NZ Comments on Apple Draft IRA - 3. Assessment of risk

3.1 Erwinia amylovora

3. Assessment of risk

In this section New Zealand presents comments on the scientific assessment of risk carried out for each of the quarantine pests. These are arranged by pest.

3.1 Erwinia amylovora

AFFA has concluded that the risk of entry, establishment and spread of Erwinia amylovora is moderate. New Zealand challenges this assessment on the basis that both the probability and consequences of entry, establishment and spread have been over-estimated. A number of factors have led to this over-estimation, including misinterpretation of published literature and selective citation of expert opinion.

Our comments largely follow the layout of the draft IRA, and provide corrections to misinterpretations of the literature, provide additional data not referenced by AFFA, and provide an alternative assessment using AFFA's terminology. In particular New Zealand is concerned that AFFA has not assessed separately the four key risks that they seek to manage. These risks are identified as:

bacterial infection of mature fruit in orchard or after harvest;
infestation of the calyx-end of the fruit;
epiphytic contamination of fruit surfaces; and
the presence of trash with imported fruit.

A separate assessment of these risks will, New Zealand believes, demonstrate that the measures proposed by AFFA are not justified.

New Zealand believes that a re-evaluation of the literature in response to our comments in the following sections will lead to an assessment that the probability of entry, establishment and spread of fire blight is negligible, that the consequences are moderate, and that the overall risk is therefore also negligible (refer Table 9 of the draft IRA).

3.1.1 Probability of entry I - importation potential

In New Zealand's view, AFFA's assessment that the probability of importation is "high" is considerably overstated for the reasons set out below. In New Zealand's view the probability of importation of fruit being infected or infested on the surface is negligible and the probability of epiphytic infestation of the calyx of any apple from New Zealand is "very low".

3.1.1.1 Importation step 1: source fruit

In assessing the risk that mature apple fruit from any orchard in New Zealand may carry Erwinia amylovora, AFFA has concluded that "at least some" fruit would carry bacteria. There is considerable published literature available that would assist AFFA to refine this estimate either in numeric terms or in the same terms as used in AFFA's own risk criteria (as specified Table 6, p. 39).

New Zealand believes that, in assessing the risk that fruit may carry E. amylovora, it is important that the relative risks of fruit infection, calyx infestation, fruit surface infestation, and trash are considered separately. These assessments of risk are very important to the selection of risk management options later in the pest risk analysis. AFFA has considered separately the mechanism for infection and infestation, but has then combined the overall probability. New Zealand believes this makes the application of phytosanitary measures non-transparent, as it is difficult to identify the relationship between pest risk and the measure imposed. AFFA has stated (p.77, l. 4) that infection and infestation are two discrete biological processes that differ with regard to their implications for the quarantine risks associated with the importation of apple fruit from New Zealand.

New Zealand requests that AFFA provides a separate assessment of the risks associated with these biological processes.

i. Fruit infection

New Zealand believes that the risk of E. amylovora fruit infection (fruit from anywhere in New Zealand) is negligible. There is a considerable amount of published information on this subject, all of which indicates that fruit infection (with E. amylovora) is an extremely rare event. New Zealand requests that AFFA assesses this risk separately to other forms of fruit infestation / infection.

Fruitlets that are infected at blossom remain undersized (AFFA, 2000) or abort. Literature is presented in the draft IRA supporting the view that fruit infection is extremely rare (Roberts et. al., 1989; Dueck, 1974a; van der Zwet et. al., 1990). Sholberg et al. (1988) tested whole core samples of mature apples from four locations in British Columbia and failed to isolate E. amylovora even though adjacent pear trees were infected and bacteria were isolated from 100% of the leaves from the same trees. E. amylovora was not detected from internal tissues from trees where 20% of wood had fire blight symptoms (Dueck, 1974). In previous correspondence with AFFA, New Zealand has indicated that infections of apple fruit have never been reported in New Zealand.

Two papers were cited by AFFA as evidence of fruit infection, van der Zwet et al. (1990) and Clark et al. (1993). This work was reporting epiphytic infestation, not internal fruit infection. AFFA has mis-reported Clark et al. (1993). The authors did not detect E. amylovora in the calyxes of any fruit samples, even within 20 cm of the inoculation site.

Van der Zwet et al. (1990) recovered E. amylovora from the cores of 2-5% of mature fruit (harvested in August) collected within 15 cm of blighted shoots. It was unclear whether the isolation of E. amylovora was associated with symptoms, as the authors reported that "symptoms were difficult to distinguish from other fruit rots". Given that all fruit sections were routinely tested (regardless of the presence of symptoms of infection) it is likely that the isolations of endophytic E. amylovora were not instances of infection (disease).

It is stated by van der Zwet et al. (1990) that E. amylovora was recovered from up to 21% of the core sections of fruit harvested from within 15 cm of visibly blighted shoots. What is not clear is the stage of maturity of these fruit. Fruit was harvested in July and August. Given that the normal fruit harvest period is between late August and early October it is highly likely that the fruit collected in July were immature fruit. This is borne out by the decline in infection between July and August (Table 3 of van der Zwet et al. (1990)), indicating a maximum recovery of 5% of tissue samples in mature fruit collected within 15 cm of visibly blighted shoots.

