condition conducive to Esca, nevertheless rainy summer was even so found more favourable to the chronic form meanwhile dry summer was found more conducive to the acute form. In France, apoplexy is known to often occur after a rain in a warm period (Dubos, 2002; Galet, 1995). Marchi et al. (2006) tended to confirm that rainfall seems positively related with manifest Esca. However, in spite of climatic variations may easily explain variable levels of disease expression or disease rates, preliminary results of a survey done from 2004 to 2007 in vineyards from Aquitaine region (Lecomte et al., 2006, 2008 a,b), showed that the appearance of leaf symptoms increased more or less progressively and regularly from the beginning of June up to the end of July (see an example in Figure 1). After this date, the rate of leaf symptom occurrence decreases, although symptoms may occur up to September. This kind of profile for Esca symptom appearance was observed whatever the vineyard and the year and was also suggested by Marchi et al., Such a profile tends to indicate that the progressive appearance of vines showing leaf symptoms could be strongly related to the progressive increase of mean temperatures in early summer (Darrieutort et al., 2007), which likely influence the fungal growth or activity in the wood tissues. Number of symptomatic vines Figure 1: Example of Esca leaf symptom occurrence during summer recorded in a vineyard located in Bordeaux area (Cénac, France, 2006). Soil may also play a significant role in the development of Esca symptoms. A survey of 22 plots representative of Bordeaux area was done in summer to look for a relationship between leaf expression and the main soil characteristics (texture, water availability and fertilizer). In preliminary results, Guerin-Dubrana et al. (2005) reported that the soil with high percentage of lemon and clay and with either high water reserve or with high nitrogen supply were the most conducive to Esca. These first results indicated a link between the disease and the vigor and/or the growth conditions. This influence of the soil, especially the sloe, was also partly reported by Surico et al. (2000), that was more conducive when the slope was more level. Robotic and Bosancic (2007) reported exactly the same result

194 174 in a vineyard from Serbia. Panon et al. (2005) also reported higher disease incidence in the heavier soils of Champagne. Information concerning the influence of training or pruning systems are still fragmentary and sometimes controversial. Most of the reports were found in the French technical literature from the last century. However, because all fungi associated with trunk diseases can enter the grapevine wood through the pruning wounds, many authors tended to point out that the number of pruning wounds, their size, their position and their concentration in the same pruning zone are key-factors that can lead to a quick development of inner necrosis (Bolay, 1979; Boubals and Mur, 1990; Dubos, 2002; Lecomte et al., 2008). One of the first study carried out about the influence of a training system, was the one performed by Lafon in 1927 who developed a pruning system previously used by a grower settled in Charentes, Mr Poussard. The principle consists in keeping the same sap route from one year to another. Wounds are on a same line and are only made on the upper part of the cordons. This system, called Guyot-Poussard, was found less conducive to Esca by Geoffrion et Renaudin (2002) in the Loire Valley and is today recommended in regions where the Guyot form is usual. Grapevine is botanically a creeper that naturally run the ground and can grow up along the trunks of other perennial plants. Current training systems often consist on developing a small fruit tree involving every year repeated pruning decisions (Lecomte et Gaudillère, 2008). This kind of training-pruning system is rather specific to viticulture, quite different from those used for fruit growing. This system is particularly severe and may have consequences on the development of inner necrosis, as exemplified by Table 1 showing that the percentages of necrotic surface, observed from transverse sections of vines, cut before uprooting, may vary considerably according to the training plus pruning systems. Moreover, wounds of the cultivated grapevine species do not heal well and short-cut pruning wounds are currently strongly unwise to avoid a too rapid drying and necrosing wood along the sap routes. Table 1. Examples of inner necrosis development according to the pruning regime. Each necrotic surface was visually assessed on transverse trunk sections cut just below the arms of 22 Cabernet Franc vines per treatment. Training system 2278 vines/ha 20 yrs-old Pruning system from 1990 to 2004 Mean % of necrotic surface per transverse cut % of vines with a necrotic area of more than 5% of the total trunk surface Lyra: opened Winter pruning 20.2 a 84 (S*) Lyra: inversed Winter pruning 13.3 a 41 (S) Lyra: inversed a, b, Minimal pruning in summer 2.3 b 9 (S) letters indicate significant differences following a Newmans-Keuls test (α=5%) and ANOVA testing. S* Significant differences between treatments after a X 2 test carried out with a distribution of necrotic surfaces in 2 classes (upper or lower than 5%). Each treatment was different from the 2 others. Since the 1990 s, demand for planting material has probably exceeded supply of high quality plant material (Waite and Morton, 2007). Grapevine has also become a speculative crop and many current plantings are produced earlier than previous ones some decades ago. A new

195 175 syndrome described as Petri disease has been identified. And very simplified training systems have also emerged like those with trunks with low diameter and very short cordons, as shown in Figure 2, leading to a concentration of large pruning wounds close to the upper part of the trunks. This kind of training system, sometimes encountered in some vineyards with high yield potential was already considered as disastrous for Eutypa dieback. In Entre-Deux- Mers, a viticultural Bordeaux region, where highly contrasted situations can be noticed as far as incidence of either Eutypa or Esca is concerned, a tentative restricted survey was done in two representative locations in order to look for putative relationship between the different agronomical characteristics and the global incidence of Esca (Table 2). This etiological approach showed important differences between plots from the same location and revealed that the differences of Esca incidence could be explained by a set of favourable factors (susceptible cultivar, intrinsic rootstock vigour, short cordons, fertilizer, plant vigour due to environment, topography, ). Table 2. Examples of highly variable damage caused by Esca in different agronomical contexts. Survey done in Bordeaux area in 2007 (Adar Créon-Cadillac- CA33). Location Tresses Bonnetan Cultivar Rootstock Planting year Cab. Sauvignon S Cab. Sauvignon Cab. Sauvignon Cab. Franc 420A 1991 Cab. Franc Training system Arm length Trunk height Guyot 5-30 cm 80 cm Guyot cm cm Guyot 5-30 cm 80 cm Guyot 5-15 cm 80 cm Guyot cm 60-70cm Fertilizer N, K N: 21 Unit/an N: 21 Unit/an N: 21 Unit/an K: 120 Unit/an No one since 2002 Other characteristics High vigor*: 111.1g Humid plot, lower part Previous crop: grass-land 2666 vines/ha Moderate vigor: 102.9g Sloping and draining soil Previous crop: grapevine 3333 vines/ha Moderate vigor: 87.8g Soil with lemon and clay Previous crop: grapevine 2666 vines/ha Moderate vigor: 86.5g Previous crop: grapevine 3333 pieds/ha Moderate vigor: 71.2g Previous crop: grapevine 4545 vines/ha * Vigor visually assessed and measured by the dry weight of 100 leaves collected on September 2007 ** Whole vines Esca-affected: vines with dead arms or entirely dead, re-trained, re-planted and symptomatic Sanitary status** ,8% ,1% ,7% ,2 % 67 23,4%

196 176 Figure 2. Examples of simplified training systems that seemed very conducive to Esca in some French vineyards as reported in Table 1 (Photographs Gaudillère and Darrieutort) Table 3. Control measures recommended for preventing trunk decline diseases, especially for Esca Before planting Planting Use controlled mother-vineyards (age limited) Use plant material of good quality Avoid the most susceptible cultivars in the most fertile soils Avoid too long immersions of roots in water Avoid too late plantings (July, ) or provide appropriated irrigation After planting No short arm Correct training of the trunks, no too early grape harvest No short-pruned wounds (close to the trunks) to avoid drying zones inside the trunks Prefer the pruning systems that avoid constant changes of sap routes (Guyot-Poussard with stumps) Prophylactic methods (prune in late winter, namely for Eutypa) Protection of pruning wounds, preferably with a paste Avoid large pruning wounds (electric shears) Avoid excessive use of fertilizers The epidemiology of Esca disease may begin very soon in the history of a young vine, notably from mother plants infected by Phaeomoniella chlamydospora (Fourie and Halleen, 2002). Different steps along the nursery process, like hydration and callusing, involve water soaking or high temperature (25-28 C), two environmental factors highly favourable to the fungal growth, among other factors well reviewed by Waite and Morton in However, the presence of wood fungi in a vine does not necessarily mean that this vine will become

197 177 diseased. This also depends on other predisposing factors during and after planting in the vineyard (Ferreira et al., 1999; Stamp, 2001; Lecomte et al., 2008). Therefore, in order to avoid a too early and rapid inner necrosis development in the wood of young grapevines, different preventive measures are now advised including recommendations applicable before, during and after planting (Table 3). Conclusion Many complex pathological scenarios could probably explain the grapevine trunk disease developments. But the high variability of situations, as far as its incidence is concerned, clearly indicates that Esca is a consequence disease resulting both from the influence of biotic factors and abiotic factors. All the favourable cultural factors mentioned above, more likely when they are combined, may largely contribute and predispose some vines to unusual levels of disease expression or of microbial development into the wood. In our opinion, control of trunk decline diseases will likely be based on the selection of tolerant cultivars and/or on managing the cultural factors than on the search of a hypothetic curative compound able to replace the sodium arsenite. Acknowledgements Sylvie Bastien, Agnès Destrac, Marie-France Costet, Jean-Pierre Gaudillère, Etienne Laveau, Jean-Michel Liminaña, Joël Ortiz and Jean-Pascal Tandonnet are gratefully acknowledged for their contribution in the studies mentioned above. References Bolay, A. 1979: Accroissement des maladies du bois dues à la transformation des souches de vigne. Bull. de l'organisation Internat. de la Vigne 52: Boubals, D. 2002: Comment sauver les vignes de Cabernet-Sauvignon, de Sauvignon Blanc et des autres cépages attaqués par les champignons parasites du bois. Prog. Agric. et Vitic. 119(18): Boubals, D. & Mur, G. 1990: Influence du mode de taille de la vigne sur l'attaque du tronc des souches par le champignon Eutypa lata - Cas du Cabernet-Sauvignon. Prog. Agric. et Vitic. 107(22): Calzarano, F. & Di Marco, S. 2007: Wood discoloration and decay in grapevines with EscaEsca proper and their relationship with foliar symptoms. Phytopath. Medit. 46(1): Cordeau, J., Dubos, B. & Dumartin, P. 1984: Tous les viticulteurs sont concernés par l'eutypiose. Cahier Technique du CIVB 122: 6-8. Darrieutort, G., Liminana, J.-M., Louvet, G. & Lecomte, P. 2007: Maladies du Bois La relation entre l Esca et le Black Dead arm se précise. Union Girondine des Vins de Bordeaux 1036: Destrac-Irvine, A., Laveau, C., Goutouly, J.-P., Letouze, A. & Guerin-Dubrana, L. 2005: L'écophysiologie de la vigne Mieux comprendre les maladies de dépérissement. L'Union Girondine des Vins de Bordeaux 1035:

198 178 Dubos, B. 2002: Le syndrome de l'esca. In Editions Féret, Maladies cryptogamiques de la vigne, 2 ed., Bordeaux: Dumot, V. 2007: Nuisibilité des maladies du bois dans le vignoble charentais. In BNIC, ed., Journée Technique de la Station Viticole, Cognac: Ferreira, J. H. S., van Wyk, P. S. & Calitz, F. J. 1999: Slow dieback of grapevine in south africa: stress-related predisposition of young vines for infection by Phaeoacremonium chlamydosporum. South Afric. Journ. of Enol. and Vitic. 20(2): Fisher, M. & Kassemeyer, H.-H. 2003: Fungi associated with Esca disease of grapevine in Germany. Vitis 42(3): Fourie, P. H. & Halleen, F. 2002: Investigation on the occurrence of Phaeomoniella chlamydospora in canes of rootstock mother vines. Austral. Plant Pathol. 31: Fussler, L., Kobès, N., Bertrand, F., Maumy, M., Grosman, M. & Savary, S. 2008: A characterization of grapevine trunk diseases in France from data generated by the National Grapevine Wood Disease Survey. Phytopathology 98: Galet, P. 1995: Précis de Pathologie Viticole (2 ed.). JF Impression, St Jean de Védas, France, 264 pp. Gaudillère, J.-P. 2003: Conduite de la vigne et résistance aux pathogènes. Journal International des Sciences de la Vigne et du Vin 37: 2. Geoffrion, R. 1977: Quelques précautions recommandées lors de la formation des jeunes vignes. Phytoma - Défense des cultures 289: Geoffrion, R. & Renaudin, I. 2002: Tailler contre l'esca de la vigne. Phytoma-LDV 554: Goutay E., 1903: Manuel de viticulture pour la région froide et tempérée. Coulet et fils éditeurs, Montpellier. Goutouly, J.-P. 2007: Influence de la physiologie de la plante dans l'expression des symptômes d'esca. Journée technique de la Station Viticole, Cognac, BNIC editor: Gräfenham, T., Zare, R. & Gams, W. 2005: Esca a disease or a case of commensalism? Phytopath. Medit. 44(1): Gu, S. L., Cochran, R. C., Du, C. Q., Hakim, A., Fugelsang, K. C., Ledbetter, J., Ingles, C. A. & Verdegaal, P. S. 2005: Effect of training-pruning regimes on Eutypa dieback and performance of Cabernet Sauvignon grapevines. Jour. of Hort. Sci. and Biotec. 80(3): Grosman J., 2008: Observatoire national des maladies du bois: bilan de 4 années d'observations, Conférence Euroviti, 17 Janvier 2008, Angers, France: Gubler, W. D., Rolshausen, P. E., Trouillas, F. P., Úrbez-Torres, J. R., Voegel, T. M., Leavitt, G. M, and Weber, E. A. 2005: Grapevine trunk diseases in California. Practical Winery and Vineyard. Jan./Feb.:6-26, Guérin-Dubrana, L., Destrac-Irvine, A., Goutouly, J.-P., Letouze, A. & Gaudillère, J.-P. 2005: Relationship between incidence of EscaEsca and black dead arm foliar symptom expression in the vineyard, ecophysiological indicators and cultural practices. Phytopath. Medit. 44(1):110. Guérin-Dubrana, L., Goutouly, J.-P., Piot J. & Maher N. 2008: Grapevine Trunk Diseases: a method to investigate relationships between internal wood decay and foliar symptoms. Journ. of Plant Pathology 90 (Supplement 2): 179. Grosman J., 2008: Observatoire national des maladies du bois: bilan de 4 années d'observations, Conférence Euroviti, 17 Janvier 2008, Angers, France: 8-17.

199 Kobes, N., Grosman, J. & Pleynet, M. 2006: L'observatoire des maladies du bois de la vigne - Bilan de 3 années d'observations. In, AFPP - 8 th Intern. Conference on Plant Diseases. Tours, 5 & 6 Décember. CD ROM: Lafon, R. 1927: Modifications à apporter à la taille de la vigne dans les Charentes. Taille Guyot-Poussard mixte et double. L apoplexie, traitement préventif (Méthode Poussard). Traitement curatif. Imp. Roumégous et Dahan, Montpellier, Larignon, P., & Dubos, B. 1997: Fungi associated with Esca disease. Eur. J. Plant Pathol. 103: Laveau, C., Letouze, A., Louvet, G., Bastien, S. & Guérin-Dubrana, L. 2009: Differential aggressiveness of fungi implicated in Esca and associated diseases of grapevine in France. Phytopath. Medit. 48(1): Lecomte, P., Leyo, M., Louvet, G., Corio-Costet M.-F., Gaudillère, J.-P., & Blancard, D. 2005: Le Black dead arm, genèse des symptômes Observations au vignoble en Bordelais et réflexions en lien avec l'esca. Phytoma-LDV 587: Lecomte, P., Darrieutort, G., Defives, A., Louvet, G., Liminana, J.-M. & Blancard, D. 2006: Observations of black dead arm symptoms in bordeaux vineyards: evolution of foliar symptoms, localisation of longitudinal necroses, questions, hypotheses. IOBC/WPRS Bull. 29(11): Lecomte, P., Darrieutort, G., Liminana, J.-M., Louvet, G., Muruamendiaraz, A., Legorburu, J., Choueiri, E., Jreijiri, F. & Fermaud, M. 2008a: Revisiting Esca symptoms in the vineyard: results of a four-year survey. Phytopath. Medit. 48(1): 175. Lecomte, P., Darrieutort, Liminana, J.-M., Louvet, G., Guerin, L., Tandonnet, J.-P., Goutouly, J.-P., Gaudillère, J.-P. & Blancard, D. 2008b: (I) Eutypiose et Esca Eléments de réflexion pour mieux appréhender ces phénomènes de dépérissement. (II) Esca de la vigne Vers une gestion raisonnée des maladies de dépérissement. Phytoma-LDV 615: & 616: Lecomte, P., Leyo, M., Louvet, G., Corio-Costet, M. F., Gaudillère, J.-P. & Blancard, D. 2005: Le Black dead arm, genèse des symptômes Observations au vignoble en Bordelais et réflexions en lien avec l'esca. Phytoma-LDV 587: Lecomte, P. & Gaudillère, J.-P. 2008: L Esca, un tribut que paye la vigne en réaction à l importante domestication de l homme, notamment par la taille? Phytoma-LDV 615: 47. Liminana, J.-M., Pacreau, G., Boureau, F., Menard, E., David, S., Himonnet, C., Fermaud, M., Goutouly, J.-P., Lecomte, P. & Dumot,V. 2009: Inner necrosis in grapevine rootstock mother plants in the Cognac area (Charentes, France). Phytopath. Medit. 48(1): Marchi, G., Peduto, F., Mugnai, L., Di Marco, S., Calzarano, F. & Surico, G Some observations on the relationship on manifest and hidden Esca to rainfall. Phytopath. Medit. 45 (Supplement): Mugnai, L., Graniti, A. Surico, G. 1999: Esca (Black Measles) and brown wood-streaking: two old and elusive diseases of grapevines. Plant Disease 83(5): Panon, M.-L.; Panigaï, L., Moncomble, D. & Boulay, M. 2005: Dossier Esca/Bda: une affaire complexe. Le Vigneron Champenois 4: Péros, J.-P. & Berger, G. 1994: A rapid method to assess the aggressiveness of Eutypa lata isolates and the susceptibility of grapevine cultivars to Eutypa dieback. Agronomie 14: Péros, J.-P., Berger, G. & Jamaux-Despréaux, I. 2008: Symptoms, wood lesions and fungi associated with Esca in organic vineyards in Languedoc-Roussillon (France). Journ. of Phytopath. 156 (5): Robotic, V. & Bosancic, R. 2007: Notes on the relationship of manifest Esca disease to vineyard slope. Phytopath. Medit. 46(1):

200 180 Sparapano, L., Bruno, G. & Graniti, A. 2000: Infection of grapevines by some fungi associated with Esca. II Interaction among Phaeoacremonium chlamydosporum, P. aleophilum and Fomitiporia punctata. Phytopath. Medit. 39: Sparapano, L., Bruno, G. & Graniti, A. 2001: Three-year observation of grapevines crossinoculated with Esca-associated fungi. Phytopath. Medit. 40 (Supplement): Stamp, J. A. 2001: The contribution of imperfections in nursery stock to the decline of young vines in California. Phytopath. Medit. 40 (Supplement): S369-S375. Surico, G., Marchi, G., Bracini, P. & Mugnai, L. 2000: Epidemiology of Esca in some vineyards in Tuscany (Italy). Phytopath. Medit. 39(1): Waite, H. & Morton, L. 2007: Hot water treatment, trunk diseases and other critical factors in the production of high-quality grapevine planting material. Phytopath. Medit. 46(1): Weber, E. A., Trouillas, F. P. & Gubler, W. D. 2007: Double pruning of grapevines: a cultural practice to reduce infections by Eutypa lata. Amer. J. of Enol. and Vitic. 58(1):

201 Session 3: Biology and population dynamics of insects and moths and modeling

202

203 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Electrospun nanofibers as novel carriers of insect pheromones: communication disruption strategy against Lobesia botrana D. F. Hein 1, M. Breuer 2, H. E. Hummel 1, A. Greiner 3, J. H. Wendorff 3, C. Hellmann 3, A. Vilcinskas 4, A. Kratt 5, H. Kleeberg 5, G. Leithold 1 1 Justus-Liebig-University Giessen, Chair of Organic Agriculture, D Giessen, 2 State Institute for Viticulture and Enology Freiburg, Department of Ecology, D Freiburg, 3 Philipps-University Marburg, Department of Chemistry, D Marburg, 4 Justus-Liebig- University Giessen, Professorship of Applied Entomology, D Giessen, 5 Trifolio-M GmbH, D Lahnau Abstract: Using organic nanofibers as dispensers for pheromones and kairomones in plant protection for disrupting insect chemical communication is a novel approach aiming at popularizing this technique in organic and integrated plant production. Expected advantages of the nanofibers are highly controlled spatiotemporal release rates of pheromones / kairomones, improved climatic stability, and mechanized application. Dispenser types used so far show deficiencies in one or more of these requirements. Mechanical application of pheromone dispensers is a new approach to reduce the costs of manual labour. Therefore the environmentally compatible, highly specific and efficient technique of mating disruption may become an alternative to the use of synthetic pesticides in integrated pest management. The electrospun nanofibers are highly elastic, which prevents break-off of smaller pieces, and polymers used are biocompatible. Due to the tiny scale of nanofibers the mass input both for pheromones and for polymeric nanofibers is extremely low. Major environmental benefits are high control specificity, very low concentrations of residues and minimal risk towards development of resistant insect strains. We tested organic polymer fibers, made from the biodegradable polymer Ecoflex (BASF), treated with Lobesia botrana (Lepidoptera: Tortricidae) sex pheromone, (E,Z)-7,9-dodecadienyl acetate, in cage tests placed inside vineyards. With this experimental setup it is possible to show the applicability of mating disruption dispensers for the use in plant protection against insect pests in vineyards. We used Isonet LE dispensers (Shin-Etsu) as a positive control. The Ecoflex nanofiber-pheromone-dispensers show a mating disruption effect which is comparable to the efficacy of the Isonet LE dispensers for at least three weeks. After that, the mating disruption effect observed tapers off to a level where it is insufficient for plant protection purposes. But, in principle, our first prototype of a nanofiber pheromone dispenser has proved its efficacy for mating disruption purposes. Modifications of the fibers, currently under development, open up the possibility of extending the disruption effect to a period of several months duration, enough to cover the entire flight period of Lobesia. Key words: electrospinning, nanofibers, nanotechnology, pheromone dispensers, Lobesia botrana, mating disruption Introduction The application of pheromones in plant protection for the control of insect pests is a sustainable alternative to the use of synthetic pesticides. In conventional crop protection, insecticides are used to hit the insect pest with a lethal dose of active ingredient. This is a wasteful and environmentally questionable strategy. The mating disruption technique, in contrast, is a pre-emptive approach, which impedes successful mating between male and female moths. As a result, female oviposition of viable eggs is minimized and prevented by purely behavioral, not toxicological modification (pheromones are nontoxic). Major 183

204 184 environmental benefits are high species specificity, very low concentrations of residues, minimal risk of development towards insect resistance, and sustainability. An optimal dispenser used for this purpose should release the pheromones at a constant rate over a long period of time (Campion 1984). Devices should be preferred which release disrupting amounts of pheromone soon after field application, hold this level, and then finally tail off (Howse et al. 1998). Dispenser types used so far show shortcomings in predictable release rates, stability against climatic factors, and mechanized application, or in a combination of these desired requirements. Using organic nanofibers as dispensers for pheromones and kairomones in plant protection for disrupting insect chemical communication is an entirely novel approach (Hein et al. 2009). Its aim is applying this technique on a wider scale in organic and integrated plant production (Greiner & Wendorff 2007, Hellmann et al. 2009). Expected advantages of nanofibers are highly controlled spatiotemporal release rates of pheromones. Mechanical application of pheromone nanofibers is a new approach reducing the costs of manual labour. The nanofibers are highly elastic which prevents break-off of smaller pieces, and polymers are selected for biocompatibility. Due to the geometrical scale of nanofibers the mass input both for pheromones and for polymeric nanofibers is very low. In our study, we used the established mating disruption strategy for the regulation of the European grapevine moth, Lobesia botrana, as a model system. At the University of Marburg Ecoflex polymer fibers prepared with the respective sex pheromone were made by electrospinning (Greiner & Wendorff 2007). Electrospinning is one of the most common techniques to produce continuous, endless fibers with diameters down to the nanometer scale. The starting material is an organic polymer solution thoroughly mixed with the respective active ingredient. This mixture is pressed through a metal nozzle which simultaneously serves as an electrode. A high electrical field of kVm -1 is applied to this nozzle. The droplets formed outside the nozzle are deformed, streched out, accelerated, and immediately deposited as fibers on the counter electrode arranged at a distance of about 1-50cm. The electrical current that flows during this electrospinning process is at the most in the microampere range (Greiner & Wendorff 2007). Thus, there is neither a high demand for electrical power ( Watt) nor a human hazard from its use. The resulting non-woven nanofibers served as pheromone dispensers in our experiments. Material and methods Pheromone dispensers Nanofiber dispensers used in our experiments were made from the commercially available, organic, biodegradable polymer Ecoflex (BASF). The incorporation of the Lobesia botrana sex pheromone (E,Z)-7,9-dodecadienyl acetate into the fibers occurs during the spinning process. Isonet LE dispensers (Shin-Etsu) served as positive controls in our semi-field experiments. A third plot without dispensers was used as a negative control. The Lobesia botrana sex pheromone load was 333g/ha in the nanofiber and 95g/ha in the Isonet plot, respectively. Experimental setup of semi-field experiments For evaluation of our nanofiber dispensers we used an experimental setup developed at the State Institute for Viticulture and Enology Freiburg (Doyé & Koch 2005, Doyé 2006). One flight cage installed in the middle of our test plot (2,000m² each) was equipped inside with two sticky traps that use freshly hatched female Lobesia botrana as a lure (GÖTZ 1941). The

205 185 final dispenser density in field plots reached one per 20m² which is in agreement with the recommended density for the commercially available material. The nanofiber dispensers were deposited as a fine layer of Ecoflex nanofibers containing the sex pheromone (E,Z)-7,9- dodecadienyl acetate. In our positive control we distributed Isonet LE dispensers (Shin-Etsu). Another plot without treatment was our negative control. We released in total 160 male Lobesia botrana (4 releases with 40 individuals per release) in the flight cage of the size 1.7 x 1.8 x 2.1m. We used insects provided by the laboratory rearing facility at the State Institute for Viticulture and Enology Freiburg. Results were obtained by counting the number of male moths caught in the sticky traps. The experiments lasted until the mating disruption effect breaks down in the nanofiber plot. The lower the numbers of catches in the treated plots the higher was the mating disruption effect of the pheromone dispensers. Zero percent recovery would be equivalent to complete control. Results and discussion For adjusting/balancing the different male catches in the control plot the number of males caught in the control plot was set as 100%. Within the first three weeks of the experiment, the efficacy of the nanofiber dispensers was on the same level as the Isonet LE dispensers. On average, 10% of the males released in the nanofiber plot and 5% in the Isonet LE variant were caught in the sticky traps baited with natural females as lure compared to 100% in the untreated control. Within the fourth week, the mating disruption effect in the nanofiber variant decreases, with 56% of the males recovered. At that point in time, the Isonet LE variant still showed a respectable recovery of 9%. The mating disruption effect in the first three weeks of the experiments with the newly developed nanofiber dispensers is excellent (highly significant to the control, P < 0.001, 9 replicates) and comparable to the effect of the Isonet LE dispensers (no significance between nanofiber plot and Isonet LE variant, P > 0.05). However, in week 4 we are finding a decrease of the efficacy in the plots treated with the nanofibers. With a recovery rate of 56% a positive difference to the control plot (100% recovery rate) still can be observed but the mating disruption effect is not sufficient any more for practical control of Lobesia botrana. In contrast, a high efficacy still can be observed with the Isonet LE dispensers. With this investigation we proved in principle the efficacy of our first prototype of a pheromone dispenser based on organic nanofibers. Future work will focus on extending the duration of the mating disruption effect for covering the entire growing season. Modified polymer fibers will be used for this approach which also could result in a reduction of the overall amount of pheromone needed for mating disruption. Thus, savings in pheromone costs will be realized as an added benefit. Currently, the nanofiber dispensers are pre-produced in the lab and then distributed in the field by manual work. To save labor expenses, we work on a method to simultaneously accomplish both the manufacture and the mechanical distribution of the dispensers within the vineyard. Mating disruption of Lobesia botrana in a vineyard with electrospun fiber dispenser serves as a model system. Our future aim is to apply this technique to other crops. Apart from other perennial crops (e.g. apple orchards) we think about the use of mating disruption in annual crops. Examples are Zea mays with the western corn rootworm Diabrotica virgifera virgifera (Hummel et al. 2009) and the pink bollworm Pectinophora gossypiella in cotton fields (Gaston et al. 1977). From our point of view farmers would not work with pre-produced dispensers in such large scale crops. Therefore we have to refine the technique in such a way

206 186 that direct electrospinning in the field becomes practicable. This would require a mechanized procedure for the application of nanofiber dispensers. Ecotoxicological investigations are in progress which will decide whether or not our nanofiber dispensers pose a danger for man, animal, and environment. Acknowledgements This paper is a result of the German interdisciplinary research project titled Organic nanofibers as novel carriers for volatile semiochemicals in integrated and ecological agriculture. The project brings together experts from various disciplines, such as nanotechnology (Philipps- University of Marburg), plant protection with pheromones (Justus-Liebig-University Giessen; State Institute for Viticulture and Enology Freiburg; Trifolio-M GmbH Lahnau), mechanization of pheromone application (Maschinenfabrik Schmotzer GmbH Bad Windsheim), risk assessment (Julius-Kühn-Institut, JKI Kleinmachnow & Berlin), and economic validation (TransMIT GmbH Giessen). Financial support from the German Federal Ministry of Food, Agriculture, and Consumer Protection is kindly acknowledged. We thank the German Federal Agency of Agriculture and Food (BLE), especially Mrs. Dorothée Hahn, for administrative support. Mrs. Elena Tessitore, State Institute for Viticulture and Enology Freiburg, is gratefully acknowledged for rearing Lobesia botrana test insects. Mr. Arno Deuker provided technical help. The donation of Ecoflex by BASF is kindly acknowledged. References Campion, D. G. 1984: Survey of Pheromone Uses in Pest Control. In: Hummel, H. E. & Miller, T. A. (eds.): Techniques in Pheromone Research. Springer-Verlag, New York: Doyé, E. & Koch, U. 2005: A reliable field test for the efficiency of mating disruption techniques. IOBC/WPRS Bull. 28(7): Doyé, E. 2006: Entwicklung eines Freilandtests zur Überprüfung der Wirksamkeit von Pheromonanwendungen im Weinbau. Dissertation University of Kaiserslautern, Biology Department. URL: _l1/WBI_Dissertation%20Doye% pdf Gaston, L. K., Kaae, R. S., Shorey, H. H. & Sellers, D. 1977: Controlling the pink bollworm by disrupting sex pheromone communication between adult moths. Science 196: Götz, B. 1941: Der Sexuallockstoff als Bekämpfungsmittel gegen die Traubenwickler im Freiland. Wein und Rebe 23: Greiner, A. & Wendorff, J. H. 2007: Electrospinning: A Fascinating Method for the Preparation of Ultrathin Fibers. Angew. Chem. Int. Ed. 46: Hein, D. F., Leithold, G., Hummel, H. E., Vilcinskas, A., Greiner, A., Wendorff, J. H., Dersch, R., Hellmann, C., Breuer, M., Beer, H., Schroer, S., Kratt, A., Kleeberg, H., Schulze, C. & Wahl, F. 2009: Nanofasern als neuartige Träger für flüchtige Signalstoffe zur biotechnischen Regulierung von Schadinsekten im integrierten und ökologischen Landbau. In: Mayer, J. et al. (eds.). Werte Wege Wirkungen: Biolandbau im Spannungsfeld zwischen Ernährungssicherung, Markt und Klimawandel. Band 1: Boden, Pflanzenbau, Agrartechnik, Umwelt- und Naturschutz, Biolandbau international, Wissensmanagement. Beiträge zur 10. Wissenschaftstagung Ökologischer Landbau Zürich, Februar 2009:

207 Hellmann, C., Greiner, A. & Wendorff, J. H. 2009: Design of pheromone releasing nanofibers for plant protection. Polymers for Advanced Technologies, published online in Wiley InterScience, DOI: /pat Howse, P., Stevens, I. & Jones, O. 1998: Insect Pheromones and their Use in Pest Management. Chapman & Hall, London. Hummel, H. E., Deuker, A. & Leithold, G. 2009: The leaf beetle Diabrotica virgifera virgifera: a merciless entomological challenge for agriculture. IOBC/WPRS Bull. 41:

208 188

209 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Control of grape berry moth larvae using parasitoids: should it be developed? D. Thiéry 1, L. Delbac 1, C. Villemant 2, J. Moreau 3 1 UMR Santé Végétale 1065 INRA, Institut de Sciences de la Vigne et du Vin de Bordeaux, IFR Muséum National d Histoire Naturelle, Département Systématique & Evolution, UMR7205 OSEB, CP50, Paris, France, 3 Équipe Écologie Évolutive, UMR CNRS 5561 Biogéosciences, Université de Bourgogne, F Dijon, France. Abstract: Besides mating disruption techniques and progress in monitoring techniques (e.g. the use of food traps against females), biological control may reveal itself very efficient at controlling grape moth populations. Parasitoids active to control grape moths are known for long in vineyards; few of them were already described in the middle of the 19th century in French vineyards and their efficiency was already recognized especially against the diapausing and the first spring generations of the moths. Rather numerous attempts to release egg parasitoids have been done in different European countries using different species of trichogrammas. The results obtained varied a lot and could not yet clearly promote the use of this technique in vineyards. We believe that a biological control based on larval parasitoids could efficiently be developed as a valuable alternative to chemical control. In the present paper, we focus on larval parasitoids among which ichneumonids and chalcidoids (Hymenoptera) dominate, and present results obtained in different French vineyards (Bordeaux vineyard, Perpignan and Montpellier area, Côtes du Rhône and Alsace). We discuss factors that may favour or reduce their efficiency as biocontrol agents. Key words: Parasitoids, biological control, viticulture, biodiversity, fitness, grape cultivars, Ichneumonids, Tachinids, Trichogrammas. Introduction Biological control in vineyard is a promising but challenging perspective, and surprisingly the development of biological control in viticulture suffers from a lack of studies. The main efforts have been done studying and attempting egg parasitoids releases (namely Trichogrammas) (see Castaneda-Samoya et al., 1993; Reda Abd el Monsef, 2004 and Hommay et al., 2010). Beside such studies, several others attempted to qualify the biodiversity of natural enemies in different European vineyards (Thiéry, 2008). This started in the middle of the 19 th century (Audouin, 1842; Jolicoeur, 1894) until more recent years (e.g. Coscolla, 1980; Marchesini & Dalla Monta, 1994; Colombera et al., 2001; Thiéry et al., 2001; Barnay et al., 2001). Also the biology, ecology or behaviour of several parasitoids occurring in vineyards have been studied both in laboratories and field during these last years focussing on Trichogrammas (Stengel et al., 1977; Le Rallec & Wajnberg, 1990; Hommay et al., 2002; Moreau et al., 2009) or larval parasitoids (Chuche et al., 2006; Xuéreb & Thiéry, 2006). Rather numerous attempts to release egg parasitoids have been done in different European countries using different species of Trichogrammas which harvested very different efficacies. Currently the release of parasitoids in vineyards is very marginal, and biological control programmes at large scale in vineyards would require research and development attempting to identify the suitable candidate species to be used like the environmental 189

210 190 conditions favouring these species. This would also require fine basic research concerning the physiology, ethology and ecology of the selected candidates. Although grape training and pesticide practices may strongly affect the natural enemies population dynamics, most of the field studies concluded that parasitism (either egg or larval) varies according to ecological factors, e.g. climatic mismatch, variable host density, lack of alternative hosts, also pointing out that variable quality of the host affect the parasitoid reproductive success. In this paper, we present several ecological parameters influencing the parasitism efficacy. Parasitoids and predators of grape moths in vineyards Natural enemies of pest in vineyards have received interest for many years. During the middle of the 19 th century, several published monographies already listed eggs or larval parasitoids, the main focus being put on the two main moth pests in French vineyards at this period, the grape berry moth (Eupoecilia ambiguella) and the leaf rolling tortrix (Sparganothis pilleriana). In a recent review, Thiéry (2008) listed more than 70 species mainly represented in Hymenoptera which shows that vineyard is far from being a no parasitoid s land. In fact parasitoids abundance and diversity may vary a lot throughout the season, between the different grape production areas but also as a consequence of large range of different ecological factors (host density, alternative hosts, grass or floral cover, hedges). Neighbouring natural landscapes probably play an important role in natural enemies population level (Genini, 2000), even though we miss accurate data from vineyards to properly consider their effect. Interestingly, also natural enemies biodiversity exists in most of the European vineyards. For example, several oophagous parasitoids as trichogrammas, or larval parasitoids as the tachinid Phytomiptera nigrina and various hymenopteras as Campoplex capitator, Dibrachys spp (cavus, affinis), Dicaelotus inflexus, Diadegma fenestralis, Itoplectis maculator, Scambus elegans or Exochus spp. (tibialis, notatus), occur and are regularly found in different European vineyards (Table 1). Several factors affect the parasitism efficacy Latitude, climate match or mismatch and vineyard location These factors are probably not enough investigated. However, in a recent elegant study, Moreau et al. (2010) sampled Lobesia botrana larvae in different vineyards from Alsace, Switzerland and east South France and compared the larval parasitoids occurrences (Table 2). This shows that both diversity and major species vary according to both latitude and climate, though grape cultivars also contribute to these differences (see below). This work matches previous studies which also suggest that parasitoids species distribution vary in France according to latitude and climate (Thiéry et al., 2001). Several species are characteristic of Mediterranean viticulture, like for example P. nigrina or also C. capitator which is classically found in most of Mediterranean viticulture countries though this species has a broader distribution also occurring for example in Switzerland. Because vineyards present a great diversity of climatic conditions, climate match or mismatch is of primary importance and conditions the success of certain parasitoids. During these 2 successive years of study, the species Exochus tibialis was exclusively found in Switzerland and Alsace, and dominant in 2003 in the Valaisan vineyards (Table 2), while P. nigrina occurred exclusively in the south of France (Côtes du Rhône). In the study conducted by Moreau et al. (2010), almost 3 times fold less parasitoids were found in 2004, as compared to 2003, but the vineyards sampled were different.

