Greedy scale, Hemiberlesia rapax (Hemiptera: Diaspididae), phenology on kiwifruit leaves and wood

New Zealand Journal of Crop and Horticultural Science ISSN: 0114-0671 (Print) 1175-8783 (Online) Journal homepage: https://www.tandfonline.com/loi/tnzc20 Greedy scale, Hemiberlesia rapax (Hemiptera: Diaspididae), phenology on kiwifruit leaves and wood R. H. Blank, G. S. C. Gill & M. P. Upsdell To cite this article: R. H. Blank, G. S. C. Gill & M. P. Upsdell (1996) Greedy scale, Hemiberlesia rapax (Hemiptera: Diaspididae), phenology on kiwifruit leaves and wood, New Zealand Journal of Crop and Horticultural Science, 24:3, 239-248, DOI: 10.1080/01140671.1996.9513958 To link to this article: https://doi.org/10.1080/01140671.1996.9513958 Published online: 22 Mar 2010. Submit your article to this journal Article views: 296 Citing articles: 6 View citing articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalinformation?journalcode=tnzc20

New Zealand Journal of Crop and Horticultural Science, 1996, Vol. 24: 239-248 0114-0671/96/2403-0239 $2.50/0 The Royal Society of New Zealand 1996 239 Greedy scale, Hemiberlesia rapax (Hemiptera: Diaspididae), phenology on kiwifruit leaves and wood R. H. BLANK G. S. C. GILL The Horticulture and Food Research Institute of New Zealand Whangarei Research Centre Private Bag 03 Whangarei, New Zealand M. P. UPSDELL New Zealand Pastoral Agriculture Research Institute Ltd Ruakura Agricultural Centre Private Bag 3123 Hamilton, New Zealand Abstract Greedy scale (Hemiberlesia rapax (Comstock)) populations at Whangarei, New Zealand were monitored at regular intervals on unsprayed kiwifruit (Actinidia deliciosa (A.Chev.) C.F. Liang et A.R. Ferguson var. deliciosa) wood and leaves from 1984 to 1989 to obtain detailed stage-specific phenological information. The relationship between relative abundance of each scale stage and Julian day or degree-day accumulation (>9.3 C) were determined using a Bayesian smoothing programme. Two scale generations, summer and winter, were readily discernible. There was close agreement between the phenology of scale on wood and leaves. However, because of senescence and abscission in autumn there was only a partial winter generation of scale on leaves. Synchrony of stage-specific events was enhanced using degree-day accumulations rather than Julian days, particularly for the winter generation. Estimates of stage-specific peaks and troughs for crawlers, white caps (first instar), yellow caps (second instar), black caps (third instar), and H96004 Received 7 February 1996; accepted 6 June 1996 mature scale are given in the paper and have a mean standard error (SE) of 10 Julian days or 38 degreedays (>9.3 C). Mean generation intervals for scale on wood and leaves were 1022 (SE 18) and 1114 (SE 53) degree-days respectively. Keywords greedy scale; Hemiberlesia rapax; phenology; kiwifruit; Actinidia deliciosa; degreeday accumulation INTRODUCTION The management of armoured scales infesting a wide range of plant hosts has generally been much improved once the phenology has been described in detail. This information can be used to enhance control, by improving the timing of insecticidal sprays (Jorgensen et al. 1981 ; Kennett & Hoffmann 1985; Ressig et al. 1985; McClain et al. 19) and biological control strategies (Yu & Luck 1988; Potter et al. 1989). In New Zealand, the production of high quality kiwifruit {Actinidia deliciosa (A.Chev.) C.F. Liang et A.R. Ferguson) for world markets requires that damage or contamination by insect pests be kept to low levels. Armoured scale insects are important quarantine pests of kiwifruit which if left uncontrolled can prevent orchard crops from being exported. A regular schedule of broad-spectrum insecticides have been used to protect fruit from the risk of scale contamination. Recent studies have shown there are opportunities for the relaxation of export standards as most of the scale found on fruit are immature or dead and therefore do not constitute a quarantine risk (Blank et al. 1993). Pest management strategies are currently being developed which monitor scale infestations on leaves to determine the need for control procedures (Steven et al. 1991; Blank etal. 1994; Steven et al. 1994). In combination, the implementation of these procedures will shift the emphasis from calendar-based prophylactic spray practices to a true pest management approach.

