The EFSA Journal (2006) 439, 1-41

Transcription

1 Opinion of the Scientific Panel on Plant Health on an evaluation of asymptomatic citrus fruit as a pathway for the introduction of citrus canker disease (Xanthomonas axonopodis pv. citri) made by the US Animal and Plant Health Inspection Service (APHIS) 1 (Question N EFSA-Q ) Opinion adopted on 15 December 2006 SUMMARY In this document, the PLH Panel has responded to a question from the European Commission DG SANCO requesting an opinion on the Evaluation of asymptomatic citrus fruit as a pathway for the introduction of citrus canker disease made by the US Animal and Plant Health Inspection Service (APHIS), and, in particular, on its conclusion that asymptomatic citrus fruit is an unlikely pathway for the spread of the disease. Citrus canker is an economically significant plant disease caused by the bacterium Xanthomonas axonopodis pv. citri (Xac) affecting most of the citrus growing areas in the world. Rigorous phytosanitary measures have insured that some areas, including Europe, are still free of the disease. These include the use of systems approaches. The current systems approach for the trade of citrus fruit requires that fruit are coming from disease-free area surrounded by a non-export buffer zone. The APHIS document proposes to modify the systems approach so that asymptomatic fruit coming from infected/contaminated areas are eligible for trade. The APHIS document describes the biology of Xac, the epidemiology of the citrus canker disease and the spread potential of the pathogen. The data were then used to evaluate the phytosanitary evidence regarding asymptomatic fruit as a pathway for the introduction of the disease into a new area. Five Events were identified in the APHIS document as necessary for the disease to be introduced into a new area through commercial fruit: Event 1: Infected/contaminated fruit is harvested. Event 2: Inoculum on infected/contaminated fruit survives the packing/treatment process. Event 3: Inoculum on infected/contaminated fruit survives shipment. 1 For citation purposes: Opinion of the Scientific Panel on Plant Health on a request from the Commission on an evaluation of asymptomatic citrus fruit as a pathway for the introduction of citrus canker disease (Xanthomonas axonopodis pv. citri) made by the US Animal and Plant Health Inspection Service, The EFSA Journal (2006) 439, of 41

2 Event 4: Infected/contaminated fruit go to a suitable area with conditions for infection. Event 5: Inoculum from infected/contaminated fruit encounters a suitable host and is able to incite disease. The PLH panel agreed with the method of analysis presented in the APHIS document, but not with the conclusion reached. The APHIS document analyses the five events that must occur for successful introduction and proposes a new systems approach to prevent entry and establishment. However, key arguments provided in the APHIS document are not supported by scientifically sound evidence: major events are not analysed on the basis of quantitative data, important statements are not always supported by adequate scientific references, and references are frequently reported in a way which do not correctly reflect the results and conclusions of the cited paper. Therefore, since no new or additional studies are presented the analysis of the evidence provided in the document does not justify a change in the regulations. None of the preventative measures in the systems approach proposed by APHIS are shown to give effective control of Xac. The Panel concluded that, where an initial inoculum load exists, the transmission of Xac in the scheme proposed by APHIS is more likely than with the current systems approach. Key words: citrus canker, Xanthomonas axonopodis pv. citri, asymptomatic citrus fruit, systems approach, quarantine 2 of 41

3 TABLE OF CONTENTS The EFSA Journal (2006) 439, 1-41 SUMMARY... 1 TABLE OF CONTENTS... 3 BACKGROUND... 3 TERMS OF REFERENCE... 4 ASSESSMENT INTRODUCTION CRITICAL REVIEW OF THE DOCUMENT... 5 Event 1 Infected/contaminated fruit is harvested [p. 8-10] General comments Specific comments... 7 Event 2 Inoculum on infected/contaminated fruit survives the packing/treatment process [p ] General comments Specific comments Event 3 Inoculum on infected/contaminated fruit survives shipment [p ] General comments Specific comments Event 4 Infected/contaminated fruit go to a suitable area with conditions for infection [p. 13] Event 5 Inoculum from infected/contaminated fruit encounters a suitable host and is able to incite disease [p ] REVIEW OF THE REFERENCES USED IN THE APHIS EVALUATION References not available Additional references Missing / Misprinted references CONCLUSIONS AND RECOMMENDATIONS DOCUMENTATION PROVIDED TO EFSA...37 REFERENCES Additional References SCIENTIFIC PANEL MEMBERS BACKGROUND 2 The Commission (SANCO E.1) was informed on April 7, 2006 that the US Animal and Plant Health Inspection Service (APHIS) has published in the Federal Register of April 6, an Evaluation of asymptomatic citrus fruit as a pathway for the introduction of citrus canker disease. Comments are due by June 5, The above evaluation concludes that it is highly unlikely that citrus canker could be introduced on asymptomatic, commercially produced citrus fruit that has been treated with disinfectant dips and subject to other mitigations (see Summary on page of the Federal Register). 2 Submitted by EC SANCO, ref. E1/VE/svi D(2006) of 41