211 191 Table 1. Non exhaustive list of parasitoids (alphabetic order) reported from literature as natural enemies of grape moths in vineyards in West European countries. This list is selected and implemented from a broader one (Thiéry, 2008). Only species reported by at least 2 references in the former list are presented here. EA = Eupoecillia ambiguella, LB = Lobesia botrana, SP = Sparganothis pilleriana, Beth = Hym. Bethylidae, Chal = Hym Chalcidoidea; Ichn = Hym. Ichneumonidae, Tach = Dip. Tachinidae. Pteromalus and Trichogramma spp regroup several species. Species Family Host orders Reported hosts in vineyardsi Parasitized instars (when known). Agrothereutes abbreviatus (F.) Ichn Lepidoptera LB EA pupae Ascogaster quadridenata (Wesm.) Ichn Tortricidae LB larvae pupae Brachymeria minuta (Wesm.) Chal Lep, Dip EA Campoplex capitator (Aub.) Ichn Tortricidae LB EA L3-L4 Diadegma fenestrale (Holm.) Ichn Lepidoptera LB SP larvae Dibrachys affinis (Masi) Chal Lepidoptera and other insects LB-EA-SP L4-L5 Dibrachys cavus (Walk.) (syn boucheanus) Chal Lepidoptera and other insects LB-EA-SP L4-L5 pupae Dicaelotus inflexus (Thom.) Ichn Lepidoptera LB pupae Dicaelotus resplendens (Holm.) Ichn Lepidoptera Elachertus affinis (Masi) Chal Tortricidae EA-LB-SP larveae Exochus tibialis (Holm.) Ichn Lepidoptera LB larvae pupae Gelis areator (Panz.) Ichn Lep, Hym LB EA larvae Goniozus claripennis (Först.) Beth Lepidoptera PS larvae Ischnus alternator (Grav.) Ichn Lepidoptera LB pupae Itoplectis alternans (Grav.) Ichn Lep, Hym LB, SP pupae Itoplectis maculator (Fabr.) Ichn Lep, Hym LB EA SP pupae Itoplectis tunetana (Schmied.) Ichn Lep, Hyma L EA pupae Phaeogenes melanogonos (Gmel.) Ichn Lepidoptera EA SP pupae Phaeogenes planifrons (Wesm.) Ichn Lepidoptera SP pupae Phytomyptera nigrina (Meig.) (= nitidiventris) Tach Lepidoptera LB larvae Pimpla spuria (Grav.) Ichn Lepidoptera LB pupae Pimpla turionellae (L.) Ichn Lepidoptera LB EA SP pupae

212 192 Species Family Host orders Reported hosts in vineyardsi Parasitized instars (when known). Pteromalus spp More than 8 species Ichn Lep, Dip, Col, Hym LB EA SP larvae pupae Scambus elegans (Woldst.) Ichn Lep, Hym LB larvae Tranosemella praerogator (L.) Ichn Lepidoptera LB EA SP larvae Trichogramma spp More than 15 species Chal Lepidoptera EA LB SP eggs Triclistus sp. Chal Lepidoptera LB larvae pupae Host density Host density is a classical major factor affecting natural enemy populations. It becomes crucial for example in specialist parasitoids which lack alternative hosts to achieve their reproductive cycle (Lane et al., 1999). The effect of host density on the parasitoids/predators population level and thus capacities to control the targeted pest is thus well documented in many biological antagonists. With most parasitoids searching for specific development stages, this host density should be viewed at the specific targeted development stage. Xuéreb & Thiéry (2006) showed for example a good correlation between the parasitism by natural populations of C. capitator and the number of L. botrana larvae per bunch (Figure 1). Also the host searching behaviour of parasitoids is often affected by host density in the patch. For example, the egg discovery rate of L. botrana by T. cacoeciae varies as a function of host density (Hommay et al., 1999). Season of the year and variation between years Except for the monovoltin leaf rolling tortrix, the grape moths (L. botrana, E. ambiguella and Argyrotenia pulchellana) accomplish at least two generations per year, and their population level may vary sometimes a lot from one generation to another. Therefore and because host density (see above) is of primary importance, parasitoids populations may suffer or profit from these variations. Variation between successive years is also a trend that has been often observed in different vineyards. This can also be illustrated by the data provided by sampling in 2003 and 2004 L. botrana larvae from different vineyards in France and Switzerland (Moreau et al., 2010) (Table 2), but also by data from Xuéreb & Thiéry (2006) (Table 3). Variation of parasitoids/predators population density may also be due only to their intrinsic annual life cycle or to their life style. Such within year variation is known for long in C. capitator. It is worth noting that this species named one century ago C. majalis meaning Campoplex of may (Audouin, 1842; Jolicoeur, 1894) can diapause in diapausing berry moths chrysalids (L. botrana and E. ambiguella)(our unpublished data). This species has traditional higher efficiency during the first generation of the moth. These observations were confirmed by Xuéreb & Thiéry (2006) (Table 3). In the same study, S. elegans was only found during 2 successive years during summer and never parasitized the spring generation of L. botrana. Intrinsic variation due to the parasitoid life cycle combined with variation of host population level within the year may also amplify such variations.

213 193 Figure 1: Host density relationship with the number of Campoplex capitator parasitizing the Lobesia botrana larvae (from Xuéreb & Thiéry, 2006, see this reference for more details). Table 2. Species diversity in larval parasitoids emerged from Lobesia botrana sampled in different French and Swiss vineyards. VS = Valais, VD = Vaud, F = France. Parasitoid species, Ichneumonidae: Et = Exochus notatus, Aa = Agrypon anxium, Cc = Campoplex capitator, Im = Itoplectis maculator, Di = Dicaelotus inflexus, Df = Diadegma fenestrale, Braconidae: As = Apanteles sp., Bethylidae: Gc = Goniozus claripennis, Tachinidae: Pn = Phytomyptera nigrina, Triclistus meridiator was found only once Y ear Locality Cultivar Et Aa Cc Im D i Pn G c D f DAs Yvorne (V D) Pinot noir Yvorne (V D) Chasselas 7 3 St P-de Clages (V S) Pinot noir St P-de Clages (V S) G amay 8 1 St P-de Clages (V S) Chasselas Tavel (F) G renache 6 1 Colmar (F) G ew urtz Colmar (F) Riesling 1 3 Total Roquemaure (F) G renache 16 1 Roquemaure (F) Syrah Sion (V S) Pinot noir Sion (V S) Chasselas 2 Nyon (V D) Chasselas Nyon (V D) Chardonnay 3 1 Total Effect of grape cultivars and larval food plant Grape cultivar, but also host plant consumed as a larva by L. botrana which is a polyphagous species (Thiéry & Moreau, 2005; Maher & Thiéry, 2006) surprisingly influenced the succeptibility of its eggs to be parasitized by Trichogramma evanescens (Moreau et al., 2009; and Figure 2). Interestingly this study also showed that the reproductive success of the

214 194 Trichogrammas emerging from the different eggs also varied like its larval growth speed. A similar result was found with eggs produced by moth fed as larvae on food supplemented with different host-plant of L. botrana and exposed to Trichogramma cacoeciae (Thiéry, Pizzol & Wanjberg, unpublished data). Table 3. Parasitism rates by natural populations of Campoplex capitator against Lobesia botrana during the first 2 successive generations of the year. See Xuéreb & Thiéry, 2006 for details on the procedure. Number of parasitized eggs (mean ± sem) B B A C A Chardonnay Chasselas Grenache Pinot Syrah Cultivar Figure 2: Effect of different grape cultivar consumed by Lobesia botrana as a larva on the parasitism rate of its eggs by Trichogramma evanescens (N of parasitized eggs per female of Trichogramma. evanescens, see Moreau et al., 2009 for experimental details).

215 195 These first studies strongly suggest that grape cultivar may also interfere with the host quality and thus their parasitism rate. However in the study done by Xuéreb & Thiéry (2006), the five cultivars tested did not significantly influence the larval parasitism by C. capitator, except Merlot and possibly Sauvignon on which L. botrana larvae were less parasitized by C. capitator in first spring generation (see Table 3). Conclusion A quick and simple answer to the question asked in the title should be yes. However successful use of natural enemies as a pest control techniques requires important efforts concerning the basic knowledge of the parasitoid or predator species and their relation with their targeting hosts. Also one key point in biological control is how to favour survival and high fecundity of the beneficial organisms. Research has thus a challenge to improve its knowledge of tritrophic interactions in vineyards which are to date not enough investigated. Viticulture has also to evolve towards practices that are more environmentfriendly, thus favouring the populations of natural enemies by offering them fairly favourable environmental conditions. Coupling releases of natural enemies naturally occurring in vineyards and vineyard/landscape management techniques in order to favour the installation and reproductive success of the beneficial organisms is an interesting way for future efficient biological control in vineyards. References Audouin, V. 1842: Histoire des insectes nuisibles de la vigne et particulièrement de la pyrale. Fortin Masson et Cie, Paris. Barnay, O., Pizzol, J., Gertz, C., Kienlen, J. C., Hommay, G., & Lapchin, L. 1999: Host density-dependance of discovery and exploitation rates of egg patches of Lobesia botrana (Lepidoptera: Tortricidae) and Ephestia kunhiella (Lepidoptera: Pyralidae) by the partasitoid Trichogramma cacoeciae (Hymenoptera: Trichogrammatidae). J. Econ. Entomol. 92: Barnay, O., Hommay, G., Gertz, C., Kienlen, J. C., Schubert, G., Marro, J. P., Pizzol, J., & Chavigny, P. 2001: Survey of natural populations of Trichogramma (Hym., Trichogrammatidae) in the vineyards of Alscae (France). J. appl. Entomol. 125: Castaneda-Samoya, O. R., Holst, H. & Ohnesorge, B. 1993: Evaluation of some Trichogramma species with respect to biological control of Eupoecillia ambiguella and Lobesia botrana Schiff (Lepidoptera: Tortricidae). Z. Pflanz. und Pflanz. 100: Chuche, J., Xuéreb, A. & Thiéry, D. 2006: Attraction of Dibrachys cavus (Hymenoptera, Pteromalidae) to its host frass volatiles. J. Chem. Ecol. 32: Collombera, S., Alma, A. & Arzone, A. 2001: Comparison between the parasitoids of Lobesia botrana and Eupoecillia ambiguella in conventional and integrated vineyards. IOBC/WPRS Bull. 24(7): Coscolla, R. 1980: Aproximacion al studio del parasitismo natural sobre Lobesia botrana Den. Y Schiff. en las comarcas viticolas. Boll. Serv. Plaga 6: Genini, M. 2000: Antagonistes de la cicadelle verte et des vers de la grappe dans le vignoble valaisan et les milieux environnants. Rev. Suisse Viti. Arbo. Hort. 32: Hommay, G., Gertz, C., Kienlen, J. C., Pizzol, J., & Chavigny, P. 2002: Comparison between the control efficacy of Trichogramma evanescens Westwood (Hymenoptera: Trichogrammatidae) and two Trichogramma cacoeciae Marchal strains against grapevine moth (Lobesia botrana Den. & Schiff.), depending on their release density. Biocontrol Sci. Techno. 12:

216 196 Hommay, G., Kienlen, J. C., Gertz, C. Bihri, C. & Pizzol, J. 2011: Factors affecting the post release dispersal of Trichogramma cacoeciae Marchal in the vineyards. IOBC/WPRS Bull. 67: Jolicoeur, H. 1894: Les ravageurs de la vigne. F. Michaud, Reims, France. Lane, S. D., Mills, N. J. & Getz, W. M. 1999: The effect of parasitoid fecundity and host taxons on the biological control of insect pests: the relationship between theory and data. Ecol. Entomol. 24: Le Ralec, A., & Wajnberg, E. 1990: Sensory receptors of the ovipositor of Trichogramma maïdis (Hym. Trichogrammatidae). Entomophaga 35: Marchesini, E. & Della Monta, L. D. 1994: Observations on natural enemies of Lobesia botrana (Den. et Schiff.) (Lepidoptera, Tortricidae) in Venetian vineyards. Boll. Zool. Agr. Bachic. 26: Moreau, J., Richard, A., Benrey, B., & Thiéry, D. 2009: The influence of plant cultivar of the grapevine moth Lobesia botrana on the life history traits of an egg parasitoid. Biol. Control 50: Moreau, J., Villemant, C., Benrey, B., & Thiéry, D. 2010: Species diversity and composition of larval parasitoids of the European grapevine moth (Lobesia botrana): the influence of region and cultivar. Biol. control 54: Maher, N, & Thiéry, D 2006: Daphne gnidium, the possible native host plant of the European grapevine moth Lobesia botrana, stimulates its oviposition. Is a host shift relevant? Chemoecology 16: Reda Abd el Monsef, A. I. 2004: Biological control of grape berry moths Eupoecilia ambiguella Hb. and Lobesia botrana Den. et Schiff. (Lepidoptera Tortricidae) by using egg parasitoids of the genus Trichogramma. Thesis Universität Giessen, 103 pp. Sentenac, G., & Thiéry, D. 2008: Les méthodes de lutte biologiques ou biotechniques contre les insectes et acariens nuisibles à la vigne. In: Conf. Int. Modiaviti, Bordeaux: Stengel, M., Voegelé, J., & Lewis, J. W. 1977: Trichogramma. V.b. Winter survival strain and the discovery of T. cacoeciae Mar. on egg masses of Ostrinia nubilalis Hubn. in the agroclimatic conditions of Alsace. Ann. Zool. Ecol. Anim. 9: Thiéry, D. 2008: Les tordeuses nuisibles à la vigne. In: Ravageurs de la vigne, Féret, Bordeaux. Thiéry, D. & Xuéreb, A. 2003: Relative abundance of several larval parasitoids of Lobesia botrana on different varieties of grapes. IOBC/WPRS Bull. 26(8): Thiéry, D. & Xuéreb, A. 2004: Vers une lutte biologique contre Eudémis (Lobesia botrana). In: Conférence Internationale Mondiaviti Bordeaux: Thiéry, D, & Moreau, J. 2005: Relative performance of European grapevine moths (Lobesia botrana) on grapes and other hosts. Oecologia 143: Thiéry, D., Xuéreb, A., Villemant, C., Sentenac, G., Delbac, L., & Kuntzman, P. 2001: Larval parasites of vineyards tortricids: a brief overview from 3 french vine growing areas. IOBC/WPRS Bull. 24(7): Thiéry, D., Yoshida, T., & Guisset, M. 2006: Phytomyptera nigrina (Meigen)( Diptera, Tachinidae) parasite of the first generation of the European grapevine moth larvae in several vineyards of the Roussillon area. The Tachinid times 19: 1-4. Xuéreb, A, & Thiéry, D 2006: Does natural larval parasitism of Lobesia botrana vary between years, generation, density of the host and vine cultivar? Bull. Entomol. Res. 96:

217 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Life history of Lobesia botrana on Daphne gnidium in a Natural Park of Tuscany A. Lucchi, L. Santini University of Pisa, Dept C.D.S.L., Sect. Entomologia agraria, Pisa, Italy alucchi@agr.unipi.it Abstract: The life cycle and the ecology of the grapevine moth (GM) Lobesia botrana (Den. & Schiff) feeding on Daphne gnidium L. (to date considered the original host for L. botrana) were investigated in a Tuscan Natural Park in the last 2 years. The studied environment was located 2 km away from the beach and hosted numerous shrubs of D. gnidium on which L. botrana has been feeding forever, in the absence of Vitis vinifera. In the mentioned environment the moth seasonal flights were monitored and the larval feeding behaviour along the whole season was defined. Key words: Grapevine moth, Tortricidae, life cycle, spurge flax, Thymelaeaceae Introduction The grapevine moth (GM) Lobesia botrana (Denis & Schiffermüller) (Lepidoptera Tortricidae) is an herbivorous species feeding on several plants growing in the Mediterranean area as grapevine, olive, rosemary, myrtle, Daphne gnidium and others (Bovey, 1966; Coscollà, 1997). Along with the grapevine, the spurge flax D. gnidium (DG) is regarded as the most important host plant and suggested as the original host for L. botrana (Marchal, 1912, Picard, 1920; Thiery, 2005). DG is an evergreen wild shrub belonging to Thymelaeaceae, a cosmopolitan family of flowering plants in the order Malvales. Plants of the family, that is composed of about 50 genera and 900 species, are known for their scented flowers and poisonous berries. DG grows in the Mediterranean area on shallow, arid, stony soils, often on hillsides. Several studies focused on DG antibacterial/antimycotic activity expressed by leaves methanol extracts and a recent survey demonstrates the insecticidal activity of their leaf ethanolic extracts (Maistrello et al., 2005). Very recently, chemical volatiles extracted from DG leaves and flowers were shown to be attractive, in laboratory conditions, on ovipositing females of GM as well as the wind tunnel preference of mated GM females for a synthetic mimic of DG volatiles over V. vinifera volatiles was proved (Tasin et al., 2010). The life history of GM on DG was investigated, at the beginning of the last century, in France. More recently, some observations on the ecology of GM on V. vinifera and DG were performed in Apulia and Sardinia (Italy) (Nuzzaci and Triggiani, 1982; Luciano et al., 1988). In both cases the association GM-DG was studied in vineyards and olive crops characterized by the presence of DG in their close surroundings as well as in forested and grassland ecosystems in the absence of V. vinifera. In such environments the moth carried out 3-4 flights per year, usually reaching higher population and showing higher parasitization rates on D. gnidium in comparison with those reached on V. vinifera. Our research has been accomplished in a wide wooded area of the Natural Park of Migliarino-San Rossore-Massaciuccoli ( This Park extends for 24,000 hectares on a coastal area including parts of the provinces of Pisa and Lucca and is 197

218 198 characterized by a typical Mediterranean climate and flora (Pinus pinea, Quercus ilex, Q. peduncolata, Fraxinus sp. etc.). The studied environment was located 2km away from the beach; it hosted numerous shrubs of DG on which GM has been feeding forever, in the absence of Vitis vinifera (the few closest vineyards are about 7km away). Material and methods Adult monitoring Two pheromone traps (Intrachem Nuova DTI, were used to monitor GM flights from August 2008 to October 2009, in two different fields located 1km far-away. The pheromone lures were monthly replaced and the captures weekly counted. Samplings DG density was expressed delimiting 2 rectangular transects of 400m 2 and noting down the number of DG shrubs growing within their perimeter. GM nests occurring on a sample of DG shrubs (n = 1000) growing in different fields of the studied area were counted in June. Then, 50 apical shoots of DG per survey (length about 20cm) were collected from May to October A total of 8 surveys were done in 4 different fields characterized by high DG density. The variation of GM larval population structure during the year has been expressed dividing the larval instars in two classes: one consisting of the first three instars, the other formed by the last two instars. GM eggs, larvae and pupae collected on the shoots and within the larval nests were individually separated in little plastic cages to obtain GM adults and parasitoids. Results and discussion In 2009, DG blooming took place from the end of June to November. Along all that period of time DG could be seen in full bloom and in fruit simultaneously. The density of DG shrubs in the whole studied area was not homogeneous but varied depending on the sun exposure and the soil features. Indeed, DG preferably spreads over sunny clearings characterized by sandy soils. To this regard, the first delimited transect hosted 182 DG shrubs whereas the second only 22. During the three generations eggs were observed only on the leaves and never on other plant organs. Every so often the eggs were laid on the lower leaf surface (280 out of 305), hardly ever on the upper leaf surface (25 out of 305). Most of the eggs observed on the upper leaf surface have been laid at the end of October (13 out of 25).

219 199 Number of males per trap per week Trap 1 Trap 2 Figure 1: Male captures of L. botrana in two areas of the National Park of Migliarino-San Rossore-Massaciuccoli in Pheromone lures have been replaced on: April 9, May 14, June 16, July 10, August 14 and September 26. In the June surveys, the average number of GM nests on DG shrubs was about 4, with a maximum of 19 nests per shrub. generation larvae fed on the tender apical leaves of the shoot, fastening some of them to build a nest in which they carried through the larval development (Figures 2a-d). Rarely if ever, pupae were observed within the larval nests. Second generation larvae kept on feeding on tender leaves that were available along the whole season, though the young larvae also penetrated the flower buds, feeding within them. In this case, leaves, flower buds and flower remnants were used by the larva to build the nest (Figure 2e). A similar behaviour was observed for the third generation larvae, that fed also on flowers and green and red fruits (Figures 2f-i). The features of the third generation nests were very similar to those of the second generation. In the first two generations one single nest was visible on the apical part of each shoot. On the contrary, during the last generation, up to 5 nests per single shoot could be often observed. In the field surveys carried out from June to October 2009, the constant presence of all the larval instars was always observed. As known for many Mediterranean vine-growing areas, also in the Tuscan vineyards the complex of GM primary parasitoids is mainly composed by several species of hymenopteran Ichneumonoidea and Chalcidoidea and by one species of Diptera Tachinidae (Bagnoli and Lucchi, 2006). Nevertheless, in the Natural Park of Migliarino-S. Rossore-Massaciuccoli, only Campoplex capitator Aubert (Ichneumonidae Campopleginae) was obtained from GM larvae along the whole season. This wasp is a larval parasitoid, widespread in the European vineyards and considered the most effective on GM. It is particularly efficient early in the season and attacks mainly third and fourth instar larvae of GM.

220 200 Figure 2: Larval activity of L. botrana on D. gnidium in the 1 st generation (a-d), 2 nd generation (e), and 3 rd generation (f-i).

221 201 In our research, two males and five females of C. capitator emerged from GM larvae of the first generation and two females emerged from larvae of the second generation, with an average parasitisation rate of about 10% for each survey. This percentage is similar to the one expressed by the wasp in the vineyards. Unexpectedly, no other parasitoids have been collected from GM eggs, larvae and pupae. For this reason, further observations could be useful to clear up the relation between GM feeding on DG and associated natural enemies. DG shrubs have been visited along the season by a number of insect species searching for nectar and/or pollen. Among these the lepidopteran Pieris rapae (L.) (Pieridae) and the hymenopteran Scolia flavifrons F. (Scoliidae) were the most frequent and particularly abundant in the month of July. Moreover, some larvae of Cryptoblabes gnidiella Milliére were found feeding on the shoots, but only in the October surveys.in summary, the following conclusions can be drawn: - Flights and phenology of GM feeding on Daphne gnidium in the Natural Park of Migliarino-San Rossore-Massaciuccoli matched with those of GM feeding on grapes in the Tuscan vineyards; - GM eggs were mainly laid on the lower leaf surface of DG; GM larvae addressed their feeding activity on different DG plant tissues along the season (leaves in the 1 st generation; leaves and flower buds in the 2 nd generation; leaves, flower buds, flowers and fruits in the 3 rd generation); - GM pupae were rarely observed within the larval nests; - The Ichneumonid Campoplex capitator was the only parasitoid obtained by GM in Acknowledgements Many thanks are due to Patrizia Mazzarisi and Elena Pozzolini (University of Pisa) for their kind help in the field surveys. We are also grateful to Dr. Bruno Bagnoli (CRA-ABP Florence) for the useful suggestions and Dr. Augusto Loni (University of Pisa) for the parasitoid identification. References Bagnoli, B. & Lucchi, A. 2006: Parasitoids of Lobesia botrana in Tuscany. IOBC/WPRS Bull. 29(11): Bovey, P. 1966: Super-famille des Tortricoidea. In: Balachowsky, Entomologie appliqué à l'agriculture. Tome II. Lépidopterés, premier volume. Masson et Cie, Paris: e Coscollà, R. 1997: La polilla del racimo de la vid (Lobesia botrana Den y Schiff.). Ed. Generalitat Valenciana, Conselleria de Agricultura, Pesca y Alimentacion, 613 pp. Luciano, P., Delrio, G., Prota, R. 1988: Osservazioni sulle popolazioni di Lobesia botrana (Den. & Schiff.) su Daphne gnidium L. in Sardegna. Atti XV Congr. naz. ital. Entomologia, L Aquila, 13-17/06/1988, pp Maistrello, L., Lopez, A., Soria, F. J., Ocete R. 2005: Growth inibitory activity of Daphne gnidium L. (Thymelaeaceae) extracts on the elm leaf beetle (Col., Chrysomelidae). J. Appl. Entomol. 129(8): Marchal, P. 1912: Rapport sur les Travaux Accomplis par la Mission de l'etude de la Cochylis et de l'eudémis pendant l'année Beranger Ed., 326 pp.

222 202 Nuzzaci, G., Triggiani, O. 1982: Note sulla biocenosi in Puglia della Lobesia (Polychrosis) botrana (Schiff.) (Lepidoptera: Tortricidae) infeudata a Daphne gnidium L. Entomologica 18: Picard, F. 1920: La lutte contre la cochylis par la choix des cépages et par la culture de plantes attractives. Progr. agric. et vitic. 41(28): Tasin, M., Bäkman A.-C., Anfora, G., Carlin, S., Ioriatti, C., Witzgall, P. 2010: Attraction of female grapevine moth to common and specific olfactory cues from two host plants. Chemical Senses 35(1): Thiéry, D. 2005: Vers de la grappe: les connaître pour s en proteger. Vigne & Vin, Bordeaux, 60 pp.