240 New Zealand Journal of Crop and Horticultural Science, 1996, Vol. 24 Greedy scale {Hemiberlesia rapax (Comstock)), latania scale (Hemiberlesia lataniae (Signoret)), and oleander scale {Aspidiotus nerii Bouche), are the most common species of armoured scale found on kiwifruit (Berry et al. 1989; Lo & Blank 1989). All three species of scale superficially resemble each other. The species can be easily confused unless correct identification procedures are followed. The life cycle of greedy scale has been described by Ferguson (1979) and Berry (1983). The mobile first instar inserts its mouthparts into plant tissue and spins a protective cover to form the settled white cap. It then moults to form the second instar or yellow cap and then the third instar or black cap. Scale continue to enlarge, although there are no further moults, and mature scale produce eggs. No males have been recorded for this species. Aerial invasion of crawlers from adjacent host plants such as taraire has been shown to be a major source of contamination of kiwifruit orchards (Blank et al. 19). The phenology of greedy scale has been described in detail from taraire (Beilshmiedia tarairi Bentham and Hooker) leaves (Blank et al. 1995a,b). This paper investigates the phenology of greedy scale on kiwifruit leaves and wood with special reference to the physiological time scale based on degree-day accumulations. MATERIALS AND METHODS Sites Armoured scale populations on unsprayed kiwifruit vines on three sites at Maunu near Whangarei, New Zealand were monitored at regular intervals over five seasons from 1984 to 1989 to obtain detailed, stage-specific phenological information. Site A was sampled in the 1984/85 season and comprised 12 abandoned 8-year-old vines grown on T-bar structures in two rows. Vines had grown together to form a dense canopy (30 x 10 m) of predominantly the 'Hayward' cultivar with small amounts of ' Bruno ' shoots originating from the rootstock. Vines at Site A were removed in September 1985 and so a second and third site within 0.5 km of Site A were monitored. These sites were also abandoned kiwifruit vines which had not received any pruning or pest management for over 2 years. Site B comprised a single 30-year-old vine of 'Gracie', acultivar which sets large amounts of fruit without the need for a pollinator (Zhang & Thorp 1986). The canopy of this vine was c. 40 x 15m and covered a rock wall and three adjacent shelter trees. At Site C both a female and male 'Hayward' vine which were at least 8 years old were monitored. These vines covered a rock wall and formed a tangled mass of canopy c. 20 x 5 m. Sites B and C were monitored for two seasons, until the removal of the kiwifruit vine at Site B in March 1989 after which only Site C was monitored (Table 1). Sampling Scale populations on leaves were monitored for five seasons and on wood for three seasons (Table 1 ). At each site, three similar sized sampling regions were established along both sides of the kiwifruit canopy giving a total of six sampling regions from which equal amounts of samples were collected. Wood samples were pruned from within the vine canopy commencing in September or October each season and continuing until July. Wood samples were monitored at 13 22-day intervals from October to April and at 28^2-day intervals in September and May June giving 12 14 assessments each season (Fig. 1). Initially, in the 1984/85 season both new wood ( 1 year old) and old wood (2 or 3 years old) were assessed. However, in later seasons only old wood was assessed because of the greater abundance of scale. A 30 cm length of wood was used as the standard sampling unit. A minimum of 12 old wood samples (2x6 sampling regions) were assessed per site at each sampling occasion (Table 1 ). An additional 6 12 old wood samples were also assessed in September, October, and July samplings. Leaf sampling commenced in November each growing season after the formation of new leaves in spring and finished in June with the onset of leaf senescence and abscission. Leaf samples were taken at 14 30-day intervals from November to April and at 22 50-day intervals from May to June giving 10-12 samples each season (Fig. 2). From -150 Table 1 Numbers of leaves and wood samples assessed for armoured scale (Hemiberlesia rapax) at three kiwifruit (Actinidia deliciosa) sites on each samphng occasion ovei five seasons. Year 1984/85 1985/86 1986/87 1988/89 1989/ Site A B C B C C C Leaves 150 Wood 12 new and 12 old 12 old 12 old 12 old 12 old -