4 In the introduction to the evaluation, APHIS mentions that in the case of citrus fruit from areas where citrus canker occurs, APHIS authorises systems approaches, in accordance with the International Standard on Phytosanitary Measures ISPM no 14: The use of integrated measures in a systems approach for pest risk management. The text continues by saying that A crucial factor in the current systems approaches for the importation of citrus fruit is free of symptoms or signs of the disease (asymptomatic). Requiring production area freedom as part of a systems approach may not be technically justified where asymptomatic fruit can be produced if asymptomatic fruit does not provide a pathway for introduction. In this respect it should be mentioned that the current Community plant health regime established by Council Directive 2000/29/EC lists citrus canker as an organism harmful to plants and fruits of Citrus, Fortunella and Poncirus and their hybrids. Moreover, citrus canker is not yet present in the Community and the Directive contains measures in the form of a systems approach, such as production freedom as well as fruit to be free from symptoms (see Annex IV, part A, section I, point 16.2, option c of Directive 2000/29/EC 3 ). A change in the strategy towards the mitigation of citrus canker in the US could also have possible influences on the Community strategy, as the US could ask the Community to also change its strategy and to bring it in line with the evaluation mentioned in the Federal Register. TERMS OF REFERENCE EFSA is requested, pursuant to Article 29(1) and Article 22(5) of Regulation (EC) No 178/2002, to provide a scientific opinion on the Evaluation of asymptomatic citrus fruit as a pathway for the introduction of citrus canker disease made by APHIS, and in particular on its conclusion that asymptomatic citrus fruit is an unlikely pathway for the further spread of the disease. ASSESSMENT 1. Introduction Citrus canker is a plant disease caused by the bacterium Xanthomonas axonopodis pv. citri (hereinafter referred as to Xac) which attacks all aerial parts of the plant. Symptoms include lesions on fruit, which reduce their commercial value. The disease is widely distributed in most citrus-growing areas. In the USA, it is present only in Florida, while Europe is free from the disease. The main pathway for the long-distance dissemination of the bacterium has been considered to be on nursery plants. Localised spread is mainly associated 3 OJ L 169/1, , p of 41

5 with inoculum on infected plants being carried by wind-driven rain. Once established, the disease is difficult to eradicate. For example, despite extensive efforts made in Florida to contain the most recent outbreak which included mass felling of citrus plants and a total ban on the movement of fruit from contaminated groves, the disease has continued to spread (Brown 2001). The bacterium is of quarantine concern in many countries. In the EU, the Council Directive 2000/29/EC declares that Member States ban the introduction into their territory of the plants and plant products listed in Annex II, Part A, when they are contaminated by the relevant harmful organisms listed in that part of the Annex. Similar phytosanitary rules have been enforced in the US until now, including listing the entire State of Florida as a quarantined area for citrus canker (7 CFR Part ). However, the US regulatory authorities have recently reevaluated the epidemiological significance of fruit harvested in contaminated areas and now propose to relax restrictions on their trade. This significant change necessitates an assessment of the scientific evidence which supports this development. The document under scrutiny entitled Evaluation of asymptomatic citrus fruit (Citrus spp.) as a pathway for the introduction of citrus canker disease (Xanthomonas axonopodis pv. citri) version March aims to demonstrate that requiring production area freedom as part of a systems approach 6 may not be technically justified where asymptomatic fruit can be produced if asymptomatic fruit does not provide a pathway for introduction 7. It consists of general introduction to the disease and the bacterium followed by a section on Phytosanitary evidence regarding asymptomatic fruit as a pathway. In the latter section five events 8 which were considered as necessary for the disease to be introduced into a new area on commercial fruit are discussed. 2. Critical review of the document The Panel thoroughly analysed the entire USDA APHIS document, made a critical review and checked the cited references. The five Events, which were used by APHIS to support their conclusions, were put in focus of the Panel s evaluation. The reasons for this were (i) the Events were considered to be of core importance to answer the question and (ii) the information provided in the general part 4 Federal Register, Volume 71, No 147, August 1, 2006, Rules and Regulations, Page Hereinafter referred to as APHIS evaluation/document 6 The international standard ISPM No. 5 defines systems approach(es) as The integration of different pest risk management measures, at least two of which act independently, and which cumulatively achieve the appropriate level of protection against regulated pests (p. 16). The international standard ISPM No 14 provides guidelines for the development and evaluation of integrated measures in a systems approach as an option for pest risk management ( ) (p. 165). A systems approach integrates pest risk management measures to meet the appropriate level of phytosanitary protection of the importing country (p.166). 7 APHIS document p. 5 8 Hereinafter referred to as Event(s) 5 of 41

6 (Introduction, Biology and Epidemiology, Spread Potential) was repeated in the discussion of each Event. For the purpose of this opinion and to help the reader, the comments are presented in the following sections in accordance with the structure of the APHIS document. Those paragraphs, statements and section names that received comments have been numbered and highlighted in bold italics and are followed by the related comments. EVENT 1 INFECTED/CONTAMINATED FRUIT IS HARVESTED [P. 8-10] GENERAL COMMENTS In Event 1, the Authors set out to analyse the disease threat posed by fruit harvested from citrus groves which are infected by or in the locality of a Xanthomonas axonopodis pv. citri outbreak. They conclude that there is a risk of fruit from such groves being infected or contaminated by the organism. They then set out the following arguments, which mitigate against the epidemiological significance of such fruit: - That groves are routinely sprayed with bactericidal copper sprays which significantly reduce the incidence of the disease; - That mature fruits which do not show symptoms of the disease are unlikely to be infected by the organism; - That epiphytic Xac populations on mature fruit decline rapidly after harvest, and - That commercial picking and sorting operations eliminate fruit, which may be infected by Xac, including those which possess uncharacteristic symptoms. The Authors conclude that canker infected fruit is likely in an infected orchard but that the occurrence of infected fruit will be epidemiologically insignificant. From the analysis of the references quoted to support these arguments the following criticism arises: - The reliability of copper sprays has not been demonstrated by scientific research. Copper sprays are one component of the systems approach but the Authors present the case that copper sprays alone are sufficient to control the disease. The efficacy of the systems approach depends on the effective implementation of all its components. The bactericidal efficacy of copper will vary with its formulation but no formulation is capable of eradicating the bacteria. - Asymptomatic mature fruit are unlikely to be a major source of inoculum for new infections however their significance has not been determined. No study has been undertaken to assess the epidemiological significance of asymptomatic fruit where disease pressure is low. There is evidence to suggest that mature fruit may be infected 9 Hereinafter: the numbers in square brackets refer to the page numbers in the APHIS document 6 of 41