223 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Current status of grapevine leafminers in north-eastern Italy C. Duso 1, A. Pozzebon 1, M. Baldessari 2, G. Angeli 2 1 Department of Environmental Agronomy and Crop Science, University of Padua, Agripolis, Legnaro (PD), Italy. 2 FEM, S. Michele all Adige, Trento, Italy. Abstract: Up to 2007, two species of grapevine leafminers were known in Italy as well as in Europe: Phyllocnistis vitegenella Clemens and Holocacista rivillei (Stainton). They are usually considered minor pests and do not require specific control measures. However, an increasing importance of these pests has been observed in north-eastern Italy in recent growing seasons. An additional leafminer species belonging to the genus Antispila was also detected in 2007 in north-eastern Italy. The origin of this species, which is new for Europe, is still unknown. The phenology, pest status, and natural control of these leafminer species have been investigated in a number of vineyards located in the Veneto and Trentino regions. The implications of leafminer occurrence for integrated pest management (IPM) on grapevines are discussed. Key words: Phyllocnistis vitegenella, Holocacista rivillei, Antispila sp., grape, leafminer, pest status, natural control Introduction Up to 2007, two leafminer species were known to occur in European vineyards, i.e. Holocacista rivillei (Stainton) (Lepidoptera: Heliozelidae) and Phyllocnistis vitegenella Clemens (Lepidoptera: Phyllocnistidae). In 2007, an additional leafminer species, belonging to the genus Antispila (Lepidoptera: Heliozelidae) was detected in north-eastern Italy (Baldessari et al., 2009a). Holocacista rivillei, native of the Mediterranean Basin, occurs in several viticultural areas of southern Europe (Mariani, 1941; Dal Rì & Delaiti, 1994). Mature larvae overwinter inside lenticular cases usually attached to the trunk. In spring, overwintered larvae pupate and adults emerge from May onwards. After mating, females lay eggs under the leaf epidermis. The larvae produce a typical mine, starting with a narrow gallery that subsequently leads to an oval blotch. Mature larvae construct a lenticular case and then abandon the leaves inside it. The cases fall and most of them hit the trunk where they remain attached to the bark. The resulting oval-shaped hole on the leaf blade identifies the evacuated mine (Marchi, 1956). The pest can develop 2-3 generations per year. Infestation levels rarely attain economic levels probably because the larvae are parasitized by a number of Eulophids (Camporese & Marchesini, 1991; Alma, 1995, Dal Rì & Delaiti, 1992). Outbreaks of H. rivillei are presumably induced by the disruption of interactions between the pest and its antagonists due to the repeated use of non-selective pesticides (Alma, 1995; De Tomaso et al., 2008). Phyllocnistis vitegenella is a Nearctic species detected in European vineyards in 1995 (Posenato et al., 1997). The species was reported first in north-eastern Italy (Vicenza province) then in other areas of Italy, in Slovenia and Switzerland (Posenato et al., 1998; Marchesini et al., 2000; Villani, 2002; Reggiani & Boselli, 2005; Seljak, 2005; M. Jermini and G. Angeli, pers. comm.). Adults overwinter inside evergreen canopies (e.g. conifers) or bark crevices. In May, adult females lay eggs on grape basal leaves and a first larval generation soon develops. The larvae produce tortuous mines more than 10cm in length. The 203

224 204 mature larvae pupate in a small silken chamber located at the distal end of the mines. Phyllocnistis vitegenella can complete four generations per year. Infestation can reach high levels in late summer, when apical leaves can be completely affected by mines. More larvae can occur on the same leaf with serious implications for physiological processes. The coalescence of mines can favor leaf rot and their drop. This species is usually kept under control by a complex of native Eulophidae (Marchesini et al., 2000). Moderate infestations are commonly encountered in vineyards contiguous to hedgerows or stand margins where adults can overwinter in great numbers. Local outbreaks are probably induced by insecticide applications that disrupt the balance between this leafminer and its natural enemies. During the summer of 2007, a new leafminer species belonging to the genus Antispila (Lepidoptera: Heliozelidae) was recorded in north-eastern Italy, initially in the Trento province (Baldessari et al., 2009a). This species has subsequently been found in the Veneto region. The mine features are similar to those produced by H. rivillei, but without the initial elongated trail. The identity of this species is still under study. Eight species in the genus Antispila are associated to grapes in North America or in Japan (van Nieukerken, com. pers.). The Japanese species have been well described, in contrast with the American species which are not frequently detected. In this paper, we present the preliminary results of observations on the phenology, pest status and natural control of grape leafminers, carried out in vineyards located in north-eastern Italy. Materials and methods Observations were carried out in two vineyards located in the Trento province (i.e., Borgo Valsugana and Borghetto), and in two vineyards located in the Treviso province (i.e., Valdobbiadene and Farra di Soligo) during 2008 and The occurrence of leafminers had been noticed by growers in previous seasons but the identity of the species and their incidence were unknown. These vineyards were under IPM; in particular they were treated with insect growth regulators (IGRs) or thiamethoxam (usually one application per year) to control Scaphoideus titanus Ball. In Borghetto, a number of plots were treated with different insecticides and an untreated control was also considered. Pest species occurrence, their phenology and infestation levels were investigated. To this end, leaves per vineyard were examined monthly during the growing season under a dissecting microscope. The number and type of mines as well as the presence of each development stage of leafminers were recorded. To assess the incidence of parasitism, a number of leaf sections with mines containing mature larvae and pupae were singly confined in test tubes. These tubes were maintained at outdoor conditions until the leafminers or their parasitoids emerged. Results and discussion Phyllocnistis vitegenella was detected in all the vineyards considered in this study but was abundant in Valdobbiadene. Antispila sp. occurred in Borgo Valsugana, Farra di Soligo and to a lesser extent, in Valdobbiadene. Holocacista rivillei was detected only in Borghetto. In the latter vineyard, H. rivillei population densities clearly increased from June to September 2008, when high infestation levels (98% of leaves with an average of 7.9 mines per leaf) were observed in untreated plots. In insecticide treated plots, infestation levels were lower (40-90% of leaves). The incidence of parasitism on H. rivillei populations was about

225 205 30%. During 2009, leafminer densities were very low, probably because two insecticide treatments were applied (against other pests) when the first generation of H. rivillei was developing. In 2008 and 2009, H. rivillei completed two generations, confirming the results obtained in previous studies carried out in the same region (Dal Rì & Delaiti, 1992). It should be stressed that climatic conditions in the Trento province (close to Alps) are quite different from those recorded in other Italian regions where the pest can develop three generations per year (Camporese & Marchesini, 1991). In Valdobbiadene, about 90% of the leaves were infested by Ph. vitegenella in the late summer of One year later, Ph. vitegenella infestation levels increased from 19% of leaves in June to 88% in September, when an average density of mines per leaf was attained. In this season the incidence of parasitism fluctuated from 7% in July to 66% in October. The identity of the parasitoid species is under study. In this locality the pest completed four generations confirming the results of studies conducted in the same region (Posenato et al., 1997; Marchesini et al., 2000). No data were available on the life cycle of Antispila sp. Observations carried out in 2008 in Borgo Valsugana, showed that mature larvae overwintered inside cases usually attached to the trunk. The larvae pupated in May and the first adults were detected in early June. The first mines were observed during the second half of June. Mature larvae form oval cases that fall, hitting trunks or training stakes. This behavior is similar to that exhibited by H. rivillei (Marchi, 1956). During 2008, the first cases were detected during the first half of July. An additional generation was completed from August to September. During this season, 97% of leaves were infested in September with an average of 6.3 mines per leaf. In Borgo Valsugana, two generations were completed during 2008 and 2009 (Baldessari et al., unpubl. data). The incidence of parasitism in this vineyard was negligible. The milder climate of Farra di Soligo allowed the completion of three generations in At the end of this season, 100% of leaves were infested by Antispila sp. with an average number of 9.1 mines per leaf. The incidence of parasitism was significant (about 77%) in July (Duso & Pozzebon, unpubl. data). The identity of parasitoids encountered in this survey is under study. Leafminers can locally be of major concern for growers, making effective insecticides a requirement. However, no insecticides are currently registered for the control of grape leafminers in Italy or indeed in Europe. Recently, it has been proved that some IGRs (e.g. flufenoxuron), pyrethroids (lambda-cyalothrin) and neonicotinoids (thiamethoxam), applied to control S. titanus, can affect H. rivillei (Baldessari et al., 2009b). Further studies are required to assess the risk associated to leafminer species, the side-effects of pesticides and the incidence of natural control on their populations. It should be mentioned that Ph. vitegenella and H. rivillei can share a number of parasitoid species (Marchesini et al., 2000). The identification of specific sex pheromones represents another promising topic for monitoring these pests under IPM tactics. References Alma, A. 1995: Ricerche bio-etologiche ed epidemiologiche su Holocacista rivillei Stainton (Lepidoptera Heliozelidae). Redia LXXVIII: Baldessari, M., Angeli, G., Girolami, V., Mazzon, L., van Nieukerken, E. J. & Duso, C. 2009a: Antispila sp. minatore fogliare segnalato in Italia su vite. L Informatore Agrario 15: Baldessari, M., Delaiti, M., Penner, F. & Angeli, G. 2009b: Difesa dalle cicaline della vite utile anche su Holocacista rivillei. L Informatore Agrario 46:

226 206 Camporese, P. & Marchesini, E. 1991: Nota preliminare sulla minatrice delle foglie di vite Holocacista rivillei (Stainton) (Lepidoptera: Incurvariidae Heliozelinae) nel Veneto. Atti XVI Congresso Nazionale Italiano di Entomologia: Dal Rì, M. & Delaiti, L. 1992: Segnalata in alcuni vigneti della Bassa Vallagarina «Minatrice fogliare della vite». Terra Trentina 5: Dal Rì, M. & Delaiti, L. 1994: Minatrice fogliare attratta dal rosso. Insetti minori in viticoltura. Terra Trentina 6: De Tomaso, B., Romito, A., Nicoli Aldini, R. & Cravedi, P. 2008: Minatrice fogliare della vite segnalata in Puglia. L Informatore Agrario 31: Mariani, M., 1941: Una infestazione di Antispila rivillei Stt. su vigneti nel Palermitano. Giornale di Scienze Naturali ed Economiche 43: Marchesini, E., Posenato, G. & Sancassani, G. P. 2000: Parassitoidi indigeni della minatrice americana della vite. L Informatore Agrario 10: Marchi, G. 1956: Ricerche su un Lepidottero Eliozelide l Holocacista rivillei (Staint.), minatore delle foglie della vite. Bollettino Istituto Entomologia Università di Bologna 22: Posenato, G., Girolami, V. & Zangheri, S. 1997: La minatrice americana un nuovo fillominatore della vite. L Informatore Agrario 15: Posenato, G., Tosi, L., Marchesini, E., Miotti, G., Malagnini, V. & Sancassani, G. P. 1998: Prime segnalazioni di Phyllocnistis vitegenella Clemens in vigneti italiani. Quaderni di Scienze Viticole ed Enologiche Università di Torino 22: Reggiani, A. & Boselli, M. 2005: Espansione nel Nord Italia della minatrice americana della vite. L Informatore Agrario 61: Seljak, G. 2005: The American leaf miner (Phyllocnistis vitegenella Clemens) of grape vines is already in Slovenia (in Slovenian). SAD, Revija za Sadjarstvo, Vinogradnistvo in Vinarstvo 16: Villani, A. 2002: La minatrice americana della vite in Friuli Venezia Giulia. Notiziario ERSA 15:

227 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Factors affecting the post-release dispersal of Trichogramma cacoeciae Marchal in the vineyard G. Hommay 1, J. C. Kienlen 1, C. Gertz 1, C. Bihry 1, J. Pizzol 2 1 UMR Santé de la Vigne et Qualité du Vin. INRA, Institut National de la Recherche Agronomique & Université de Strasbourg, 28 rue de Herrlisheim. BP 20507, Colmar, France; 2 UR 880, Unité de Recherches Intégrées en Horticulture, INRA PACA 400 route des Chappes BP Sophia Antipolis, France Abstract: Several releases of 20,000 and 40,000 individuals of Trichogramma cacoeciae were made at a central point in a vineyard in order to follow the dispersal of trichogrammes and the distribution of parasitism on Ephestia kuehniella egg-cards. The dispersal of trichogrammes essentially occurred along the release row and over the nearest rows, since the vine rows formed a plant screen that channelled their dispersion. The daily captures of trichogrammes increased with the number of trichogrammes released and decreased over time. It increased with minimal temperature and solar radiation, but did not depend on other meteorological factors. The number of parasitised egg-cards decreased with the distance from the release point and increased with the amount of wind received at the control points. The doubling of the number of trichogrammes released influenced parasitism by increasing both the discovery rate and the exploitation rate of egg-cards. Key words: Trichogramma cacoeciae, dispersal, weather conditions, inundative release Introduction The dispersal of trichogrammes (Trichogramma spp.) depends on the activity characteristics of the species and ecotype and on their physiological state, as well as on meteorological conditions, on the localisation and density of the hosts (Forsse et al., 1992) or of their chemical cues and on the type of crop in which they are released (Stein, 1961; Hendricks, 1967). Wind, temperature and light intensity are the main meteorological variables that influence the dispersal of trichogrammes and the distribution of parasitism (Stein, 1961; Kot, 1964, 1979; Hendricks, 1967; Yu et al., 1984; Smith, 1988; Greatti & Zandigiacomo, 1995; Glenn & Hoffmann, 1997; Fournier & Boivin, 2000). The activity of trichogrammes is greatly reduced at low temperatures (Forsse et al., 1992). Trichogrammes exhibit threshold flight temperatures which vary with the light intensity (Kot, 1979). Temperatures of C increase the active flight of trichogrammes (Kot, 1964), but higher temperatures have proven to be detrimental to their activity (Fye & Larsen, 1969; Glenn & Hoffmann, 1997). Trichogrammes disperse both passively with the wind and actively by flying (Smits, 1982) or by walking over short distances on the vegetation (Keller, 1987). This active dispersal is particularly involved in the search for a suitable host and can vary according to the structure of the vegetation (Fye & Larsen, 1969; Yu et al., 1984; Pizzol et al., 1997). The development of biological control using trichogrammes requires knowledge of their dispersal to improve the release method. Little research has been done on this subject in the vineyard (Babi & Voegelé, 1990; Castaneda-Samayoa et al., 1993; Glenn & Hoffmann, 1997). We were therefore interested in the dispersal of Trichogramma cacoeciae Marchal and in the resulting distribution of parasitism according to meteorological conditions, within the framework of 207

228 208 biological control experiments of the grapevine moth Lobesia botrana Denis & Schiffermüller. Several methods have been used in the field to evaluate the dispersal of Trichogrammes: radioactive marking (Stern et al., 1965), distribution of sticky traps (Hendricks, 1967) or of egg-cards of the rearing host (Stein, 1961; Fye & Larsen, 1969; Smith, 1988; Castaneda- Samayoa et al., 1993; Glenn & Hoffmann, 1997; Fournier & Boivin, 2000), and the recording of egg parasitism of the natural host around the release points (Babi & Voegelé, 1990; Greatti & Zandigiacomo, 1995). More recently, the post-release dispersal was studied on other parasitoids, using mark-release-recapture with immunological markers (Hougardy & Mills, 2006). In this work, funnels containing water and a surfactant were tested to study the dispersal of trichogrammes from a single release point. Dispersal was evaluated using the number of trichogrammes recovered in funnels and parasitism of egg-cards. Material and methods Trichogrammes The trichogrammes released belonged to the species T. cacoeciae (T c). The strain originated from parasitised eggs of L. botrana collected in the Rouffach vineyard (Haut-Rhin, eastern France) in The trichogrammes were produced at INRA Antibes on eggs of the Mediterranean flour moth, Ephestia kuehniella Zeller, that had been sterilised by ultraviolet radiations. E. kuehniella (E k) eggs parasitised by T c were delivered to Colmar in refrigerated parcels. Upon arrival, trichogrammes continued their development in the laboratory at temperatures varying from 15 to 25 C so as to obtain emerging adults at the scheduled release dates. Emergence rate under laboratory conditions was determined for each shipment at the production site. Methods The experimental plot, located in Wintzenheim (Haut-Rhin) and planted with Pinot blanc vines, has been conducted under organic pest management for over 20 years. Vine rows were planted in a north-south direction with a spacing of 1.95m and a distance of ca 0.95m between two vine stocks. A release of 20,000 or 40,000 T c emerging from E k eggs took place at 9:00 am at a central point of the plot (four and eight glass tubes respectively, containing a card with ca eggs, plugged with cotton and placed between 1.1 and 1.3m in height). A streak of honey laid inside each tube ensured the feeding of young adults. The tubes were shaded by the vine leaves. Tubes were opened and left in place throughout the experiment, because of the spreading of T c emergence, which generally took three days. The experiment was renewed six times for each release dose between June and September, from 1999 to According to a method used by G. Sentenac (Institut Français de la Vigne et du Vin, Beaune, France), 18cm diameter white funnels containing 800ml of water, supplemented with a few drops of surfactant (dish washing liquid), were placed next to the release point and at 48 points arranged all around (7 rows x 7 vine stocks). Funnels were attached at 1.3m in height on the second straining wire, which corresponded to the main fruit-bearing area. Dispersals were recorded at 9:00 am, 1, 2, 3, 4 and 7 days after release. The spacing chosen along the row was longer than between rows because previous experiments had shown a greater dispersal along a row than across rows. After each recording, the content of each funnel was emptied into a numbered bottle and the liquid was replaced immediately. The bottles were brought back to the laboratory and their content was filtered through sieves to recover the trichogrammes, which were examined under a binocular microscope. Six control recordings

229 209 without T c releases were made from May to July to determine the size of natural populations. Trichogrammes were counted and males were placed in tubes containing 70% ethanol and later determined according to their genital apparatus by Dr B. Pintureau (INRA Villeurbanne). At the end of the experiment, the release tubes were closed and egg-cards were brought back to the laboratory and incubated at 20 C to control residual emergence. In 2001, cards bearing from 100 to 500 sterilised E k eggs (mean ± sd = 172 ± 69 eggs), 15 days old at the most, were distributed near each of the 49 funnels and replaced at the same time as the water. Ephestia eggs were stuck on a 6 x 1cm yellow cardboard over an area of 1cm 2, using a non-repellent glue. The scales of dead L. botrana were stuck on the opposite end of a series of 49 cards. This series was compared with a series of 49 cards with only E k eggs. The cards brought back from the vineyard were isolated individually into a glass tube, plugged with cotton and kept at 20 C in the laboratory, in order to assess the number of parasitised eggs. Trichogrammes emerging from parasitised cards were later controlled to verify whether or not they belonged to the species released. Discovery rate (proportion of cards exhibiting at least one parasitised egg) and exploitation rate (proportion of parasitised eggs in the cards discovered by the parasitoids) were calculated. Records of minimal and maximal temperatures ( C), solar radiation (kj/m2), total precipitation (mm), maximal wind speed (km/h) and wind direction (16 different cardinal directions) at 2m above the ground, were collected in Colmar (located about 5km from the experimental plot) during the experiments. Data analysis The number of T c collected in funnels and their distance to the release point, were analysed by analysis of variance (ANOVA) according to the number of T c released, the number of days after the release and the row number. For the first four consecutive days of records, the daily number of T c collected and their distance to the release point were analysed by ANOVA, according to the meteorological conditions of the day before the recording day. Because of the high number of catches at the release point, the central trap was excluded of the calculations. The different recording periods were considered to be replicates. For 2001, the same analyses were made on the number of parasitised cards. Additional ANOVA was made on the daily number of T c collected and of parasitised cards according to the amount of wind received by the respective traps (max wind speed x angle of the trap with the daily prevailing wind). Results and discussion The capture of a small number of trichogrammes ( 0Tc release dose, Figure 1), during prerelease monitoring, leads us to believe that there is a negligible influence of natural trichogrammes on the results of dispersal experiments. Moreover, the number of trichogramme males not belonging to the released species (Trichogramma evanescens Westwood and Trichogramma semblidis (Aurivillius) were identified) did not exceed 2 % of the trichogrammes trapped after the different releases, with a maximum of 8 % observed at the end of August. Control of emergence of the released parasitoids made in the Antibes laboratory at 25 C from cards issued from the same batches showed that 90% of the T c emerged over a period of three days, and that 5-7% emerged during the next two days. Except for the release on 3 July 2000, where a third of the T c was drowned inside the tubes because of heavy rainfall accompanied by violent wind, 90-95% of the T c emerged out of the parasitised eggs within the experiment period. Some of the T c that remained in the tubes at the end of the recordings survived in the laboratory for up to 14 days after release.

230 210 ANOVA showed that the number of days following the release had a negative effect on the number of trichogrammes trapped (Figure 1). The size of captures was greater in the release row than in the nearby rows (Figure 2). The number of trichogrammes trapped was significantly higher with 40,000 than with 20,000 T c released, but there was an interaction with the effect of the release dates. Minimal temperature and solar radiation had a significant positive effect on the daily captures of trichogrammes, while the other meteorological factors had no significant incidence. Neither the amount of wind received by the traps nor the distance from the release point had a significant effect on the number of trichogrammes trapped. Mean number of Tc trapped release dose 0 Tc 20,000 Tc 40,000 Tc D+1 D+2 D+3 D+4 D+7 day after release Figure 1: Mean number of T. cacoeciae trapped per release, according to the number of days after the release and the number of T. cacoeciae released at a central point in a vineyard. Mean number of Tc trapped vine row release dose 20,000 Tc 40,000 Tc Figure 2: Mean number of T. cacoeciae trapped per release (release point excluded), according to the vine row, where row 4 is the release row. When 20,000 T c were released, the daily mean distance travelled increased until the seventh day of recording (Figure 3). For 40,000 T c released, the daily mean distance was greater and increased during the first three days and then decreased. ANOVA showed that

231 211 none of the factors tested significantly affected the daily mean distance covered by the parasitoids. Mean distance travelled (cm) ,000 Tc 40,000 Tc D+1 D+2 D+3 D+4 D+7 Day after release Figure 3: Daily mean distance travelled by individuals of T. cacoeciae, according to the number of T. cacoeciae released. Monitoring of E k egg-cards made in 2001 showed a similar evolution of the number of parasitised eggs and of the number of trichogrammes trapped. The percentage of parasitised cards showed no difference between cards with or without L. botrana scales (5.5 and 5.0%). ANOVA showed that the number of parasitised cards depended on the number of T c released and on the row number, but not on the number of days following the release. The number of parasitised cards declined from the release row to the most remote rows. The number of daily parasitised cards was significantly higher for 40,000 than for 20,000 T c released (t test, P < 0.05, n = 124). ANOVA on the daily number of parasitised cards showed that this number decreased with the distance from the release point and increased with the amount of wind received by the control points. The other meterological factors had no significant effect. The most remote parasitised cards were located 10.48m from the release point as soon as the first day after release. The discovery rate (Dr) of cards was low, with a mean of 4.2% for the set of recording points, but 81.5% at the release point. Dr was significantly higher for 40,000 T c released (6.1%) than for 20,000 (2.3%) (Khi2 test, P < 0.001) and it significantly decreased with the distance from the release point (d), if the release point was excluded (Dr = d, r 2 = 0.576, P < 0.001, n = 15). The exploitation rate of eggs on cards was reduced, with a mean of 19% for the set of points and of 27.9% at the release point. The exploitation rate was significantly higher for 40,000 T c released (18%) than for 20,000 (15.5%) (Khi2 test, P < 0.001) and showed no correlation with the distance from the release point. Experiments conducted in the vineyard with T. cacoeciae and based on egg parasitism around a release point showed that parasitoids dispersed better along the release row than on the nearby ones (Babi & Voegelé, 1990; Castaneda-Samayoa et al., 1993). This was confirmed in our observations. The comparison of the number of Trichogrammes caught in funnels and the number of parasitised cards showed that the use of funnels was an suitable method for monitoring T c dispersal. The dispersal of T c in the vineyard occurred essentially along the release row and on the closest rows, because the vine rows formed a plant screen that channelled their dispersion. In the same way, in apple orchards (Yu et al., 1984) and in forests (Smith, 1988), Trichogramma minutum Riley dispersed regularly within a tree, but

232 212 downwind between trees. In a cotton field, Trichogrammes released at canopy level dispersed more evenly than those released above the canopy which spread downwind (Hendricks, 1967). The mean number of trichogrammes trapped with funnels gradually decreased with the number of days following the release. Trichogrammes probably foraged first around their release point and then disappeared outside the trapping network (probably firstly on plants growing between vine rows and secondly beyond the trapping area). Minimal temperatures had a positive effect on the number of trichogrammes recovered, probably due to the stimulation of their activity with rising temperature. Contrary to the number of parasitised cards, the numbers of trichogrammes trapped did not depend on the distance to the release trap and on the amount of wind received by the traps. Below a certain speed threshold, trichogrammes are able to fly against the wind (3m/sec according to Schread & Garman (1933), 1.9m/sec for T. minutum according to Steenburgh (1934)). Stronger winds cause a passive dispersal, whereas violent winds force the trichogrammes to hide under leaves or bark and to stop activity (Kot, 1964). Moreover, wind speed and direction can change during the day, which would explain this absence of direct relationship. Similarly to the observations of Fournier & Boivin (2000), the influence of solar radiation on the dispersal of trichogrammes could be observed. Several parasitised egg-cards were found at the greatest control distance (10.5m), as soon as the first day after release. In a preliminary dispersal experiment with eggcards (unpublished data) on a radius of 45.2m around the release point, a dispersal distance of 34m across the rows was observed as soon as the second recording day and the maximal dispersal distance of 45.2m was recorded 12 days after the release. According to Kot (1964), the maximal dispersal radius of T. evanescens and of T. cacoeciae was 40-50m and 60-70m, respectively, in most cases. Therefore, it was possible that T c scattered beyond the trapping area after a few days, causing underestimation of the distance travelled. The kairomones left by the host increase the searching activity of trichogrammes and prolong their retention in the target area (Lewis et al., 1976; Smits, 1982; Zaborskiet al., 1987). Lepidopteran wing scales left on egg laying sites are considered to be a source of kairomones which increase searching and parasitism of several species of trichogrammes (Gross et al., 1981, 1984; Noldus & van Lenteren, 1983). Thus, preliminary marking of E k eggs with L. botrana scales in the laboratory increased the discovery and exploitation rates by T. cacoeciae (Barnay et al., 1999), with a more pronounced effect on the discovery rate. Contrary to this result, the presence of L. botrana scales did not change the level of parasitism of E k egg-cards in the vineyard under our experimental conditions. Pre-release stimulation by host kairomones (Gross et al., 1981) would clarify their influence on dispersal in the vineyard. Hougardy and Mills (2006) suggested that to enhance local establishment in a fragmented environment, parasitoids should be given access to hosts prior to release. Thus, trichogrammes could be supplied with kairomones inside the release capsule. The effect of kairomones could be used in the vineyard to improve the efficacy of trichogrammes by maintaining them in the host distribution area and by increasing their discovery and exploitation rates. The discovery rate of eggs by T. cacoeciae was directly density-dependent in the laboratory, while the exploitation rate was inversely density-dependent (Barnay et al., 1999). The efficacy of parasitism relies essentially on the capacity of trichogrammes to discover eggs, which appeared very limited in this experiment. The discovery rate of E k egg-cards was only 4.2% on average for the set of recording points, while the exploitation rate was 19%. In a similar study on T. cacoeciae dispersal, Castaneda-Samayoa et al. (1993) observed 14.7% exploitation rate on egg-cards of Sitotroga cerealella Olivier. However, in our experiment, the mean discovery rate was 81.5% and the mean exploitation rate was 27.9% at the release

233 213 point. Thus, increasing the number of release points would have probably more effects than increasing the number of T c per release capsule. In an experiment on inundative release, increase of the release density from 400 to 800 points/ha made it possible to obtain a better parasitism of L. botrana with T. evanescens, but not with T. cacoeciae (Hommay et al., 2002). However, in another comparison (unpublished data), we observed that parasitism of T. cacoeciae increased with 600 release points/ha, and additional investigations are necessary to verify if higher densities could improve its efficiency. Acknowledgements We are very grateful to Dr B. Pintureau (INRA Villeurbanne) for the identification of male trichogrammes. Many thanks to A. Stentz, for providing access to his vine plot and for assistance on the project, to Dr L. Lapchin (INRA Antibes) and Dr E. Duchêne (INRA Colmar) for help in statistical analysis, and to G. Wagman for the revision of the English text. References Babi, A. & Voegelé, J. 1990: Dispersion des trichogrammes T. cacoeciae Marchal et T. daumalae Dugast et Voegelé (Hym.: Trichogrammatidae) dans le vignoble. In: ANPP (ed.), Deuxième conférence internationale sur les ravageurs en agriculture. Versailles, December 1990, pp Barnay, O., Pizzol, J., Gertz, C., Kienlen, J. C., Hommay, G. & Lapchin, L. 1999: Host density dependance of discovery and exploitation rates of egg patches of Lobesia botrana (Lepidoptera: Tortricidae) and Ephestia kuehniella (Lepidoptera: Pyralidae) by the parasitoid Trichogramma cacoeciae (Hymenoptera: Trichogrammatidae). J. Econ. Entomol. 92: Castaneda-Samayoa, O., Holst, H. & Ohnesorge, B. 1993: Evaluation of some Trichogramma species with respect to biological control of Eupoecilia ambiguella Hb. and Lobesia botrana Schiff. (Lep., Tortricidae). Z. Pflanzenkr. Pflanzenschutz 100: Forsse, E., Smith, S. M. & Bourchier, R. S. 1992: Flight initiation in the egg parasitoid Trichogramma minutum: effects of ambient temperature, mates, food, and host eggs. Entomol. Exp. Appl. 62: Fournier, F. & Boivin, G. 2000: Comparative dispersal of Trichogramma evanescens and Trichogramma pretiosum (Hymenoptera: Trichogrammatidae) in relation to environmental conditions. Environ. Entomol. 29: Fye, R. E. & Larsen, D. E. 1969: Preliminary evaluation of Trichogramma minutum as a released regulator of lepidopterous pests of cotton. J. Econ. Entomol. 62: Glenn, D. C. & Hoffmann, A. A. 1997: Developing a commercially viable system for biological control of light brown apple moth (Lepidoptera: Tortricidae) in grapes using endemic Trichogramma (Hymenoptera: Trichogrammatidae). J. Econ. Entomol. 90: Greatti, M. & Zandigiacomo, P. 1995: Postrelease dispersal of Trichogramma brassicae Bedzenkko in corn fields. J. appl. Entomol. 119: Gross, H. R., Lewis, W. J., Donald, A. & Nordlund, D. A. 1981: Trichogramma pretiosum: effect of prerelease parasitization experience on retention in release areas and efficiency. Environ. Entomol. 10:

234 214 Gross, H. R., Lewis, W. J., Beevers, M. & Nordlund, D. A. 1984: Trichogramma pretiosum (Hymenoptera Trichogrammatidae): Effects of augmented densities and distribution of Heliothis zea (Lepidoptera Noctuidae) hosts eggs and kairomones on field performances. Environ. Entomol. 13: Hendricks, D. E. 1967: Effect of wind on dispersal of Trichogramma semifumatum. J. Econ. Entomol. 60: Hommay, G., Gertz, C., Kienlen, J. C., Pizzol, J. & Chavigny, P. 2002: Comparison between the control efficacy of Trichogramma evanescens Westwood and of two Trichogramma cacoeciae Marchal strains against grapevine moth (Lobesia botrana Den. & Schiff.), depending on their release density. Biocontrol Science and Technology 12: Hougardy, E. & Mills, N. J. 2006: The influence of host deprivation and egg expenditure on the rate of dispersal of a parasitoid following field release. Biol. Contr. 37: Keller, M. A. 1987: Influence of leaf surfaces on movements by the hymenopterous parasitoid Trichogramma exiguum. Entomol. Exp. Appl. 43: Kot, J. 1964: Experiments in the biology and ecology of species of the genus Trichogramma Westwood and their use in plant protection. Ekologia Polska - Seria A. 15: Kot, J. 1979: Analysis of factors affecting the phytophage reduction by Trichogramma Westw. species. Polish Ecological Studies 5: Lewis, W. J., Jones, R. L., Gross, H. R. & Nordlund, D. A. 1976: The role of kairomones and other behavioral chemicals in host-finding by parasitic insects. Behav. Biol. 16: Noldus, L. P. J. J. & van Lenteren, J. C. 1983: Kairomonal effects on searching for eggs of Pieris brassicae, Pieris rapae and Mamestra brassicae of the parasite Trichogramma evanescens Westwood. Meded. Fac. Landbouww. Rijksuniv. Gent 48: Pizzol, J., Ciociola, A., Marro, J. P. & Tronchetti, F. 1997: Dispersion de deux espèces de trichogrammes (Trichogramma evanescens et Trichogramma voegelei) en culture de tomate en serre. In: ANPP (ed.), Quatrième conférence internationale sur les ravageurs en agriculture. Montpellier, December 1997, pp Schread, J. C. & Garman, P. 1933: Studies on the parasites of the oriental fruit moth I. Trichogramma. Bull. Conn. Agric. Exp. Stn 35: Smith, S. M. 1988: Pattern of attack on Spruce budworm egg masses by Trichogramma minutum (Hymenoptera: Trichogrammatidae) released in forest stands. Env. Entomol. 17: Smits, P. H. 1982: The influence of kairomones of Mamestra brassicae L. on the searching behaviour of Trichogramma evanescens Westwood. In: Les trichogrammes (INRA, ed.). Les colloques de l'inra n 9. Antibes, April 1982, pp Steenburgh, M. 1934: Trichogramma minutum Riley as a parasite of the oriental fruit moth (Laspeyresia molesta) in Ontario. Nat. Res. Counc. Canada 10: Stein, W. 1961: Die Verteilung des Eiparasiten der Gattung Trichogramma embryophagum cacoeciae (Htg.) in den Baumkronen nach seiner Massenfreilassung zur Bekämpfung des Apfelwicklers. Z. Pflanzenkr. Pflanzenschutz 68: Stern, V. M., Schlinger, E. I. & Bowen, W. R. 1965: Dispersal studies of Trichogramma semifumatum (Hymenoptera: Trichogrammatidae) tagged with radioactive phosphorus. Ann. Entomol. Soc. Am. 58: Yu, S. K., Laing, J. E. & Hagley, E. A. C. 1984: Dispersal of Trichogramma spp. (Hymenoptera: Trichogrammatidae) in an apple orchard after inundative releases. Environ. Entomol. 13: Zaborski, E., Teal, P. E. A. & Laing, J. E. 1987: Kairomone-mediated host finding by spruce budworm egg parasite, Trichogramma minutum. J. chem. Ecol. 13:

235 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Mating disruption field trials to control the vine mealybug Planococcus ficus A. Cocco 1, M. Coinu 1, A. Lentini 1, G. Serra 2, G. Delrio 1 1 Dipartimento di Protezione delle Piante, sezione Entomologia agraria, Università degli Studi di Sassari, Sassari,Italy, 2 Istituto per lo Studio degli Ecosistemi sede di Sassari, CNR, Li Punti, Sassari, Italy. acocco@uniss.it Abstract: A mating disruption experiment against the vine mealybug, Planococcus ficus, was carried out during 2008 in 3 commercial vineyards in northwestern Sardinia. Within each vineyard, 2 adjacent plots of variable size (0.2, 0.4, and 0.5ha, respectively) were established as mating disruption and control plot. Plastic dispensers (Suterra Inc., USA) with 100mg each of the synthetic sex pheromone were deployed in mid-may, before the first flight of adult males, at a rate of 625 dispensers per hectare. The effectiveness of the mating disruption technique was evaluated by comparing the number of males captured in pheromone traps, the leaf population density, the percentage of ovipositing females, and the crop damage. Male flights were monitored with traps baited with 0.01mg of the sex pheromone. The percentage of females with ovisacs and their fecundity were estimated by collecting adult females after each peak male flight and rearing them individually under laboratory conditions. The mealybug density was assessed from June to September by counting every 2 weeks nymph and female mealybugs on basal leaves of vines per plot. The crop damage was evaluated at harvest by assessing the percentage of damaged bunches on the same vines. In all the vineyards, the number of males captured in mating disruption plots was lower by far than in control plots, while the percentage of females with ovisac was significantly different only in the first generation. Females collected from the field after the first male flight produced more eggs than those collected after the second and third flight, with no significant difference between treatments. The density of mealybugs on leaves in mating disruption plots was lower than that in control plots, but the difference was not significant. Also, the percentage of damaged bunches at harvest did not differ between treatments. Mating disruption experiments to control the vine mealybug produced mixed results. This could be due to the limited size of experimental plots and/or the suspected parthenogenetic reproduction of Sardinian populations of P. ficus. Key words: Pseudococcidae, vine mealybug, sex pheromone Introduction The vine mealybug Planococcus ficus (Signoret) (Homoptera: Pseudococcidae) has become a key pest of vineyards in Sardinia (Lentini et al., 2008). Improvements in cultivation techniques, such as increased irrigation and nitrogen fertilization and diffusion of vigorous grapevine clones, led to a build-up of P. ficus populations (Dalla Montà et al., 2001). Vine mealybug infestations can determine in some grape-growing regions higher economic losses than attacks of the European grapevine moth Lobesia botrana (Den. et Schiff.) (Lepidoptera: Tortricidae). Heavy infestations of P. ficus can severely lower crop quality and yield. Mealybugs feed on phloem and excrete honeydew promoting the growth of sooty mold fungi that reduces photosynthetic activity in leaves. The presence of mealybugs in wine grapes lowers their quality for wine-making and renders table grapes unmarketable. Moreover, P. ficus can transmit viral diseases, such as the grapevine leafroll associated virus 215

236 216 (GLRaV) and is therefore considered a primary pest even at low population densities (Engelbrecht & Kasdorf, 1990). In Sardinia, P. ficus control programs rely on multiple in-season applications of organophosphate insecticides. The effectiveness of chemical applications is often limited because the majority of mealybugs are in concealed locations (under the bark or beneath the bud scale) (Godfrey et al., 2003). For these reasons, integrated pest management programs using selective and low-impact tools need to be developed to control the vine mealybug. The synthetic sex pheromone of P. ficus has been identified, commercially produced, and used to develop mating disruption programs to control the pest (Walton et al., 2006). The present paper presents the results of a mating disruption experiment conducted against the vine mealybug in northwestern Sardinia. Material and methods The study was performed during 2008 in 3 commercial vineyards ranging from 0.4 to 6.8ha; vine cultivars were Vermentino, Cannonau, and Chardonnay-Cabernet, while training systems were Tendone, Guyot, and Guyot-spur pruned, respectively. In all 3 vineyards, 2 adjacent plots of 0.2, 0.4, and 0.5ha, respectively, were established as mating disruption and control plot. Plastic dispensers (Suterra Inc., USA) loaded with 100 mg of the active ingredient (S)- lavandulyl senecioate were deployed at a rate of 625 dispensers per hectare. Dispensers were applied in mid-may, before the first flight of males, and hung to trellis wires at bunch height. With the aim to evaluate the effectiveness of the mating disruption technique, we compared (i) mealybug population density, (ii) population dynamics, (iii) age structure, (iv) percentage of ovipositing females and mean fecundity, and (v) crop damage at harvest on control and pheromone-treated plots. Population density was monitored in 2-3 transects per plot, with a minimum spacing of 10-20m each others and from plot edges. Each transect consisted of 10 adjacent vines in the same row. In each sampled vine, the first 2 leaves of the basal cane of the first spur were flagged and all the mealybug stages (first and second instars, third instars, females, females with ovisac, and ovisacs with eggs) were recorded every 2 weeks from June to September. Male flights were monitored from June to October using bottle traps baited with 0.01mg of the sex pheromone (Isagro Italia, Italy). Traps consisted of Plexiglas containers (8cm tall, 6 cm diameter) with the pheromone dispenser hung underneath the cap and 4 holes (1.5cm diameter) on the upper part of the wall to allow male access. To capture the attracted males, traps were filled with 100ml of soapy water. Pheromone traps were hung to a trellis wire inside the canopy of the fifth vine of each transect. In the vineyard where only 2 transects/plot were established, one more pheromone trap was used. Captured males were counted every week, while pheromone dispensers were replaced every 4 weeks. The age population structure was evaluated in 10 randomly selected vines per plot on 3 dates (25 June, 23 July, and 5 September). All mealybugs on trunk, cordon, canes, leaves, and grape bunches were counted for 5 min (Geiger & Daane, 2001) and categorized as described above for leaves. Fecundity was estimated by collecting approximately 50 females per plot after each peak male flight. Females were placed individually on Plexiglas containers (3cm tall, 3cm diameter) and fed with 3 grapes. Containers were secured with a paper napkin and a rubber band to prevent males to access and crawlers to escape. Mealybugs were reared under laboratory conditions at 22 C until death, after which the percentage of ovipositing females and their fecundity were determined.