Van der Zwet et al. (1990) also found E. amylovora in the internal tissues (core) of fruit sampled from blighted orchards in a number of regions of the USA. It is difficult to determine the percentage of fruit with E. amylovora as the data are presented as numbers of isolations from the upper core, core, and lower core and it is not stated whether these were the same, or different, fruit. The percentage of fruit with E. amylovora present was therefore between 1.5% (5/320) and 4.4% (14/320).⁵

This information is relevant to the assessment of the risk that fruit from anywhere in New Zealand may be infected. There is a <5% chance that fruit from within 15 cm of a blighted shoot would contain E. amylovora in the internal fruit tissues. Fruit from elsewhere on a tree would not be infected at all because "symptomless fruit from branches distant from the site of infection on the same tree will not be endophytically infected" (AFFA, 2000). The overall risk of fruit infection (or the presence of endophytic E. amylovora) is therefore considerably less than 5% unless every fruit was harvested from within 15 cm of blighted shoots.

Aldwinckle (pers. comm., 2000) pointed out (in answering AFFA's questionnaire) that endophytic infection could only be caused by resurgent calyx infection or multiplication of internally translocated E. amylovora cells (from a shoot or blossom infection) within the fruit. He said "I know of no published report or even rumour of either. I regard either as extremely unlikely to occur. Therefore I regard endophytically infected fruit as non-existent to date". New Zealand believes that published research supports Aldwinckle's views.

New Zealand asks that AFFA review the assessment of risk and suggests that on the basis of the available scientific evidence an appropriate assessment is that the probability of fruit infection is negligible.

ii. Infestation of immature apple fruit

In reviewing the literature it is evident that AFFA has made a number of errors when interpreting the conclusions of some authors. When these errors are corrected and additional literature not examined by AFFA is considered, the probability of infestation of apple calyxes is much lower than that assessed by AFFA. New Zealand believes that a more accurate assessment is that there is a "very low" probability.

Much of the fire blight research carried out in New Zealand has been done on immature fruit (fruitlets) because it is accepted that E. amylovora is more likely to be detected at this stage than it is when the fruit mature. As is correctly pointed out in the draft IRA the level of infestation declines significantly between fruitlet stage and maturity. When considering the risk associated with fruitlet infestation, this decline in infestation prior to maturity should be taken into account. In the literature cited in the draft IRA, the levels of immature calyx infestation ranged from 21% in orchards with 1-2 strikes per tree to 50% in a "severely blighted orchard". In this worst-case situation, the level of infestation had declined to 3% of mature fruit at harvest time.

In reviewing the data from orchards without symptoms AFFA has perpetuated a typographical error in Clark et al. (1993). The published Table 1, Orchard F cites 87/100 infested immature fruit. In the original publication (Hale and Clark, 1990) the correct figure is presented 87/1000 fruit, i.e., an 8.7% infestation - not 87%. AFFA has been notified of this by MAF.

When this is taken into account it is clear that levels of infestation of calyxes of immature apples range from approximately 0-9% in orchards without fire blight symptoms (but in close proximity to blighted trees) to 50% in orchards with severe fire blight.

New Zealand requests that AFFA correctly reviews all available literature, and believes that a more accurate assessment is that infestation of immature apple fruit has a very low probability.

iii. Infestation of mature fruit calyxes

The available literature relating to calyx infestation of mature fruit is considerable, however the data presented in the E. amylovora datasheet has only been selectively referenced in the risk assessment. In addition, clarification is required of some of the conclusions attributed to authors cited in this section.

Hale et al. (1987) recovered E. amylovora from 3% of the calyxes of mature apples from a severely blighted orchard by isolation (not PCR). As is stated in the draft IRA, PCR does not differentiate between live and dead cells. McManus and Jones (1995) did not recover live bacteria, so the 75% infestation level cited here is not relevant to

the draft IRA. The draft IRA states that Sholberg et al. (1988) recovered bacteria from 100% of mature fruit. This is incorrect. The authors recovered bacteria from 100% of apple and pear leaves (i.e., apple leaves and pear leaves), but no apple fruit. Van der Zwet et al. (1990) did recover bacteria from the calyx of mature apple fruit from orchards free of fire blight when severe fire blight was present in the area. New Zealand has been advised that a blighted orchard was located <10m from the fire blight free orchard in West Virginia (Roberts, pers. comm., 2000). The level of infestation reported by van der Zwet et al. (1990) is 2 infested fruit out of 40 sampled, or 5% of fruit.

It would be helpful if the work by van der Zwet et al. (1990) was more fully reported and discussed in the draft IRA. In all, 560 fruit were tested and the calyxes of only 5 fruit (<1%) were found to be infested. From the 4 blighted orchards a total of 280 fruit were tested and the calyxes of 3 fruit were infested and from the four apparently healthy orchards the calyxes of only 2 fruit were infested. In the data sheet AFFA reports an 8% recovery, however this is from all parts of the fruit. It is important that the affected part of the fruit is clearly identified. AFFA incorrectly reports that "bacterial numbers exceeded 103 cfu / fruit in the calyxes of fruit harvested from blight free orchards". The level of infestation was <50cfu.

Although not mentioned in the risk assessment, further literature is cited in the datasheet. It is stated that Covey (1975) recovered E. amylovora from mature apple fruit. This is incorrect. Covey's work was on pears.