7 without expressing symptoms. The concept of asymptomatic fruit is not clearly defined. Quantitative data are needed to assess the risk of asymptomatic fruit. - The evidence presented to support the contention that epiphytic populations of Xac decline rapidly is tenuous in that the Authors quote only one reference, which is derived from an abstract. Not withstanding this, it is entirely probable that the ability of the organism to survive epiphytically is low. This would be typical of many plant pathogenic bacteria. The Authors do not address the issue of survival of the organism on sites where bacteria are protected (wounds, etc.), which may be of a major epidemiological significance. The Authors present no evidence to support their claim that commercial handling of fruit eliminates diseased fruit. It is unlikely that all diseased fruit can be eliminated by these methods. Complete removal of fruit with lesions, which may measure less than 1 mm in diameter, is unlikely. Skaria et al. (2003) noted concern among citrus growers and packers about what is termed delivery problems, including bruises, cuts and infections that occur at harvest and packing but which may not manifest themselves until delivery to the consumer. Blasco et al. (2003) (Abstr) tested a machine vision system for automatic quality grading of fruit, which achieved a blemish detection success rate of 86% on apples. Miller and Drouillard (2001) (Abstr) tested a multiple feature analysis for machine vision grading citrus in Florida and achieved overall correct classification of 98.5% for grapefruit and orange and 98.3% for tangerine. It is therefore likely that, if citrus fruit were permitted for export from areas infected with Xac, infected fruit would enter into commerce, all be it at low levels. 1.2 SPECIFIC COMMENTS [Page 8] The first event is assumed to be likely if fruit originates in areas where the disease occurs, but the magnitude of the hazard at this stage will depend in large part on the proportion of infected fruit and the nature of the contamination. Groves in infested areas may have citrus canker infected fruit at varying levels depending on the prevalence of inoculum, the susceptibility of the variety, climatic, environmental, and cultural conditions. Presence of the organism on fruit may be associated with lesions, injuries, or blemishes, or it may be epiphytic (contamination). The prevalence of infected or contaminated fruit will depend primarily on the variety, environmental conditions, and field treatment regimes. - This section does not include supporting references. - The term variety should be replaced by the term citrus species and cultivars. - The assumption that infected/contaminated fruit originates in areas where the disease occurs ignores the fact that long distance spread of the disease has been observed (Das, 2003). A definition of the term area is therefore required here. Additionally, spread by contaminated boxes, clothing, etc. may increase the potential spread of the organism beside that associated with weather phenomena. 7 of 41

8 - The Authors mention disease level but it is not clear if this refers to disease severity or disease incidence. The differential susceptibility of citrus species to Xac as a factor influencing infection prevalence is not listed. - That the disease may be associated with blemishes suggests that the organism may reside in fruit blemishes other than those characteristic of Xac. [Page 9] Copper protective sprays are a key tool for reducing the incidence of Xac infestation in the field (Stall et al., 1980; Leite and Mohan, 1990; Gottwald et al., 2002a). Well-timed field treatments significantly reduce the incidence of the disease and therefore the level of inoculum and the number of symptomatic fruit in the field. Commercial growers for the fresh fruit market routinely apply copper sprays to control citrus canker (Graham and Gottwald, 1991; Gottwald et al., 2002a; Graham et al., 2004b) and as a general measure to reduce the incidence of other bacteria and fungi that adversely affect the quality and shelflife of citrus fruit. - The statement Well-timed field treatments significantly reduce the incidence of the disease does not include the level of reduction and therefore is insufficient to make a conclusion on the efficacy of this treatment. The statement is not referenced. - Copper is less effective when (i) the disease severity is high and there is therefore a high inoculum pressure (Stall et al., 1980; Stapleton and Medina, 1984), (ii) the citrus species and cultivars are susceptible to Xac (Stall et al., 1980; Stall and Seymour, 1983; Stapleton and Medina, 1984), (iii) there is an Asian citrus leaf miner (Phyllocnistis citrella Stainton) infestation (Schubert et al., 2001), (iv) fruits are at the susceptible stage of growth (2-6 cm) (Stall et al., 1980), (v) strains resistant to copper have developed (Canteros, 2004), and (vi) bacterial invasion occurs quickly, for example as water soaks the leaves when rain falls during windy conditions (Stall et al., 1980). - Gottwald et al. (2002a) mention the role of copper sprays but they reported that these sprays are effective as a stand-alone method on disease resistant or moderately resistant cultivars only. Additionally, they concede that this is based on work done in Brazil, which they consider to have a climate less conductive to citrus canker than Florida. - Gottwald et al. (2002a) and Graham et al. (2004b) are review articles but they do not take into account the USDA s own assessment that copper compounds, which have been used to treat citrus canker symptoms, are now known to mask evidence of the disease (Anon., 1999). - Graham et al. (2004b) mentioned the occurrence of copper resistance in xanthomonad populations and the negative environmental consequences of excessive copper use. Copper resistance was also noted by Canteros (2004) and by Schubert et al. (2001). 8 of 41