Blank et al. Greedy scale phenology on kiwifruit 241 1 0, 3001 6-4- T3 OO I o CO 4- CO o CO 3-2- CO 05 250-200- 150-100- ^ 100-I CO Ü CO 8CH ea 4a 2a 1984/85 1985/86 1986/87 1987/88 1988/89 A S O N D J F M A Month M 0- O N J F M Month Fig. 1 Abundance of greedy scale {Hemiberlesia rapax): A, total live scale; and B, live white caps on old kiwifruit (Actinidia deliciosa) wood over three seasons. Fig. 2 Abundance of greedy scale (Hemiberlesia rapax): A, total live scale; and B, live white caps on kiwifruit (Actinidia deliciosa) leaves over five seasons. leaves/site were sampled randomly from the vine canopy (Table 1). Leaf and wood samples were held at 5 C and assessed for scale stage and mortality under a binocular microscope within 1-4 days. During the course of this study 283 adult scale were collected from both kiwifruit leaves and wood from all sites for identification using procedures described by Lo & Blank (1989). Crawler activity on wood was monitored using tape traps at Site C from October 1985 to August 1989. Tape traps were made by completely encircling the wood with white tape (Nitto 20 mm wide) with the sticky side towards the wood. The two ends of tape were joined to form a flap for ease of handling during removal. After attachment, grease (Shell Snow White Petrolatum) was applied thinly over the tape to trap crawlers. A total of 16 traps were used with eight traps positioned on both sides of the vine canopy. Tape traps were attached above and within 40 mm of mature scale on wood. Tape traps were removed and replaced with new traps in the same position at regular intervals during each season. Intervals ranged from 13 to 23 days over November- April and from 19 to 56 days during May-October giving 10 14 assessments per season. Tapes were examined under a microscope for greedy scale crawlers which were distinguished by their yellow oval shape, size, antennae, and lack of eyespots. Crawlers from a laboratory colony of identified greedy scale were used as a reference. Crawlers adhering to both the outer greased surface and the inner sticky surface of the tape were counted. Degree-day accumulations were recorded using an Omnidata Growing Degree-Day Accumulator (Model TA51-PC), or biophenometer, located at Site B from October 1985 onwards. The temperature probe of the biophenometer, which had been checked for calibration, was positioned inside a white, louvred, wooden meteorological box attached to a post 1 m above the ground underneath the kiwifruit canopy. The biophenometer measured temperature

242 New Zealand Journal of Crop and Horticultural Science, 1996, Vol. 24 every 10 min and calculated a new degree-day total. The upper temperature limit of 40 C was never exceeded. The developmental threshold from settlement to first reproduction of 9.3 C was used as the base temperature to calculate degree-day accumulation (Blank et al. 1995b). Degree-day accumulations for each sampling date were estimated by interpolating between the 9 and 9.5 C channels of the biophenometerto approximate 9.3 C. Degree-day accumulations for the 1984/85 season were estimated from maximum and minimum daily temperatures recorded from the nearby (5 km) Whangarei Base Hospital meteorological station with adjustments based on statistically significant correlations established over four seasons when both data sets were available. The relationship between the relative abundance of each scale stage and Julian