237 217 Crop damage was estimated before harvest by rating the first 2 bunches of flagged canes of each transect using the following indexes: 0 means no mealybug damage; 1 means the presence of honeydew, bunch salvageable; 2 means the presence of honeydew and mealybugs, bunch partially salvageable; 3 means the total bunch loss (Geiger & Daane, 2001). Population density on leaves was compared throughout the season by using repeated measure analysis of variance (ANOVA). Mealybug density is presented as treatment means (± SEM) of 3 vineyards. Proportion of P. ficus stages, proportion of ovipositing females, and bunch damage were evaluated with the Fisher s exact test. Mealybug fecundity was compared with the t-test by pooling the number of eggs laid by treatment. Data were log(x+1) transformed to normalize the variance when needed. Statistical analyses were performed using the software SAS (SAS Institute, 2002). Results and discussion In all the control plots, male trap catches showed 4 peaks, on June, July, September, and October (Figure 1). Trap catches in mating disruption plots were lower than in control plots throughout the season. The total male catches were 11 times higher in control plots than in pheromone-treated plots. 300 Control Mating disruption Males/trap/week (n) JUNE JULY AUGUST SEPTEMBER OCTOBER Figure 1. Mean number of Planococcus. ficus males captured in pheromone traps in control and mating-disruption plots. Mealybugs were recorded on flagged leaves throughout the sampling period, from June to September (Figure 2). The highest population density was detected on early July in both pheromone-treated and control plots (with a mean number of 9.9 and 16.5 mealybugs per leaf, respectively). From late July to September, mealybug density decreased steadily, lowering to 0.4 and 2.1 mealybugs per leaf in treated and control plots, respectively, in mid September. No season-long statistical difference in mealybug density was observed between treatments (F = 0.34; df = 1.12; P = ). The 5-min counts on random vines showed that the percentage of females over all sampled stages did not differ statistically in both treatments on 25 June and 23 July (P = and , respectively) (Figure 3). Instead, on 5 September, the proportion of females in pheromone treated plots was significantly higher than that in control plots (P < ).

238 218 Mealybugs/leaf (n ± SEM) Control Mating disruption 0 JUNE JULY AUGUST SEPTEMBER Figure 2. Vine mealybug population density on flagged leaves in both control and mating disruption plots. Treatments are not significantly different by repeated measures ANOVA (P > 0.05). Percentage of P. ficus stages A B C b a Control Mating disruption I-II instars III instars Females Females with ovisac+ovisacs a a a a I-II instars III instars Females Females with ovisac+ovisacs a b a a I-II instars III instars Females Females with ovisac+ovisacs a a b a a a a a a b a a Figure 3. Proportion of Planococcus. ficus developmental stages from 5-min counts of field populations on 25 June (A), 23 July (B), and 5 September (C). Pairs of bars with different letters are significantly different by Fisher s exact test (P < 0.05).

239 219 On 25 June, the percentage of females with ovisac + ovisacs over all sampled stages in control and mating disruption plots was 33 and 7%, respectively, with significant difference between treatments (P < ). The percentage of ovipositing females (intended as the number of females with ovisac and ovisacs over females and females with ovisac and ovisacs) recorded on 25 June in control plots (67.7%) was significantly higher than these in mating disruption plots (27.9%) (P < ) (Figure 4). On 23 July, 12.9 and 17.6% of ovipositing females were observed in mating disruption and control plots, respectively, with no significant difference between groups (P = ). Observations during the third generation (5 September) showed that the proportion of ovipositing females in control and pheromone-treated plots (10.6 and 7.7%, respectively) was not different (P = ). Ovipositing females (%) a Control Mating disruption b a a a a 25 June 23 July 5 September Figure 4. Proportion of Planococcus ficus ovipositing females observed during 5-min counts of randomly sampled vines. Pairs of bars with different letters are significantly different by Fisher s exact test (P < 0.05). Females collected from the field after the first male flight produced more eggs than those collected after the second and third flight, respectively (Table 1). Fecundity of females collected from control and mating disruption plots was not statistically different in all the 3 generations (P = ; P = ; P = , respectively). All females of the first and second generations oviposited, while only a portion of females collected in the third generation produced ovisacs (83 and 68% of females from control and mating disruption plots, respectively). Table 1. Mean fecundity of Planococcus. ficus ovipositing females collected from the field and reared under laboratory conditions. Treatment Mean no. of eggs/female (± SEM) a I generation II generation III generation Mating disruption ± ± ± 7.3 Control ± ± ± 6.4 a Treatment means within a column are not significantly different by t-test (P > 0.05).

240 220 Crop damage at harvest was negligible; in fact, 90 and 87.8% of grape bunches were rated as not infested in control and pheromone-treated plots, respectively (Table 2). None of the sampled grape bunches was rated as having severe damage. No significant difference in crop damage was observed between treatments (P = ). Table 2. Bunch damage ratings at harvest in mating disruption and control plots. Treatment a Bunch damage (%) None Low Moderate Severe Mating disruption Control a Treatment means are not significantly different by Fisher s exact test (P > 0.05). Pheromone dispensers lowered significantly male captures in pheromone traps in mating disruption plots throughout the monitoring season (5 months). However, the decrease in trap catches was not related to a statistically significant reduction of population density and crop damage. Observations on population dynamics on leaves showed a reduction in mealybug density in pheromone-treated plots with respect to control plots, but the difference was not significant. Timed observations on vines pointed out differences on proportion of ovipositing females only on 25 June. Egg production of females collected from mating disruption and control plots was not significantly different. This could be due because a portion of males was not disrupted by the synthetic pheromone in treated plots and mated the virgin females. Instead, studies carried out in California vineyards showed that females collected from plots protected with 3 or 4 applications of microencapsulated pheromone produced fewer ovisacs and eggs than in control plots (Walton et al., 2006). Moreover, bunch damage was significantly reduced at harvest in pheromone-treated plots. The slight difference in the proportion of ovipositing females and crop damage between treatments might be due to the small size of the experimental plots and the short distance between them. Because control and mating disruption plots were adjacent and ranged from 0.2 to 0.5ha, pheromone cues might have been drifted by wind from mating disruption to control plots. Therefore, mealybug monitoring and sampling data would have been affected and differences between treatments would have been reduced. Another possible reason for the mixed results of the mating disruption field trial is the suspected parthenogenetic reproduction of Sardinian populations of P. ficus. References Dalla Montà, L., Duso, C. & Malagnini, V. 2001: Current status of scale insects (Hemiptera: Coccoidea) in the Italian vineyards. Boll. Zool. agr. Bachic. Ser. II, 33: Engelbrecht, D. J. & Kasdorf, G. G. F. 1990: Transmission of grapevine leafroll disease and associated closteroviruses by the vine mealybug Planococcus ficus. Phytophylactica 22: Geiger, C. A. & Daane, K. M. 2001: Seasonal movement and distribution of the grape mealybug (Homoptera: Pseudococidae): developing a sampling program for San Joaquin valley vineyards. J. Econ. Entomol. 94:

241 Godfrey, K., Ball, J., Gonzalez, D. & Reeves, E. 2003: Biology of the vine mealybug in vineyards in the Coachella Valley, California. Southwest. Entomol. 28: Lentini, A., Serra, G., Ortu, S. & Delrio, G. 2008: Seasonal abundance and distribution of Planococcus ficus on grape vine in Sardinia. IOBC/WPRS Bull. 36: SAS Institute 2002: SAS procedure s guide, version 9. SAS Institute, Cary, NY. Walton, V. M., Daane, K. M., Bentley, W. J., Millar, J. G., Larsen, T. E. & Malakar-Kuenen, R. 2006: Pheromone-based mating disruption of Planococcus ficus (Hemiptera: Pseudococcidae) in California vineyards. J. Econ. Entomol. 99:

242 222

243 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Biological characteristics of Heliococcus bohemicus and Ericydnus sipylus in controlled conditions A. Fleisch 1, L. Pérez 1, P. Kreiter 1, A. V. Odor 1, G. Sentenac 2 1 INRA U.E.L.B., 1382 Route de Biot, Valbonne, France. 2 IFV - Unité de Beaune, 6 rue du 16ème Chasseurs, Beaune, France Abstract: Heliococcus bohemicus is an increasing pest in vineyard in the north-east of France. A biological control was initiated in 2007 and A study in laboratory, to measure the two antagonist biological characteristics, the mealybug and one of its parasitoids Ericydnus sipylus, was undertaken. The description of the mealybug life cycle is led under the climatic conditions: 25 C and 60% R. H. It appears that H. bohemicus has three larvae instars. Females did not lay off eggs and the larvae hatched after 23.5 days on average. The development of first and second instars lasted near 34 days and for the third instars it was near 14 days. The biological characteristics of the parasitoid were studied at 21 C and 50% R. H. Under these conditions the pre-imaginal development lasted 35 days and the length of adult emergence was 57 days on average. The fertility is equal to 15 individuals. Others parameters are studied and are explained in the article. The method of the mass rearing is developed. The pest insect is developed on potato tubers turned green by light. Key-words: Pseudococcidae, biological control, mass rearing, parasitoid Introduction Heliococcus bohemicus Šulc (Hemiptera, Pseudococcidae) was discovered for the first time in France in 1927 on Virginia creeper (Parthenocissus quinquefolia, L.) (Goux, 1934). Sforza et al. reported this species on Vitis vinifera L. in France in On V. vinifera, H. bohemicus has long been regarded as a secondary pest and as safe. Furthermore, this mealybug was also identified as a vector of the grapevine leaf-roll associated viruses-1 and -3 and the grapevine virus A (Sforza et al., 2003, Zorloni et al., 2006). These viruses induce a significant loss of performance and ultimately death of the plant. In the aim to reduce these damages, a biological control against H. bohemicus should be engaged. The first step of this program was to produce the mealybug and its parasitoïds in laboratory, in order to study their biological characteristics. With females of H. bohemicus sampled in French vineyard, we have developed a mass rearing method. An inventory of the natural enemies of this mealybug was undertaken. Two main species of parasitoids were found inside the mummies sampled in the Bourgogne and Alsace vineyards: Anagyrus szodensis Erdös and Ericydnus sipylus Walker (Hymenoptera, Encyrtidae) (Sentenac & Kuntzmann, 2003). After preliminary studies, E. sipylus appeared easier to rear in controlled condition. We have thus chosen this species to develop a biological control of this virus vector; in the aim to produce and release it in vineyards to increase the population of parasitoids. The parasitoïd life history traits were quantified. The quality of the parasitism is discussed function of the different sizes of the host. The host size is an important factor for the biological characteristics of entomophagous insects, especially for parasitoids (Chow & Heinz, 2005; Karamanoua and Copland, 2000). 223

244 224 Material and methods Origins of the biological material The mealybugs originated from Bourgogne French vineyards. They were sampled in different plots, to increase the genetics heterogeneity. The E. sipylus population came from an inventory of H. bohemicus natural enemies, realized in the Bourgogne region, in France. The potatoes, on which we breed the mealybugs, came from a French grower (Sisteron). Insects rearing methods The rearing of H. bohemicus was conducted in confined cages of 48 x 37.5 x 57cm, with thin canvas walls (meshes of 20µm) to aerate the cage. The tuber maturation occurs during 40 to 60 days at 4 C in obscurity. Then the tubers were placed at 10 C in a dark climatic room for one month before their transfer into a chamber at 22 C and 70 to 80% relative humidity. When the germs were long and thick enough, they were exposed to the light during 4 to 5 days. The climatic conditions for the mealybug rearing are 25 C, 60% of relative humidity (R.H.) and a 8:16 D:L.. Furthermore, we introduce a healthy vine in cage to allow the mealybugs to climb and to nourish themselves on the plant. We observed a population increase when these two nutritive sources are available. Periodically old potatoes were removed and replaced by new ones. At regular intervals, mealybugs were isolated in the aim to serve as host for the E. sipylus rearing. For that, in an aerated plastic box of 12.5 x 14 x 19cm a soil of absorbent paper and new healthy potato tubers were introduced. In addition 10 to 20 fertilized females of H. bohemicus taken in the rearing cages are introduced. Then, the box is placed in a climatic room, at 21 C, 75% R.H. and with a 8:16 D:L. Two hundred E. sipylus were randomly collected in old rearing-boxes (to mix the population) and were introduced in the enclosure with new hosts. To make sure the parasitoids nourishment, the lid of the box is covered with ad libitum fine honey droplets. History life traits of H. bohemicus Development cycle and all stage durations. We isolated 40 neonates of H. bohemicus at the same time. Each larva was filed on a potato sprout. Then, we placed a small aerated cap with a bronze grid with meshes of 20µm, and checked every day the larvae development on the potato tuber. We could determine the precise instars number and the duration of each one. Fertility. We tried to quantify the H. bohemicus average progeny. For that, we isolated 12 mealybug mature females in small pills. Every day, we checked out the neonates under the mature females. So, we obtained the parturition time average, larva number average born per day and in whole the H. bohemicus fertility. Sex-ratio and larval mortality. We have isolated each day the offspring of six mature females. The first instar larvae are placed on a new potato tuber turning green. At the third larval instar, the sex differentiation occurs. That because the males are in a wax cocoon and haven t a third mobile instar. We count the mature males and females. The H. bohemicus in third instars are considerated as females. To calculate the sex-ratio, we use the following ( Number of males) equation: 100 ( Number of third instars + Number of females + Number of males) At the end, we have the number of neonates and the number of survivors, then the larval mortality too. History life traits of E. sipylus. This study was carried out in a climatic room, at 21 C, 75% R.H. and with a 8:16 D:L. To collect the parasitoids, we isolated mummies from the rearing

245 225 and we placed them into an aerated glass box with honey droplets placed on the lid. Each day of emergence, the mature wasps were collected and grouped in a box during 48 hours. In all experiments, we introduced two-days-old parasitoids couples, period needed for fecundation and pre-oviposition period allowing obtaining gravid females. The egg-laying duration and fertility. the objective was to determine if 2 days-old gravid female can lays eggs immediately in contact with H. bohemicus. In addition, we measured the laying period duration. We selected one couple of E. sipylus from the rearing room. We introduced the couple in an aerated box with honey droplets deposited on the lid. The boxes contained 30 mealybugs of different sizes. Each day, we changed the boxes to offer females new unparasitized mealybugs. The boxes were examined daily. We could not distinguish the female and the male without an observation under binocular microscope. The transfer was replicated until parasitoids death. When mummies were observed, we deduced that the laying occurred the day of contact between the parasitoid and the host. All the emergence days were recorded to know the egg-laying duration. We replicated this with ten females. History life traits of E. sipylus in relation on the mealybug size. The H. bohemicus different instars were morphologically similar. To differentiate the development stage, we have chosen to classify the mealybugs in size categories (respectively groups 1, 2, 3 and 4 of size less than 1mm, between 1 and 2mm, between 2 and 3mm and more than 3mm). A parasitoid couple was introduced in a glass box during 21 days. In the box, we introduced a sprouted potato tuber with 30 H. bohemicus of the same size. We repeated this procedure 30 times for each mealybug size group. A daily examination was made. The period between the parasitoid oviposition and the first emergence indicates the necessary minimal time to obtain the next generation of mature parasitoids (the intergeneration time or IGT ). The offspring was removed at the emergence day. One hundred of them were collected randomly during emergence period. They were sexed to determine the sex-ratio of the E. sipylus progeny for each size group of host. The equation used for the sex-ratio is: ( Number of males) ( Number of females + Number of males) 100 Fifteen males and 15 females of E. sipylus were removed at their emergence days. They were isolated in a glass tube with a honey droplet on the cap. We recorded the death date to calculate the parasitoid longevity among the host size groups. We measured the both sexes longevity into a same host size group. The maximum observation time corresponded to the double of the period necessary for the first parasitoid emergence. After this period, we couldn t know if neonates were issued from the first parasitoids generation or from the second one. To find out how long this time was, we recorded the emergence number per couple so we obtained the observable fertility. All of this data were collected with the 4 H. bohemicus size groups. Independently, we have realized a comparison of the female longevity in mealybug presence or absence. We used females of E. sipylus who emerged from different host size. We placed the parasitoid at its emergence day in a plastics cylinder (with honey droplets). We introduced 30 mealybugs from all size groups mixed. The death day is recorded for the 30 E. sipylus used here. We grouped all the data for the size group experiment and we compared them with data acquired with the presence of mealybugs. Statistical analyses The data were analysed with the R freeware (GNU project). The Kruskall-Wallis test was carried out to determine if the four group medians differ among themselves. If a significant difference appeared at p < 0.05, we completed the analysis with a Mann-Whitney test, to

246 226 compare the samples in pairs. The total comparisons number realized (equal at 6) was higher, so the risk to accept the H0 hypothesis wrongfully increases, too. We apply the Bonferroni correction on the threshold α. So, we obtain a significant level of α B = 0.05/6 [number of comparison] = (or 8.33x10-3 ). Results and discussion History life traits of H. bohemicus Development cycle stages. Only 28 new-borns survived up to the second instars (Table 1). It is the first time thatthird instars of H. bohemicus were obtained. The few data concerning the pre-parturition period give us some idea of the period between the first instars and the second generation first instars. The total biological cycle average lasts approximately 154 days. Table 1. Development cycle of H. bohemicus, in controlled conditions Variable duration (in day) Individuals Mean (in day) Standard deviation (in day) First instars N= ± Second instars N=16 38 ± Third instars N= ± 3.08 Pre-parturition period N= ± 5.5 Fertility characteristics. The neonate number varied between 43 and 404 newborns in total. The larvae number produced per day is relatively close for the different females. That is consistent with the mean per female (Table 2). Note that the progeny number corresponds to the less time of parturition, i.e. 11 days. Table 2. Fertility characteristics of H. bohemicus N=12 Mean Standard deviation Parturition duration (in days) ± 7.11 Number of larvae per day 8.61 ± 2.92 Total fertility per female ± Sex-ratio and larval mortality. At the parturition time end of the six females, we have collected a total of 1344 newborns. After 3 months only 34.74% survived. Among these 467 mealybugs, only 285 reached the mature development stage. We obtain 53 third instars individuals (who are future female), 110 mature females and 72 mature males. i.e. 30.9% of males. The egg-laying duration of E. sipylus All the females have laid at the first contact day with H. bohemicus. The maximum laying period is 35 days. A quarter of the total offspring emerges at the 5 first emergence days.

247 227 Respectively, the 50%, the 75 % and 90% of the newborns emerge after 9 days, 13 days and 16 days. After 16 days, the majority of the descendants have emerged. On average one female produces 68 eggs per oviposition period of 18.5 days. History life traits of E. sipylus Intergeneration time. The E. sipylus inter-generation time varies significantly (Kruskall wallis, p = 10-4 ) among the four groups (Table 3). The data concerning the group 1 are significantly different for the second and third size groups (in that order; p = 1.2 x 10-3 and p = 10-4 ) (Figure 1). No difference appears between the others two-per-two comparisons (for the comparisons groups 1vs4, groups 2vs3 and groups 3vs4, respectively p=3.8 x 10-2 ; p= and p=4.7 x 10-2 ) for the larval development time of E. sipylus. The figure 1 represents the four host size groups IGT distribution. Table 3. Average of the IGT among the four size groups (in days) Host-size group N M and σ 38.6 ± ± ± ± 3.17 N = Parasitoid number for each modality / M = Means of IGT (days) / σ = standard deviation 55 Days Mealybugs size groups 3 4 Figure 1. Box-plots representation of IGT of E. sipylus by mealybug size groups We can see heterogeneity of the IGT between different hosts sizes (Table 3). The minimal IGT for all size groups is 29 days. For the first size of mealybug, half of the parasitoids have an IGT around 37 days. Concerning hosts measuring between 2 and 4mm, the next generation appears after 33 days for 75% of each population. All the parasitoids from the second size group have an IGT less than 37 days (excepting the extreme values). 50% of the fourth group develop themselves in 35 days. Observable fertility. Parasitoid female fertility is variable between the different hosts sizes (p = 4.06 x 10-3 ). We note for size 1, that the offspring number per female average is low, less than 3 individuals (Figure 2). The comparisons between the parasitoids fertility lying in H. bohemicus group 1 and from mealybugs from size 2 and 3 show that the variable is

248 228 significantly different (respectively p = 5.23 x 10-4 and p = 8.28 x 10-3 ). Between the mode 1 and mode 4, we are close to a significant difference, with p = 9.7 x Other comparisons (between the second group and the third group, p = ; between size group 3 and size group 4 p = , and p = between group 2 and group 4) were not significantly different. The females that lay eggs in hosts of size 2, size 3 or size 4, are more fertile than for the first mealybugs size group. For these entire size groups, 75% of the E. sipylus females have more than 4 descendants and up to 38 descendants at maximum (for the size group 4). For the smallest mealybugs, we observe that 100% of the parasitoids have fertility less than 10 newborns. The size groups with whom we obtain the best fertility rate are the second one and the third one. 40 Number of neonates Mealybugs size groups Figure 2: Box-plots representation of the fertility by size range of mealybugs Sex-ratio. The sex-ratio of the groups 1, 2, 3 and 4 are respectively 69.38%; 73.33%; 50.28% and 51.50% in average. For the size group 1 and 2, the male number is superior to females. There are respectively two males and three males for one female from the mealybugs size group 1 and the second one. The Kruskall-Wallis statistical value is not significant at α = 0.05 (p=0.111) among the size groups (Figure 3). 100 Sex-ratio Mealybugs size groups Figure 3. Box-plots representation of the impact of the host size on the E. sipylus sex ratio

249 229 Longevity depending to sex. We compile all emergence data whatever the host size. The females live on average 53 days and the males stay alive 26 days. Females longevity is significantly longer than males one (p = 2.2 x ). Three quarters of females live more than 30 days. While all males dead before 60 days and approximately 75% of them survive less than 32 days (Figure 4) Days Parasitoid sex (1 = female; 2 = male) Figure 4. Box-plots representation of E. sipylus longevity depending to sex. Female longevity depends to the host size. Among the four mealybugs size groups, the female longevity differ significantly (p = 9.6 x 10-6 ) (Figure 5). The survive duration of the females emerged from the host groups 1 and 2 is not different (p = ). However the E. sipylus female longevity from the mealybug size group 3 varies significantly of the females from the size groups 1 and 2 (respectively p = 6.8 x10-5 and p = 4.19 x 10-4 ). The comparisons of the group 4 and the groups 2 (p = 1.81 x 10-2 ) and 3 (p = ) do not show significant difference. The females emerging from the group 1, have a shorter longevity than the group 4 one (p = 9 x 10-4 ). The females, who emerge from hosts of size 3, have the longest survive period with a mean of 76 days. The longest survive times are measured in the groups 3 and 4 with 115 and 117 days, correspondingly. 75% of the females from the groups 3 and 4 have longevity at least of 42 days, whilst 75% or more of the E. sipylus from the groups 1 and 2 subsist less than 50 days Days Mealybugs size groups Figure 5. Box-plots representation of female longevity by size range of mealybugs.

250 230 Male longevity depending to the host size. The males longevity does not differ significantly among the hosts size groups (p = 0.604). The longevity means is approximately the same, with 23 days (Figure 6). Nevertheless, we observe males from the groups 3 and 4 that can survive more than 45 days. They represent less than 5% of this population Days Mealybugs size groups 4 Figure 6. Box-plots representation of male longevity by size range of mealybugs Into a same host size group, we compared the longevity between females and males. Among the first group there was no significant difference for the males and females longevity (p = ). The survive time between the both sexes differ significantly for the group 2, the group 3 and the group 4 (respectively p = 1.6 x 10-2 ; p = 1 x 10-5 and p = 8 x 10-4 ). The longevity of females depends to the availability of hosts. A significant difference was found in the females longevity not exposed to H. bohemicus (p = 5.94 x10 11 ). The females have a longer longevity without accessible host for their egg-laying than the other ones (Figure 7). Females in test tub with mealybugs survive 10.8 days on average. When females are alone in a tub, their longevity is equal at 53.5 days on average Days With Without Presence (with) or absence (without) of mealybugs Figure 7. Box-plots representation of female longevity in the presence or not of mealybugs. The biological characteristics of parasitoids are often studied as part of a biological control (Chong and Oetting, 2007a), on the one hand in rearing management and other hand

251 231 to better target crop releases periods. However, the development stage and the host size, often correlated for the mealybugs, play a major role in the pest biological control. Karamaouna and Copland (2000) highlight the host role in the Pseudaphycus flavidulus Brethes (Hymenoptera, Encyrtidae) choice behaviour. When a choice is offered, this parasitoid prefers laying in the second instar of Pseudococcus viburni Signoret (Hemiptera, Pseudococcidae). Hence it is important to know the different parasitoid preferred development stage or size for oviposition, to favour the complementarity among them. The pre-imaginal development time (to egg to adult) is depending on the host size (Ya-Hui and Bao-Ping, 2007) especially among Lepidoptera parasitoids. In 2005, Lysiphlebus testaceipes Cresson (Hymenoptera, Braconidae) fertility was studied on different aphids sizes by Silva et al. E. sipylus having laid in the mealybug size group 4, i.e. older mealybugs, can have their fertility decreased by a more mature host immune system and by a greater eggs encapsulation rate (Blumberg, 1988;). Chow and Heinz, (2005) measure the host size impact on the sex ratio of Diglyphus isaea Walker (Hymenoptera, Eulophidae). Despite the significant difference absence, we note that females having laid in small hosts produce an offspring, in male favour. Females come from largest hosts with more nutritional resources for their development. This seems in accordance with the work carried out by Quicke (1997). Hegazi et al (2007) revealed that the host size influence on the longevity of Microplitis rufiventris Kokujev (Hymenoptera, Braconidae), a parasitoid of Spodoprera littoralis Boisduval (Lepidoptera, Noctuidae). Longevity significant differences between host sizes can be explained by the different available resources quantity during the pre-imaginal development of the parasitoids. Biological life traits at 21 C were not encouraging for keeping E. sipylus like an efficient biological control agent (a low fertility, a high sex-ratio in males favour and an important development time for a parasitoid). Nevertheless the important female longevity can compensate these characteristics. However, these biological life traits measurements help us to increase the parasitoid rearing yields. We have not measured host-feeding occurrence on young stages, if this were the case this longevity would be an advantage in biologic control. We measured however that the female parasitoid longevity in host absence is significantly greater than the E. sipylus longevity with mealybug. We can suppose on an oosorption phenomenon existence (Godfray, 1994). The oosorption allows to the gravid female to survive longer in the absence of host, thanks to some mature egg nutrients re-absorption. The female keeps an egg asset to lay whenever it encounters a favourable host. For efficiency in vineyard, the development time, the fertility and longevity are equally important. This study led to show that there is indeed a difference of parasitoid biological characteristics depending on the mealybug host size. Offspring emerging from host measuring between 2 and 3mm are still having the best biological potential, with a sex ratio of 50% and a fertility average of 9 individuals by female. According to these results, a release should be carried out when this type of population is maximal. But it is the case in mid-september, at the maturity of berries beginning. This seems later in the season, because E. sipylus feed mainly of flower nectar. It would be wise to release the parasitoids at the spring beginning, when the host measure between 1 and 2mm. Despite a higher sex ratio this mealybug population allows the parasitoids to install a reservoir population before autumn releases. Acknowledgements Thanks are due to FRANCE-AGRIMER for funding this research.

252 232 References Blumberg, D. 1988: Encapsulation of eggs of the encyrtid wasp, Metaphycus swirskii, by the hemispherical scale, Saissetia coffeae: Effects of host age and rearing temperature. Entomologia Experimentalis et Applicata 47(1): Chong, J. H. and Oetting, R. D. 2007: Functional response and progeny production of the Madeira mealybug parasitoid, Anagyrus sp. nov. nr. Sinope: the effect of host stage preference. Biological Control 41(1): Chow, A. and Heinz, K. M. 2005: Manipulation of sex ratios in mass rearing of Diglyphus isaea (Walker), an ectoparasitoid of agromyzid leafminers. Bulletin OILB/SROP 28(1): Godfray, H. C. J. 1994: Parasitoids: Behavioral an evolutionary ecology. Princeton University Press, Chichester, 470 pp. Goux L. 1934: Note sur les Coccides de la France, Contribution à l étude du genre Heliococcus avec description de deux nouvelles espèces. Bull. Soc. Ent. France 11: Hegazi, E. M., El-Aziz, G. M. A. and El-Shazly, A. Y. 2007: Influence of host age and host deprivation on longevity, egg load dynamics and oviposition rate in Microplitis rufiventris. Insect Science 14(6): Karamaouna, F. and Copland, M. J. W. 2000: Host suitability, quality and host size preference of Leptomastix epona and Pseudaphycus flavidulus, two endoparasitoids of the mealybug Pseudococcus viburni, and host size effect on parasitoid sex ratio and clutch size. Entomologia Experimentalis et Applicata 96(2): Quicke, D. L. J. 1997: Parasitic wasps. Chapman & Hall, London, 470 pp. Sentenac, G. and Kuntzmann, P Etude des Cochenilles et des antagonistes qui leur sont associés dans des vignobles en Bourgogne et en Alsace de 2000 à Bull. OILB/SROP 26(8): Sforza, R. and Greif, C. 2000: Les cochenilles et l enroulement de la vigne Données de phytopathologie et d éthologie. Phytoma, la défense des végétaux 532: Sforza, R., Boudon-Padieu, E. and Greif, C. 2003: New mealybug species vectoring Grapevine Leafroll associated Viruses -1 and -3. European Journal of Plant Pathology 109: Silva, R. J., Bueno, V. H. P. and Sampaio, M. V. 2008: Quality of different aphids as hosts of the parasitoid Lysiphlebus testaceipes (Cresson) (Hymenoptera: Braconidae, Aphidiinae). Neotropical Entomology 37(2): Ya-Hui, L. and Bao-Ping, L. 2007: Effects of Helicoverpa armigera (Noctuidae, Lepidoptera) host stages on some developmental parameters of the uniparental endoparasitoid Meteorus pulchricornis (Braconidae, Hymenoptera). Bulletin of Entomological Research 98: Zorloni, A., Prati, S. and Bianco, P. A. 2006: Transmission of Grapevine virus A and Grapevine leafroll-associated virus 3 by Heliococcus bohemicus. Journal of Plant Pathology 88(3):

253 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Survey on Scaphoideus titanus egg distribution on grapevine B. Bagnoli, E. Gargani CRA-ABP Centro di ricerca per l'agrobiologia e la Pedologia, Florence, Italy Abstract: In the context of the Euphresco-Propscaph project, a survey on Scaphoideus titanus egg distribution on grapevine has been under way since January Samples of one and two to fouryear old vine wood of the Trebbiano Toscano variety from an untreated vineyard in Latium (Central Italy) and samples of one and two-year old of the Kober 5BB rootstock from an abandoned field in Veneto (North Italy) have been analysed. The vine samples, collected during the winter pruning, were stocked at 4 C till the beginning of the experimentation, when the same materials were cut into pieces of different lengths and put into rearing boxes (cm 29x20x12) maintained at 24 C, 75% RH and LD 16:8 photoperiod to obtain the first instar S. titanus larvae. In the middle of May in two Latium untreated vineyards, several original capture devices were put on the plants, in particular around the trunk, the cordon, the cane of one-year old and the buds of the vines to survey the emergence of S. titanus larvae. Results confirmed that S. titanus females prefer to oviposit in the bark of two or more year old wood ( larvae per cm 2 ), even if very few specimens were obtained from the one-year old wood ( larvae per cm 2 ) as well. The field experiments with the capture devices evidenced for the first time that even the bark of the trunk is a preferential site of the cicadellid oviposition as well. Key words: Scaphoideus titanus, vine propagation material, spreading, oviposition sites, Italy Introduction Scaphoideus titanus Ball (Cicadellidae, Deltocephalinae), found for the first time in Europe and Italy respectively in 1958 (Bonfils & Schvester, 1960) and 1963 (Vidano, 1964), is now widely distributed in many European vine regions well above and below the 45 th parallel (Alma, 2004; Bagnoli et al., 2008). Many authors assume that the vector of Flavescence dorée phytoplasma, spread to Europe mainly due to the marketing of propagating material infested with the overwintering eggs (Alma, 2004; Steffek et al., 2007). This hypothesis is supported by two main considerations: first of all invasive European populations of the Cicadellid are characterized by low levels of genetic diversity and differentiation (Santoni et al., 2004; Bertin et al., 2007; Papura et al., 2007); furthermore the leafhopper tends to not disperse over long distances (Lessio & Alma, 2004, 2006). Since November 2008 Propscaph project has attempted to answer the Euphresco call regarding Evaluation of the risk of spread of Scaphoideus titanus, the vector of Grapevine Flavescence dorée, with commercial grapevine propagation material. In the context of this project the Italian partners (CRA-ABP, CRA-PAV, CRA-VIT) were concerned with WP4 whose objective was the evaluation of the presence and distribution of S. titanus eggs on grapevine propagation plant material collected from nurseries and untreated vineyards located in different geographical areas. This paper reports the studies performed in untreated or abandoned vineyards on the preferred oviposition sites of S. titanus females. 233

254 234 Material and methods Surveys were performed through laboratory and field trials. Laboratory Due to the great difficulty to ascertain the presence and density of S. titanus eggs in the vine bark directly, we opted for an indirect evaluation based on the finding and counting of the newly hatched larvae. The tested material was represented by samples of one and two to fouryear old wood of the Trebbiano Toscano variety from an untreated vineyard in Latium (Central Italy) and samples of one and two-year old canes of the Kober 5BB rootstock from an abandoned field in Veneto (Northern Italy). In both vineyards an abundant population of S. titanus was observed during the preceding summer. The vine wood material was collected during the winter pruning period in January-February and then it was stored in climatic chambers at 4 C, from the time of the collection to the beginning of the experiments. To set up the trials, at the end of February the one-year old canes of Trebbiano Toscano and the two-year old canes of Kober 5BB were cut into pieces about 20cm long, carrying two gems, while the samples of two to four-years old of cordon of Trebbiano Toscano were cut into pieces of different lengths. In a second moment (in June), to further investigate the distribution of eggs on the vine bark, samples four cm long of the internode and node zone, from one-year old canes of Trebbiano Toscano and from two-year old canes of Kober 5BB, were prepared. The test material was placed into rearing cages that were set up using transparent plastic boxes (i.e. 29x20x12cm) similar to that described by Chuche & Thiéry (2009) and Mazzoni et al. (2009). The rearing cages were kept in climatic chambers set at 24 C, 70-75% RH and LD 16:8 photoperiod. For easier counting and collecting of the newly hatched larvae, a fresh grape vine leaf was put into each cage so as to attract the first instar larvae (Figure 1). S. titanus larvae were monitored and removed from the rearing cages by a mouth aspirator at least three times a week. Observations ended when no more hatching had occurred for two weeks. Figure 1. Rearing cages.