9 - The routine use of copper bactericides for the control of citrus canker is not mentioned by Graham and Gottwald (1991). Their paper concerned the use of copper sprays to limit the spread of the organism within orange groves. They found that, when used on their own, copper compounds only slightly reduced the spread of the organism. In addition they did not demonstrate that there was a significant difference in the spread of the organism between copper-treated and untreated groves Citrus canker is mainly a leaf spotting and fruit blemishing disease. Symptoms of citrus canker on fruit are visually obvious except when uncharacteristic lesions may be associated with injuries or blemishes. Despite decades of commercial and scientific observations, there is no evidence of the existence of undetectable lesions on mature fruit or latency which would result in lesions being expressed after harvest. Mature (not expanding) asymptomatic fruit without injuries or blemishes are not known to express symptoms (Graham et al., 1992b). - In this section the Authors refer to decades of commercial and scientific observations to support their claim that undetectable lesions on mature fruit do not exist. However, they do not give any of these sources of information. - The statement attributed to Graham et al. (1992b) that Mature (not expanding) asymptomatic fruit without injuries or blemishes are not known to express symptoms was not made in their paper. However, the results presented by Graham et al. (1992b) show that lesions did develop on mature Valencia orange inoculated with 10 8 cfu/ml. It is noteworthy that this work was done using a Xanthomonas species other than Xac. - It is not entirely correct that there is no evidence of absence of latency on fruit (Vernière et al., 2003) Resistance of fruit to infection by citrus canker is correlated with its stage of maturity. The younger the fruit, the less resistant it is to citrus canker (Graham et al., 1992b). Persistence of the citrus canker pathogen depends on where the canker lesions are located (Schubert et al., 2001). Fruit and leaves that are removed from trees sustain bacterial populations in lesions for 1-2 months in the presence of antagonistic microorganisms (Schubert et al., 2001; Timmer et al., 1996). - Xac was reported to survive for up to 4 months (120 days) in the section of the APHIS document entitled Biology and Epidemiology. - The Authors state that, according to Graham et al., (1992b), fruit susceptibility is inversely related to its age. However, Graham et al. (1992b) state that very young fruit (<20mm diameter) are more resistant to infection than older fruit. In addition, Koizumi and Kuhara (1982, in Gottwald et al., 2002b) stated that mature fruit can be infected but without subsequent lesion development. 9 of 41

10 - Schubert et al. (2001) did not state that the longevity of the citrus canker pathogen was related to the location of the lesion but they suggested that inoculum associated with exposed, symptomless citrus leaves can survive and remain infective for at least several months, perhaps located within older, more disease-resistant leaf and stem tissues. - Schubert et al. (2001) reported that the decline in bacteria populations on fallen tissue was due to the action of saprophytic and not of antagonistic micro-organisms, as stated in the APHIS document Xac is considered a relatively labile bacterium. Epiphytic populations of Xac on citrus tissue drop 3-5 orders of magnitude in 24 hours during experimental testing (Timmer et al., 1996). Timmer et al. (1996) also states, We detected epiphytic [Xac] on asymptomatic plants, but the occurrence of epiphytic populations was not related to subsequent appearance of symptoms, and additionally, Our evidence indicates that [Xac] is highly unlikely to persist on hosts or non-hosts in the absence of symptoms for long periods. Populations decline rapidly even within the lesions of infected fruit after harvest (Civerolo, 1981). Researchers in Argentina sprayed asymptomatic fruit with a bacterial suspension of 10 6 cfu/ml resulting in non-recovery of inoculated bacteria after five days at room temperature under lab conditions (Risk Assessment, 2005). The survival of epiphytic Xac populations on asymptomatic fruit under natural conditions would be expected to be further reduced due to natural environmental conditions being harsher than controlled laboratory conditions. - Civerolo (1981) reported that the bacterial population declines rapidly in dropped and not in harvested fruit, as claimed by the Authors. Fruit drop occurs as a result of high levels of citrus canker and the dropped fruit may not be at the same developmental stage as harvested fruit. In the APHIS document section Biology and Epidemiology, it is mentioned that the bacteria survive up to 120 days in fruits dropped on the soil. - Risk Assessment (2005) refers to an undated abstract by Belasque and Rodriguez Neto. There is therefore no way of evaluating their methods or the conclusions drawn. - Stapleton (1986) reported that bacteria swabbed onto wounded citrus fruit survived for two weeks in the open air. This is significantly longer that the 5 days reported in Risk Assessment (2005). It should also be noted that several authors have reported the importance of the bacteria penetrating the plant tissue through either stomatal openings or wounds (Gottwald and Graham, 1992; Graham et al., 1992a and b; Koizumi and Kuhara, 1982 in Gottwald et al. 2002b). Clothing, pruning equipment and boxes are also recognised as means of disease spread. The significance of the survival potential of bacteria on the fruit surface is therefore questionable. - According to Timmer et al. (1996), occasional infections may occur from epiphytic populations of Xanthomonas campestris pv. citri (=X. axonopodis pv. citri), although there is little evidence that casual epiphytic bacteria are an important inoculum source of 41