In a severely blighted orchard, Hale et al. (1987) isolated bacteria from 3% of the calyxes of mature fruit. In a survey of 60,000 fruit over three seasons, Clark et al. (1993) did not recover E. amylovora from the calyxes of fruit from orchards free of fire blight symptoms. This is supported by extrapolation of other data (Hale et al., 1987) demonstrating that the percentage of infested fruit declines by 94% between fruitlet stage and harvest. Taking the reported (Hale et al., 1987) 6.7-8.7% infestation of fruitlets in orchards free from fire blight symptoms and applying this factor, New Zealand sees that only 0.05% of the fruit would have been infested at harvest.

Many authors (Hale et al. 1987; Dueck, 1974; Dueck and Morand, 1975; van der Zwet et. al., 1990) have reported a reduction in E. amylovora populations towards harvest. Dueck (1974) reported that "Apparently orchard conditions are unfavourable for the survival of E. amylovora toward the end of the growing season". Roberts et al. (1989) proposed a mechanism for the reduction in E. amylovora populations towards harvest. These authors believe that biotic factors such as naturally occurring biological control may explain the lack of recovery of E. amylovora from mature fruit. Strains of E. herbicola are known antagonists of E. amylovora and may act by antibiosis and pre-emptive site occupation (Roberts et al. 1989). This may explain why van der Zwet et al. (1990) were able to recover E. amylovora from a much higher proportion of surface disinfested fruit than from fruit from the same orchard sampled on the same dates that had not been surface sterilised.

The highest reported percentage of mature apple fruit with calyx infestation (as recovered by bacterial isolation) is 5% (van der Zwet et. al., 1990), however it is not clear whether the fruit harvested by these authors were all mature, as they were the same variety taken from the same orchard over a 3 month period. On the basis of these data, the risk that calyxes of apple fruit from orchards apparently free from fire blight is at most 5% (from US data⁶) or 0.05% considering New Zealand data⁷. Roberts et al. (1998) concluded that 0.35% of mature fruit (regardless of fire blight status) would be infested.

New Zealand believes the evidence suggests that it is very unlikely that mature fruit calyxes would be infested, and according to Table 6, this represents a very low probability.

iv. E. amylovora on the surface of mature fruit

AFFA has not provided an assessment of the probability that E. amylovora would occur on the surface of harvested fruit, however it is noted that E. amylovora is sensitive to UV light. In other sections of the draft IRA the fragility of E. amylovora is discussed. AFFA has stated that infestation of the calyx-end of mature fruit is of particular relevance to the draft IRA. It would appear, therefore, that the probability of E. amylovora being present on the surface of harvested fruit is considerably lower than in the calyx. The impact of these processes on the probability of survival of bacteria on the surface of fruit should be assessed.

It is New Zealand’s view that if assessed separately, the probability of E. amylovora occurring on the surface of fruit would be negligible, i.e., the event would almost certainly not occur.

v. Trash

In the very brief assessment of the risk that "trash" associated with apple fruit may be infected with E. amylovora it has been concluded that symptomless leaves and stems may be infected. No data have been presented to estimate the likelihood of this occurring, and no assessment has been made of how this "trash" might become associated with apple fruit. Despite this, phytosanitary measures are being imposed to manage the possibility of a risk.

New Zealand requests that AFFA undertake an assessment of the risk of trash being infected with E. amylovora.

(vi) Conclusions: Importation step 1: source fruit

AFFA has concluded that "it would be likely that at least some of these fruit would be either infested or infected with E. amylovora". Given that infection and infestation have been given separate consideration in the risk assessment, and that in the risk management section of the draft IRA AFFA has identified bacterial infection of mature fruit, infestation of the calyx-end of the fruit, epiphytic contamination of fruit surfaces and the presence of trash with imported fruit as risks that require management, New Zealand believes that these risks should be assessed separately. By doing so the application of phytosanitary measures to manage the risks becomes more transparent.

If this is done, the assessment would find that the probability of fruit being infected or infested on the surface is negligible ("the event would almost certainly not occur"). The risk of epiphytic infestation of the calyx of any apple from New Zealand is <3%, and New Zealand suggests that this represents "very low probability".

New Zealand requests that AFFA reviews the risk of bacterial infection of mature fruit, infestation of the calyx-end of the fruit, epiphytic contamination of fruit surfaces and the presence of trash separately. If this is done, New Zealand considers that the probability of fruit being infected or infested on the surface is negligible and the probability of epiphytic infestation of the calyx is very low.

3.1.1.2 Importation step 2: packing house procedures

(i) Cold storage

AFFA states that "Research regarding the efficacy of cold storage, as a means of reducing the viability of E. amylovora or the number of viable bacteria on fruit, is inconclusive". New Zealand believes that the available literature supports a view that cold storage acts to reduce E. amylovora populations on apple fruit.

AFFA has previously expressed concern that the number of fruit used by Hale and Taylor (1999) was insufficient. New Zealand wishes to point out that the numbers of fruit used by van der Zwet et al. (1990) were less than those used by Hale and Taylor (1999), however AFFA appears comfortable in accepting the paper by van der Zwet et al. (1990) in this and the preceding sections. New Zealand believes that AFFA should consider all available literature using common criteria for assessing the validity of conclusions reached by authors.

Hale and Taylor (1999) showed that after 25 days of cold storage the number of fruit with detectable levels of E. amylovora reduced in all cases, regardless of inoculation level. Naturally infested fruit reduced from 2% infestation at harvest to zero after cold storage (from trees with less than 5 strikes per tree).