255 235 Field To gain further knowledge about the distribution of S. titanus eggs on the different woody parts of vine, a specific survey was undertaken in two vineyards of the Trebbiano Toscano and Cabernet Sauvignon varieties in Latium. In May-June, several original capture devices were applied to 18 plants for catching S. titanus larvae hatched from eggs laid in the bark of the trunk or the cordon or the node and internode area of the cane. With few exceptions, every plant examined was treated with four or five sleeves, replaced approximately every two weeks (Figure 2). Figure 2. Yellow plastic sleeves, with inner sticky surface, used to monitor the S. titanus larvae emerged from the bark of the different woody parts of grapevine. Results and discussion Laboratory The test results are presented in summary form in Tables 1 and 2. Table 1. S. titanus first instar larvae hatched from one or more year old wood in the rearing cages (b.p.: basal part of the cane; c.p.: central part; d.p.: distal part). Age of the vine wood Boxes Pieces per box Larvae per cm 2 "Trebbiano Toscano" 1 year (b.p.) year (c.p.) year (d.p.) years ± 0.09 "Kober 5BB" 1 year years ± 0.03

256 236 As shown in Table 1, the bark of two or more year old canes (cordons) confirmed to be the best site for S. titanus female oviposition. However S. titanus larvae emerged also from one-year old cane samples collected in the untreated vineyard of the Trebbiano Toscano variety and in the abandoned field of the Kober 5BB rootstock. Regarding the distribution of S. titanus eggs in the bark along the cane, the highest concentration of eggs was detected indirectly, in the form of hatched larvae, in the area surrounding the node that, in the two year old Kober 5BB canes, showed an egg density more than twice that of the internode area (0.18 larvae per cm 2 compared to 0.07 larvae per cm 2 ). Node site, because its particular morphology, is probably more preferred than the internode, even in one-year old canes (Table 2). Table 2. S. titanus first instar larvae emerged from internode and node area. Part of the cane Boxes Pieces per box Larvae per cm 2 "Trebbiano Toscano", one-year old wood, central part of the cane internode node "Trebbiano Toscano", one-year old wood, distal part of the cane internode node "Kober 5BB", two-year old wood internode ± node ± Field As expected, the bark of the two to four-year cordon resulted to be the most affected by S. titanus egg population, however a newly hatched larva was detected by a sleeve applied on a one-year old cane as well. This confirms what has been previously reported about the susceptibility of the young cane bark to the leafhopper oviposition. The use of the sleeves also allowed us to highlight that even the bark of the trunk is an important oviposition site for S. titanus females. As shown in Table 3, 0.09 larvae per cm 2 emerged from the trunk compared with 0.11 from the cordon. This data help us to understand why the early larval stages usually inhabit the basal shoots of the vine: they do not necessarily reach the trunk shoots after falling off the cordon but they can be already present in the basal area of the plant if they have emerged from the trunk bark. Table 3. S. titanus first instar larvae emerged from the trunk and from the cordon. Sleeve type Sleeves Larvae per cm 2 For trunk (23 cm long) For cordon (13 cm long)

257 237 Conclusions Our survey result of greatest economic importance is undoubtedly the ascertain of the susceptibility of the bark of one-year old grapevine cane to host S. titanus eggs. From this comes the need to adopt appropriate phytosanitary measures to avoid the risk of spread of the insect through the grape vine propagation materials. The bark of two or more year old canes (cordons) confirmed to be the best site for S. titanus female oviposition and specific laboratory experiments allowed to point out that the highest concentration of eggs was in the node surrounding area. From an ethological point of view, a result of considerable interest is the detection, for the first time, that even the bark of the trunk is a preferred site for egg laying. Acknowledgements We are very grateful to Dr. Valerio Mazzoni (IASMA Istituto Agrario di San Michele all'adige Fondazione Edmund Mach) for his valuable advice on setting up the S. titanus breeding boxes. We would also like to thank Luca Tirinnanzi and Riccardo Frosinini (CRA-ABP Centro di ricerca per l Agrobiologia e la Pedologia) for their cooperation during the field and laboratory tests. References Alma, A. 2004: The genus Scaphoideus in the world. The diffusion of S. titanus in Europe. Third European Hemiptera - Congress, St. Petersburg, 8-11 June Abstracts: 3-5. Bagnoli, B., Ferretti, L., Trivellone, V., Nuccitelli, L., Pasquini, G. 2008: Accertamento della presenza di Scaphoideus titanus nel Lazio. Atti del "Workshop" 4 Incontro nazionale sulle malattie da fitoplasmi, Roma maggio Petria 18(2): Bertin, S., Guglielmino, C. R., Karam, N., Gomulski, L. M., Malacrida, A. R., Gasperi, G. 2007: Diffusion of the Nearctic leafhopper Scaphoideus titanus Ball in Europe: a consequence of human trading activity. Genetica 131: Bonfils, J. & Schvester, D. 1960: Les cicadelles (Homoptera Auchenorrhyncha) dans leurs rapports avec la vigne dans le Sud-Ouest de la France. Annales des Épiphyties 11(3): Chuche, J. & Thiéry, D. 2009: Cold winter temperatures condition the egg-hatching dynamics of a grape disease vector. Naturwissenschaften 96(7): Lessio, F., & Alma, A. 2004: Dispersal patterns and chromatic response of Scaphoideus titanus Ball (Homoptera: Cicadellidae), vector of the phytoplasma agent of grapevine Flavescence dorée. Agricultural and Forest Entomology 6: Lessio, F. & Alma, A. 2006: Spatial distribution of nymphs of Scaphoideus titanus (Homoptera: Cicadellidae) in grapes and evaluation of sequential sampling plans. J. Econ. Entomol. 99(2): Mazzoni, V., Prešern, J., Lucchi, A., Virant-Doberlet, M. 2009: Reproductive strategy of the Neartic leafhopper Scaphoideus titanus Ball (Hemiptera: Cicadellidae). Bulletin of Entomological Research 99:

258 238 Papura, D., Van-Helden, M., Giresse, X., Salar, P., Danet, J.-L., Foissac, X., Malembic-Maher, S. 2007: Genetic structure of Scaphoideus titanus populations and genetic diversity of the epidemic strains of flavescence dorée phytoplasma: the situation in France. Bulletin of Insectology 60(2): Santoni, R., Alma, A., Bonizzoni, M., Parisi, M., Malacrida, A. R., Gomulski, L. M., Gasperi, G. 2004: Variabilità genetica di popolazioni di Scaphoideus titanus Ball (Homoptera Cicadellidae) analizzata con l uso di marcatori RAPD. Atti XIX Congresso nazionale italiano di Entomologia, Catania, giugno 2002: Steffek, R., Reisenzein, H., Zeisner, N. 2007: Analysis of the pest risk from Grapevine flavescence dorée phytoplasma to Austrian viticulture. EPPO/OEPP Bulletin 37(1): Vidano, C. 1964: Scoperta in Italia dello Scaphoideus littoralis Ball, cicalina americana collegata alla Flavescence dorée della vite. L Italia Agricola 101(10):

259 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Preliminary study of the aggregative behaviour of Scaphoideus titanus larvae J. Chuche 1, A. Boursault 1,2, D. Thiéry 1 1 INRA, UMR 1065 Santé Végétale, Villenave d'ornon, France, 2 present address, INRA, UMR Biologie et Gestion des Adventices, Dijon, France. thiery@bordeaux.inra.fr Abstract: The leafhopper Scaphoideus titanus is the vector of the Flavescence dorée. In this study, we investigated the aggregative behaviour of the larvae. We conclude from different experiments that larval aggregation occurs at the plant scale and the age and colour of the food source could be factors cueing an aggregation. These aggregation patterns should be studied in more details in order to gain knowledge in the epidemiology of Flavescence dorée and eventually to develop control strategies based on interindividual epideictic regulation. Key words: Scaphoideus titanus, grapevine, aggregation Introduction Flavescence dorée, a grapevine yellows caused by phytoplasmas of the 16SrV-group, is one of the current major threats in European viticulture. The spread of the disease at a great scale is mainly due to human activities. Nevertheless, the behavior of the vector Scaphoideus titanus, especially its movements, is responsible of the propagation of this yellows disease from one infected vine stock to other grapevine plants in the same or nearby plots. Aggregative patterns have been observed at vineyard scales for S. titanus larvae (Lessio & Alma, 2006) as well as for imagos (Bosco et al., 1997). Aggregative patterns of juvenile stages may influence the dispersive behavior and in the case of insect vectors, aggregation can influences some decisive parameters (e.g. acquisition and inoculation rate) for the disease spread (Zhang et al., 2000). No hypothesis has yet been formulated to explain these aggregative patterns and if aggregative behaviours exist at different spatial scales (intra stock or intra plot). Field studies do not allow distinguishing between larval aggregation due to their own behaviour and aggregation due to a hypothetical aggregative egg-laying behaviour of females. In this study, we first tried to confirm if an active aggregative behavior of the 1 st larval instar of S. titanus may exists and secondly to investigate several stimuli that may lead to such an aggregation. Material and methods Insect rearing Scaphoideus titanus larvae were obtained as described in Caudwell et al. (1970). Egg hatchings were obtained by placing 2 year old woody canes inside plastic hatching cages in a climatic chamber under a 16:8 (L:D) photoperiod, at 23 ± 1 C, and 65 70% RH. In order to collect neonate larvae, six cutted leaves of Cabernet-Sauvignon, maintained each in a glass 239

260 240 tube with water, were added in the cage, twenty days after eggs were removed from the cold room. Leaves were replaced when they began to wither. All tests were performed with 1 st larval instar (L1). Aggregation test To test if the L1 tend to aggregate, we performed choice tests between four identical grapevine cuttings that were placed in each corner of an Altuglass cage (60 x 60 x 60cm). A number of larvae between 140 to 350 individuals were placed in the centre of the cage at equal distance of each plant. The number of larvae on each plant was numbered after 9 hours. Six repetitions were made. Intra-plant distribution of larvae To study the aggregation behaviour of neonate larvae at plant scale we provided to larvae two 8/9 leaves grapevine cuttings as only food source in each hatching cage. Leaves were classified in 4 categories according to their size and position on the plant: 1) small leaves on the top which were the youngest leaves with leaf area was smaller than11 cm², 2) small leaves on the bottom which corresponds to the oldest leaves with leaf area was smaller than 23cm², 3) large leaves in intermediate position with leaf area > 60cm² and 4) we added an extra cutting category, the buds shoots because with the humidity of the hatchery cages, woody canes which carried eggs were budbursted. The numbers of L1 on each leaf were checked daily during 33 days and the leaf area index of each leaf was measured with a LAI meter. Three cage replicates were made. Response to colour Visual stimuli are generally important cues for the Homoptera. To test the influence of the colour on food choice, we placed on the internal side of hatching cage lid 4 coloured traps (8 x 23.6cm) representing grapevine organs (brown: bark, green: chlorophyllian organs, red and yellow: mature berries an/or symptoms of Flavescence dorée on leaves) sprayed with insect glue without any food resource,. Larvae glued on the traps were daily counted during 28 days. After each monitoring, traps were randomly rotated to avoid any position effect. Three cage replicates were made. Odour attraction To check the occurrence of an aggregative olfactory stimulus produced by conspecifics, we extracted 8,000 neonate larvae in cold methylene chloride during 24h. After the extraction, the solvent was evaporated under a nitrogen flow and the residues were taken up in pure water. Then, in the same experimental setup as described above the agregation test consisted in four identical plants offered in choice situation: two were sprayed with the extract, the 2 others only with pure water. The volume of sprayed extract was adjusted from 5 to one hundred larvae equivalent per leaf and 6 cage replicates per concentration were made. Results and discussion Aggregation test As often observed in vineyards, the larvae tended to aggregate in our choice experiments. In spite of the occurrence of 4 identical food sources, on average, almost the half of the larvae were found feeding on the same plant (Figure 1). This result demonstrates that neonate larvae distributed by themselves among the four plant in a nonrandom pattern. Thus, aggregation of larvae in vineyards could be to some extend not solely attributed to ovipositing spatial patterns.

261 241 Percentage of larvae a b c d 1st 2nd 3rd 4th Favorite plant Figure 1. Percentage of larvae per grapevine cutting. (1 st : plant that received most of the larvae in each choice test; 2 nd : plant receiving second most of the larvae, 3 rd : plant receiving third most of the larvae, 4 th : plant receiving least of the larvae. Boxplots with different letters are significantly different under the Friedman Anova and LSD post hoc tests at 1 % threshold.). Intra-plant distribution of larvae Density (larvae per cm²) a a b b Bu Sh Li Sd Density (number of larvae per cm² of leaf) Figure 2. Larval density on different leaf classes. (Bu: bud shoots; Li: large and intermediate leaves; Sd: small and bottom leaves; Sh: small and high leaves. Boxplots with different letters are significantly different under the Friedman Anova and LSD post hoc tests at 1 % threshold.).

262 242 Densities of larvae were not uniform (Figure 2). Higher densities were observed on the youngest organs: the buds shoots and the apical leaves. The larvae prefer feeding on younger and growing leaves than on older ones. Attraction by coloured surrogates In the present experiment, the major part of the larvae was caught by yellow traps (Figure 3). These results are in contradiction with Lessio and Alma (2004) who caught more imagos with red traps. On the other hand, these results could be related to the L1 preferences for the youngest organs. Indeed, young leaves are yellowish and have higher nitrogen content (Mooney& Gulmon, 1982) which in turn increases the fitness of other leafhoppers (Rossi & Strong, 1991). Moreover, yellow wavelengths are known to be attractive for sap-sucking insects (Saxena et al., 1974; Prokopy & Owens, 1983; Todd et al., 1990). Number of insects caught per hour a b b b Yellow Brown Red Green Color Figure 3. Colour choice by neonate larvae. Boxplots with different letters are significantly different under the Friedman Anova and LSD post hoc tests at 1 % threshold. Attraction by odours Whatever the concentration of the extract, the larvae showed no preference between plant sprayed with extract or with water (Figure 4). Our extract was neither attractive nor deterrent. As a conclusion, we have demonstrated that larval aggregation in S. titanus occurs regardlessly of adult behaviour even if female egg laying behaviour could also contribute to aggregation. The aggregation pattern also occurs at a plant scale and the age and the colour of the leaves appears to be a key factors in the feeding site choice. This work is a first step towards understanding the aggregation behaviour of the Flavescence dorée vector and how it influences the epidemiology of the disease. Our results seem to demonstrate a "non social" aggregation for S. titanus but do not allow to establish yet the stimuli gathering the larvae.

263 243 Mean percentage of insects per plant extract Extrait water Solvant [5 L1] [25 L1] [50 L1] [100 L1] Testing densities Figure 4. Response to conspecifics odour. ([X]: concentration of X larvae equivalent per leaf.) Acknowledgements This project and first author were supported by a Ph.D. grant from the Conseil Interprofessionnel des Vins de Bordeaux, the region Aquitaine and the scientific Dept. INRA SPE. References Bosco, D., Alma, A., Arzone A. 1997: Studies on population dynamics and spatial distribution of leafhoppers in vineyards (Homoptera: Cicadellidae). Ann. Appl. Biol. 130: Caudwell, A., Kuszala, C., Bachelier, J. C., Larrue, J. 1970: Transmission de la Flavescence dorée de la vigne aux plantes herbacées par l allongement du temps d utilisation de la cicadelle Scaphoideus littoralis Ball et l étude de sa survie sur un grand nombre d espèces végétales. Ann. Phytopathol. 2: Lessio, F. & Alma, A. 2004: Seasonal and daily movement of Scaphoideus titanus ball (Homoptera : Cicadellidae). Environ. Entomol. 33: Lessio, F. & Alma, A. 2006: Spatial distribution of nymphs of Scaphoideus titanus (Homoptera: Cicadellidae) in grapes, and evaluation of sequential sampling plans. J. Econ. Entomol. 99: Mooney, H. A. & Gulmon, S. L. 1982: Constraints on leaf structure and function in reference to herbivory. BioScience 32: Prokopy, R. J. & Owens, E. D. 1983: Visual detection of plants by herbivorous insects. Ann. Rev. Entomol. 28: Rossi, A. M. & Strong, D. R 1991: Effects of host plant nitrogen on the preference and performance of laboratory populations of Carneocephala floridana (Homoptera: Cicadellidae). Environ. Entomol. 20:

264 244 Saxena, K. N., Ghandi, J. R., Saxena, R. C. 1974: Patterns of relationship between certain leafhoppers and plants. I. responses to plants. Entomol. Exp. Appl. 17: Todd, J. L., Harris, M. O., Nault, L. R Importance of colour stimuli in host finding by Dalbulus leafhoppers. Entomol. Exp. Appl. 54: Zhang, X. S., Holt, J., Colvin, J. 2000: A general model of plant-virus disease infection incorporating vector aggregation. Plant Pathol. 49:

265 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Vibrational signals associated to the mating behaviour of Hyalesthes obsoletus Signoret (Hemiptera Fulgoromorpha) V. Mazzoni 1, G. Anfora 1, C. Ioriatti 2, M. Virant-Doberlet 3, A. Lucchi 4 1 IASMA Research and Innovation Centre, Fondazione Edmund Mach, San Michele all Adige (TN), Italy, 2 IASMA, Centre for Technology Tranfer, Fondazione Edmund Mach San Michele all Adige (TN), Italy; 3 National Institute of Biology, Večna Pot 111, SI-1000 Ljubljana, Slovenia; 4 University of Pisa, Dept. CDSL, Sect. Entomologia Agraria, Pisa, Italy. valerio.mazzoni@iasma.it Abstract: Hyalesthes obsoletus Signoret is a vector of the stolbur phytoplasma (group 16SrXII-A), causal agent of the grapevine disease Bois Noir. In planthoppers the intraspecific communication is based on substrate-borne vibrational signals. The aim of this study was to investigate the mating behaviour of H. obsoletus. Vibrational signals were detected from nettle cuttings by laser vibrometer. Both male and female can start the mating process by the emission of calling signals. The male calling signal is a single short syllable (Male Syllable 1) whereas the female s is a train of pulses of variable length. Initially, the male/female duet is based on the alternating emission of calling signals (recognition phase). In a second step, the male starts emitting long syllable trains (male courtship signal) in which the female occasionally inserts single pulses (courtship phase). The male courtship signal is made of two different types of syllables (Male Syllable 2 and 3), the switch from one to another often triggered by the emission of a female pulse. When partners are in close vicinity the male emits a specific pre-copula signal that precedes the copula (pre-copula phase). Rivalry tests, either pairs of males or trios, showed a strong competition between males, expressed in the form of chorusing, rivalry signals and aggressiveness. Keywords: vibrational communication, hemiptera, planthopper, phytoplasma vector Introduction The planthopper Hyalesthes obsoletus Signoret (Hemiptera: Cixiidae) is the vector of the stolbur group (16SrXII-A) phytoplasmas (Lee et al. 1998) associated with Bois noir (BN), a disease spread in most of the European grape-growing areas (Maixner 1994, Sforza et al. 1998, Alma et al. 2002). Although in Europe the main planthopper host plants are nettle (Urtica dioica L.) and bindweed (Convolvulus arvensis L.), adults of H. obsoletus can occasionally feed on grapevine (Vitis vinifera L.), thus transmitting the phytoplasma (Langer and Maixner 2004, Bressan et al. 2007). Recently, the role of olfactory cues in host plant selection has been investigated (Sharon et al. 2005, Riolo et al. 2009). On the other hand, the knowledge of intra-specific communication is still lacking. In leafhoppers and planthoppers there is no evidence of a role of either pheromones or vision in sexual communication and only species- and sex-specific substrate-borne vibrational signals are used for partner recognition, courtship and mate location (Čokl and Virant-Doberlet 2003). Our work aims to describe some aspects of the mating behaviour of H. obsoletus in order to highlight the role of substrate-borne acoustic signals in intra-specific communication and pair formation. 245

266 246 Material and methods H. obsoletus adults were collected from nettles in a grassland area at S. Piero a Grado (Pisa, Italy, N, E) by sweep netting on July 2008 and Single individuals and pairs were placed on nettle leaves and their vibrational signals recorded with a laser vibrometer (PDV 100, Polytec GmbH, Waldbronn, Germany). Signals were digitized with 48 khz sample rate and 16-bit depth and stored directly onto a hard drive of a PC computer using the data acquisition device Plug.n.DAQ (Roga Instruments, Waldalgesheim, Germany) and Adobe Audition 1.0 (Adobe Systems Inc.). Signal recordings were analyzed for temporal and spectral features using the computer software program Raven The behaviour of H. obsoletus was recorded with a Canon HV30 camera together with vibrational signals. Recordings were made in a climatic conditioned room (T = 25 ±1 C; HR = 60±5%). Tests Individuals were placed on distinct leaves of a cut nettle stem (distance between leaves: cm). We conducted the following experiments: 1. Single males (n = 21) were placed on a nettle leaf and vibrational signals and behaviour were recorded for 30 minutes. 2. Single females (n = 28) were placed on a nettle leaf and vibrational signals and behaviour were recorded for 30 minutes. 3. One male and one female (n = 20) were put on two distinct leaves and vibrational signals and behaviour were recorded either until the male reached the precopula position (see below) and attempted to copulate or for 30 minutes in cases when planthoppers did not show any mating behaviour. 4. Two males (n = 20) were put on two distinct leaves and vibrational signals and behaviour were recorded for 30 minutes. 5. Two males and one female (n = 20) were placed on three distinct leaves and vibrational signals and behaviour were recorded for 30 minutes. Results and discussion Temporal and spectral characteristics of the vibrational signals recorded from H. obsoletus males and females are reported in table 1. Four different syllables constitute the male mating language. The Male Syllable 1 (MSy1) is emitted as a single signal or as a pair in the context of a male calling signal (MCS) and as a component of the male precopula signal (MpCS). The Male Syllable 2 and 3 (MSy2 and MSy3) are emitted in trains (male syllable trains, MST) of variable duration, where sequences of the two syllables can alternate each other. MST mostly made of MSy2 are possibly connected with male territoriality, whereas those made mainly of MSy3 with courtship. The Male Syllable 4 (MSy4) is always emitted in association with MSy1, thus forming the MpCS. The female emits pulses (Female Pulse, FP) either spontaneously in trains in the context of a female calling signal (FCS) or as a single pulse when emitted as reply to a MST. Either male or female can initiate the mating communication with the calling signal. A stereotyped behavioural sequence with a progressive development of the mating process was observed with three distinct phases: (1) recognition, (2) courtship and (3) precopula. Regardless of which gender initiated a duet, the recognition phase is characterized by short FCS alternated with MCS (mean phase duration: 37.0 ± 40.9s). After an initial stationary duet,

267 247 the male approaches the female. It follows the courtship phase (206.0 ± 189.4s) when the male starts emitting MST, mostly made of MSy3, while the female inserts one pulse between male syllables without a regular rhythm or latency. The emission of FP, in reply to MTS often elicits the switch from MSy2 into MSy3 sequences. The mating process continues with the precopula phase (15.0 ± 12.3s), with male and female in close vicinity. In this phase the female stops replying while the male starts to emit MpCS accompanied by quick leg tapping. The male moves around the female while maintaining body contact with her. At the final stage the male positions himself slightly behind the female, before the copulation attempt. Rivalry tests, either males alone or trios, showed a strong competition between males, which consists in chorusing, rival signalling and aggressiveness. In case of chorusing, a male overlaps his call with the neighbour, by alternating his pulses with those of the rival. Rival signals are irregular trains of pulses that are aimed to interrupt another male s call. The aggressive behaviour was recorded only in presence of a female and consisted of close and prolonged contacts between males, accompanied with a continuous emission of rivalry noise, which in this context can be interpreted as a threatening signal. The general pattern found in H. obsoletus has the following unique features: (i) pair formation begins with the emission of either male or female vibrational signals; (ii) the mating duet is divided into two distinct sections with females being the more vibrationally active gender in the recognition phase and males in the courtship and precopula stages; (iii) the structure of the male syllable train in the courtship phase is regulated by the female reply; (iv) when in close proximity to the female males emit specific precopula signals, made of two regularly repeated distinct syllables and accompanied by vibrational signals induced by legtapping and (v) males can adopt three different strategies in rivalry: emission of proper rivalry signals, chorusing and physical aggressiveness. To conclude, the results of the present study show that in H. obsoletus the vibrational signals are necessary for mate recognition, to initiate the male search and to assure the female acceptance, and also to regulate male intraspecific competition for territorial occupation. Table 1. Temporal and spectral features (mean ± SD) of male and female vibrational signals of H. obsoletus. In the columns for each signal are reported dominant frequency (Dom. Freq), duration and syllable repetition time (SRT). Signal Dom. Freq. (Hz) Duration (ms) SRT (ms) Msy1 576 ± ± ± 71 Msy2 396 ± ± ± 47 Msy3 447 ± ± ± 55 Msy4 379 ± ± ± 80 FP 287 ± ± ± 57 Acknowledgements This study was funded by Interneuron Project (Fondazione E. Mach), Host Project (Autonomous Province of Trento), Fondi Ateneo of Pisa University and the research project V (Slovenian Research Agency).

268 248 References Alma, A., Soldi, G., Tedeschi, R. & Marzachì, C. 2002: Ruolo di Hyalesthes obsoletus Signoret (Homoptera: Cixiidae) nella trasmissione del Legno Nero della vite in Italia. Petria 12: Bressan, A., Turata, R., Maixner, M., Spiazzi, S., Boudon-Padieu, E. & Girolami, V. 2007: Vector activity of Hyalesthes obsoletus living on nettles and transmitting a stolbur phytoplasma to grapevines: a case study. Ann. Appl. Biol. 150: Čokl, A. & Virant-Doberlet, M. 2003: Communication with substrate-borne signals in small plant-dwelling insects. Annu. Rev. Entomol. 48: Langer, M. & Maixner, M. 2004: Molecular characterisation fo grapevine yellows associated phytoplasmas of the stolbur-group based on RFLP-analysis of non-ribosomal DNA. Vitis 43: Lee, I. M., Gundersen-Rindal, D. E., Davis, R. E. & Bartoszyk, I. M. 1998: Revised Classification Scheme of Phytoplasmas based on RFLP Analyses of 16S rrna and Ribosomal Protein Gene Sequences. Int. J. Syst. Bacteriol. 48: Maixner, M. 1994: Transmission of German grapevine yellows (Vergilbungskrankheit) by the planthopper Hyalesthes obsoletus (Auchenorryncha: Cixiidae). Vitis 33: Riolo, P., Minuz, R. L., Anfora, G., Rossi Stacconi, M. V., Isidoro, N. & Romani, R. 2009: Olfactory responses of Hyalesthes obsoletus to host plants volatile compounds. In Abstracts of the XIX Congress of the European Chemoreception Research Organization, September 2009, Villasimius, Italy. Sforza, R., Clair, D., Daire, X., Larrue, J. & Boudon-Padieu, E. 1998: The role of Hyalesthes obsoletus (Hemiptera: Cixiidae) in the occurrence of bois noir of grapevines in France. J. Phytopathol. 146: Sharon, R., Soroker, V., Wesley, S. D., Zahavi, T., Harari, A. & Weintraub, P. G. 2005: Vitex agnus-castus is a preferred host plant for Hyalesthes obsoletus. J. Chem. Ecol. 31:

269 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp A beneficial species becomes a pest - the common earwig Forficula auricularia (Linnaeus 1758) C. Huth 1, K.-J. Schirra 1, A. Seitz 2, F. Louis 1 1 Agricultural Service Centre Palatinate, D Neustadt an der Weinstraße, Germany; 2 Johannes Gutenberg University, Institute of Zoology, D Mainz, Germany Abstract: The common earwig Forficula auricularia has been classified as a beneficial predator in vineyards. Amongst others the insect feeds on grape pests like different tortricids. In recent years within many regions of the viticultural area of the Palatinate the individual densities increased to an extremely high level. Earwigs may cause direct damages such as contamination of the grapes with faeces, eroded berries and transfer of pathogens. The chemical agent 2-methyl-1,4-benzoquinone, released from the abdominal glands while earwigs are menaced and likewise contained in faeces, may have a negative influence on the wine quality. All these facts constitute a deterioration of grape quality for wine-growers. The high number of earwigs in the grapes after vintage causes a negative image by consumers. This study was carried out to investigate possible relations between the population dynamics of earwigs and specific environmental conditions in vineyards. The main focus of the research project was focussed on chemical, ecological and biological strategies to reduce the population densities. Another important point of survey was to study the life cycle of earwigs especially in vineyards. For sampling purposes in the cluster zone a special life trap made of bamboo tubes has been developed. This type of trap showed the highest catch rate of the four trapping types tested. For monitoring the ground dwelling earwigs pitfall traps were used. Key words: Forficula auricularia, common earwig, pest, bamboo tube, life cycle, population ecology, population control, insecticide SpinTor Introduction The common earwig is omnivorous, feeding on arthropods, plants, fruit, detritus and during scarcity of food resources himself. In orchards the earwig is predominantly classified as a beneficial. To a large extent, the insect feeds on adults of the woolly aphid Eriosoma lanigerum (Hausmann 1802), on adults of the green apple aphid Aphis pomi (Degeer 1773), on eggs of the apple mussel scale Lepidosaphes ulmi (Linnaeus 1758) and on eggs of the codling moth Cydia pomonella (Linnaeus 1758) (Chant & Mcleod 1952, Asgari 1966, Glen 1975). Some apple growers regard the earwig as a pest, because they bite holes in fruit, after the skin has already been damaged by other insects and birds (Fox-Wilson 1942), but also feed on sound apples (Massee 1954). Growers of peach, apricot and cherry have similar problems since earwigs are able to bite through the filiform skin without a preliminary damage (Dahlbender & Hensel 2006). Furthermore faeces accumulate on the fruit when earwigs are very abundant (Dahlbender & Hensel 2006). In a survey Croxall et al. (1951) found a close correlation between the level of earwig damage and the incidence of the brown rot Monilinia fructigena (Honey 1945). In viticulture the earwig is also classified as a beneficial predator feeding on larvae and pupae of grape pests like the European grapevine moth Lobesia botrana (Denis & Schiffermüller 1775), the vine moth Eupoecilia ambiguella (Hübner 1796) and the grape 249

270 250 leafroller Sparganothis pilleriana (Denis & Schiffermüller 1775) (Schirra & Louis 1995, Mohr 2005). Since 2005 the classification as a beneficial species is discussed critically, due to the high number of individuals in grapevines and grapes during the vegetation period and especially before harvest (Mohr 2005). Over the past years in some of the German wineproducing regions like in the Palatinate, Rheinhessen, Rheingau, Baden and Württemberg the population densities of earwigs increased to a high level with subsequent damages. One damage, which may influence the grape health, is the huge amount of faeces in between grape berries, on grape leaves and between canes. Faeces may contain infectious units of grey mould Botrytis cinerea (Persoon 1994, Taher et al. 2009). Thus, faeces are a source for grape diseases. Faeces also attract different species of flies, which use it for egg deposition. It contains the defense agent 2-methyl-1,4-benzoquinone, which in high concentrations may have a negative influence on wine smell and flavour. The agent has a smoky smell and chemical flavour like germicide. Earwigs transfer pathogens like grey mould from grape to grape, because conidia and mycelia of fungi may attach on mouthparts and bodies of the insects. Earwigs feed on berries with preliminary damages. Other insects like wasps and fruit flies are attracted to grape juice and fruit pulp. It is known that fruit flies transfer acetic acid rot (Mohr 2005). By using grape harvesters during vintage high amaounts of earwigs are collected with the grapes. The outcome of this is a decline of grape quality for wine-growers and consumers. During the process of vintage it is possible that the earwigs may release the chemical agent 2- methyl-1,4-benzoquinone from the abdominal glands into the grapes. Whether the defense agent in grapes has an influence on wine smell and flavour is examined at present. On account of these problems on May 2007 a research project was started at the Agricultural Service Center Palatinate in Neustadt/Weinstraße, financed by the FDW (Forschungsring des Deutschen Weinbaus). Until April 2010 the following aspects are of special interest and will be investigated: to develop a life trapping method for monitoring earwigs in vineyards to analyse the spatial and temporal variation of population size in vineyards to examine the effects of different land-use managements on population size in vineyards, like mechanic soil cultivation, mulching of the vegetation cover and spraying of pesticides to observe the population ecology of earwigs in vineyards with special focus on the life cycle in vineyards (migration, mating, breeding) and biotic and abiotic environmental factors, which may influence the population densities to minimize the population size by chemical, ecological and biological strategies according to normal wine growing praxis to analyse the chemical influence of the defence agent 2-methyl-1,4-benzoquinone in high concentrations on the flavour and the smell of wine to establish rearing in the laboratory Material and methods Sampling For sampling earwigs in vineyards traps transects were built. A trap transect requires an observer to move along a fixed path and to count the population densities along the path at the same time (Buckland et al. 1993). In general one trap transect per grapevine row consisted of 30 pitfall traps (Barber 1934) and 30 self made bamboo traps.