days or degree-day accumulations (>9.3 C) were determined using a Bayesian smoothing programme, Flexi (Upsdell 1994). A full description of the use of this Bayesian smoothing approach to describe the phenology of greedy scale stages on taraire leaves at a nearby (5 km) site is given by Blank et al. (1995b). The important considerations were that Flexi gave different weights to each assessment depending on the numbers of scale sampled, and combined information from several assessments to produce a graphical description of the information together with confidence intervals which may be tabulated. Flexigraphs and tables of stage-specific events with standard errors were prepared combining sites for each season. From this data the overall mean Flexigraphs and tables of stage-specific peaks and troughs were prepared. RESULTS Greedy scale comprised 98.9% of the armoured scale identified from the leaves and wood of kiwifruit vines at all sites. The remainder of the armoured scale (1.1%) were oleander scale with three specimens found on wood from two sites. On 12 of the 13 sampling dates in the 1984/85 season more live scale were found on old than new wood and on six occasions this difference was significant (P < 0.05). Overall, there was a 7-fold increase in the amount of scale found on old compared to new wood. As there was no difference in the phenology of scale from different ages of wood, these data have been combined for presentation. 100 60 0 = 60 40 "Í 2 o o 0 o = 60 80-40- 20-40- 20 0 25-Aug 27-Oct 29-Dec 2-Mar 4-May 6-Jul 7-Sep Date Fig. 3 Relative abundance of greedy scale (Hemiberlesia rapax) development stages: A, white cap; B, yellow cap; C, black cap; D, mature on kiwifruit wood (Actinidia deliciosa) and the mean plotted against calendar date for three seasons. The seasonal abundance of total live scale and live white caps on wood and leaves (before analysis by Flexi) is shown in Figs 1 and 2 respectively. Total live scale numbers on both wood and leaves increased in spring and autumn as a result of recruitment of immature scale into the population with subsequent declines in summer and winter (Figs 1 and 2). The two generations of scale can be clearly seen from the abundance of white caps on both wood and leaves, although the position and

Blank et al. Greedy scale phenology on kiwifruit 243 -- 1984/85 ---1987/88 -- 1985/86 1988/89-1986/87 mean 60-40- 20-0 60-40- 20 0 0 300 600 0 1200 1500 1800 2100 Degree-days > 9.3 C 30-Oct 10-Dec 20-Jan 3-Mar 13-Apr 24-May 5-Jul Date Fig. 4 Relative abundance of greedy scale (Hemiberlesia Fig. 5 Relative abundance of greedy scale (Hemiberlesia rapax) development stages: A, white cap; B, yellow rapax) development stages: A, white cap; B, yellow cap; C, black cap; D, mature on kiwifruit (Actinidia cap; C, black cap; D, mature on kiwifruit (Actinidia deliciosa) wood for three seasons and the mean plotted deliciosa) leaves for five seasons and the mean plotted against degree-days from a starting date of 1 August. against calendar date. intensity of peaks and troughs varied between seasons. The proportion of each scale stage present during each season and the means for the three seasons for the wood and the five seasons for the leaves are presented as Flexigraphs for both Julian days (expressed as calendar dates) and physiological time (expressed as degree-days) (Figs 3 6). The two generations of scale are easily discernible for all stages, except for the black cap and mature stages on leaves, which were unable to complete the winter generation because of senescence of the leaves. The pattern of peaks and troughs for scale on the wood generally showed good consistency between seasons. Flexigraphs using degree-day accumulations improved synchrony compared to Julian days, especially for the winter generation. The greatest variability between seasons generally occurred in early spring or late winter when