11 - The final sentence may also be questioned in that, it is recognised that bacteria will have difficulty in surviving on unprotected sites. Of more epidemiological concern is the scenario where bacteria are located into protected sites where their longevity and their subsequent potential to cause infection on other host material is significantly increased. The Authors also do not take into account the extracellular polysaccharides produced by Xac, which protect the bacterial cells from desiccation in air (Goto, 1985 in Das, 2003; Goto and Hyodo, 1985) In commercial operations, diseased, damaged, disfigured, and blemished fruit are culled in both the field and packinghouse. Because the survival of bacteria on non-attached citrus fruit is significantly reduced, the likelihood of spread of citrus canker through bacteria on harvested commercial fruit is considered very low (Graham et al., 1992b). Furthermore, no infection was observed when citrus fruit artificially inoculated with Xanthomonas axonopodis pv. citri were used as an inoculum source, even under ideal environmental conditions (E. Civerolo, personal communication on unpublished research). - The Authors present no data to support their contention that culling is an efficient measure to eliminate diseased fruit. The occurrence of asymptomatic infected fruits decreases the efficiency of the field culling and there are no quantitative data on the efficiency of field culling to discard fruits with symptoms. In addition, copper treatments may decrease the efficiency of field culling by masking symptoms on fruits (Anon., 1999). - The statement that In commercial operations, diseased, damaged, disfigured, and blemished fruit are culled in both the field and packing house indicates that field culling alone is not efficient enough to prevent infected fruit being harvested. Field culling and packing house processing are not perfect and there is a probability of fruit with symptoms to go undetected. - Graham et al. (1992b) did not study the infective potential of diseased or contaminated fruit. They did not address the issue attributed to them in the APHIS document and they did not attempt to quantify the likelihood of spread of citrus canker from harvested contaminated fruit. - The Authors do not explain the meaning of the term non-attached fruit. - There is evidence of survival of Xac for 120 days on fallen fruit (see section of the APHIS document entitled Biology and Epidemiology ). - Civerolo (pers. comm.) is quoted as reporting that artificially inoculated fruit were not an infection source. However, no data are presented to support this critical contention Conclusions for the first event: 11 of 41

12 The presence of canker infected fruit in an infected orchard is likely; however the prevalence of diseased fruit and of healthy fruit with epiphytic Xac will be epidemiologically insignificant. Xac is likely to be present in groves if active infections occur within the export grove or in nearby groves from which the bacteria may be introduced by wind driven rain. - The meaning of the term epidemiologically insignificant is unclear. In situations such as that in Florida, where the disease is widespread, the epidemiological role of infected fruit may be insignificant relative to the potential for spread of other inoculum sources. However in a Xac free environment, the introduction of the organism, either in a lesion or epiphytically on fruit surface, cannot reasonably be considered unimportant. Evidence: [Page 10] Under conditions of wind-driven rain (greater than 8 m/sec) bacteria are dispersed within trees and from tree to tree (Timmer et al., 2000). The bacteria have been dispersed from 32 meters to a few hundred meters to several miles (Stall et al., 1980; Gottwald et al., 1988, 1992, 2002b; Timmer et al., 2000). In Florida, the pathogen spreads less than 1900 ft from the source 95% of the time (Gottwald, 2002a; Gottwald et al., 2001; Compton and Fagan, 2000; McElroy, 2000). - Wind-driven rain (greater than 8m/sec) is not essential for the dispersal of bacteria within trees as the bacteria can be either washed-off from the higher parts and infect lower parts of the tree or be splash-dispersed onto the tree canopy by the irrigation water. - Concerning bullet point 3, it is assumed that Gottwald (2002a) should be Gottwald et al., (2002a). Gottwald et al. (2002a) and Gottwald et al. (2001) are a review article and a letter to the editor, respectively. Compton and Fagan (2000) is a press release from the Florida Department of Agriculture. In all cases these reports refer to conditions specific to Florida Under the systems approach, the prevalence of citrus canker diseased fruit will be low. - The APHIS document does not propose the full implementation of the systems approach in particular the fact that the fruit must be grown in a canker-free planting 12 of 41

13 stocks surrounded by a citrus-free buffer zone. See also general comments, section 1.1 above. Evidence: Copper protective sprays are a key tool for reducing the incidence of Xac infestation in the field (Stall et al., 1980; Leite and Mohan, 1990; Gottwald et al., 2002a). Prophylactic sprays of copper oxychloride or other copper-containing compounds provide protection against initial infection in canker-endemic areas during growth flushes and fruit development (2-6 cm diameter) (Stall et al., 1980; Leite and Mohan, 1990; Graham et al., 2004a; Muraro et al., 2001; Das, 2003). - For bullet point 1 see section 1.1 above. - Although copper sprays have some protective effect, according to Graham et al. (2004b), their efficacy varies from moderate to high, depending on the disease pressure. When the disease pressure was low, there was no significant difference in canker development between treated and untreated plots. However, copper sprays had a significant benefit in reducing fruit drop and subsequent losses of fruit destined for juicing. Graham and Leite (2004a) also mentioned the possibility of development of Xac resistant strains following repeated copper treatments Infected fruit express detectable symptoms and are likely to be culled or not harvested - Comments made in Section above are also valid for this statement. Evidence: Citrus canker is mainly a leaf spotting and fruit blemishing disease. Symptoms of citrus canker on fruit are visually obvious except when uncharacteristic lesions may be associated with injuries or blemishes (No reference cited in document). Mature (not expanding) asymptomatic fruit without injuries or blemishes are not known to express symptoms (Graham et al., 1992b). In commercial operations, diseased, damaged, disfigured, and blemished fruit are culled in both the field and packinghouse (No reference cited in document). - The Authors present no scientific data to support bullet point of 41