Van der Zwet et al. (1990) found that E. amylovora persisted in cold stored fruit to a greater extent than that reported by Hale and Taylor (1999). The fruit in which E. amylovora was most persistent was collected from 1-3 cm from a blighted canker, and were endophytically infected (as stated by the authors). These results cannot be compared directly to those of Hale and Taylor (1999). The environmental conditions in the cortex of an apple are more favourable to E. amylovora survival than in the calyx. AFFA states that "van der Zwet et al. (1990) found bacteria retained in the calyx-end of mature fruit were not affected by cold storage", however the paper does not support this statement. The authors did not test the calyxes of the fruit and, as reported above, the authors acknowledge that most of the bacteria recovered were from endophytic infections. The levels of E. amylovora did not change during cold storage and, as the authors did not test the fruit at harvest, it cannot be determined whether the percentage of infected fruit found in cold stored fruit differed from the level that would have been present in the fruit prior to cold storage.

Sholberg et al. (1988) also reported a decline in the number of bacteria (as opposed to percentage infested fruit) in the calyxes and stems (not differentiated) of artificially inoculated apple fruit during cold storage at 2 - 4oC. These authors suggest that "possibly cold storage alone could be used as a method to assure countries free of fire blight that the apples they are importing are free of epiphytic E. amylovora".

McLarty (1926 - cited in Roberts et al. 1998) inoculated fruit in the calyx and cold stored the fruit. At the end of cold storage the author was unable to recover bacteria from any of the fruit.

The literature supports New Zealand’s position that commercial cold storage acts to reduce the risk that calyxes of mature the fruit are infested with E. amylovora.

ii. Water dump

AFFA has not considered the effect of immersing fruit in water as a means of reducing surface populations of E. amylovora. All fruit is immersed in water as part of normal pack house procedures and the fruit float to the grading and packing line. This process takes at least 5 minutes. In the same trial that was carried out for approval of the chlorine dip treatment for apples to Japan, a number of fruit were also immersed in the water dump before chlorine was added. This water treatment almost completely eliminated the bacteria from the surface of these fruit. In conducting laboratory experiments with E. amylovora it is necessary to use buffered saline to protect the bacteria from the osmotic effects of water. It is likely, therefore that the water dump has some effect on reducing epiphytic populations of bacteria on the surface of fruit.

New Zealand requests that AFFA review the practice of immersing fruit in a water dump as a measure to reduce epiphytic populations of bacteria on the surface of fruit.

(iii) Risk of cross contamination on the packing line

AFFA states that it is conceivable that 'clean' apples may be contaminated through contact with infected/infested apples, but provides no data to support this assumption. Accordingly, it appears that AFFA has given weight to a mere possibility in its assessment of probability at this step. This is inappropriate. No estimate of probability has been assigned. There is no literature demonstrating that cross contamination occurs. Despite this, AFFA is proposing phytosanitary procedures to reduce this perceived risk. AFFA has assigned a "high" likelihood that an infested/infected apple would remain so after pack house procedures.

New Zealand requests that AFFA assess the risk of cross contamination and provide justification for imposing the trade restrictive phytosanitary measures concerned (namely disinfestation of fruit and sanitation of the packing line).

3.1.1.3 Importation step 3: storage and transport

Nachtigall et al. (1985) reported survival of E. amylovora during cold storage, however these authors injected high levels of bacteria (109cfu) deep into the cortex of apple fruit, rather than inoculating the calyx of fruit. Injection of very high numbers of bacteria into the cortex causing wounding does not replicate the normal mode of fruit infection. Similarly, van der Zwet et al. (1990) reported the survival of endophytic populations of E. amylovora. In neither case, however did these authors report an increase in populations. Hale and Taylor (1999) and Sholberg et al. (1988) reported a reduction in populations of epiphytic bacteria during cold storage. It is reasonable to conclude that cold storage (in New Zealand and during transport) would result in a reduction in E. amylovora infection / infestation of fruit.

New Zealand questions AFFA’s allocation of a "high" probability to the likelihood of E. amylovora surviving storage and transport.

3.1.1.4 Importation step 4: on-arrival inspection

In this section AFFA states that "There are no published data regarding the likelihood that an apparently healthy apple will develop visible symptoms during the period of transport and storage". This is most likely because this is such a rare event. The preceding section cited some studies assessing this risk. However New Zealand agrees that it is virtually certain that on-arrival inspection would not detect infected or infested apples.

New Zealand considers that the proposed on-arrival inspection measure has no justification.

3.1.1.5 Conclusions: importation potential

AFFA has concluded that the importation potential is "high" - i.e., that the event is likely to occur. This assessment is one step greater than "moderate" which is the situation where there is a 50% likelihood of the event occurring (the event would occur with an even probability). This is a considerable overestimation of probability, due to misinterpretation of the literature, and failure by AFFA to consistently apply its methodology at each step along the importation pathway.

The probability of a mature symptomless apple fruit from any orchard in New Zealand being infected or infested with E. amylovora is very unlikely (considerably less than 5%). These levels of infection / infestation are further reduced during packing, cold storage and transport. When all of these events are considered, New Zealand believes that the importation potential of infected and surface contaminated fruit is negligible, and that the importation potential of calyx infested fruit is very low.

New Zealand believes that the importation potential is very low and requests that AFFA re-assess the probability of importation.