244 New Zealand Journal of Crop and Horticultural Science, 1996, Vol. 24 -- 1984/85 1987/88 1985/86 1988/89 ---1986/87 mean 0 300 550 800 1050 1300 1550 1800 2050 2300 Degree-days > 9.3 C Fig. 6 Relative abundance of greedy scale (Hemiberlesia rapax) development stages: A, white cap; B, yellow cap; C, black cap; D, mature on kiwifruit (Actinidia deliciosd) leaves for five seasons plotted against degreedays from a starting date of 1 August. stage-specific events were more difficult to detect as all stages were present but in low numbers (Figs 1 and 2). The other main cause for lack of synchrony was where bimodal peaks occurred. For example, the 1986/87 summer generation ofwhite caps formed a bimodal peak which was also reflected in the summer yellow cap bimodal peak. Synchrony of the stage-specific events on leaves was also improved using degree-day accumulations particularly for the winter (second) generation of white caps and yellow caps (Figs 5 and 6). The mean Julian days (expressed as calendar dates) and degree-day accumulations for the stage-specific peaks and troughs are summarised in Tables 2 5. s refer to the minimum proportion of scale preceding the maximum or peak proportion. On some occasions one season's data was omitted, even though it had little effect on overall means, because of insufficient data points either at the commencement or end of a season to fully validate an event. For example, if the last sampling point in a season was also the lowest data point then Flexi would extrapolate using the following season's data points to determine the position of the trough. However, sometimes because of a long interval between sampling dates or low numbers of live scale over the winter, this estimate would have a high standard error indicating that this was a poor estimate of this stage-specific event. The 1984/85 white cap and mature summer generation as well as black cap winter generation on wood were not clearly shown and although Flexi was able to discern peaks and troughs these were also omitted from the estimates. Exclusion of these events had little effect on the overall generation intervals which are presented in Tables 6 and 7. The overall estimate of generation interval derived from scale on leaves suffers from a high white cap peak-peak estimate caused by the bimodal nature of the winter generation. The loss of black cap and mature peak-peak estimates because of senescence of leaves also reduces confidence in this estimate compared to the generation interval derived from wood. Despite these considerations, the scale generation intervals and standard errors are remarkably similar, showing that scale development was similar on both leaves and wood. The phenology of crawlers was similar between seasons and clearly shows the two crawler release periods of the two mature scale generations (Fig. 7). The peak crawler activity in spring occurred on 13 December (SEM 6.5 days), followed by a trough on 22 February and a summer peak on 17 April. In most seasons the intensity of the spring crawler activity appeared higher than the summer peak, but this may simply reflect the abundance of mature scale near traps. DISCUSSION These phenology studies dealt solely with greedy scale as the closely related latania and oleander

Blank et al. Greedy scale phenology on kiwifruit 245 Table 2 Calendar date estimates (mean + SE in days) for greedy scale {Hemiherlesia rapax) stagespecific peaks and troughs over all sites for three seasons on kiwifruit (Actinidia deliciosa) wood at Whangarei, New Zealand. White cap Yellow cap Black cap Mature a Estimate for two seasons. 30 Sep ± 7 a 8 Nov ± 8 10 Dec ± 8 25 Jan ± T Summer generation 17 Dec ± 23 Jan ± 9 19 Feb 1 9 5 Apr± 9a 6 a 3Feb± 15 24 Mar± 12 19 Apr ± 20 a 9 Aug ± 9 a Winter generation 9 Apr±10 10 8Jun ±1 14 28 Sep ±1 10 a 6 Nov i 4 Table 3 Degree-day accumulation (>9.3 C) estimates (mean ± SE) for greedy scale (Hemiberlesia rapax) stage-specific peaks and troughs over all sites for three seasons on kiwifruit (Actinidia deliciosa) wood at Whangarei, New Zealand. White cap Yellow cap Black cap Mature a Estimate for two seasons. 