14 - For bullet point 3 see Section above Fruit contaminated by epiphytic populations of Xac is unlikely to have a significant role in disease spread. - The Authors present no references to support their statement that contaminated fruit are unlikely to have any epidemiological significance. However, Schubert et al. (2001) reported that some Citrus and Poncirus species, which can serve as hosts of the organism under experimental conditions, may be troublesome in relation to eradication or suppression programmes. Evidence: Xac is considered a relatively labile bacterium (No reference cited in document). Epiphytic populations of Xac on citrus tissue drop 3-5 orders of magnitude in 24 hours during experimental testing (Timmer et al., 1996). Timmer et al. also states, we detected epiphytic [Xac] on asymptomatic plants, but the occurrence of epiphytic populations was not related to subsequent appearance of symptoms, and additionally our evidence indicates that [Xac] is highly unlikely to persist on hosts or non-hosts in the absence of symptoms for long periods. Researchers in Argentina sprayed asymptomatic fruit with a bacterial suspension of 10 6 cfu/ml resulting in non-recovery of inoculated bacteria after five days at room temperature under lab conditions (Risk Assessment, 2005). - No reference is given to support bullet point 1. - For bullet point 2, see comment in bullet point 4 in Section above. - The reference in bullet point 3 Risk Assessment (2005) refers to an undated abstract by Belasque and Rodriguez Neto. There is therefore no way of evaluating their methods or the conclusions drawn. EVENT 2 INOCULUM ON INFECTED/CONTAMINATED FRUIT SURVIVES THE PACKING/TREATMENT PROCESS [P ] 2.1 GENERAL COMMENTS The Authors claim that the treatments and packing house processing will remove infected fruit and eliminate inoculum associated with asymptomatic and contaminated fruit. They present the following arguments to support their position: 14 of 41

15 - Symptomatic fruit will be removed by field culling and mechanical and/or visual screening. - Physical treatments applied to the fruit (washing, brushing and drying) will reduce bacterial populations. - The use of disinfectants (SOPP 10 and chlorine individually or in combination) will reduce bacterial populations to the point of complete elimination, and - The application of a finishing wax will reduce the likelihood of any remaining bacteria causing new infections. They conclude that the likelihood of bacterial populations on infected/contaminated fruit surviving the packing house and treatment process is very low. The Authors do not present any verified scientific data to support the accuracy and reliability of infected fruit removal by field culling and mechanical/visual packing house inspections. Although it is likely that these methods remove a substantial amount of symptomatic fruit, they cannot reasonably be expected to remove all such fruit. This is especially true when lesions are very small (i.e. <1 mm diam.) or copper compounds have been used in field treatments (masking the symptoms). There is evidence that suggests that blemished fruit do enter into commerce despite such controls (see comment bullet point 6 to statement 2.2.2). The Authors did not take into account the likelihood of a significantly higher amount of Xac on fruits collected from infected areas than those collected in orchards free of Xac. Physical treatments (washing, brushing and drying) most likely reduce bacterial populations on individual fruit but the significance of this has not been established. Moreover, the effect of physical treatments on bacterial populations within canker lesions is unclear. Finally, it is not known what proportion of the harvested citrus fruit is subjected to these treatments. The efficacy of disinfectant treatments appears quite variable and does not achieve the complete eradication claimed by the Authors. In some cases, disinfectant treatments only reduced bacterial populations by 77% (Stapleton, 1986). Additionally, bacterial populations were found to survive at chlorine concentrations of 900 ppm, well in excess of the 200 ppm used commercially (Stapleton, 1986). The efficacy of disinfectant treatments is dependant on a number of factors including ph, disinfectant concentration, presence of organic matter, and frequency of renewal of the disinfectant solution. Even when these factors are optimized, bacteria have still been shown to survive. The efficacy of wax treatments in reducing bacterial populations has not been extensively studied and only one reference (Stapleton, 1986) was presented by the Authors. This reference did not establish that waxing significantly reduced the risk of disease spread from contaminated fruit. 10 Sodium o- phenylphenate 15 of 41