3.1.2 Probability of entry II- distribution potential

It is New Zealand's view that the distribution potential of E. amylovora is negligible. A number of factors in the distribution pathway (discussed below) serve to reduce risk. In particular AFFA has not properly taken into account the effect of exposure to the environment and the complete lack of a vector or means of transfer of E. amylovora from imported fruit to a host.

3.1.2.1 Distribution step 1: storage and distribution of apple fruit

AFFA states "there is no published information to suggest that cold storage and/or procedures involved in the sorting and distribution of imported apples would lead to a decrease in the number of viable bacteria". In previous sections New Zealand has suggested that the published information (Hale and Taylor, 1999; McLarty, 1926; Sholberg et al. 1988) indicates that cold storage results in both a reduction in the number of viable bacteria infesting or infecting fruit, as well as reducing the proportion of fruit with detectable levels of infection/infestation.

New Zealand questions whether the assessment of "high" probability is appropriate, and asks AFFA to reconsider the basis of this assessment.

3.1.2.2 Distribution step 2: discarded waste

There is no information to suggest that large numbers of E. amylovora are likely to be present in endophytically infected apples. The consumption of apple fruit is, in itself, a process that further reduces the likelihood that infected or infested fruit will be discarded into the environment. The number of discarded whole fruit (those discarded because of spoilage) is likely to be extremely low, so on the majority of sound fruit, any surface populations would be consumed, thus removing them as a risk factor.

Apple peelings and apple cores would be disposed of in a number of ways. Given that Australia has a high urban population, the majority of this waste would be disposed of in waste disposal units or buried in land fills. Only a small portion would be discarded into the environment. New Zealand requests that AFFA undertakes an assessment of the effect of waste disposal patterns and incorporates this into the risk analysis. Roberts et al. (1998) estimated that in Japan the probability of an imported apple fruit being discarded near a host was 0.0025. An estimate could be made in a similar way in Australia.

New Zealand submits that the process of consumption and disposal of waste significantly reduces the risk of fruit carrying E. amylovora being discarded into the environment. This is an important step on the importation pathway and New Zealand asks that it be carefully considered and an assessment of probability be assigned to this step.

New Zealand requests that AFFA assign a probability of risk to distribution through discarded waste.

3.1.2.3 Distribution step 3: exposure to the environment

Given that E. amylovora only multiplies epiphytically on the stigmas of hosts it is important that the fate of epiphytic and endophytic populations of E. amylovora are considered separately here.

i. Epiphytic populations

The number of E. amylovora in the calyx or on the surface of a discarded fruit (or peeling or core) will not increase during exposure to the environment. Epiphytic E. amylovora has been shown to multiply predominantly on the stigma of a host. The factors cited by AFFA in the draft IRA are all likely to lead to a rapid reduction in the survival of these bacteria. Hale and Taylor (1999) were unable to recover E. amylovora from fruit exposed to room temperature for 14 days (following a 25 day period of cold storage). This experiment was conducted under reasonably "clean" conditions with no exposure to the environment.

AFFA sought expert opinion on the likelihood of epiphytic E. amylovora surviving and multiplying in the environment, but for some reason did not cite these opinions. "Calyx populations are 1) infrequent, and 2) very low. They will be extremely low in the few infected calyxes of apple cores. They will decline to zero very quickly" (Aldwinckle, pers. comm., 2000). Beer (pers. comm., 2000) was unaware of any evidence for survival of E. amylovora in fruit calyxes after harvest or discard into the environment and suggested that any bacteria would survive only a few days. Pusey (pers. comm., 2000) said that the probability of E. amylovora surviving in the calyx of a discarded apple is near zero, as the conditions are unfavourable for E. amylovora and it rapidly dies. Zeller (pers. comm., 2000) stated that low concentrations of E. amylovora could survive in the calyx after harvest, but these could not have the capacity to initiate infection of a host. It seems reasonable, therefore, to conclude that bacteria present in the calyx of an apple, or on the surface, are very unlikely to survive exposure to the environment.

New Zealand requests that AFFA re-assess this step taking into account all information available, including the expert opinions.

(ii) Endophytic populations

As with epiphytic E. amylovora, there are no literature reports to support the multiplication of E. amylovora in discarded apple tissues. A number of environmental factors that would serve to reduce E. amylovora populations are discussed in the draft IRA. AFFA sought expert opinion on the effect of environmental factors on E. amylovora survival and the longevity of E. amylovora on discarded waste. The responses cited below indicate that endophytic (and epiphytic) E. amylovora would survive for only a short period of time. It is unclear why these expert responses were not fully considered in the risk assessment.

Aldwinckle (pers. comm., 2000) indicated that E. amylovora in infected or infested apple cores (regardless of the level of E. amylovora present) discarded into the environment would not have the potential to survive and be a source for spread of fire blight. Beer (pers. comm., 2000) stated that it was very unlikely that E. amylovora in infected/ infested apple cores (regardless of the level of E. amylovora present) discarded into the environment would have the potential to survive and be a source for spread of fire blight. Paulin (pers. comm., 2000) responded that these events could occur with low probability in suitable conditions and with very very low probability in unfavourable conditions. Pusey (pers. comm., 2000) concurred, stating that E. amylovora in infected/infested apple cores had no potential to survive and be a source of fire blight spread. Wimalajeewa (pers. comm., 2000) stated that E. amylovora could only multiply in a discarded apple core if nutrients became available with the breakdown of core tissue, but that it would not survive the fierce competition of saprophytic microflora invading the core. Zeller (pers. comm., 2000) said that it was very unlikely that that E. amylovora in infected/ infested apple cores (regardless of the level of E. amylovora present) discarded into the environment would have the potential to survive and be a source for fire blight spread.