271 251 The pitfall traps were dug in the topsoil of a grapevine row next to an adjacent vine having a bamboo trap. Pitfall traps were used for investigating the temporal and spatial pattern of earwig distribution on the ground surface. The bamboo traps were vertically fixed with wire at the grapevine trunk to sample earwigs in the cluster zone. This type of trap showed the highest catch rate of four different traps tested. Bamboo traps consisted of three bamboo tubes connected with wire or hot glue. Each tube had an inner diameter of about 1cm and a length of 20cm. The tubes were open at the lower end and closed by the node as rain protection at the upper end. The bamboo traps were easy to clean and could be used several times. Both types of traps were used from May to September in all experimental areas. In 2008 and 2009 for example 500 pitfall traps and 500 bamboo traps were evaluated per week. Results and discussion The data ascertained during the earwig monitoring are currently evaluated statistically. For example, about earwigs were caught in 2007 and in 2008 within the trial sites. Populations densities in vineyards are significantly dependent on abiotic factors like air temperature, soil type (Figure 1) and intensity of soil management. sandy clay loamy sand mean number of individuals (n = 4) / grapevine catch range (t+7) xxx * annual mean 2008 Figure 1. Mean number of earwigs in the cluster zone of two Riesling vineyards planted on different soil types. *: significant differences between the mean number of individuals (Student-t-Test, α = 0.05), Neustadt-Mußbach Furthermore the population densities in vineyards are significantly dependent from biotic factors like density of vegetation cover, mode of vegetation and root penetration (nest-building) as well as grape characteristics like vine variety (loose and tight clusters), weight of grape clusters, surrounding of the grapes and percentage of rot in grapes (Table 1). Due to the fact that the data evaluation is not yet concluded the results presented in this paper are still partial. Figure 2 shows the comparison of the annual mean number of earwigs in the cluster zone of different grape varieties between 2007 and It appears that 2007 the

272 252 earwig numbers were higher compared to A reason for this may be that the high population densities in 2007 led to a high intraspecific competition with significant resource shortage. The resource shortage had a negative effect on the reproduction in the subsequent year. Another reason could have been the climatic situation during summer Due to the numerous rainy and cold days the earwigs migrated from the cluster zone into the ground. Table 1. RDA forward selection (redundancy analysis) based on the number of earwigs (species variables) in 300 grapes with different characteristics (environmental variables). The p-value (RDA forward selection, α = 0.05) shows the influence of the environmental variables at earwig infestation in grapes. Species variables environmental variables p-value individuals/grape grape with loose clusters 0,0020 individuals/grape grape with tight clusters 0,0020 individuals/grape Grape weight (g) 0,0010 individuals/grape Grapes with leaves 0,0010 individuals/grape % ratio of rottenness/grape 0,0310 annual mean number of individuals / grape variety Helios (white grape) Riesling (white grape) Goldmuskateller (white grape) 81 * Pinotin (red grape) 98 * Reberger (red grape) 77 * 34 Spätburgunder (red grape) trap transects with grape varieties Figure 2. Comparison of the annual mean number of earwigs in the cluster zone of different grape varieties between 2007 and 2008, Neustadt-Mußbach. In addition to the monitoring in spring, summer and autumn surveying and mapping of earwig nests were carried out in winter to get to know the life cycle of the Common Earwig in vineyards of the Palatinate. For the first time these investigations described the complete cycle within vineyards. The cycle can be divided in two significant phases (Figure 3). Overwintering and breeding take place from early October to Mid-May (Figure 3). At the middle of September earwigs migrate from the foliage into the ground. At the beginning of October the females already begin with the building of nests. The average nest depth in the

273 253 ground is between 4 and 7cm, the maximum nest depth reaching about 20cm. At the middle of November the females oviposit in nests, which take place under laboratory and outdoor conditions as well. A nest contains up to 60 eggs. During wintertime the females practice active maternal care (Lamb 1976). They do not hibernate but are relatively inactive when temperatures are low (Lamb 1976). The length of the incubation period depends on ground temperature. At the middle of April the first nymphs (L 1 -larvae) hatch. About two weeks later the first moult takes place (L 2 -larvae). The nymphs stay in the nest of the female until their first moult. At the beginning of May the second moult takes place (L 3 -larvae) and the nymphs leave the nest. At the middle of May the third moult takes place (L 4 -larvae) and the nymphs forage for food on soil surface and begin to migrate into the foliage of grapevines at the end of May or at the beginning of June. The grapevine pollen is the first food attraction for earwig larvae on vines. There is a strong correlation between the beginning of blossom and migration into the foliage. In the shelter of the foliage the fourth moult takes place and the young adults hatch. Overwintering and breeding preferably take place within areas having ground vegetation and without soil cultivation. These ranges are undisturbed and the roots are integrated into the nest structure. Moving takes places from early June to Mid-September (Figure 3). During the whole summer the nocturnal earwigs spend the daytime in dark and tight places in the grapevine, like beneath the bark, in grapes, under leaves, between canes and in woody poles. Mating starts at the end of July. foliage/ cluster zone moving mating foliage/ cluster zone migration from topsoil to cluster zone Vineyard migration from cluster zone to topsoil early Mid- L 4 June September Mid- May early Oktober L 4 L 3 topsoil overwintering breeding topsoil Figure 3. Life cycle of the Common Earwig in vineyards of the Palatinate. The morphological drawings were made by Kulzer (1996).

274 254 For earwig control different mechanical, biological and chemical strategies were tested on the ground and in the foliage of vineyards (Table 2). Table 2. Mechanical, biological and chemical strategies estimated for earwig control on the ground and in the foliage of vineyards until 2008, t: start of treatment, c: concentration of agent. Control Strategy Location Development Stage Effect bottom area overwintering adults, t: larvae (stage L c: 3 kg/a, 6 kg/a 3, L 4 ) granulate material PERLKA calcium cyanamide insect parasitic nematodes Nematop 50 Heterorhabditis bacteriophora soil cultivation by removing vegetation cover for one winter Earth wall covers with black plastic film between grapevine rows grape cluster thinning with harvester insect repellent ENVIRepel essence of garlic liming the foliage with Hydrocal SUPER 85 hydrated lime (ph 12,6) one-sided and both-sided leaf stripping in the foliage night spraying with the insecticide Reldan 22 agent: Chlorpyrifos-methyl night spraying with the insecticide Confidor WG 70 agent: Imidacloprid night spraying with the insecticide SpinTor agent: Spinosad day spraying with the insecticide SpinTor agent: Spinosad bottom area t: c: 50 Mio/100 m² bottom area t1: t2: soil surface t: foliage t: foliage t1-t3: c: 0,5% foliage t1-t4: c: 30 kg/ha, 60 kg/ha foliage t: foliage t1: t2: , c: 0,2% foliage t1: t2: , c: 0,03% foliage t1: t2: , c: 0,01% foliage t1: c: 0,01% overwintering adults, larvae (stage L 3, L 4 ) eggs, overwintering adults moving adults, moving larvae (stage L 4 ) moving adults moving adults moving adults moving adults moving adults moving adults moving adults moving adults no reducing effect on earwigs no reducing effect on earwigs no reducing effect on earwigs no reducing effect on earwigs no reducing effect on earwigs no reducing effect on earwigs no reducing effect on earwigs no reducing effect on earwigs short reducing effect to 50% of efficiency short reducing effect to 76% of efficiency long reducing effect to 94% of efficiency short reducing effect to 67% of efficiency The insecticide SpinTor was very efficient against earwigs in the foliage of grapevines. One week after night application the population densities were 94% lower compared to the densities in the untreated control (Figure 4). Day application reached an efficiency of 67%. These results show that a night application of SpinTor was more effective against the earwig due to the fact that earwigs are nocturnal insects leaving their daytime shelter for food searching during the night. Therefore SpinTor reaches the earwigs better at night than at daytime: The daytime shelter habitats avoid a direct contact between the agent and the insects.

275 255 mean number of individuals / grapevine (n=20) a b c a ab b a ab b a 55 ab 17 b effectiveness ABBOTT (%) untreated control SpinTor-day spraying SpinTor-night spraying day spraying effectiveness night spraying effectiveness a, ab, b, c - groups of significance, Tukey (HSD): a = 0,05 catch range (t+7) Figure 4. Efficiency of the insecticide SpinTor on earwigs: A single day spraying in comparison with a single night spraying, Riesling-grapevines, Neustadt-Mußbach References Abbott, W. S. 1925: A method of computing the effectiveness of an insecticide. Journal of Economic Entomology 18: Asgari, A., 1966: Untersuchungen über die im Raum Stuttgart-Hohenheim als wichtigste Prädatoren der grünen Apfelblattlaus (Aphidula pomi Deg.) auftretenden Arthropoden. Zeitschrift für angewandte Zoologie 53(1): Barber, H. 1931: Traps for cave-inhabiting insects. Journal of Elisha Mitchell Science Society 46: Buckland, S. T., Anderson, D. R., Burnham, K. P. & Laake, J. L. 1993: Distance Sampling: Estimating Abundance of Biological Populations. London. Chapman and Hall, 446 pp. Chant, D. A. & McLeod, J. H. 1952: Effects of certain climatic factors on the daily abundance of the European earwig Forficula auricularia L. (Dermaptera: Forficulidae) in Vancouver, British Columbia. The Canadian Entomologist 84: Croxall, H. E., Collingwood, G. A. & Jenkins, J. E. E. 1951: Observations on brown rot (Sclerotinia fructigena) of apples in relation to injury caused by earwigs (Forficula auricularia). Annals of Applied Biology 38(4): Dahlbender, W. & Hensel, G. 2006: Regulierung des Ohrwurms. Mitteilungen DLR Rheinpfalz Kompetenzzentrum Gartenbau Juniausgabe: 1-6. Fox-Wilson, G. 1942: The invasion of houses by earwigs and ants. Annals of Applied Biology 29(3): Glen, D. M. 1975: The effects of predator on the eggs of codling moth Cydia pomonella, in a cider-apple orchard in South-West England. Annals of Applied Biology 80: Kulzer, L. 1996: The European Earwig Forficula auricularia, Order Dermaptera, Family Forficulidae. Scarabogram, New Series 198: 2-4. Lamb, R. J. 1976: Dispersal by nesting earwigs, Forficula auricularia (Dermaptera: Forficulidae). The Canadian Entomologist 108: Massee, A. M. 1954: The Pests of Fruit and Hops. Crosby, Lockwood and Son Ltd., London.

276 256 Mohr, H. D. 2005: Farbatlas Krankheiten, Schädlinge und Nützlinge an der Weinrebe. Stuttgart (Hohenheim), Eugen Ulmer KG, 320 pp. Phillips, M. L. 1981: The ecology of the Common Earwig Forficula auricularia in apple orchards. Thesis, University of Bristol, Department of Agriculture and Horticulture Long Ashton Research Station. Schirra, K. J. & Louis, F. 1995: Auftreten von natürlichen Antagonisten des Springwurmwicklers Sparganothis pilleriana in der Pfalz. Deutsches Weinbau-Jahrbuch 1995: Steinmann, H. 1993: The Animal Kingdom Dermaptera/Eudermaptera II. Walter de Gruyter, Berlin, New York: 711 pp. Taher, M., Breuer, M. & Zebitz, A. 2009: The European earwig, Forficula auricularia L. (Dermaptera, Forficulidae) a vector of phytopathogenous fungi of grapevines? Posterbeitrag, DGaaE-Tagung Vogt, H., J. Just & Grutzmacher, A. 2009: Field test of the impact of four insecticides on European earwig, Forficula auricularia, in an apple orchard. DGaaE-Nachrichten 23(1): 31 Walker, K. A., Jones, T. H. & Fell, R. D. 1993: Pheromonal basis of aggregation in European Earwig, Forficula auricularia L. (Dermaptera: Forficulidae). Journal of Chemical Ecology 28:

277 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Effects of some insecticides on Kampimodromus aberrans: laboratory and field studies P. Tirello, S. Vettore, A. Pozzebon, M. Lorenzon, C. Duso Department of Environmental Agronomy and Crop Science, University of Padua, Agripolis, Legnaro (PD), Italy Abstract: Kampimodromus aberrans is an important predatory mite in European vineyards and fruit orchards. Pesticides can exert profound effects on K. aberrans populations that are influenced by the pesticide experience of predatory mites. Resistance to organophosphates and dithiocarbamates has been suggested for some Italian populations of K. aberrans but this phenomenon has not been shown in the laboratory. The effects of three different pesticides (chlorpyriphos-ethyl, methoxyfenozide and flufenoxuron) on K. aberrans were investigated in field and laboratory conditions. In field tests, predatory mites were released at the beginning of February on defined plots. Kampimodromus aberrans densities were evaluated on insecticide treated plots and on control plots by examining leaf samples in the laboratory from June to August. The lethal and sub-lethal effects of the same pesticides were tested in the laboratory using a residual bioassay. Field experiments showed some effects of the three pesticides on predatory mite populations. Results of bioassays were partially correlated with those obtained in the field. Implications for IPM are discussed. Key words: Kampimodromus aberrans, insecticides, resistance, Integrated pest management, grape Introduction Kampimodromus aberrans is one of the most important predatory mites occurring in European vineyards and fruit orchards. The biological and ecological peculiarities of this predator, such as wide food range, persistence in conditions of prey scarcity, and tolerance to warm and dry conditions, make K. aberrans particularly effective in preventing spider mite infestations (Duso, 2006). Its potential as a beneficial is often limited by its susceptibility to several pesticides (Girolami, 1981; Ragusa, 1983; Marchesini, 1989; Baldessari et al., 2005; Tosi et al., 2006). A large number of pesticides are used to control various pests and diseases in vineyards. Chemical control is often required against berry moths, mealybugs and leafhoppers. Among the latter, Scaphoideus titanus Ball is actually one of the most important pests in several European regions. It is the objective of mandatory chemical control because of its role as vector of phytoplasmas, in particular Flavescence dorée, to grapes. For these reasons, predatory mites occurring in vineyards are exposed to a significant number of pesticide applications. For a number of new compounds, knowledge of the side-effects of pesticides on beneficial mites is limited. Moreover, the susceptibility of predatory mite populations to traditional pesticides (e.g., organophosphates and carbamates) is not homogeneous: strains susceptible or resistant to the same compound can occur in the same area (Ioratti & Balloid, 1987; Maixner, 1992; Cross & Berrie, 1994; Angeli et al., 1997; Barbar et al., 2007; Bonafos et al., 2007). Pyrethroids are usually associated to negative effects on phytoseiids (Aliniazee & Cranham, 1980; Abou-Awad & El-Banhawy, 1985), but resistant strains have recently been found (Bonafos et al., 2007). 257

278 258 During recent decades some products with new modes of action have been developed by agrochemical industries. Growth regulators, such as flufenoxuron and methoxyfenozide, appeared harmless to predatory mites in vineyards (e.g., Mori et al., 1999; Tosi et al., 2006; Scannavini et al., 2006). Some neonicotinoids resulted toxic towards phytoseiids (James, 2003; Poletti et al., 2007) while others appeared safer to predatory mites (Tosi et al., 2006; Poletti et al., 2007). Here we present the results of field and laboratory trials on three insecticides largely used in European vineyards (i.e., chlorpyriphos, methoxyfenozide and flufenoxuron) on a K. aberrans strain probably resistant to organophosphates. Material and methods Field experiments Field experiments were conducted in the experimental station of CRA - Viticulture located at Spresiano (Treviso province, Veneto region), Italy. The experimental design comprised four treatments: methoxyfenozide, chlorpyriphos, flufenoxuron and control with four replications of four plants per treatments. At the end of winter, approximately 100 females of K. aberrans were released per plant, using two-year old branches (see Duso, 1989 for a description of the release method). This strain of K. aberrans had been collected in a commercial vineyard treated with dithiocarbamates (e.g. mancozeb) and organophosphates (e.g. fenitrothion). Insecticides were applied in mid-june at the highest doses admitted for grapevines. In particular, the following compounds and doses were used: chlorpyriphos - Dursban 75WG 70g/hl, methoxyfenozide - Prodigy 40ml/hl, and flufenoxuron - Cascade 50DC 150ml/hl). Leaf samples were collected every days from June 4 th to August 13 th to assess mite population variations. Leaves were transported to the laboratory where they were examined under a dissecting microscope to assess mite identity and their numbers. Data were analyzed using a Repeated Measures General Linear Mixed model and F test (α = 0.05) was performed to evaluate the effect of treatment, time and their interaction. The effects (E F ) of insecticides were estimated as the change in population levels in treated plots related to the change in the control plots during the observation period, using Henderson and Tilton s formula (1955). Laboratory tests Predatory mites were mass-reared on artificial arenas using a grape leaf placed on a wet cotton layer as a substrate. A cotton barrier was created on the edges of leaves to prevent mite escape. Kampimodromus aberrans individuals were collected from leaves and transferred onto rearing arenas using a fine brush. Small pieces of PVC were placed on each arena as oviposition sites. Typha latifolia pollen was used as food source and was replenished every two days. We evaluated the effect of chlorpyriphos, methoxyfenozide and flufenoxuron on the mortality, fecundity and hatching rate of K. aberrans. The same pesticide doses used in field experiments were adopted. The experimental units were similar to rearing arenas and were constituted by a grape leaf section. Prior to the experiments, leaf sections were immersed in the pesticide solutions for 30s (water was used in control treatments). When pesticide residues completely dried out, two coeval females of K. aberrans were gently transferred onto each leaf section. Female mortality was recorded three days after treatments, and fecundity was recorded daily for four additional days. After seven days, the surviving females and young stages were removed and the eggs were monitored until they hatched out completely in the

279 259 control. The experimental units were maintained in a climatic chamber at 25 ± 2 C, 60 ± 10% RH and a photoperiod of 16L: 8D. We estimated the overall toxicity of insecticides using the following formula: E L = 100% - ((100% - M) R 1 R 2 ) M = corrected mortality (Abbott, 1925) R 1 = corrected fecundity R 2 = corrected hatching rate Results and discussion In the field experiments we observed significant effects (p < 0.05) of the three insecticides on predatory mite populations. No differences were observed among insecticide treatments. The highest reduction of K. aberrans populations was found in chlorpyriphos treated plots (E F = 65.5%), while reductions at lower levels were observed in methoxyfenozide (E F = 56.9%) and flufenoxuron (E F = 28%) treated plots. The laboratory tests showed no significant effects of chlorpyriphos, methoxyfenozide and flufenoxuron on the survival of predatory mites. The fecundity of K. aberrans was lower in the chlorpyriphos treatment compared to the other treatments. Egg hatching rates were lower in chlorpyriphos and methoxyfenozide treatments compared to other treatments. In term of overall toxicity, chlorpyriphos (E L = 89) resulted the most toxic insecticide, while methoxyfenozide and flufenoxuron were associated with lower toxicity levels (E L = 37 and E L = 23.6, respectively). Regarding chlorpyriphos, the effects on fecundity and hatching rate observed in the laboratory could explain the reduction in predatory mite densities observed in the field. Concerning methoxyfenozide, the reduction in K. aberrans populations observed in the field could be related to the pesticide effects on predator hatching rate. Flufenoxuron did not affect markedly predatory mites parameters in the laboratory, confirming the moderate effects seen in the field. It should be stressed that K. aberrans populations were released in that season and their densities were relatively low. Further experiments will be designed to further our knowledge on the impact of these insecticides on K. aberrans populations. Acknowledgements We would like to thank Dr. M. Borgo (CRA Viticulture) and his co-workers for their assistance in the field trials. References Abou-Awad, B. A. & El-Banhawy, E. M. 1985: Comparison between the toxicity of synthetic pyrethroids and other compounds to the predacious mite Amblyseius gossipi (Mesostigmata: Phytoseiidae). Exp. Appl. Acarol. 1: Aliniazee, M. T. & Cranham, J. E. 1980: Effect of four synthetic pyrethroids on predatory mite, Typhlodromus pyri and its prey Panonychus ulmi on apples in south-east England. Environ. Entomol. 9: Angeli, G., Forti, D. & Maines, R. 1997: Effetti collaterali di fitofarmaci di interesse fruttiviticolo verso gli acari fitoseidi. Informatore agrario 14:

280 260 Angeli, G. & Ioratti, C. 1994: Susceptibility of two strains of Amblyseius andersoni Chant. (Acari: Phytoseiidae) to dithiocarbamate fungicides. Exp. Appl. Acarol. 18: Baldessari, M., Angeli, G. & Girolami, V. 2005: Effects of plant protection products on Kampimodromus aberrans Oudemans: the dietary effect of airborne pollen. IOBC/WPRS Bull. 28(7): Barbar, Z., Tixier, M. & Kreiter, S. 2007: Assessment of pesticide susceptibility for Typhlodromus exhilaratus and Typhlodromus phialatus strains (Acari: Phytoseiidae) from vineyards in the south of France. Exp. Appl. Acarol. 42: Bonafos, R., Serrano, E., Auger, P. & Kreiter, S. 2007: Resistance to deltamethrin, lambdacyhalothrin and chlorpyriphos-ethyl in some populations of Typhlodromus pyri Scheuten and Amblyseius andersoni Chant. (Acari: Phytoseiidae) from vineyards in the south-west of France. Crop Protection 26(2): Cross, J. V. & Berrie, A. M. 1994: Effects of repeated foliar sprays of insecticides or fungicides on organophosphate-resistant strains of the orchard predatory mite Typhlodromus pyri on apple. Crop protection 13(1): Duso, C. 1989: Role of the predatory mites Amblyseius aberrans (Oud.), Typhlodromus pyri Scheuten and Amblyseius andersoni (Chant) (Acari, Phytoseiidae) in vineyards. I. The effects of single or mixed phytoseiid population releases on spider mite densities (Acari, Tetranychidae). J. Appl. Entomol. 107: Duso, C. 2006: Il controllo biologico ed integrato degli acari fitofagi associati alla vite. In: La difesa della vite dagli artropodi dannosi, eds. Ragusa and Tsolakis: Girolami, V. 1981: Danni, soglie di intervento, controllo degli acari della vite. Atti "III incontro su la difesa integrata della vite" Latina, Italy, 3-4 dicembre 1981: Henderson, C. F. & Tilton, E. W. 1955: Test with acaricides against brown wheat mites. J. Econ. Entomol. 48: Ioratti, C. & Balloid, M. 1987: Determinazione della tossicità di 15 insetticidi su un ceppo di Amblyseius andersoni Chant. (Acari: Phytoseiidae). Vignevini 5: James, G. D. 2003: Toxicity of imidacloprid to Galendromus occidentalis, Neoselius fallacis and Amblyseius andersoni (Acari: Phytoseiidae) from hops in Washington State, USA. Exp. Appl. Acarol. 31: Maixner, M. 1990: Investigations on insecticide resistance in the predatory mite Typhlodromus pyri Scheuten (Acari: Phytoseiidae) on grapevines in the wine region Mosel-Saar-Ruwer. Mitteilungen aus der Biologischen Bundesanstalt fur Land- und Forstwirtschaft Berlin-Dahlem 257: 118 pp. Marchesini, E. 1989: Effetti collaterali antiperonosporici diversi su Kampimodromus aberrans. Informatore agrario 45(12): Mori, N., Posenato, G., Sancassani, G., Tosi, L. & Girolami, V. 1999: Insetticidi per il controllo delle cicaline nei vigneti. Informatore Agrario 55(15): Poletti, M., Maia, A. H. N. & Omoto, C. 2007: Toxicity of neonicotinoid insecticides to Neoselius californicus and Phytoseiulus macropilis (Acari: Phytoseiidae) and their impact on functional response to Tetranychus urticae (Acari: Tetranychidae). Biol. Control 40(1): Ragusa, S. 1983: Effetti collaterali di fitofarmaci su acari fitoseidi del nocciolo. Atti 13 Congr. Naz. It. Ent., Sestriere, Torino: Scannavini, M., Cavazza, F., Boselli, M., Melandri, M., Pradolesi, G., Cavallini, G. & Milanesi, L. 2006: Difesa contro la tignoletta con methoxyfenozide. Informatore Agrario 62(8): Tosi, L., Farinazzo, E., Posenato, G. & Girolami, V. 2006: Effetti collaterali di insetticidi su Kampimodromus aberrans. Informatore agrario 26:

281 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Development and reproduction of the predatory mites Kampimodromus aberrans, Typhlodromus pyri and Amblyseius andersoni on different food sources M. Lorenzon, A. Pozzebon, C. Duso Department of Environmental Agronomy and Crop Science, University of Padua, Agripolis, Legnaro, Padova, Italy Abstract: Kampimodromus aberrans, Typhlodromus pyri and Amblyseius andersoni, important predatory mites in European vineyards, were reared in the laboratory on five potential food sources: the herbivores Panonychus ulmi and Eotetranychus carpini, the non-specialized feeder Tydeus caudatus, the pollen of Typha latifolia, and the mycelium of grape downy mildew Plasmopara viticola. Experiments were conducted using grape leaves as a substrate. The developmental times, survival and fecundity of predatory mites were determined on each food type. None of these species developed on T. caudatus. Developmental times of K. aberrans and T. pyri were faster on pollen than on spider mites or P. viticola mycelium. The fecundity of K. aberrans and T. pyri was higher on pollen than on other foods. Amblyseius andersoni developed faster on pollen than on P. ulmi or P. viticola mycelium, and showed a low fecundity on P. viticola mycelium. Key words: Phytoseiidae, generalist predators, feeding habits, biological parameters, biological control Introduction Kampimodromus aberrans (Oudemans), Typhlodromus pyri Scheuten and Amblyseius andersoni (Chant) are common predatory mites in European vineyards (Duso & Liguori, 1984; Schruft, 1985; Maixner, 1990; Tixier et al., 1998; Pereira et al., 2003). They can keep the populations of tetranychid mites Panonychus ulmi (Koch) and Eotetranychus carpini (Oudemans) at non-damaging levels (Duso, 1989; Camporese & Duso, 1996). Generalist phytoseiids can persist on plants when their main prey is absent or scarce by feeding on other foods (McMurtry & Croft, 1997). On perennial plants, such as grapevines, the persistence of predatory mites is fundamental if tetranychid and eriophyoid outbreaks are to be prevented. Those predators that have a wide range of food sources can maintain their populations at high levels and ensure a better biological control of herbivores (Symondson et al., 2002). The range of food sources for generalist phytoseiids can include alternative prey, particularly insects like thrips, coccids and whiteflies (e.g., Engel & Ohnesorge, 1994; Schausberger, 1998; Nomikou et al., 2001), and mites like eriophyids (Overmeer, 1985), tarsonemids, tenuipalpids (McMurtry et al., 1984), or other phytoseiids (Schausberger, 2003). Tydeids were also proposed as potential prey for phytoseiids. Homeopronematus anconai (Baker) was a food source for Metaseiulus occidentalis (Nesbitt) (Flaherty & Hoy, 1971) and has favoured predator persistence in California vineyards when the main prey Tetranychus pacificus McGregor was scarce (Knop & Hoy, 1983). Tydeus caudatus Dugès was considered a suitable prey for T. pyri in the laboratory (Calis et al., 1988). This tydeid is very common in European vineyards where it is a prey of the phytoseiid mite Paraseiulus talbii (Athias- 261

282 262 Henriot) (Camporese & Duso, 1995). However, there is still very little known about the relationships between generalist phytoseiids and T. caudatus. Several studies suggest that among alternative foods, pollen is suitable for many phytoseiid species (e.g., Dosse, 1961; Overmeer, 1985; McMurtry & Rodriguez, 1987; McMurtry et al., 1991; Addison et al., 2000; Duso et al., 1997, 2004; Kasap, 2005; Roda et al., 2003). Pollen often favours faster developmental times and greater oviposition rates than tetranychids (Abdallah et al., 2002; Duso & Camporese, 1991; Kasap, 2005). However, the abundance of pollen in crop perennial ecosystems is inconstant throughout the season (Eichhorn & Hoos, 1990; Duso et al., 1997; Addison et al., 2000). An additional possible alternative food source for predatory mites is represented by plant pathogenic fungi (Daftari, 1979; Zemek & Prenerovà, 1997; Pozzebon & Duso, 2008; Pozzebon et al., 2009). Grape downy mildew (GDM) Plasmopara viticola (Berk & Curtis), the most important grape disease in temperate humid areas, often spreads in late summer when the use of fungicides has ceased. A correlation between the spread of GDM foliar symptoms and the abundance of T. pyri and A. andersoni has been found in north-italian vineyards (Duso et al., 2003). Knowledge of the influence of food sources on phytoseiid life history parameters is not exhaustive for many species. In this paper, the effects of five food sources on the development and reproduction of K. aberrans, T. pyri and A. andersoni are briefly reported. Material and methods Kampimodromus aberrans, T. pyri and A. andersoni were collected from vineyards and reared in laboratory conditions (24 C, 70±10% r.h., 16L:8D of photoperiod) for at least three generations before experiments were carried out. To obtain coeval eggs, the females were transferred from laboratory colonies onto new leaves and allowed to lay eggs for 12 hours. The eggs were then transferred singly to experimental units consisting of a section of grape leaf placed on a wet layer of cotton within a plastic box. The experimental units were maintained in climatic chambers at constant conditions (24 C, 70±10% r.h., 16L:8D of photoperiod). To determine developmental times, the units were monitored every 12 hours. The presence of an exuvium was the criterion used to prove successful moulting to the next developmental stage. When adults were obtained, one male and one female obtained on the same food source were put together and oviposition was monitored every 24 hours. We compared five food sources: the mites P. ulmi, E. carpini, and T. caudatus, Typha latifolia L. pollen, and P. viticola mycelium. The cattail pollen had previously been collected, dried and stored in a freezer and was supplied every day (about 0.5mg of pollen per day) to phytoseiids with a fine paint brush. Panonychus ulmi and E. carpini were collected from infested vineyards. Tydeus caudatus was reared in the laboratory. Mite individuals were supplied every day. The grape downy mildew was propagated and maintained on grape leaves in laboratory conditions and was offered using infected leaf sections. Predatory mites were transferred onto new leaves when the original leaves started to deteriorate. Results and discussion In the control, the protonymphs died before moulting to deutonymphs. Similar results were obtained when T. caudatus was offered as prey. On the other food sources the predatory mites completed their development.