16 2.2 SPECIFIC COMMENTS [Page 10] The EFSA Journal (2006) 439, The second event is the likelihood of inoculum surviving the packing and treatment process. Harvested fruit is transported to the packing house where it is subjected to one or more visual inspections. Field culling would be expected to have removed nearly all symptomatic and blemished fruit prior to arrival. - The APHIS document does not describe in detail what training, if any, is given to citrus fruit pickers and the staff of packing houses for detecting and culling fruit with canker symptoms. Schubert et al. (2001) detail the training given to field inspectors of APHIS as part of the citrus canker control programme. Field culling is based on visual inspection and, as it happens with all human eye inspections, there is a percentage of failure, which is not mentioned in the above statement. Moreover, inspection must be carried out by well-trained staff and this is unlikely in most citrus groves at harvest. It is also worth mentioning that, according to Canteros (2004), symptoms will be visible to an untrained person only days after infection. Furthermore, as field culling is not only a time-consuming but also a costly procedure, there is no guarantee that this process is followed in all citrus groves at harvest. - Symptomatic fruit that have been treated with copper compounds in the citrus groves may go undetectable, as copper masks symptoms (Anon., 1999). Moreover, according to Schubert et al. (2000) copper sprays should be avoided near infested areas, as copper only suppresses infection rather than eradicating the disease, thereby making canker detection more difficult for the inspectors. - Field culling cannot remove asymptomatic fruit. - Mature, asymptomatic but infected fruit may also be undetectable during field culling as, according to Vernière et al. (2003) and Graham et al. (1992b), the incubation period (time between infection and appearance of symptoms) for mature oranges and limes inoculated with 10 8 cfu/ml of Xac was 56 days and for mature grapefruit was 18 days. During these periods infected citrus fruit may be undetectable during field culling. - According to Koizumi and Kuhara (1982) in Gottwald et al. (2002b), mature citrus fruit could be infected without developing lesions. These fruit will also be undetectable during field culling. - No scientific data are presented to support the statement that Field culling would be expected to have removed nearly all symptomatic and blemished fruit Visual and/or mechanical inspections remove any remaining symptomatic or blemished fruit as well as misshapen, off-color, damaged and otherwise low quality fruit of 41

17 - Visual inspection for the detection of Xac symptoms must be carried out by welltrained inspectors, as citrus fruit will usually have blemishes, injuries or lesions caused by parasitic and/or non-parasitic pests and disease. Some of these symptoms are similar to those caused by Xac and would be difficult for untrained staff to differentiate. - There is no evaluation of the efficacy of each procedure (visual or mechanical inspection). As with all inspections it is certain that both procedures have some degree of error. - The APHIS paper does not detail what official controls are in place at packing houses to ensure the elimination of symptomatic fruit. - Observations made on fruit imported into some EU Member States would question the efficiency of these visual and/or mechanical inspections, as these fruit quite often have blemishes, injuries or lesions. - Visual and/or mechanical inspections in the packing houses will not remove asymptomatic fruit that carry epiphytic bacteria (contaminated fruit). - Skaria et al. (2003) noted concern among citrus growers and packers about what is termed delivery problems, including bruises, cuts and infections which occur at harvest and packing but which may not manifest themselves until delivery to the consumer. - Blasco et al. (2003) (Abstr.) tested a machine vision system for automatic quality grading of fruit, which achieved a blemish detection success rate of 86% on apples. Miller and Drouillard (2001) (Abstr.) tested multiple feature analysis for machine vision grading of citrus in Florida and achieved overall correct classification of 98.5% for grapefruit and orange and 98.3% for tangerine. It is therefore likely that, if citrus fruit were permitted for export from areas infected with citrus canker, infected fruit would invariably enter into commerce, all be it at low levels. [Page 10/11] The efficacy of this process varies slightly depending on the facility and system, but APHIS experience in the United States, Argentina, Japan and other countries shows that it can be highly effective for removing symptomatic fruit. - No information is given on the existence of such facilities globally or specifically in all packing houses in Florida. - That this process varies implies that its efficacy may also vary. - This process cannot remove asymptomatic fruit contaminated with Xac. [Page 11] 17 of 41

18 2.2.4 Citrus fruit is washed when it enters the packing house as a normal part of any packing process (see also Figure 4). This washing usually includes mechanical brushing and detergent (at the very least) which will reduce epiphytic bacterial populations (Schubert et al. 1999) by 1 to 2 orders of magnitude. - That the washing usually includes mechanical brushing and detergent use does not mean that this is a standard procedure followed in all exporting packing houses. No precise estimation of the bacterial population that remains on the fruit surface after washing is given. Moreover, it is not quite clear what a reduction by 1 to 2 orders of magnitude means, as the bacterial population that remains on the fruit surface after the washing is highly dependent on the initial epiphytic bacterial population. - The statement that washing reduces epiphytic bacteria might be true. However, washing does not eliminate bacterial population as, according to Stapleton (1986), both intact and wounded fruit surfaces may provide bacteria with protective sites, allowing them to avoid the effects of packing house treatment The drying process (with hot air at 58 0 C for 2.5 minutes) after washing further reduces populations of epiphytic Xac (Schubert et al., 1999). - Schubert et al. (1999) provided no data on the efficacy of the drying method on the level of reduction. Moreover, the experiment was performed with X. axonopodis pv. citrumelo (another citrus pathogen), not Xac. - The drying process is not performed when a water-based wax treatment is applied (Schubert et al., 1999). - According to Goto (1985) (in Das, 2003) It is well known that Xac bacteria produce extracellular polysaccharides (EPS) both in culture and in host tissues and that the bacterial cells on canker lesions are embedded in a dense matrix of EPS (Goto, 1985 in Das, 2003). These molecules exhibit great protective effects against dilution effect in water and desiccation in air providing benefits for the bacterial ecology (Goto, 1985 in Das, 2003). Therefore, it is possible that the drying process may not be so effective in reducing epiphytic Xac The application of a finishing wax further reduces the likelihood that any remaining bacteria will be available to cause new infections (Stapleton, 1986). - No precise estimation of the reduction in bacterial population by the finishing wax is given. - No information is given on the concentration of the finishing wax used in packing houses. According to Stapleton (1986), the effectiveness of the finishing wax in inhibiting bacterial growth depends on its concentration. He showed that the wax 18 of 41