On the basis of published literature and expert opinion New Zealand challenges AFFA’s statement that "the likelihood that E. amylovora would survive in the environment for a sufficient period, and be able to either multiply or persist in sufficient numbers to be transferred to a host in a receptive state is low". New Zealand believes that this risk is very low.

3.1.2.4 Distribution step 4: vectors and other means of transfer

AFFA has concluded that it is "extremely unlikely that viable E. amylovora would be transferred from either infested or infected apple tissue to an appropriate site on a susceptible host". MAF agrees with this statement. AFFA has not cited any published information on vectors or mechanisms for the transfer of E. amylovora from an infected / infested apple fruit to a receptive host. This is because no author has been able to find such a link. Because of the lack of vectors or means of transfer, the risk posed by imported apple fruit is negligible.

AFFA has referred to its previous risk assessment (AQIS, 1998) where it was suggested that browsing insects, or mites may be able to transfer bacteria to a receptive flower or wounded twig. That assumption or suggestion was not supported by any published data.

In this step it is also important to consider "infectious dose" or inoculum potential. Not only is it necessary for there to be a mechanism to transfer bacteria from an apple to a receptive host, there also needs to be a sufficient number of bacteria present in order for infection to occur. There is published research indicating that >104cfu of E. amylovora is required to initiate infection under field conditions (Hale et al. 1996). This is supported by other literature (Thomson et al. 1999) indicating that the rate of infection is lower at low inoculum levels that at higher levels. Hale et al. (1996) surface-inoculated apple fruit with E. amylovora (108cfu/ml) and placed them immediately adjacent to blossom clusters. None of the blossom clusters became infected.

It appears that the bacteria must multiply on the stigma (the only site where epiphytic multiplication has been recorded) to a high level before being carried to the hypanthium by rain (Thomson, 1986) where infection may occur if bacteria are present in sufficient numbers. Blossoms with 103-107 cfu failed to develop infection (Thomson, 1986) and Beer and Norelli (1975) found that inoculum levels of 102 per blossom failed to initiate infection, but flowers with populations of 107 bacteria were likely to be diseased 5 days later. Bacterial multiplication is time-and-climate dependent, and a low dose such as that likely to be present on an infested apple fruit (<102cfu) is very unlikely to be sufficient to result in infection of a host.

Much is made by AFFA of the work of Hildebrand (1937) demonstrating infection of apple blossom with a single bacterium. It is important to note that Hildebrand (1937) injected the single bacterium directly into the nectary of an apple flower. As has been discussed in the previous paragraph, it is generally accepted that bacteria are first deposited onto the stigmatic surface where (under favourable circumstances) they multiply to a level where bacteria can enter the nectarthodes. The work by Hildebrand (1937) does not, therefore, represent a natural infection process. In the same experiments Hildebrand (1937) was not able to initiate infection when a single bacterium was transferred to stigmas or anthers, presumably because the single bacterium was not able to survive and produce a sufficient population to be transferred into the nectarthodes by natural means. No other author has reported a single bacterium infecting a flower.

The expert opinion sought by AFFA in this section confirms that there is no known mechanism by which E. amylovora could be transferred from a contaminated apple to a receptive site, and that this risk is extremely low. However, AFFA did not ask these experts to express their answers in the same terms as those used to assess risk (Table 6 of the draft IRA).

AFFA has concluded that it is "extremely unlikely that viable E. amylovora would be transferred from either infested or infected apple tissue to an appropriate site on a susceptible host". However, AFFA has then changed the assessment of probability from negligible to low on the basis of "uncertainty". It is difficult to understand how AFFA can consider that there is any "uncertainty" surrounding this step. AFFA does not cite any contradictory opinion in the assessment of this step.

New Zealand does not believe that it is appropriate to change the risk assessment from "negligible" to "low" on the basis of "conservatism". ALOP should be used as a yardstick for evaluation of possible measures to mitigate risk during the risk management stage. It is premature, however, to introduce the concept of ALOP when making an objective assessment of the risks based on the available scientific evidence at the risk assessment stage of the draft IRA. The introduction of "conservatism" at this step of the risk assessment is inappropriate.

New Zealand believes that the published literature (or lack of it) and expert opinion supports a conclusion that there is negligible risk of transfer of bacteria from an imported apple to a receptive host.

3.1.2.5 Conclusions: distribution potential

As has been discussed above, there are a number of factors in the distribution and consumption of apple fruit that would considerably reduce the risk that an imported apple will carry E. amylovora, be discarded into the environment and that bacteria will survive to be distributed to a host. New Zealand challenges AFFA's assertion that the distribution potential is low.

New Zealand believes an accurate interpretation of the literature and expert opinion is that the distribution potential is negligible. New Zealand requests that AFFA re-assess this conclusion.

3.1.3 Conclusions: probability of entry

AFFA concludes in this section that the probability is low, based on a high importation potential and a low distribution potential. New Zealand believes a more accurate assessment is that the probability of entry is negligible, based on a very low importation potential and a negligible distribution potential.