283 263 The developmental times of both K. aberrans females and males ranged from 156 to 204 hours. For K. aberrans females and males, pollen and E. carpini induced the shortest developmental times, while P. viticola induced the longest. Female longevity ranged from 6.5 to 42 days, male longevity from 4 to 27 days depending on the food source. Adult longevity was relatively high on pollen, low on P. viticola and intermediate on tetranychids. Total fecundity lasted from 11 to 27 eggs, and was higher on pollen than on spider mites, while no eggs were laid by mites confined on leaves infected by P. viticola. Regarding T. pyri, development was completed in hours by females, and in hours by males. The type of food significantly affected female developmental times. Pollen induced a faster development than P. viticola, while tetranychid mites were associated to intermediate values. Female longevity ranged from 16 to 35 days, male longevity from 7 to 19 days. The lowest values were obtained on P. viticola mycelium. Total fecundity lasted from 4 to 29 eggs, and the highest levels were observed when females were fed with pollen. Amblyseius andersoni females completed their development in hours, males in hours. Developmental times on P. viticola and P. ulmi were longer than on pollen, while E. carpini were associated to intermediate values. Female longevity ranged from 16 to 41 days, male longevity from 5 to 13 days. The lowest longevity values were observed on P. viticola mycelium. Total fecundity (7-28 eggs) was higher on pollen or on tetranychids than on P. viticola mycelium. Results suggest that T. caudatus does not represent a food source for these predatory mites. The suitability of pollen for generalist predatory mites is confirmed, and it should be stressed that they obtained a higher performance on pollen than on tetranychids. Plant pathogens were confirmed as an alternative food for A. andersoni and T. pyri but only as a supplementary food for K. aberrans. Amblyseius andersoni developed faster and laid more eggs than T. pyri or K. aberrans when fed with pollen or other food sources. However, field observations suggest that A. andersoni populations do not persist in vineyards where tetranychid numbers decline (Duso, 1989; Duso & Vettorazzo, 1999). In these situations, A. andersoni is frequently displaced by K. aberrans or T. pyri despite its greater aggressiveness (Zhang & Croft, 1995; Croft & Croft, 1996). Food availability in time and space and its quality should be considered more in depth when the ecology of generalist predatory mites is investigated in perennial agro-ecosystems. References Addison, J. A., Hardman, J. M. & Walde, S. J. 2000: Pollen availability for predaceous mites on apple: spatial and temporal heterogeneity. Exp. Appl. Acarol. 24: Calis, J. N. M., Overmeer, W. P. J. & Van Der Geest, L. P. S. 1988: Tydeids as alternative prey for phytoseiid mites in apple orchards. Proceedings International Symposium on Crop Protection, University of Gent, 40: Camporese, P. & Duso, C. 1995: Life history and life table parameters of the predatory mite Typhlodromus talbii. Entomol. Exp. Appl. 77: Camporese, P. & Duso, C. 1996: Different colonization patterns of phytophagous and predatory mites (Acari: Tetranychidae, Phytoseiidae) on three grape varieties: a case study. Exp. Appl. Acarol. 20: Croft, B. A. & Croft, M. B. 1996: Intra- and interspecific predation among adult female phytoseiid mites (Acari: Phytoseiidae): effects on survival and reproduction. Environ. Entomol. 25:

284 264 Daftari, A. 1979: Studies on feeding, reproduction and development of Amblyseius aberrans (Acarina: Phytoseiidae) on various food substances. J. Appl. Entomol.-Zeitschrift für Angewandte Entomologie 88: Dosse, G. 1961: Über die Bedeutung der Pollennahrung für Typhlodromus pyri Scheuten (= tilllae Oud.) (Acari: Phytoseiidae). Entomol. Exp. Appl. 4: Duso, C. 1989: Role of predatory mites Amblyseius aberrans (Oud.), Typhlodromus pyri Scheuten and Amblyseius andersoni (Chant)(Acari, Phytoseiidae) in vineyards. I. The effect of single or mixed phytoseiid population release on spider mite densities (Acari, Tetranychidae). J. Appl. Entomol. 107: Duso, C. & Camporese, P. 1991: Developmental times and oviposition rates of predatory mites Typhlodromus pyri and Amblyseius andersoni (Acari: Phytoseiidae) reared on different foods. Exp. Appl. Acarol. 13: Duso, C. & Liguori, M. 1984: Ricerche sugli Acari Tetranichidi della vite in Veneto: aspetti faunistici ed incidenza degli interventi fitosanitari sulle popolazioni degli acari fitofagi e predatori. Redia 67: Duso, C., Malagnini, V. & Paganelli, A. 1997: Indagini preliminari sul rapporto tra polline e Kampimodromus aberrans (Acari: Phytoseiidae) su Vitis vinifera L. Allionia 35: Duso, C., Malagnini, V., Paganelli, A., Aldegheri, L., Bottini, M. & Otto, S. 2004: Pollen availability and abundance of predatory phytoseiid mites on natural and secondary hedgerows. Biocontrol 49: Duso, C., Pozzebon, A., Capuzzo, C., Bisol, P. M. & Otto, S. 2003: Grape downy mildew spread and mite seasonal abundance in vineyards: evidence for predatory mites Amblyseius andersoni and Typhlodromus pyri. Biol. Control 27: Duso, C. & Vettorazzo, E. 1999: Mite population dynamics on different grape varieties with or without phytoseiids released (Acari: Phytoseiidae). Exp. Appl. Acarol. 23: Eichhorn, K. W. & Hoos, D. 1990: Investigations in population dynamics of Typhlodromus pyri in vineyards of Palatina, F. R. Germany. IOBC/WPRS Bull XIII/7: Engel, R. & Ohnesorge, B. 1994: Die Rolle von Ersatznahrung und Mikroklima im System Typhlodromus pyri Scheuten (Acari, Phytoseidae) - Panonychus ulmi Koch (Acari, Tetranychidae) auf Weinreben I. Untersuchungen im Labor. J. Appl. Entomol.- Zeitschrift für Angewandte Entomologie 118: McMurtry, J. A., Badii, M. H. & Johnson, H. G. 1984: The broad mite, Polyphagotarsonemus latus, as a potential prey for phytoseiid mites in California. Entomophaga 29: McMurtry, J. A. & Croft, B. A. 1997: Life-styles of phytoseiid mites and their role in biological control. Ann. Rev. Entomol. 42: McMurtry, J. A., de Moraes, G. J. & Johnson, H. G. 1991: Arrestment responses of some phytoseiid mites to extracts of Oligonychus punicae, Tetranychus urticae and pollen. Isr. J. Entomol : McMurtry, J. A. & Rodriquez, J. 1987: Nutritional ecology of phytoseiid mites. In: Slansky, F., Rodriquez, J. (eds.): Nutritional Ecology of Insects, Mites and Spiders. Wiley & Sons, New York: Nomikou, M., Janssen, A., Schraag, R. & Sabelis, M. W. 2001: Phytoseiid predators as potential biological control agents for Bemisia tabaci. Exp. Appl. Acarol. 25: Overmeer, W. P. J. 1985: Alternative prey and other food resources. In: Helle, W., Sabelis, M. W. (eds.): Spider Mites. Their Biology, Natural Enemies and Control. Elsevier, Amsterdam:

285 Pereira, J. A., Torres, L., Espinha, I. & Ferragut, F. 2003: Contribution to the knowledge of phytoseiid mites associated with vineyards in the Regiào Demarcada do Douro (porto wine region). Acarologia 18: Pozzebon, A. & Duso, C. 2008: Grape downy mildew, an alternative food for generalist predatory mites occurring in vineyard. Biol. Control 45: Pozzebon, A., Loeb, G. M. & Duso, C. 2009: Grape powdery mildew as a food source for generalist predatory mites occurring in vineyard: effects on life history traits. Ann. Appl. Biol. 115: Kasap, I. 2005: Life-history traits of predaceous mites Kampimodromus aberrans (Oudemans) (Acarina: Phytoseiidae) on four different types of food. Biol. Control 35: Knop, N. F. & Flaherty, D. L. 1971: Biological control of Pacific mites and Willamette mites in San Joaquin Valley vineyards: part III. Role of tydeid mites. Res. Popul. Ecol. 13(1): Knop, N. F. & Hoy, M. A. 1983: Biology of a tydeid mite, Homeopronematus anconai (n. comb.) (Acari: Tydeidae), important in San Joaquin Valley vineyards. Hilgardia 51(5): Roda, A., Nyrop, J. & English-Loeb, G. 2003: Leaf pubescence mediates the abundances of non-prey food and the density of the predatory mite Typhlodromus pyri. Exp. Appl. Acarol. 29: Schausberger, P. 1998: Survival, development and fecundity in Euseius finlandicus, Typhlodromus pyri, and Kampimodromus aberrans feeding on the San José scale Quadraspidiotus perniciosus. J. Appl. Entomol. 122: Schausberger, P. 2003: Cannibalism among phytoseiid mites: a review. Exp. Appl. Acarol. 29: Schruft, G. 1985: Grape. In: Helle, W. & Sabelis, M. W. (eds.): Spider Mites. Their Biology, Natural Enemies and Control. Elsevier, Amsterdam: Symondson, W. O. C., Sunderland, K. D. & Greenstone, M. H. 2002: Can generalist predators be effective biocontrol agents? Ann. Rev. Entomol. 47: Tixier, M. S., Kreiter, S., Auger, P. & Weber, M. 1998: Colonization of Languedoc vineyards phytoseiid mites (Acari: Phytoseiidae): influence of wind and crop environment. Exp. Appl. Acarol. 22: Zemek, R. & Prenerovà, E. 1997: Powdery mildew (Ascomycotina: Erysiphales) an alternative food for predatory mite Typhlodromus pyri Scheuten (Acari: Phytoseiidae). Exp. Appl. Acarol. 21: Zhang, Z.-Q. & Croft, B. A. 1995: Interspecific competition and predation between immature Amblyseius fallacis, Amblyseius andersoni, Typhlodromus occidentalis and Typhlodromus pyri (Acari: Phytoseiidae). Exp. Appl. Acarol. 19:

286 266

287 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Susceptibility and sensibility of grape cultivars to the leafhopper Empoasca vitis D. Fornasiero 1, A. Pozzebon 1, F. Pavan 2, C. Duso 1 1 Department of Environmental Agronomy and Crop Science, University of Padua, Agripolis, Legnaro (PD), Italy. 2 Department of Biology and Plant Protection, University of Udine, Udine, Italy. Abstract: The leafhopper Empoasca vitis (Goethe) (Homoptera: Cicadellidae) is a serious pest in European vineyards. The assessment of economic damage caused by E. vitis was associated with no definite results probably because of a different susceptibility and sensibility of grape cultivars to the pest. Empoasca vitis abundance and symptoms associated with the activity of this pest were evaluated during three years on 10 grape cultivars grown in collections located in north-eastern Italy. Empoasca vitis densities were estimated from May-June to September by sampling leaves every 7-10 days. At vintage time, the percentage of leaf surface with symptoms caused by E. vitis feeding was estimated. Carménère, Sauvignon Blanc and Tocai Friulano showed to be very susceptible and very sensible, whereas Cabernet Sauvignon, Chardonnay, Verduzzo Friulano and Pinot Gris showed an opposite behaviour. However, such correspondence was influenced by meteorological conditions and plant stress factors. These results may be useful for establishing economic thresholds and sampling plans for E. vitis in vineyards. Key words: Empoasca vitis, grape cultivar, susceptibility, sensibility, economic thresholds Introduction The leafhopper Empoasca vitis (Goethe) (Homoptera, Cicadellidae) is a serious pest of grapevine in Europe (Vidano, 1958; Baggiolini et al., 1968; Moutous & Fos 1973; Cerutti et al., 1988; Baillod et al., 1990). Adults and nymphs can feed on vascular tissues, mainly on the phloem (Vidano, 1963; Carle & Moutous, 1965; Tavella & Arzone, 1992) causing a reduction in photosynthesis, mesophyll conductance and transpiration rate (Candolfi et al., 1993). Typically leaf margins become reddish or yellowish and the symptoms progressively invade the leaf lamina that can partially or completely dry out (Vidano, 1963; Baggiolini et al., 1972). The severity of symptoms is affected by the cultivar (Trentini, 1962; Baggiolini et al., 1968; Touzeau, 1968; Vidano et al., 1988; Cerutti et al., 1989; Dal Rì, 1992; Pavan et al., 1992; Pavan et al., 2000; Pavan & Picotti, 2009), dry conditions and plant stress factors (Schvester et al., 1962; Vidano, 1963; Pavan et al., 1992, 2000). The assessment of economic damage caused by E. vitis was associated with no definite results probably because of a different susceptibility and sensibility of grape cultivars to the pest (Pavan et al., 2000). We compared the susceptibility (probability of a plant to be attacked by herbivores) and the sensibility (degree of symptoms manifestation) of 10 grape cultivars to E. vitis. In particular, four red cultivars (Carménère, Cabernet Sauvignon, Merlot, and Refosco dal peduncolo rosso) and six white cultivars (Chardonnay, Pinot Gris, Rhine Riesling, Sauvignon Blanc, Tocai Friulano, and Verduzzo Friulano) were considered. Observations were carried out in collections of grape cultivars located in North-eastern Italy during three years. 267

288 268 Material and methods Estimation of Empoasca vitis population densities In the first experimental year, 40 shoots per cultivar were chosen and the leafhopper nymphs density was weekly estimated from June to September on the five basal leaves of each shoot. In the remaining years, the leafhopper population was estimated by sampling 20 leaves per cultivar, every 7-10 days from May to September. Estimation of Empoasca vitis foliar symptoms At vintage time, the percentage of symptomatic leaf area (changes in colour and/or drying out) caused by E. vitis feeding activity was estimated. The five basal leaves of 40 shoots in the first year, and the 10 basal leaves of 10 shoots in the other two years were respectively sampled. The percentage of leaf surface with symptoms was visually estimated according to eight damage classes (0, 1, 5, 10, 25, 50, 75, 100) on the basis of reference patterns. Data analysis Cumulated densities of E. vitis nymphs and the percentages of symptomatic leaf area observed on each grape cultivar were normalized in relation to the more infested (susceptibility) or damaged cultivars. The sensibility of a cultivar was evaluated comparing the level of the two percentages; for example, a cultivar relatively sensible showed a percentage of damage higher than that of infestation. Results and discussion The red grape cultivar most susceptible to E. vitis was Carménère, while the less susceptible was Cabernet Sauvignon. Among white grape cultivars, Sauvignon Blanc and Tocai Friulano were the most susceptible in contrast with Chardonnay, Verduzzo Friulano and Pinot Gris. Carménère and Sauvignon Blanc showed to be very sensible, while Cabernet Sauvignon, Chardonnay and Pinot Gris showed a low incidence of symptoms despite leafhopper densities were moderate. Meteorological conditions (i.e. temperature and relative humidity) and plant stress factors (i.e. water stress) can modify the degree of sensibility of a cultivar to E. vitis. In particular, Sauvignon Blanc, Rhine Riesling and Verduzzo Friulano showed to be very sensible only during the second year, that was characterized by a hot and dry summer. The understanding of the susceptibility of different grape cultivars to E. vitis infestation is an important factor for the improvement of IPM strategies in viticulture. Based on the results here reported we would suggest that the most infested cultivars can be used as early indicators of E. vitis infestation. According to the degree of sensibility of red grape cultivars considered, the economic threshold should be lower for Carménère, higher for Cabernet Sauvignon and intermediate for Merlot and Refosco dal Peduncolo Rosso. In the white grape cultivars, the economic threshold has to be lower for Sauvignon Blanc and Tocai Friulano, higher for Chardonnay and Pinot Gris, intermediate for Verduzzo Friulano and Rhine Riesling. Acknowledgements This work has been carried out during the Doctoral project in viticulture, oenology and marketing supported by the Province of Treviso. We thank Dr. Michele Borgo (CRA -

289 269 Viticulture, Conegliano, Treviso, Italy) for giving us the possibility to carry out these investigations using the CRA grapevine collections. References Baggiolini, M., Canevascini, V., Tencalla, Y., Caccia, R., Sobrio, G., & Cavalli, S. 1968: La cicadelle verte, Empoasca flavescens F. (Homopt., Typhlocybidae), agent d altérations foliaires sur vigne. Rech. Agron. Suisse 7: Baggiolini, M., Canevascini, V., & Caccia, R. 1972: La cicadelle verte (Empoasca flavescens F.), cause d importants rougissements du feuillage de la vigne. OEPP/EPPO Bull. 3: Baillod, M., Jermini, M., & Schmid, A. 1990: Essais de nuisibilité de la cicadelle verte, Empoasca vitis Goethe sur le cépage Merlot au Tessin et le cépage Pinot noir en Valais. IOBC/WPRS Bull. 13: Candolfi, M. P., Jermini, M., Carrera, E., & Candolfi-Vasconcelos, M. C. 1993: Grapevine leaf gas exchange, plant growth, yield, fruit quality and carbohydrate reserves influenced by the grape leafhopper, Empoasca vitis. Entomol. Exp. Appl. 69: Carle, P., & Moutous, G. 1965: Observations sur le mode de nutrition sur vigne de quatre espèces de cicadelles. Ann. Epiphyties 16: Cerutti, F., Baumgärtner, J., & Delucchi, V. 1988: Ricerche sull ecosistema vigneto in Ticino: I. Campionamento delle popolazioni di Empoasca vitis Goethe (Hom., Cicadellidae, Typhlocybinae). Mitt. Schweiz. Entomol. Ges. 61: Cerutti, F., Delucchi, V., Baumgärtner, J., & Rubli, D. 1989: Ricerche sull ecosistema vigneto in Ticino: II. La colonizzazione dei vigneti da parte della cicalina Empoasca vitis Goethe (Hom., Cicadellidae, Typhlocybinae) e del suo parassitoide Anagrus atomus Haliday (Hym. Mymaridae), e importanza della flora circostante. Mitt. Schweiz. Entomol. Ges. 62: Dal Rì, M. 1992: Dynamique des populations et évaluation des dégâts provoqués par Empoasca vitis Goethe. IOBC/WPRS Bull. 15(2): 40. Moutous, G., & Fos, A. 1973: Influence des niveaux de populations de cicadelles de la vigne (Empoasca flavescens Fab.) sur le symptôme de la grillure des feuilles. Ann. Zool. Ecol. Anim. 5: Pavan, F., Picotti, P., & Girolami, V. 1992: Strategie per il controllo di Empoasca vitis Göthe su vite. Inf. tore agr. 48: Pavan, F., Stefanelli, G., Villani, A., Gasparinetti, P., Colussi, G., Mucignat, D., Del Cont Bernard, D., & Mutton, P. 2000: Danni da Empoasca vitis (Göthe) (Homoptera: Cicadellidae) in vigneti dell Italia nord-orientale e soglie d intervento. Frustula entomol. 21: Pavan, F., & Picotti, P. 2009: Influence of grapevine cultivars on the leafhopper Empoasca vitis and its egg parasitoids. BioControl 54(1): Schvester, D., Moutous, G., Bonfils, J., & Carle, P. 1962: Etude biologique des cicadelles de la vigne dans le sud-ouest de la France. Ann. Epiphyties 13(3): Tavella, L., & Arzone, A. 1992: Aspetti nutrizionali in Zyginidia pullula (Boheman), Empoasca vitis (Goethe) e Graphocephala fennahi Young (Homoptera: Auchenorrhyncha). Boll. Zoll. agr. Bachic. 24: Touzeau, J. 1968: La cicadelle Empoasca flavescens et le grillage de la vigne dans le Sud- Ouest de la France. Phytoma 200: Trentini, R. 1962: La cicalina della vite e dei fruttiferi. Inf. fitopatologico 12(10):

290 270 Vidano, C. 1958: Le cicaline italiane della vite. Hemiptera Typhlocybinae. Boll. Zool. agr. Bachic. 1: Vidano, C. 1963: Alterazioni provocate da insetti in Vitis osservate, sperimentate e comparate. Ann. Fac. Sci. agr. Univ. Torino 1: Vidano, C., Arnò, C., & Alma, A. 1988: On the Empoasca vitis intervention threshold on vine (Rynchota, Auchenorrhyncha). In: Vidano, C. e Arzone, A. (eds.): Proc. 6 th Auchenorrhyncha Meeting., Turin, Italy, 7-11 Sept. 1987:

291 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Can Harmonia axyridis affect the taste of European wines? C. Linder, F. Lorenzini, P. Kehrli Research Station Agroscope Changins-Wädenswil ACW, CH-1260 Nyon, Switzerland Abstract: The multicoloured Asian ladybeetle Harmonia axyridis Pallas is now established in Central Europe and is likely to spread soon over the whole continent. In the United States, H. axyridis moves to vineyard shortly before harvest. When disturbed or crushed during vinification, the ladybeetles release alkylmethoxypyrazines, which can strongly affect the quality of white and red wines. In Europe, wine growers are beginning to worry about the impact of H. axyridis on the quality of their processed wines. In this study we artificially contaminated harvested grapes of the variety 'Chasselas' and 'Pinot noir'', the two main varieties of Switzerland, with three different densities of H. axyridis and determined the impact of crushed ladybeetles on the taste of processed wines. The addition of ladybeetles did neither affect fermentation nor the basic chemical properties of processed wines. However, at the organoleptic level the contamination of grapes had a significant effect. Contaminated 'Chasselas' wines were of lower fruitiness, fineness, structure, acidity, equilibrium and exhibited a "rancid oil" odour. Moreover, the quality of contaminated 'Pinot noir' was rated low, in particular, its fruitiness, spiciness, fineness, structure, softness, robustness as well as intensity and quality of tannins was poor. Even though H. axyridis is rarely present in vineyards of Western Switzerland, our results emphasise that there is a risk of wine contamination. Thus, the evolution of the multicoloured Asian ladybeetle in European vineyards should be carefully monitored in order to anticipate future problems. Key words: Asian ladybeetle, Vitis vinifera, organoleptic tests Introduction The multicoloured Asian ladybeetle Harmonia axyridis Pallas (Coleoptera, Coccinellidae) is now established in Central Europe and is likely to spread soon over the whole continent (Brown et al., 2008). In the United States, H. axyridis moves to vineyards shortly before harvest, when natural prey gets scarce. There they aggregate in grape clusters and start to feed on previously injured berries (Galvan et al., 2006). When ladybeetles are disturbed or crushed during vinification, they release a yellow fluid that contains alkylmethoxypyrazines, which creates an unpleasant odour and tastes nasty (Pickering et al., 2005). In Europe, wine growers are beginning to worry about the impact of H. axyridis on the quality of their processed wines. In this study we artificially contaminated the vinification of 'Chasselas' and 'Pinot noir', the two main varieties of Switzerland, with three different densities of H. axyridis and determined the impact of crushed ladybeetles on the taste of processed wines. Material and methods In 2007, individuals of H. axyridis (HA) were collected in a sunflower field. Collected ladybeetles were kept in a climate chamber at 10 C until vinification. Harvested grapes were contaminated with three different densities of living adults of HA (0, 1 and 5 HA/kg grapes). After a standard microvinification wines were bottled in March At filling, their basic chemical properties were analysed and two months later, a panel of 12 trained tasters 271

292 272 estimated their organoleptic quality in a series of altogether 33 wines. Tasters rated various organolpetic criteria on a scale from 1 (bad/weak) to 7 (excellent/high). Data obtained were analysed using FIZZ software. Results and discussion The addition of HA did neither affect the start nor the duration of the fermentation. The basic chemical properties, such as ph, total acidity, tartaric acid, malic acid, volatile acidity and ethanol, did not differ between the three wine groups contaminated by different densities of ladybeetle. At the organoleptic level, contamination of grapes with HA had a significant effect on the rating of processed 'Chasselas' wines as well as 'Pinot noir' wines (Figure 1). A 'Chasselas' Structure Acidity Bitterness B 'Pinot noir' Intensity tannins Quality tannins Acidity Bitterness Colour intensity 2 Structure 2 Dryness Overall note 1 Colour intensity 1 Spiciness Quality / Fineness Quality / Fineness Equilibrium Softeness Fruitiness Fruitiness Shade Robusteness Overall note Control 1 HA/ kg 5 HA/ kg Control 1 HA/ kg 5 HA/ kg Figure 1. Organoleptic results of A) 'Chasselas' and B) 'Pinot noir' artificially contaminated with Harmonia axyridis. Notes from 1 (=weak, bad) to 7 (=high, excellent). In 'Chasselas', contaminated wines were of significant lower fruitiness, fineness, structure, acidity and equilibrium. The colour of the noncontaminated 'Chasselas' was more intense than in the two HA variants. Tasters did not appreciate the overall note of the contaminated 'Chasselas' and were astounded by their "rancid oil" odour. In 'Pinot noir', the noncontaminated wine was judged superior to the two ladybeetlecontaminated wines for most organoleptic criteria. However, statistically significant differences could mainly be established between the control and the heavily contaminated wine. The quality of the heavily contaminated 'Pinot noir' was rated very low by the panel, in particular, its fruitiness, spiciness, fineness, structure, intensity and quality of tannins, softness and robustness. Altogether, this led to a very poor appreciation of the overall note. However, colour intensity, shade, acidity, dryness and bitterness were not noticeably affected by the addition of ladybeetles.

293 273 Our results confirm the previous findings in the United States (Pickering et al., 2007; Galvan et al., 2007a). Taking into account the average weight of a grape cluster of 300g for 'Chasselas' and of 180g for 'Pinot noir', one can estimate that the "ladybug taint" can be noticed at densities as low as 0.3 HA and 0.2 HA per cluster, respectively. According to Galvan et al. (2007b), these values correspond to 18% and 12% of grape clusters colonized by at least one HA. However, in 2009, shortly before vintage, not a single HA was observed in or on grape clusters. This even though nearly 5'000 clusters in more than 10 Swiss vineyards were checked and the abundance of HA significantly increased over the last two years (Figure 2) weekly catches of H. axyridis may june july august sept. oct. Figure 2. Captures of Harmonia axyridis in a light trap over the summer 2008 and 2009 in Nyon, Switzerland. In conclusion, our results emphasise that there is a real risk of wine contamination. Thus, the evolution of the multicoloured Asian ladybeetle and its potential spread in European vineyards should be carefully monitored in order to anticipate future problems. Acknowledgements We thank J. Derron for initiating this study, P. Cuénat and E. Zufferey for the vinification of wines and S. Breitenmoser for the collection of processed ladybeetles.

294 274 References Brown, P. M. J., Adriaens, T., Bathon, H., Cuppen, J., Goldarazena, A., Hägg, T., Kenis, M., Klausnitzer, B. E. M., Kovář, I., Loomans, A. J. M., Majerus, M. E. N., Nedved, O., Pedersen, J., Rabitsch, W., Roy, H. E., Ternois, V., Zakharov, I. A., Roy, D. B. 2008: Harmonia axyridis in Europe: spread and distribution of a non-native coccinellid. Biocontrol 53: Galvan, T. L., Burkness, E. C. & Hutchinson, W. D. 2006: Wine grapes in the Midwest: reducing the risk of the multicolored Asian lady beetle. Public Univ. of Minnesota Extension Service, St. Paul, MN. Galvan, T. L., Burkness, E. C., Vickers, Z., Stenberg, P., Mansfield, A. K. & Hutchinson, W. D. 2007a: Sensory-based action threshold for multicolored Asian lady beetle-related taint in winegrapes. Am. J. Enol. Vitic. 58(4): Galvan, T. L., Burkness, E. C. & Hutchinson, W. D. 2007b: Enumerative and binomial sequential sampling plans for the mulitcolored Asian lady beetle (Coleoptera: Coccinellidae) in wine grapes. J. Econ. Entomol. 100 (3): Pickering, G. J., Lin, Y., Reynolds, A., Soleas, G., Riesen, R. & Brindle, I. 2005: The influence of Harmonia axyridis on wine composition and aging. J. Food Sci. 70(2): Pickering, G. J., Ker, K. & Soleas, G. J. 2007: Determination of the critical stages of processing and tolerance limits for Harmonia axyridis for "ladybug taint" in wine. Vitis 46(2):

295 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Feeding behaviour of Lobesia botrana on leaves and shoots of grapevine A. Lucchi 1, E. Pozzolini 1, G. Anfora 2, V. Mazzoni 2, M. Tasin 2, E. Leonardelli 2, C. Ioriatti 3 1 Dept C.D.S.L., Sect. Entomologia agraria, University of Pisa, Pisa, Italy; 2 IASMA Research and Innovation Center, FEM, S. Michele a/a, Italy; 3 IASMA Technology Transfer Center, FEM, S. Michele a/a, Italy. alucchi@agr.unipi.it Abstract: The larval feeding behaviour of the grapevine moth Lobesia botrana (GM) on grapevine leaf and shoot tissues in the absence of grapes is described. GM larvae were obtained from eggs laid on the grapevine leaves in a field cage covering two rows of a Chardonnay vineyard in which all the bunches were previously removed. Larvae developed till pupation feeding only on leaves and shoots and showed the characteristic leaf rolling habit of the Tortricidae. All of them were able to pupate and to give adults that mated in laboratory conditions. Females laid fertile eggs and gave rise to a new generation of larvae. Pupal weight, number of eggs laid, and longevity of leaf/shoot fed insects were compared with those of the insects reared on berries and artificial diet. Key words: European Grapevine Moth, Tortricidae, larval feeding, pupal weight, fecundity Introduction Females of the grapevine moth (GM), Lobesia botrana (Den. & Schiff.) (Lepidoptera: Tortricidae) (Figure 1b), are attracted to and lay their eggs on Vitis vinifera L. and on a number of other plants (Marchal, 1912; Picard, 1920; Thiéry, 2005). On grapevine, GM larvae feed primarily on the flowers as well as on green and ripe berries. In the wine-grape the damage results from the increased susceptibility of damaged bunches to rot rather than from the direct loss of product from insect feeding (Ioriatti et al., 2008). As known, the quality of the larval diet strongly influences Lepidoptera fitness in terms of pupal mass, mating success, reproductive output, fecundity, egg size and egg hatchability (Moreau et al., 2006 and references therein). With regards to L. botrana, different grape cultivars can have different nutritional quality for larvae and alternative plants could be more suitable than grapevine and other Vitaceae for the reproductive performance of the species (Moreau et al., 2007). At the same time GM larvae fed on diets containing different parts of berries performed differently in terms of developmental time and mortality (Moreau et al., 2006). The feeding of GM larvae on leaves and shoots has been reported by Silvestri (1912) about one Century ago. He stated that if grapes are not available, larvae of L. botrana can feed on tender shoots and leaves of grapevines. Two years later Luestern (1914) included grapevine leaves between the vegetables used to feed GM larvae, suggesting the capability of a leaf-based diet to fully allowing the species development in laboratory conditions. More recently Thiéry (2005) recalled the habit of Roerich to feed and grow L. botrana larvae in the laboratory with non host plants as lettuce, bay and apple fruits. 275

296 276 In July 2008 a remarkable GM oviposition was recorded on leaves of grapevine deprived of bunches in semi-field experiments (Anfora et al., 2009). Moreover, GM eggs on leaves and secondary shoot grapes have been at times observed in several Italian vineyards, mainly at the end of the season. Damage on leaves and shoots is not usually considered economically and ecologically relevant, but the knowledge of the possible performance of GM larvae feeding on leaves in the field would be of great interest in order to understand the potential survival rate of the species and might open new questions for the scouting activity of GM on grapevine. Our experiments were designed to provide an accurate description of L. botrana larval feeding behaviour on leaf and shoot tissues. To assess the suitability of those organs in allowing the larval development and the successful fulfilment of GM life cycle, leaf/shoot fed specimens were compared for some of the fitness traits with those fed with either berries or artificial diet. Material and methods Moths were selected from a permanent colony of L. botrana reared on an artificial diet at the IASMA-FEM Research Centre (San Michele a/a, Trento). The colony was maintained on a artificial diet, with the following composition: 750ml water, 90g wheat germ, 40g casein, 25g freeze-dried alfa alfa, 30g sugar, 18g yeast, 15g agar, 12,5g Wessen salt, 10g ascorbic acid, 7.5g vitamins (Vanderzahnt), 2.5ml maize oil, 2.5ml propionic acid, 2g sorbic acid, 1.25g cholesterol, 1.25g tetracycline, 1ml linoleic acid. The semi-field experiment was conducted at San Michele a/a in a thin meshed net tunnel cage (length 15m, width 4m, median height 2.5m) that covered two grapevine rows of a Chardonnay vineyard (Figure 1a). All the grape clusters within the cage were previously removed. Mated females (n=100) were launched in a single release during the second week of July Samples of mature larvae and pupae developed on leaves and shoots were collected and individually inserted in cilindrical plastic containers (H 10cm; Diam. 5cm ). Each pupa was weighed with an ultra-sensitive electrobalance (Kern-Balinger, Germany). Once adults emerged, couples were formed and all the eggs laid were counted. During this period the moths were fed with sugar solution and maintained at 25 ± 1 C, 60 ± 10% RH with a photoperiod of light/dark (LD) 15: 8 h + 1 of dusk. Pupal weight was compared with that of other individuals of the same colony reared on the artificial diet and on green berries of the cultivar Chardonnay collected in the field at the phenological phase beginning of berry touch (stage 35, Eichhorn and Lorenz 1977). Moreover pupal weights were also compared with those of a wild strain collected in the field from grapes of the cultivar Chardonnay (phenological phase beginning of berry touch ). Female fertility, number of eggs laid, and longevity (from the pair formation) of leaf/shoot fed insects were compared with those of the insects reared on artificial diet. Nonparametric oneway ANOVA (Kruskal-Wallis test) followed by the Bonferroni-Dunn post-hoc multiple comparison test was used to assess differences in the average pupal weight. Fisher exact test was used to evaluate the differences in fertility and the Wilcoxon-Mann-Whitney test for unpaired samples was used to determine significant differences in the numbers of eggs laid and the longevity of the pairs (Siegel and Castellan, 1988). The software used for the statistical analyses was KyPlot 2.0 beta (Koichi Yoshioka).