19 product Citrus Lustr 266 at concentrations lower than 105 ppm had no effect on growth of the T20Rr strain of Xac in solid or liquid shake culture. - Stapleton (1986) also demonstrated that scanning electron micrographs of Mexican lime fruit surface treated with wax showed that a relatively heavy coating was formed that became extensively cracked after application compared with the natural fruit cuticle. The cracking prevented complete sealing of the fruit surface by the wax layer. - According to Stapleton (1986), there is no difference in the bacterial populations from the wax treated and untreated lots When the washing process includes a disinfectant such as 200 ppm chlorine or SOPP, Xac populations are reduced to undetectable levels (Brown and Schubert, 1987; Graham and Gottwald, 1991; Obata, 1969). - Graham and Gottwald (1991) is a secondary source. The primary source is Brown and Schubert (1987). - In contrast to the APHIS statement, the treatment with 200 ppm chlorine is effective in reducing the number of bacteria but not to undetectable levels as, according to Brown and Schubert (1987), 200 cfu of Xanthomonas campestris pv. vesicatoria (X. axonopodis pv. vesicatoria) were detected on fruit even after having been sprayed with 200 g/ml chlorine for 2 min. - Brown and Schubert (1987) gave little information on the comparability of X. campestris pv. vesicatoria and Xac. - The rating of 0 in Brown and Schuber (1987; Table 5, SOPP soap treatment) does not mean that no bacteria were detected, as rating 0 represented 0-5 cfu of bacteria. According to Tables 1, 2 and 5 from Brown and Schubert (1987), X. c. pv. vesicatoria bacteria were still detected at ratings >0 (zero) on citrus fruit following treatment with 200 ppm chlorine or 20,000 ppm SOPP or 200 ppm chlorine dioxide. - The cited reference by Stapleton (1986) also contradicts the above statement, as in Figure 1A it is shown that total bacteria were reduced by 77% (23% survived) to 99% after dip-tank or in vitro NaOCl treatment ( ppm Cl) of Mexican limes. - Stapleton (1986) recovered live bacteria from commercial dip-tank solutions and found that 2.7x x10 3 cfu/cm 2 of epiphytic bacteria survived dip treatment containing Cl at concentrations above the recommended 200 ppm level. - Stapleton (1986) reported that the increase of the recommended 200 ppm Cl concentration to improve fruit disinfection would not be effective, as according to the results of his experiments, Cl concentrations as high as 900 ppm did not eradicate lime epiphytic bacteria. Furthermore, increase of Cl concentration above 500 ppm may cause fruit damage (Schubert et al., 2000; in: 2000 Florida Citrus Pest Management Guide: Citrus Canker, p ). Stapleton (1986) also noted that if fruit treatments 19 of 41

20 are desired for regulatory purposes to minimize the possibility of pathogen dissemination on fruit, the effects of disinfecting agents more active than NaOCl should be tested, both in dip-tanks and when added to protective wax formulations. - The effectiveness of the chlorine treatment seems to depend on the amount of organic matter present on the fruit surface (Dychdala, 1983). This seems to be confirmed by the results of Brown and Schubert (1987) according to which, much of the applied chlorine (41 μg/cm 2 out of 45.4 μg/cm 2 of fruit surface) was inactivated by the large amount of organic matter commonly present on Florida citrus fruit. - That the Xac population was reduced to undetectable levels means that no bacteria were cultured after the treatment. However, Cubero and Graham (2002) (Abstr.) indicated that a proportion of Xanthomonas axonopodis pv. citrumelo population appeared viable by LIVE/DEAD kit, but non-culturable for several days. Therefore, it is possible that bacteria can enter a viable but non-culturable state, which can be induced by adverse conditions that result in a starvation phase from which bacteria may be resuscitated (Cubero and Graham, 2002). - For Cl treatments to be effective, they must be applied under certain conditions, such as (i) the ph of the water should be maintained within a range of 6 to 7.5, (ii) the concentration of Cl should be maintained within a range of 150 to 250 ppm, and (iii) the wash water/sodium hypochloride solution should be changed daily or more frequently, as dirty water reduces free/available Cl (Schubert et al., 2000). However, there is no evidence that these procedures (guidelines) are followed by all exporting packing houses. - Schubert et al. (2001) reported on the use of SOPP or chlorine as post-harvest treatments of fruit from groves certified as being canker free but in canker affected regions. Notwithstanding this, such fruit were not permitted for export to citrus producing areas free of the disease. No new evidence is presented to show why these restrictions are no longer considered necessary by APHIS. - Obata (1969) is not listed in the literature cited. Instead there is a reference of Obata et al. (1969) The post harvest treatments such as chlorine and SOPP guarantee the complete eradication of epiphytic Xac on fruits without symptoms (Canteros, 2004). - The above statement contradicts Brown and Schubert (1987) (see comments to statement under point bullet point 4). - Canteros (2004) reported on the efficacy of combined disinfectant treatments on Xac, but his results were based on artificially inoculated fruit and in vitro tests. - It cannot be concluded from Canteros (2004) that complete eradication is guaranteed. Canteros (2004) is a review article that does not include such a statement. According to Canteros (2004), both treatments together (SOPP+Sodium hypochloride) killed high 20 of 41