46 ± 33 a 389 ± 44 685 ± 42 927 + 55 a Summer generation Degree-day accumulations 679 + 43 a 958 ±31 1225 ± 33 1320 ± 56 a 1101165 1427 ± 29 1673 ± 31 a 19±70 a Winter generation 1589 + 25 1974 ± 76 203+ 3 a 379±48 Table 4 Calendar date estimates (mean ± SE in days) for greedy scale {Hemiberlesia rapax) stagespecific peaks and troughs over all sites for five seasons on kiwifruit (Actinidia deliciosa) leaves at Whangarei, New Zealand. < summer generation Winter generation White cap Yellow cap Black cap Mature 15 Aug ± 20 a 24 Oct ±1 10 23 Nov ± 5 28 Dec ± 8 a Estimate from four seasons. 8 Nov i 18 Jan ± 12 Feb ± 19 Mar'± 10 9 6 10 6 Feb ± 8 17 Mar± 9 26 Apr± 17 a 19Jun ± 17 5 25 Apr ± May ±1 6 a 11 Table 5 Degree-day accumulation (>9.3 C) estimates (mean ± SE) for greedy scale (Hemiberlesia rapax) stage-specific peaks and troughs over all sites for five seasons on kiwifruit (Actinidia deliciosa) leaves at Whangarei, New Zealand. White cap Yellow cap Black cap Mature Summer generation 2098 ± 43 a 321±59 594 ± 94 883 ± 69 Degree-day accumulations 409 + 103 958± 59 1210± 29 1484± 56 1152± 40 1451 ± 58 1795+ 121 a 2023 ± 52 a Estimate for four seasons. b Winter generation peaks could not be determined because of loss of leaves. Winter generation 1717± 72 a 1975 ± 42 _b _b

246 New Zealand Journal of Crop and Horticultural Science, 1996, Vol. 24 Ilers/tra >. x> a> Q- 0.5 Q. ö 0.8 0.6 0.7-0.4-0.3-0.2 0.1 7/ 1985/86.7-1986/87 1987/88 1988/89 1-Sep 31-Oct 31-Dec 2-Mar 1-May 1-Jul 31-Aug Date Fig. 7 Greedy scale (Hemiberlesia rapax) crawler activity on kiwifruit (Actinidia deliciosa) wood assessed using sticky tape traps over four seasons. scales were either not found or were only found in small numbers. Earlier studies have also shown that greedy scale was dominant on kiwifruit in the Whangarei district (Lo & Blank 1989; Blank et al. 1992). This study has been successful in measuring the variation in greedy scale phenology between seasons. The study was not designed to compare sites, but Flexi was robust enough to accommodate the possible increased variability due to more than one site having to be used. Crawler activity on the wood closely reflected peaks and troughs of white cap production. Crawlers settle to form white caps within 2 5 days (Berry 1983). This study has demonstrated that the phenology of white caps can be used to predict crawler phenology, and so avoid the need for regular monitoring of sticky tape traps. There was close agreement between the phenology of scale on leaves and wood using both Julian days and degree-day accumulations. Deviations were caused by the initiation of new leaves in spring or senescence of leaves in autumn. Kiwifruit wood was useful for establishing the complete phenology of both summer and winter scale generations, whilst leaves gave information on the phenology of primarily the summer generation. In commercial orchards leaves are easier to sample than wood during the growing season. It is therefore important that we have been able to demonstrate that predictions of scale phenology based on leaf data will reflect the whole kiwifruit plant. In many kiwifruit orchards the main source of scale may be derived from aerial invasion of crawlers from adjacent host plants (Blank et al. 19). In these circumstances the phenology of the scale settling on kiwifruit will reflect that of the original host plant rather than scale derived from kiwifruit. The phenology of greedy scale has been extensively studied on taraire which were also located in the Maunu region and were within 5 km of these kiwifruit study sites (Blank et al. 1995b). The estimate for the mean generation interval for greedy scale on taraire leaves was 1056 (SE 34) degree-day accumulations (>9.3 C) which is close to the estimates of 1022 (SE 18) and 1114 (SE 53) for kiwifruit wood and leaves respectively. The use of a biofix based on the appearance of a key