The latter conclusion concurs with the published opinion cited by AFFA. Other published opinion not cited by AFFA includes: Smith et al. (1997) "it is widely accepted that fruits present an insignificant risk in practice"; Thomson (1992) "it seems very remote that contaminated fruit could be responsible for establishing new outbreaks"; Schroth et al. (1974) "not an important feature of the disease cycle"; Dueck (1974) "the risk of disseminating fire blight bacteria on symptomless mature apples is considered negligible". It also concurs with New Zealand's assertion that mature fruit are not a pathway for the introduction of fire blight.

New Zealand believes that the probability of entry is negligible based on a very low importation potential and a negligible distribution potential.

3.1.4 Probability of establishment

The probability of establishment of fire blight in the major pome fruit producing areas of Australia has been assessed as "high". New Zealand believes that this risk has been over-estimated, and should at most be "moderate" - an event likely to occur with even probability.

In estimating the risk of establishment it is necessary to first determine whether the establishment is likely to result from a single or multiple introduction. AFFA has concluded that the unrestricted distribution of apples from New Zealand is unlikely to result in the infection of susceptible hosts in Australia (in previous sections New Zealand argues that this risk is, in fact, negligible). From this it therefore seems reasonable to conclude that the introduction would involve a single host or a limited number of hosts in one introduction area.

It is also important to consider where this introduction would likely take place. AFFA has concluded that fire blight would be likely to establish in one of the pome fruit growing regions. However if there was indeed a risk of the entry of fire blight through trade in apple fruit, the initial introduction would be most likely to occur in the metropolitan areas where the majority of the apple consumers are located.

The assessment of establishment potential should then involve a consideration of the likelihood of establishment (perpetuation) in the area where introduction is most likely to occur. AFFA has assessed climatic conditions and concluded that climate may be suitable for perpetuation of fire blight in most areas of Australia. Other factors have not been considered. These include the likelihood that the initial infection will be removed by pruning before spread can occur, or the high probability that the disease will be detected and eradication activities undertaken in a metropolitan area before the disease can spread to another area. The only empirical evidence of this occurring in Australia is the confirmation of fire blight in the Royal Botanic Gardens, Melbourne. This entry was detected after it had been present for some time and was eradicated. In this instance fire blight apparently did not perpetuate.

At this point New Zealand wishes to register concern that AFFA is casting doubt on the confirmation of E. amylovora from the Royal Botanic Gardens, Melbourne (refer p202, l.45 of the draft IRA). Jock et al. (2000) reported the isolation of E. amylovora from plant tissues from Melbourne, and that pathogenicity was confirmed on young apple trees and Koch's postulates established. Similar testing in New Zealand (Taylor and Hale, 1998) also confirmed the presence of E. amylovora in samples provided by Australia.

New Zealand requests that AFFA takes these factors into consideration and reviews the assessment of a high probability of establishment. New Zealand believes "moderate" is a more accurate assessment of risk.

3.1.5 Probability of spread

The probability of spread has been over-estimated. New Zealand believes that an appropriate assessment of spread is "moderate", not high.

In assessing the probability of spread, AFFA has not taken into account the activities that would be taken to prevent spread after the disease was first detected. In the assessment of consequences, AFFA has factored-in the cost of inter-state movement controls and eradication activities. Given the statement that these activities would occur and indeed have been factored into the assessment of economic impact, their impact on reducing both the rate and extent of spread should also be considered. For example, it is very unlikely that an introduction in Sydney would spread to Western Australia.

Regardless of eradication and control activities, AFFA has made no attempt to estimate the rate of spread of fire blight from either single or multiple introductions. These factors are important in assessing the consequences of entry, establishment and spread. There seems to be a view that a single introduction of fire blight will result in the disease instantaneously spreading to all pome fruit growing areas of Australia. Even under favourable circumstances the spread would take many years. For example, fire blight appears to have been present in the Royal Melbourne Botanic Gardens for a number of years without spreading. Europe has maintained for many years special protected zones where fire blight does not occur. New Zealand suggests that the risk of fire blight to all areas of Australia has been over-estimated and therefore requests that AFFA re-examine this.

New Zealand requests that AFFA reviews the risk of spread to all areas of Australia. New Zealand believes that an appropriate assessment of spread is moderate, not high.

3.1.6 Conclusions: probability of entry, establishment and spread

It has been concluded by AFFA that the overall unrestricted probability of entry, establishment and spread of E. amylovora is low. This is based on the combined probabilities of entry (low), establishment (high) and spread (high). New Zealand has indicated that these probabilities have been grossly over-estimated. It is New Zealand's position that the combined probability is negligible, based on an assessment that the probability of entry is negligible, the probability of establishment is moderate and the probability of spread is moderate. This assessment falls within Australia's ALOP as expressed in Table 6 of the draft IRA and no phytosanitary measures are necessary to manage this level of risk.

New Zealand believes that there is a negligible probability of the unrestricted importation of apple fruit resulting in the entry, establishment and spread of fire blight in Australia.

3.1.7 Economic consequences

As has been discussed previously, New Zealand is concerned that a number of terms used by AFFA in determining economic impact are not clearly defined. This makes difficult the task of assessing AFFA's conclusion that the economic consequences of the introduction of fire blight into Australia would be extreme.