297 277 Results and discussion GM larvae emerged from eggs laid on the lower leaf surface and developed till pupation by feeding only on tender leaves (Figure 1c) and shoots (Figure 1d). When feeding on the leaves the larvae showed the characteristic leaf rolling habit of the Tortricidae: leaves were pierced (Figure 1c) and progressively tied together with silken threads to build a shelter (Figure 1e-h); tender green shoots were tunnelled after penetration near the buds (Figure 1d). If feeding on fully developed leaves, the larvae chewed on the leaf blade causing numerous sub-circular holes on it (Figure 1g). All GM larvae feeding only on leaf and shoot tissues were able to pupate and to give adults. The pupal average weight of laboratory strain insects fed on leaf/shoot tissues and green grape berries was significantly lower than that of the laboratory strain insects fed on the artificial diet, but not different from wild-strain insect pupae fed on green grape berries (Table 1). Table 1. Average weight of two strains of Lobesia botrana pupae reared under different diets. N = number of tested pupae. Values followed by different letters are significantly different (P<0.001). Kruskal-Wallis test followed by the Bonferroni-Dunn post-hoc multiple comparison test. Laboratory strain Wild strain leaf berry artificial diet Berry N Average weight (mg) 6.22 b 5.64 b a 6.32 b The percentage of fertile pairs did not significantly differ between the two groups of insects (P=0.22), while the female longevity tended to be longer for those fed with artificial diet (P=0.07). As expected, since pupae weighed less, the leaf/shoot fed insects laid significantly less eggs than those fed with artificial diet (P<0.05) (Table 2). Insects fed on artificial diet showed the best performance for all the considered traits (Tables 1 and 2) Table 2. Effect of artificial and leaf/shoot diets on fertility, number of eggs laid and longevity (from the pair formation) of the Lobesia botrana laboratory strain (F = Fischer exact test; WMW = Wilcoxon-Mann-Whitney test). Diet Statistics artificial leaf Test p % fertile pairs F = 0.22 average number of egg/pair WMW < 0.05 average female longevity (day) WMW = 0.07 number of egg/day per female WMW < 0.01

298 278 Figure 1. An overview on the development of L. botrana on leaves and shoots of grapevine

299 279 Our results confirm the capability of GM to complete the life cycle and to give rise to a fertile offspring by feeding only on grapevine leaves and shoots, as reported by Silvestri in Therefore, L. botrana leaves and shoots seem able to supply a nutritive value comparable with that of the usual food of larvae in the field (flowers and berries). On the contrary, it was showed that larvae of the Codling moth Cydia pomonella (L.) are not able to pupate when feeding exclusively on a leaf-based diet (Pszczolkowski et al., 2002). The difference observed between this two related species could probably depend on the different threshold of body weight that the larva has to reach at the end of its development. In Lepidoptera, indeed, the body weight of the mature larva must exceed a species-specific threshold to allow pupation (Nijhout 1981). This threshold usually varies between 25% and 50% of the mean maximum body weight (Slansky and Scriber, 1985). Hence, the preference for green berries versus leaf/shoot tissues in natural conditions should rely on other unknown factors. On the other hand, our data suggest that leaf/shoot tissues could represent for GM a useful pabulum to complete the larval development and then pupate, when climate conditions are still favourable but grapes were already harvested. Acknowledgements We are grateful to Luisa Mattedi and Mauro Varner for kindly sharing information about their Sicilian field observations. References Anfora, G., Tasin, M., De Cristofaro, A., Ioriatti, C., Lucchi, A. 2009: Synthetic grape volatiles attract mated Lobesia botrana females in laboratory and field bioassays. J. Chem. Ecol. 35: Eichhorn, K. W. & Lorenz, D. H. 1977: Phänologishche Enwicklungsstadien der Rebe. Nachrichtenbl. Dtsch. Pflanzenschutzdienstes (Braunschweig) 29: Ioriatti, C., Lucchi, A., Bagnoli, B. 2008: Grape areawide pest management in Italy. In: Koul, O., Cuperus, G. W., Elliott, N. (eds.): Areawide pest management: theory and implementation. CABI: Luestner, G. 1914: Das Verhalten der Raupen des einbindigen und bekreuzten Traubenwicklers (Cochylis ambiguella Hb. und Polychrosis botrana Schiff.) zu den Weinbergunkräutern und anderen Pflanzen. Z. f. Weinbau u. Weinbehandl. 1: Marchal, P. 1912: Rapport sur les Travaux Accomplis par la Mission de l'etude de la Cochylis et de l'eudémis pendant l'année Beranger Ed., 326 pp. Moreau, J., Benrey, B., Thiéry, D. 2006: Assessing larval food quality for phytophagous insects: are the facts as simple as they appear? Funct. Ecol. 20: Moreau, J., Benrey, B., Thiéry, D. 2007: Grape variety affects female but also male reproductive success in wild European grapevine moths. Ecol. Entomol. 32: 1-7. Moreau, J., Arruego, X., Benrey, B., and Thiéry, D. 2006: Differences in nutritional quality of parts of Vitis vinifera berries affect fitness of the European grapevine moth. Entomol. Exp. Appl. 119: Nijhout, H. F. 1981: Physiological control of moulting in insects. Am. Zool. 21: Picard, F. 1920: La lutte contre la cochylis par la choix des cépages et par la culture de plantes attractives. Progr. agric. et vitic. 41(28):

300 280 Pszczolkowski, M. A., Matos, L. F., Brown, R., Brown, J. J. 2002: Feeding and development of Cydia pomonella (Lepidoptera: Tortricidae) larvae on apple leaves. Ann. Entomol. Soc. Am. 95: Siegel, S., Castellan, N. J. 1988: Non parametric statistics for the behavioural sciences, 2nd ed. Singapore Mc Graw-Hill. Silvestri, F. 1912: Contribuzioni alla conoscenza degli insetti dannosi e dei loro simbionti. III. La Tignoletta dell uva (Polychrosis botrana Schiff.) con un cenno sulla Tignola dell uva (Conchylis ambiguella Hb.). Boll. Lab. Zool. Gen. e Agr., Portici 6: Slansky, F., Scriber, J. M. 1985: Food consumption and utilization. In: G. A. Kerkut and L. I. Gilbert (eds.): Comprehensive insect physiology, biochemistry and pharmacology, vol. 4, Pergamon Press, Oxford, UK: Thiéry, D. 2005: Vers de la grappe: les connaître pour s en protéger. Vigne & Vin, Bordeaux, 60 pp.

301 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Attractiveness of different colours to Scaphoideus titanus Ball (Hemiptera: Cicadellidae) adults V. Mazzoni 1, F. Trona 1, C. Ioriatti 2, A. Lucchi 3, A. Eriksson 3, G. Anfora 1 1 IASMA Research and Innovation Centre, Fondazione Mach, San Michele all Adige (TN), Italy; 2 IASMA Centre for Technology Transfer, Fondazione Mach, San Michele all Adige (TN), Italy; 3 University of Pisa, Dept. CDSL, Sect. Entomologia agraria, Italy Abstract: Laboratory and field tests were performed to ascertain a possible role of visual cues for Scaphoideus titanus during the location of a host plant. In laboratory, two choice colour tests were made in a Y-tube to compare the attractiveness of blue (peak wavelength: 475nm), green (521nm), yellow (573nm) and red (650nm). The observed preference order was yellow > red > green > blue for males and red > yellow > green > blue for females. In field, nine groups of three coloured (yellow, red, blue) sticky traps were positioned in an organic vineyard and replaced once a week, for two summers (2008 and 2009). In both seasons the yellow traps captured significantly more individuals (either males or females) and had the highest sex rate (percentage of males on the total captures) followed by red and blue. Our investigation suggests that, despite a partial different response between laboratory and field tests, S. titanus adults are influenced by visual cues present in their habitat. The possible reasons of the different responses observed in lab and field are discussed. Key words: hemiptera, leafhopper, choice tests, y-tube, sticky traps Introduction The leafhopper Scaphoideus titanus Ball (Hemiptera: Cicadellidae) is the insect vector of the Flavescence dorée, a phytoplasma grapevine disease quarantined in Europe. When mating, S. titanus adults communicate by means of substrate-borne vibrational signals (Mazzoni et al., 2009a), whereas the host plant recognition and identification by nymphs are partially mediated by volatile cues (Mazzoni et al., 2009b). The role of vision in environmental orientation has been poorly investigated. Studies on Auchenorrhyncha showed that several foraging leafhoppers and planthoppers discriminate using primarily visual stimuli, even combined with olfaction (Bullas-Appleton et al., 2004; Patt & Sétamou, 2007; Todd et al., 1990a, b). In the case of S. titanus, Lessio & Alma (2004) employed sticky traps of 4 colours (yellow, red, blue and white). In their specific experimental conditions, they found that all were extremely more effective in capturing males than females and, in particular, red traps captured significantly more males than other colours while no differences were detected for females. The aim of the present study was to investigate the role of vision in orientating S. titanus adults. Laboratory tests were made to isolate the colour factor from other possibly interactive factors (e.g. odours and background colours). In addition, field studies were made to verify the laboratory results, in which the insects had to choose among the colours with completely different environmental conditions. 281

302 282 Material and methods Insects for laboratory tests In the winter 2008, two year-old grapevine canes were collected from organic farms at Villazzano (Trento, Italy) and stored in a cold room at 4 C. Eggs were hatched in climatic chamber (25±1 C, L16:D8, RH:75±5%). Nymphs were kept in plastic cylinders (10x5cm) at 25±1 C, 65±5% relative humidity, and 16:8 (L:D) photoperiod. Adult males and females were removed from the nymphal culture on the day of emergence and kept separated by gender. Grapevine leaves were provided as food source: a grapevine leaf disk was laid over a 1-cm layer of technical agar solution [0.8% (wt vol)] at the bottom of a plastic cylinder and replaced twice a week. All tests were done with virgin males and females. Colours The chosen colours were blue, green, red and yellow for laboratory tests and blue, red and yellow for field tests. The spectral reflectance (range: nm) of each employed colour was measured with a EPP-2000 spectroradiometer (Stellarnet). Peaks of reflectance were: blue = 475nm (46%), green = 521nm (54%), yellow = 573nm (56%) and red = 650nm (56%). Laboratory tests To observe the behaviour of S. titanus when exposed to colour stimuli, two-choice tests were performed in a Y-tube (stem and arms, 40cm; diameter, 7cm; side arms at 60 ). The tube was kept vertical since in preliminary tests the insect showed a noteworthy negative geotropism. Coloured disks (6.9cm i.d.) were inserted in each arm 10cm before the fork and kept still with an iron wire fixed at the upper mouth of the same arm. Groups of 2-3 insects of the same gender were inserted from the base of the stem. Tests were conducted in a dark room (25±1 C, 60±5%) and each coloured disk was enlightened with a fluorescent light (~300 lux). Two-choice tests were made by crossing any colour with each others. In tests involving blue, 20 males and 10 females were released; in all other tests 20 males and 20 females were released. Results were analyzed by log-likelihood ratio (G-test), after Williams correction. The statistics included also individuals that did not make a choice. Field tests Nine groups of three differently coloured (blue, red and yellow) sticky traps (Rebell, Andermatt, Switzerland) of 15x8cm were homogeneously positioned in a Chardonnay vineyard trained on Trentino Pergola system at Villazzano (Trento, Italy), from 10 th of July to 30 th of September 2008 and Traps were hung at the grapevine supporting iron wire (height 120 cm) and weekly replaced. Within each group the traps were separated by a distance of 50cm. Each group was distant from the closest one of a range of 50-70m. A Friedman test (non-parametric two-way Anova) followed by Bonferroni-Dunn multiple comparison tests was used to discriminate among colours and genders in terms of number of captured individuals (Siegel & Castellan, 1988). The sex rate (percentage of captured males) was measured along each season, by calculating the average of each month (four samplings) for each colour.

303 283 Results and discussion Laboratory tests The two-choice tests in Y-tube showed that S. titanus responds differently to red, yellow, green and blue colours (Tab. 1). Blue disks were never significantly preferred by the tested group of specimens, while green prevailed only over blue. The yellow disks were significantly more attractive to males whereas females were more attracted by the red colour. Table 1. Results from the two-choice tests conducted on males (grey cells) and females (white cells) of S. titanus. Inside each cell are indicated the first letter of the preferred colour and the statistical level of significance after G-test (* < 0.05; ** < 0.01; *** < 0.001; ns = not significant). For each cell n = 20 except those cells involving blue where n = 10. Blue Green Red Yellow Blue G ** ns Y ** Green G * R *** ns Red R *** R * R *** Yellow Y *** Y * Y * Field tests The numbers of captured individuals and seasonal trends were similar between the two years of study. Yellow sticky traps collected always the highest number of both males and females, followed by red and blue traps (Tab. 2). Male and female captures were in similar quantity in 2008, while in 2009 the females slightly prevailed. During each season the sex rate (percentage of males on the total of captures) greatly varied: the highest value was found at the first seasonal sampling, then it constantly decreased until the end of the surveying period (values around 50% were recorded in August), exactly mirroring the insect life biological cycle. In general, yellow traps always kept the highest and blue traps the lowest values of sex rate. Table 2. The rank mean (after Friedman s Test) of S. titanus males and females captured by red, yellow and blue sticky traps in 2008 and 2009 and the mean value of sex rate (Sr) of individuals captured for each month of each year. Different letters indicate significant difference among the rank means of the same year (Bonferroni-Dunn multiple comparison test). Year Gender Blue Red Yellow Males 21.3 a 28.5 bc 36.7 d Females 23.0 a 26.3 b 29.1 c 2008 Sr July Sr August Sr September Males 23.5 a 27.2 b 36.6 d Females 28.1 bc 30.2 c 37.4 d 2009 Sr July Sr August Sr September

304 284 The results are partially different between field and laboratory, in particular for the female preference, to red in Y-tube tests and to yellow in field tests. This difference could be explained by the following three factors: 1) the different physiological state of the females in the field (mated and virgin) and in the lab (virgin), 2) the presence of a background of colours in the field, 3) the variable light conditions during the day in the field. Moreover, the peculiarity of experimental conditions, such as the grapevine training system, the pedoclimate, and the monitoring method, may strongly affect the field results and determine remarkable discrepancies also between field tests if conducted in different localities. Further research will focus on the above mentioned factors. Nonetheless, these results clearly suggest that S. titanus adults are influenced and oriented by the visual cues present in their habitat. An improved understanding of colour attraction is an important knowledge to take into account when creating a monitoring system of S. titanus. Acknowledgements This study was funded by Interneuron Project (Fondazione E. Mach), Host Project (Autonomous Province of Trento) and Fondi Ateneo of Pisa University. References Bullas-Appleton, E. S., Otis, G., Gillard, C. & Schaafsma, A. W. 2004: Potato leafhopper (Homoptera: Cicadellidae) varietal preferences in edible beans in relation to visual and olfactory cues. Environ. Entomol. 33: Lessio, F. & Alma, A. 2004: Dispersal patterns and chromatic response of Scaphoideus titanus Ball (Homoptera Cicadellidae), vector of the phytoplasma agent of grapevine flavescence dorée. Agric. Forest Entomol. 6: Mazzoni, V., Presĕrn, J., Lucchi, A., & Virant-Doberlet, M. 2009a: Reproductive strategy of the Nearctic leafhopper Scaphoideus titanus Ball (Hemiptera: Cicadellidae). B. Entomol. Res. 99: Mazzoni, V., Anfora, G., Trona, F., Lucchi, A., De Cristofaro, A. & Ioriatti, C. 2009b: Study on the role of olfaction in host plant detection of Scaphoideus titanus (Hemiptera: Cicadellidae) nymphs. J. Econ. Entomol. 102: Patt, J. M. & Sétamou, M. 2007: Olfactory and visual stimuli affecting host plant detection in Homalodisca coagulata (Hemiptera: Cicadellidae). Environ. Entomol. 36: Siegel, S. & Castellan, N. J. 1988: Nonparametric statistics for the behavioral sciences, 2 nd ed., McGraw-Hill, Singapore: 399. Todd, J. L., Harris, M. O. & Nault, L. R. 1990a: Importance of color stimuli in host-finding by Dalbulus leafhoppers. Entomol. Exp. Appl. 54: Todd, J. L., Phelan, P. L. & Nault, L. R. 1990b: Interaction between visual and olfactory stimuli during host-finding by leafhopper, Dalbulus maidis (Homoptera: Cicadellidae). J. Chem. Ecol. 16(7):

305 Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp The impact of a number of insecticides on Empoasca vitis populations in north-eastern Italy A. Pozzebon, M. Pederiva, R. Moret, C. Duso Department of Environmental Agronomy and Crop Science, University of Padua, Agripolis, Legnaro (Padova), Italy. Abstract: The green leafhopper Empoasca vitis Goethe (Homoptera: Cicadellidae) is an important pest of grapes in Europe. Its feeding activity may cause a reduction in photosynthesis, mesophyll conductance and transpiration rate with implication for yields. For these reasons insecticides are widely used to control this pest with potential effects on beneficial arthropods. Field experiments were conducted to evaluate the impact of a number of insecticides (chlorpyriphos, flufenoxuron, thiamethoxam, and indoxacarb) on E. vitis populations. Pesticide side-effects on the predatory mite Typhlodromus pyri were assessed. Thiamethoxam was the most effective towards E. vitis. Pest population levels were slightly reduced by chlorpyriphos, suggesting that this strain was resistant to organophosphates. Indoxacarb and flufenoxuron were associated to intermediate effects. The predatory mite populations were affected especially by thiamethoxam. Key words: Empoasca vitis, grape, chemical control, Typhlodromus pyri, pesticide side-effects Introduction The green leafhopper Empoasca vitis Goethe (Homoptera Cicadellidae) is an important pest of grapevine in Europe (Vidano, 1958; Baggiolini et al., 1968; Moutous & Fos 1973; Cerutti et al., 1988; Pavan et al., 1988; Baillod et al., 1990). Empoasca vitis feeding activity may cause a reduction in photosynthesis, mesophyll conductance and transpiration rates (Vidano, 1963; Tavella & Arzone, 1992; Candolfi et al., 1993) with potential implications for yields (Pavan et al., 1998). For these reasons, insecticides are widely used to control this pest. However, the use of insecticides non selective towards beneficial mites can induce spider mite outbreaks (McMurtry et al., 1970; Girolami, 1981). The efficacy on the target pests as well as the impact on beneficials are important features to be known for insecticide use in IPM strategies (Dent, 1991). The results of field trials with insecticides frequently used to control E. vitis are here presented. In these experiments, we considered the efficacy of pesticides towards green leafhoppers and their side-effects on Typhlodromus pyri Scheuten, a key predator of phytophagous mites in European vineyards (Maixner, 1990). Material and methods Experiments The effect of insecticides on E. vitis and T. pyri were investigated in three experiments carried out in two vineyards located in the Veneto region, Italy, during Vineyards comprised Prosecco variety grapevines. In Experiment 1 the effects of two insecticides were compared: chlorpyriphos 75% (70g/hl) and flufenoxuron 4.7% (100ml/hl). In Experiment 2 285

306 286 the effects of three insecticides were compared: chlorpyriphos 75% (70g/hl), thiamethoxam 25% (20g/hl), and indoxacarb 30% (15g/hl). In Experiment 3 the effects of three insecticides were compared: chlorpyriphos 75% (70g/hl), thiamethoxam 25% (20g/hl), and flufenoxuron 4.7% (100ml/hl). In each experiment, an untreated control was included. Experimental design was a randomized block with four replications per treatment. Timing of application was determined according to E. vitis population dynamics observed in each experimental site. A total of 120 leaves per treatment were observed in the field to evaluate E. vitis abundance variations. Predatory mite abundance was evaluated in the laboratory by analyzing 100 leaves per treatment under a dissecting microscope. The observations on the effects of insecticides started the day of application and continued for about one month after pesticide applications. Data analysis The effects of insecticides on E. vitis and T. pyri were evaluated with a Restricted Maximum Likelihood Repeated Measures model (α = 0.05). Data were checked for normality assumption and the number of E. vitis motile forms per leaf as well as the number of phytoseiids per leaf were log (x+1) transformed. The effect (E) of insecticides was estimated as the change in population levels in treated plots related to the change in the control plots during the observation period, using Henderson and Tilton s formula (1955). Results and discussion Experiment 1 Insecticides were applied on June 9 th. In pre-treatment observations, E. vitis densities attained about one motile form per leaf in all treatments. Leafhopper abundance was significantly affected by insecticide applications (P < 0.01). The use of flufenoxuron resulted in a remarkable reduction of E. vitis population densities (E = 70.95), while low effects were associated to the chlorpyriphos application (E = 9.30). A significant effect of both insecticides (P < 0.05) was observed on predatory mite abundance. Typhlodromus pyri population densities were reduced by flufenoxuron (E = 47.37) and chlorpyriphos (E = 37.38) applications. Experiment 2 Insecticides were applied on July 7 th. Prior of insecticide applications, E. vitis densities reached about two motile forms per leaf in all treatments. After insecticide applications, leafhopper densities were different among treatments (P < 0.01). A significant reduction in E. vitis densities was associated to thiamethoxam (E = 89.36) and indoxacarb (E = 40.65), while chlorpyriphos effects were very low (E = 6.35). A significant effect of insecticide use was observed on T. pyri populations densities (P < 0.05). Chlorpyriphos and thiamethoxam applications reduced beneficial mite populations (E = and 44.20, respectively) while indoxacarb was associated to negligible effects (E = 1.24). Experiment 3 Flufenoxuron was applied on July 2 nd while thiamethoxam and chlorpyriphos were applied on July 8 th. In pre-treatment observations, E. vitis densities attained about one motile form per leaf in all treatments. Insecticide applications significantly affected leafhopper abundance (P < 0.01). Empoasca vitis population densities were reduced by thiamethoxam (E = 94.40), flufenoxuron (E = 90.03), and to a lesser extent chlorpyriphos (E = 30.45). A significant effect of insecticides (P < 0.05) was observed on predatory mites' abundance. Typhlodromus

307 287 pyri densities were reduced by thiamethoxam (E = 50.87), flufenoxuron (E = 24.61), and chlorpyriphos (E = 30.82) applications. Among tested insecticides, thiamethoxam and flufenoxuron were more effective than indoxacarb in controlling E. vitis. Pest population levels were slightly reduced by chlorpyriphos applications. Failure of chlorpyriphos in controlling E. vitis, as observed in the present study, strongly suggests that E. vitis strains resistant to organophosphates (OPs) are spreading in north-eastern Italy (Mori et al., 1999; Girolami et al., 2001). OPs have been considered as principal compounds in E. vitis management but the occurrence of resistant strains suggests that alternatives to these insecticides should be considered. In terms of E. vitis control, our results confirm previous data on the high efficacy of thiamethoxam and flufenoxuron as well as the intermediate effects of indoxacarb (Mori et al., 1999; Delaiti et al., 2005; Lavezzaro et al., 2006; Posenato et al., 2006). On the other hand, indoxacarb showed negligible effects on T. pyri, while a population reduction was observed when the other insecticides were applied. According to the IOBC Working Document on Selectivity of Pesticides (Boller et al., 2005), thiamethoxam resulted moderately harmful (Reduction 51-75%) while flufenoxuron, chlorpyriphos and especially indoxacarb can be considered as slightly harmful (reduction 0-50%) to predatory mites. None of these insecticides compromised the persistence of T. pyri populations in vineyards confirming previous results on indoxacarb, flufenoxuron and thiamethoxam (Delaiti et al., 2005). OPs applications were often associated to detrimental effects on T. pyri (Zacharda & Hluchy, 1991; Duso, 1994), however resistance to OPs in T. pyri has been widely reported (e.g., Maixner, 1990; Bonafos et al., 2008). In the present investigation, T. pyri densities resulted reduced by chlorpyriphos applications even if no extinction of predators was observed suggesting that this strain may show some resistance to OPs. References Baggiolini, M., Canevascini, V., Tencalla, Y., Caccia, R., Sobrio, G. & Cavalli, S. 1968: La cicadelle verte, Empoasca flavescens F. (Homopt., Typhlocybidae), agent d altérations foliaires sur vigne. Rech. Agron. Suisse 7: Baillod, M., Jermini, M. & Schmid, A. 1990: Essais de nuisibilité de la cicadelle verte, Empoasca vitis Goethe sur le cépage Merlot au Tessin et le cépage Pinot noir en Valais. IOBC/WPRS Bull. 13: Boller, E. F., Vogt, H., Ternes, P. & Malavolta, C. 2005: Working Document on Selectivity of Pesticides. Bonafos, R., Serrano, E., Auger, P. & Kreiter, S. 2007: Resistance to deltamethrin, lambdacyhalothrin and chlorpyriphos-ethyl in some populations of Typhlodromus pyri Scheuten and Amblyseius andersoni (Chant) (Acari: Phytoseiidae) from vineyards in the southwest of France. Crop Protection 26: Candolfi, M. P., Jermini, M., Carrera, E. & Candolfi-Vasconcelos, M. C. 1993: Grapevine leaf gas exchange, plant growth, yield, fruit quality and carbohydrate reserves influenced by the grape leafhopper, Empoasca vitis. Entomol. Exp. Appl. 69: Cerutti, F., Baumgärtner, J. & Delucchi, V., 1988: Ricerche sull ecosistema vigneto in Ticino: I. Campionamento delle popolazioni di Empoasca vitis Goethe (Hom., Cicadellidae, Typhlocybinae). Mitt. Schweiz. Entomol. Ges. 61: Delaiti, M., Angeli, G., Sandri, O., Tomasi, C. & Ioriatti, C. 2005: Nuovi insetticidi per il contenimento della cicalina verde della vite. L Informatore Agrario 25:

308 288 Dent, D. 1991: Insect Pest Management. CAB International, Wallingford, UK, 583 pp. Duso, C. 1994: Comparison between field and laboratory testing methods to evaluate the pesticide side-effects on the predatory mites Amblyseius andersoni and Typhlodromus pyri. OILB/SROP Bull. 17(10): Girolami, V. 1981: Danni, soglie di intervento, controllo degli acari della vite. Atti "III incontro su la difesa integrata della vite" Latina, Italy, 3-4 dicembre 1981: Girolami, V., Mori, N., Marchesini, E. & Duso, C. 2001: Organophosphate resistance in grape leafhoppers and IPM strategies. Redia 84: A1-A17. Henderson, C. F. & Tilton, E. W. 1955: Test with acaricides against brown wheat mites. J. Econ. Entomol. 48: Lavezzaro, S., Morando, A. & Gallesio, G. 2006: Un quadriennio di prove di lotta contro la cicalina verde della vite in Piemonte. Atti delle Giornate Fitopatologiche 2006, Vol. I: Maixner, M. 1990: Investigations on insecticide resistance in the predatory mite Typhlodromus pyri Scheuten (Acari: Phytoseiidae) on grapevines in the wine region Mosel-Saar-Ruwer. Mitteilungen aus der Biologischen Bundesanstalt für Land- und Forstwirtschaft Berlin-Dahlem 257: 118 pp. McMurtry, J. A., Huffaker, C. B. & Van de Vrie, M. 1970: Ecology of tetranychid mites and their natural enemies: A review. 1. Tetranychid enemies, their biological characteristics and the impact of spray practices. Hilgardia 40: Mori, N., Posenato, G., Sancassani, G., Tosi, L., & Girolami, V. 1999: Insecticides for the control of cicadellids in vineyards. L Informatore Agrario 55(15): Moutous, G. & Fos, A. 1973: Influence des niveaux de populations de cicadelles de la vigne (Empoasca flavescens Fab.) sur le symptôme de la grillure des feuilles. Ann. Zool. Ecol. Anim. 5: Pavan, F., Pavanetto, E., Duso, C. & Girolami, V. 1988: Population dynamics of Empoasca vitis (Göethe) and Zygina rhamni (Ferr.) on vines in northern Italy. In: C. Vidano and A. Arzone (eds.): Proc. 6th Auchenorrhyncha Meeting, Turin, Italy, 7-11 September 1987: Pavan, F., Stefanelli, G., Villani, A., Gasparinetti, P., Colussi, G., Mucignat, D., Del Cont Bernard, D. & Mutton, P. 1998: Danni da Empoasca vitis (Goethe) (Homoptera: Cicadellidae) in vigneti dell Italia nord-orientale e soglie di intervento. Frustula Entomologica 21: Posenato, G., Marchesini, E. & Mori, N. 2006: Efficacia di thiamethoxam su Empoasca vitis a confronto con lambda-cyhalothrin, abamectina, indoxacarb e chlorpyrifos. Atti delle Giornate Fitopatologiche 2006, Vol. I: Tavella, L. & Arzone, A. 1992: Aspetti nutrizionali in Zyginidia pullula (Boheman), Empoasca vitis (Goethe) e Graphocephala fennahi Young (Homoptera Auchenorrhyncha). Boll. Zoll. agr. Bachic. 24: Vidano, C. 1958: Le cicaline italiane della vite. Hemiptera Typhlocibinae. Boll. Zool. agr. Bachic. 1: Vidano, C. 1963: Alterazioni provocate da insetti in Vitis osservate, sperimentate e comparate. Ann. Fac. Sci. agr. Univ. Torino 1: Zacharda, M. & Hluchy, M. 1991: Long-term residual efficacy of commercial formulations of 16 pesticides to Typhlodromus pyri Scheuten (Acari: Phytoseiidae) inhabiting commercial vineyards. Exp. Appl. Acarol. 13(1):

309 293Integrated protection and production in viticulture IOBC/ wprs Bulletin Vol. 67, 2011 pp Small insect enclosure field cages: A simple method to assess mating disruption F. Briand, C. Vergely, P.-J. Charmillot, P. Kehrli Station de recherche Agroscope Changins-Wädenswil ACW, CP 1012, CH-1260 Nyon, Switzerland. patrik.kehrli@acw.admin.ch Abstract: Mating disruption is an effective and sustainable integrated pest management technique that is now established in a broad range of cropping systems. For example, it is a popular alternative to insecticides for controlling the grape berry moth Eupoecilia ambiguella Hb. (Lepidoptera, Tortricidae), a main pest of European vineyards. A downside of this environmental friendly pest control technique is the challenging assessment of its efficiency in the field. In order to obtain reliable results, newly processed pheromone dispensers have to be tested at a large spatial scale where pest densities, crop variety, cultural practices, microclimate etc. can vary considerably among trial sites. The realisation of statistically sound results therefore asks for numerous independent replicates, which are time-consuming, space-demanding and expensive. The aim of this work was to facilitate the evaluation of the efficiency of mating disruption. We tested the suitability of small insect enclosure field cages for a pre-evaluation of the impact of sex pheromones on insect mating. Our enclosure field cages consisted of a cubic metal frame of 35cm side length that was covered with gauze tissue. Cages were installed in the centre of pheromone treated and untreated vineyards. Thereafter, various numbers of grape berry moth couples were exposed for 1, 2 or 3 days. The impact of pheromones was assessed by counting the number of mated females, e.g. the number of dissected spermatophores. Our data showed that mating was significant lower in pheromone treated compared to untreated vineyards. Moreover, there were significant differences among pheromone dispensers tested and mating disruption generally increased with the amount of pheromone diffused. Overall, the exposure of eight couples for one night seemed to be optimal for comparing different control schemes. Our study showed that small insect enclosure field cages offer a fast, simple, cheap and reliable method to measure the impact of pheromones on insect mating. Data obtained in such trials may therefore help predicting the value/use of setting-up large-scale field trials. In conclusion, small insect enclosure field cages may facilitate the development and assessment of semiochemical lures. Key words: Integrated pest management, mating disruption, Vitis vinifera, vine moth Introduction Insecticides extensive use of favoured the development of resistant insect pests and harmed beneficial insects. Social awareness to the drawbacks of classical insecticides urged the biotechnological industry to develop safer and more ecological friendly alternatives. One of these alternatives was the exploration of pheromone-mediated mate-finding systems. Gaston et al. (1967) were one of the firsts, which were able to confirm that pre-mating communication between sexes could be disrupted by permeating the atmosphere with synthetic sex pheromones. Today, the validity of manipulating and interfering with insect olfactory systems via the broadcast of synthetic pheromones has been demonstrated in many insect species and mating disruption has been established as an effective and sustainable integrated pest management measure in a broad range of cropping systems (Cardé & Minks, 1995). 289

310 290 A downside of this environmental friendly pest control technique is the laborious development process as well as the challenging assessment of its effectiveness. Electrophysiological responses of antennal receptor neurones are a useful first step to identify the basic chemical components of sex pheromones. Thereafter, these components have to be synthesised and their attractiveness has to be tested in wind tunnels or/and in baited monitoring traps. The next step is to evaluate the effectiveness of developed pheromone dispensers in the field. This is usually done by assessing pest densities in fields that were treated with pheromone dispensers compared to fields that remained unprocessed. To achieve reliable results, the plot size of such pheromone treated fields has to be of significant size, quite frequently of up to four hectares. Because pheromone dispensers have to be evaluated at a sizeable scale, preconditions between tested fields are rarely alike. The abundance of pests, the crop variety, the cultural practices, the microclimate, the soil etc. can vary significantly. The realisation of statistically sound results therefore asks for a lot of independent repetitions, which demands time, space and money. Several alternative methods were proposed for a preliminary assessment of pheromone dispensers. One of these methods is the exposure of tethered, virgin females in pheromonetreated and untreated fields over a defined period. Another method is the use of large insect enclosure field cages (Doye & Koch, 2005). These cages are set-up in pheromone-treated and untreated fields and a defined number of males are released within each cage. To assess the effectiveness of pheromone dispensers, they assessed the number of males captured. A similar approach was also taken by other authors (e.g. Schmidt & Seabrook, 1979). However, these authors cages were significantly smaller and they all exposed a defined number of insect males and females in their cages. Thereafter, they evaluated the effectiveness of mating disruption by dissecting preserved females and assessing their mating status. The great advantage of this approach is the direct measurement of pheromones' impact on mating. However, even though such small enclosure cages were used in the past, they are not commonly employed for testing newly processed pheromone dispensers. The refinement of small enclosure cages in order to obtain a reliable, standardised, quick and cheap screening method for the preliminary test of pheromone dispensers in the field seems therefore be a welcome asset for the biotechnological industry. Aiming to develop such screening cages we made us of the grape berry moth, Eupoecilia ambiguella (Hübner). This Tortricid is a major lepidopteran pest insect in European vineyards. In general, this moth is controlled by the application of insecticides, but mating disruption has been widely implemented over the last decade and several types of pheromone dispensers are available on the market (Ioriatti et al., 2004). In this study we present a generic approach on how to construct and test small insect enclosure field cages, which permit a preliminary evaluation of newly developed pheromone dispensers under standardised semifield conditions. Material and methods Insects Grape berry moths used in this study originated from a permanent laboratory culture at Agroscope Changins-Wädenswil. Moths were reared on a semi-artificial diet in a climate chamber (16:8 h L:D cycle, 70 ± 10% RH and 22 C). Pupae were sexed and males and females were separated. After emergence, moths were kept in cylindrical plastic until their sexual maturity. After this period, a defined number of E. ambiguella couples was exposed in the enclosure field cages for a fixed number of nights. Moths were exposed at the end of the

311 291 afternoon and they were recovered in the morning. Moths collected were killed and preserved in ethanol. In order to determine the mating status of preserved females, their bursa copulatris were dissected to confirm the presence or absence of spermatophores. In the case that at least one single spermatophore was present, females were classified as mated, otherwise they were scored as unmated. Test of field cages All field trials were conducted in three different vineyards around Nyon, Switzerland. The distance among them was between 500 and 1100 meters and they were of about 3 hectares. The goal of the field trial was 1) to examine if grape moths mate inside of small insect enclosure field cages and 2) to test if these cages are suited to measure the effect of sex pheromones on mating. The core of these enclosure field cages consisted of a cubic metal frame of 35cm side length each. The frame was covered with a mosquito net of 1,500µm mesh size (Figure 1). Two vineyards were either equipped with Isonet-LE or Isonet L-Plus pheromone dispensers manufactured by Shin-Etsu Chemical Co. Ltd. (Tokyo, Japan) and the third vineyard served as a reference. Enclosure field cages were put up in the middle of the foliage. Five couples of E. ambiguella were exposed in these enclosure field cages for a single night from June to August Figure 1. Insect enclosure field cage consisting of a cubic metal frame of 35cm side length, covered with a mosquito net of 1,500µm mesh size. Refinement of field cages For a refinement of these insect enclosure field cages, the optimal length of time of insects' exposure as well as the optimal number of exposed moth couples were tested. In 2008, cages were also set-up in two vineyards either equipped with Isonet-LE or Isonet L-Plus dispensers and a third vineyard that served as a reference. In a first trial, five couples of E. ambiguella were exposed within cages for either one, two or three nights. In a second trial, 1, 2, 5, 8, 12 and 20 couples of E. ambiguella were exposed within a single cage for one night. Both trials were conducted between May and June 2008.