stage-specific event could be used to bring together the timing of phenological events of scale on different host plants. A biofix based on plant phenology may also be useful for forecasting scale phenological events. The commencement of budbreak or flowering on kiwifruit or first leaf growth flush in spring on taraire may be a useful Table 6 Mean degree-day accumulation (>9.3 C) estimates between greedy scale {Hemiberlesia rapax) stage-specific generation times over all sites for three seasons on kiwifruit (Actinidia deliciosa) wood at Whangarei, New Zealand. Degree-day accumulations -trough -peak White cap 1055 910 Yellow cap 1038 1016 Black cap 998 1023 Mature 1073 1061 Overall mean ± SE 1022 ± 18 Table 7 Mean degree-day accumulation (>9.3 C) estimates between greedy scale (Hemiberlesia rapax) stage-specific generation times over all sites for five seasons on kiwifruit (Actinidia deliciosa) leaves at Whangarei, New Zealand. Degree-day accumulations -trough White cap 946 Yellow cap 1130 Blackcap 1080 Mature 1201 Overall mean ± SE 1114 ± 53 -peak 1308 1017

Blank et al. Greedy scale phenology on kiwifruit 247 indicator of events. The forecasting of phenological events of other diaspidids have been based on flowering of cherry, apples, and pears{kozar 19). The ability to forecast stage-specific events of greedy scale opens up new opportunities for enhancing pest management strategies on kiwifruit. Phenological predictions using either calendar dates or degree-day accumulations may be used to target scale monitoring activities to determine the need for spray applications or, directly by targetting spray applications against stage-specific events. For example, spray applications may be targetted against immature scale of both generations so that the enhanced activity of the insecticides against these stages can be utilised (Blank & Olson 1987, 19) and so prevent establishment of scale on the fruit. Alternatively, a narrow range mineral oil could be used which will prevent settlement of crawlers on fruit for 16-24 days (Blank et al. 1995c). The use of novel compounds, such as insect growth regulators which interfere with the moulting process of immature scale, would be enhanced by targetting against specific immature stages (Paloukis 1986). Further research is required to investigate the use of phenological models of greedy scale to time critical management decisions such as scale pest management assessments and targetted spray applications. ACKNOWLEDGMENTS Our thanks to Mr and Mrs P. Hale, Mr I. J. Gilmer, and Mr B. McKenzie for allowing us to sample their kiwifruit vines, and to Mrs J. M. Kelly, Mrs N. B. Tapper, and others who assisted with the scale assessments. This work was supported by the Foundation for Research, Science and Technology Contract CO6315. REFERENCES Berry, J. A. 1983: Aspects of the ecology and control of greedy scale Hemiberlesia rapax (Comstock). Unpublished MSc thesis, University of Auckland, New Zealand. 112 p. Berry, J. A.; Morales, C. F.; Hill, M. G.; Loforth, B. J.; Allan, D.J. 1989: The incidence of three diaspidid scales on kiwifruit in New Zealand. Proceedings of the 42nd New Zealand Weed and Pest Conference: 182-186. Blank, R. H.; Gill, G. S. C.; Olson, M. H. 1994: Relationship between armoured scale infested kiwifruit leaves and fruit. Proceedings of the 4 7th New Zealand Plant Protection Conference: 304-309. Blank, R. H.; Gill, G. S. C.; Olson, M.H. 1995a: Seasonal abundance of greedy scale (Hemiptera: Diaspididae) and associated parasitoids on tarairc (Beilschmiedia tarairi). Journal of economic entomology 88: 1634-1640. Blank, R. H.; Gill, G. S. C.; Olson, M. H.; Upsdell, M. P. 1995b: Greedy scale (Hemiptera: Diaspididae) phenology on taraire based on Julian date and degree-day accumulations. Environmental entomology 24: 1569-1575. Blank, R. H.; Holland, P. T.; Gill, G. S. C.; Olson, M. H.; Malcolm, C. P. 1995: Efficacy and persistence of insecticide residues on fruit of kiwifruit to