Nevertheless, New Zealand questions AFFA's conclusion that the economic impact would be "highly significant at the national level", "of significant national concern" and that "economic stability, societal values or social wellbeing would be seriously affected in more than one geographic region". It is more probable that, at most, the impact of fire blight would be moderate, i.e., "the impact is likely to be recognised at a national level, and significant within the affected geographic regions. The impact is likely to be highly significant to directly affected parties"⁸.

AFFA has presented no data to indicate how fire blight might seriously affect societal values or social wellbeing, other than a very brief mention of unemployment effects. We are therefore left to conclude that the major impacts of concern to AFFA are significance to directly affected parties and economic stability of geographic regions. In the following sections the basis of AFFA's conclusions are challenged.

New Zealand requests that AFFA clearly define the terms used to assess economic consequences and present data to support any conclusions reached.

3.1.7.1 Impact on directly affected parties

For the purposes of this discussion we have assumed the directly affected parties to be the growers of fire blight host material (apples, pears, and the nursery and amenity horticulture industries).

Several Australian studies have been cited which estimate production losses ranging from 1-20% for apples and 3-50% for pears depending on the extent to which the disease establishes and spreads across all regions and states of Australia. The worst case scenario assumes a 100 percent probability of the disease occurring in all regions simultaneously and ideal environmental conditions for fire blight to thrive in. The estimated loss of production for this scenario is estimated to be 50 percent for pears and 20 percent for apples. At no point does the draft IRA provide a rationale for why such an extreme scenario should be regarded as likely. New Zealand contends that it is more appropriate to consider the most likely scenario, rather than the worst case.

In assessing the economic impact of fire blight in Australia, AFFA has not followed the international guideline which states "In order to estimate the potential economic importance of the pest, information should be obtained from areas where the pest currently occurs. For each of those areas, note whether the pest causes major, minor, or no damage." AFFA has not presented data from other countries that have fire blight. For example, New Zealand has the world's highest per hectare production of apples (Bhati and Rees, 1996) and has a successful pear export industry, despite the presence of fire blight.

It is acknowledged that there is a recent instance (Southwest Michigan, USA) where the impact of fire blight has been severe, however the circumstances surrounding that incident are unique and do not apply to Australia's production systems. The combination of extreme climatic conditions and the presence of a large proportion of abandoned or poorly maintained orchards resulted in this a disease problem. This situation bears little resemblance to the effect of a possible fire blight outbreak in Australia.

AFFA states that fire blight is likely to affect access to markets in pomefruit producing countries where fire blight is absent. Bhati and Rees (1996) report that Australia exports to the European Union, Malaysia, Singapore, Indonesia, and Hong Kong. The AAPGA website reports that Australia exports to the United Kingdom, Hong Kong, Taiwan and India. None of these countries imposes restrictions on apples from countries where fire blight occurs.

New Zealand submits that the impact on directly affected parties has been over-estimated by AFFA. It is acknowledged that the impact may be significant to the directly affected parties, and at most highly significant to the directly affected parties. In the terms used by AFFA (p.46 of the draft IRA) this equates to at most a moderate impact on these affected parties.

New Zealand requests that AFFA follow the international guideline for economic consequences and include in the draft IRA information on the impact of fire blight in other countries.

3.1.7.2 Impact nationally or on geographic regions

For the overall economic impact to be assessed as high or extreme, AFFA has stated that it is necessary for there to be a serious effect on the economic stability of one or more geographic areas and to be significant at the national level. AFFA has not clearly assessed the effect of fire blight on the economic stability, or specified the meaning of "significance at the national level".

To examine the significance at a national level, the value of lost production could be compared against both the value of agriculture in Australia and Australia's Gross Domestic Product. The value of all agricultural commodities in Australia (1997/98) was $28 billion and Australia's GDP is around A$520 billion. The consequences of fire blight, in this frame of reference, are insignificant at both the national agriculture and GDP level. Using AFFA's definition, the economic consequences of the disease would be classified as 'low'; "not likely that the impact will be recognised at the national level".

A similar approach should be taken to examining the impact of fire blight on economic stability of geographic areas. It has been suggested that fire blight would have the most serious impact in the Goulburn Valley. The Goulburn Valley creates over $1 billion of food production and processing, based on dairy, vegetable and fruit products (Corporate Strategy Consulting, 1997). On this basis the maximum expected cost to the apple and pear industry represents a 6 percent to 10 percent reduction of the Goulburn Valley's output. This is unlikely to have a serious effect on the economic stability of the region. The yield losses in other regions (1-3%) would presumably have an even lower impact on the economic stability of those areas.

New Zealand submits, therefore, that AFFA has presented no data to suggest that fire blight would affect the economic stability of any region, or of the national economy. It is therefore New Zealand's position that the economic impact of fire blight would, at most, be moderate.

New Zealand suggests that as the economic stability of the regional or national economy would not be affected, the economic consequences of fire blight, are moderate.

3.1.8 Unrestricted risk estimate

New Zealand considers that the unrestricted risk estimate of the entry, establishment and spread of fire blight is negligible, based on a negligible probability of entry, establishment and spread combined with moderate economic impact.

⁵ Table 1. van der Zwet et al. (1990)
⁶ van der Zwet et al. (1990)
⁷ Hale et al. (1987)
⁸ "the individual, or group of individuals who experience the pest" p.46 of the IRA

Contact for Enquiries

MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND

Fax: +64 4 894 0721
Contact this person