prevent greedy scale (Hemiptera: Diaspididae) crawler settlement. New Zealand journal of crop and horticultural science 23: 13-23. Blank, R. H.; Olson, M. H. 1987: Differential toxicological responses of greedy scale stages to diazinon. Proceedings of the 40th New Zealand Weed and Pest Control Conference: 161-164. Blank, R. H.; Olson, M. H. 19: The toxicity of methidathion against armoured scale pests on kiwifruit. Proceedings of the 43rd New Zealand Weed and Pest Control Conference: 243-246. Blank, R. H.; Olson, M. H.; Gill, G. S. C. 1992: Armoured scale, Hemiberlesia lataniae and H. rapax (Hemiptera: Diaspididae), infestation of kiwifruit rejected for export at two packhouses from 1987 to 1991. New Zealand journal of crop and horticultural science 20: 397-405. Blank, R. H.; Olson, M. H.; Gill, G. S. C. 1993: An assessment of the quarantine risk of armoured scale (Hemiptera: Diaspididae) fruit infestations on kiwifruit. New Zealand journal of crop and horticultural science 21: 139-145. Blank, R. H.; Olson, M. H.; Lo, P. L. 19: Armoured scale (Hemiptera: Diaspididae) aerial invasion into kiwifruit orchards from adjacent host plants. New Zealand journal of crop and horticultural science 18: 81-87. Ferguson, A. M. 1979: Greedy scale, Hemiberlesia rapax (Comstock), life cycle. DSIR information series no. 105/31. Jorgensen, C. D.; Rice, R. E.; Hoyt, S. C.; Westigard, P. H. 1981: Phenology of the San Jose scale (Homoptera: Diaspididae). Canadian entomologist 113: 149-159. Kennett, C. E.; Hoffmann, R. W. 1985: Seasonal development of the California red scale (Homopte.a: Diaspididae) in San Joaquin Valley citrus based on degree-day accumulation. Journal of economic entomology 78: 73-79. Kozar, F. 19 Forecasting and monitoring infestations. Pp. 335-340 in: Armoured scale insects: their biology, natural enemies and control. Vol. B. Rosen, D. ed. Amsterdam, Elsevier.

248 New Zealand Journal of Crop and Horticultural Science, 1996, Vol. 24 Lo, P. L.; Blank, R. H. 1989: A survey of armoured scale species (Hemiptera: Diaspididae) in kiwifruit orchards. New Zealand entomologist 12: 1-4. McClain, D. C.; Rock, G. C.; Stinner, R. E. 19: San Jose scale (Homoptera: Diaspididae): simulation of seasonal phenology in North Carolina orchards. Environmental entomology 19: 916-925. Paloukis, S. 1986: Evaluation of two new insecticides for the control of scale insects on fruit trees in northern Greece. Pp. 179-183 in: Proceedings of the Fifth International Symposium of Scale Insect Studies, Naples, Italy. Supplement Bullettino del Laboratorio di Entomologia agraria. Potter, D. A.; Jensen, P.; Gordon, F. C. 1989: Phenology and degree-day relationships of the obscure scale (Homoptera: Diaspididae) and associated parasites on pin oak in Kentucky. Journal of economic entomology 82: 551-555. Reissig, W. H.; Weires, R. W.; Onstad, D. W.; Stanley, B. H.; Stanley, D. M. 1985: Timing and effectiveness of insecticide treatments against San Jose Scale (Homoptera: Diaspididae). Journal of economic entomology 78: 238-248. Steven, D.; Blank, R. H.; Tomkins, A. R. 1991 : Monitoring scale to reduce sprays in kiwifruit. New Zealand journal of agricultural science 26: 84. Steven, D.; Tomkins, A. R.; Blank, R. H.; Charles, J. G. 1994: A first-stage integrated pest management system for kiwifruit. Proceedings Brighton Crop Protection Conference: 135-142. Upsdell, M.P.I 994: Bayesian smoothers as an extension of non-linear regression. The New Zealand statistician 29: 66-81. Yu, D. S.; Luck, R. F. 1988: Temperature-dependent size and development of California red scale (Homoptera: Diaspididae) and its effect on host availability for the ectoparasitoid, Aphytis melinus DeBach (Hymenoptera: Aphelinidae). Environmental entomology 17: 154-161. Zhang, J.; Thorp, T.G. 1986: Morphology of nine pistillate and three staminate New Zealand clones of kiwifruit (Actinidia deliciosa (A.Chev.) C.F. Liang et A.R. Ferguson var. deliciosa). New Zealand journal of botany 24: 589-613.