In light of this evidence, the supplemental RMA considered five risk management options for the interstate movement of commercially packed citrus fruit from areas quarantined for citrus canker:

After considering the evidence presented in the updated PRA and the supplemental RMA and the conclusions of those documents, we determined that currently available scientific evidence provides additional certainty that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the spread of Xcc. Therefore, no mitigations beyond treatment with an APHISapproved disinfectant are necessary. Accordingly, we proposed to implement Option 2.

Several commenters acknowledged that the risk associated with the interstate movement of regulated fruit from a quarantined area is low but stated that, if there is any risk associated with allowing fruit to move from areas quarantined for citrus canker into commercial citrus producing States, such movement should be prohibited. These commenters stated that citrus canker has been a destructive and costly disease in Florida, one which spurred an eradication attempt that was ultimately unsuccessful, and that other commercial citrusproducing States do not want to be at risk for the introduction and establishment of the disease. One commenter recommended that we err on the side of caution in making changes to the regulations and stated that further research should be done before fruit from quarantined areas is allowed into commercial citrusproducing States.

Two of these commenters proposed additional risk mitigation measures to address the risk they perceived to be associated with fruit moved interstate from an area quarantined for citrus canker. Both stated that such fruit should not be allowed to move into the eight county Citrus Zone in south Texas. These commenters cited the suitability of Texas' climate to citrus canker establishment (as demonstrated by previous outbreaks of citrus canker in Texas), the susceptibility of grapefruit (a common citrus crop in Texas) to citrus canker, and citrus canker's effect on young citrus trees. One of these commenters additionally requested that fruit destined for Texas originate only from groves that have been certified as being free of citrus canker for more than a year, based on a survey.

Another commenter, responding to some of these commenters, stated that no agricultural trade between States and countries anywhere in the world could be conducted if minimal risk is unacceptable and that the proposed rule would mitigate the risks to the point that risks are negligible.

Our goal in restricting the interstate movement of plants, plant products, and other articles is not to achieve zero risk, which, as the last commenter noted, cannot be achieved in agricultural trade. Rather, we seek to impose restrictions on the interstate movement of such articles that are commensurate with the risk they pose and that mitigate the risk associated with their interstate movement. Based on all the available scientific evidence, the updated PRA and supplemental RMA concluded that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc. We received several comments on the two new publications that led us to prepare the updated PRA and supplemental RMA, as well as comments on the updated PRA and supplemental RMA themselves. These comments are discussed in further detail later in this document. However, they did not change our conclusion that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the spread of Xcc. Accordingly, this final rule implements Option 2 as proposed.

We are not retaining the current prohibition on the distribution of fruit from a quarantined area to commercial citrusproducing States, and we are not adding the additional mitigations requested by two of the commenters. Based on our determination that fruit is not an epidemiologically significant pathway, we have determined that those additional mitigations are unnecessary to prevent the spread of citrus canker via the interstate movement of fruit from quarantined areas. As noted, it is impossible to eliminate all risk associated with the interstate movement of fruit from quarantined areas; given the conclusions of the updated PRA and the supplemental RMA, following the recommendation that we prohibit the movement of fruit into commercial citrusproducing States unless all risk is eliminated would impose an unnecessary restriction on the movement of fruit.

Under section 412(a) of the Plant Protection Act (7 U.S.C. Sec. 7712), the Secretary of Agriculture may prohibit or restrict the interstate movement of any plant or plant product if the Secretary determines that the prohibition or restriction is necessary to prevent the dissemination within the United States of a plant pest or noxious weed. Based on our supplemental RMA, APHIS has concluded that commercially packed citrus fruit treated with an APHISapproved disinfectant is not an epidemiologically significant pathway for the dissemination of citrus canker within the United States. Accordingly, APHIS has determined that it is not necessary to prohibit the interstate movement of regulated fruit that is commercially packed and treated with an APHISapproved disinfectant from an area that is quarantined for citrus canker in order to prevent the dissemination within the United States of a plant pest. This determination is based on the findings of the updated PRA and the supplemental RMA referred to earlier in this document and our judgment that the application of the measures we proposed will prevent the dissemination of plant pests within the United States.

One commenter who was opposed to allowing the interstate movement of citrus fruit from a quarantined area to commercial citrusproducing States stated that California, a commercial citrusproducing State, is the home of three of the most important resources of citrus germplasm in the United States: The National Clonal Germplasm Repository for Citrus and Dates (NCGRCD), a U.S. Department of Agriculture Agricultural Research Service (ARS) facility supplying budwood worldwide; the Citrus Clonal Protection Program, University of CaliforniaRiverside (UCR), the first citrus germplasm program in the world supplying budwood to California, Arizona, and Texas; and the UCR Citrus Variety Collection, perhaps the most diverse citrus collection in the world dating back to 1907. The commenter stated that certified diseasefree budwood and a broad genetic basis for
[[Page 54434]]
variety development and improvement are the foundation of every successful, profitable, and sustainable citrus industry in the world and that those three germplasm resources are the only ones in the United States (if not the world) that have not been exposed to citrus canker or other devastating citrus diseases such as citrus greening. The commenter stated that taking a ``calculated'' risk to expose these invaluable resources to one of the worst citrus diseases in the world, citrus canker, based on limited field and packinghouse practices that will not be inspected for compliance is unacceptable. This commenter also stated that the Florida citrus industry funded a project to ``rescue'' Florida citrus germplasm by moving it to citrus canker and citrus greeningfree California in the NCGRCD facilities.

As we have determined that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of citrus canker, we do not expect that these facilities will be exposed to citrus canker as a result of the implementation of this final rule.

However, it should be noted that germplasm facilities are devoted to the preservation of the germplasm within the facilities and thus are protected against potential sources of pest and disease introduction. Indeed, potentially infected germplasm from foreign countries is imported into these same facilities for screening purposes, which is a much more likely pathway for the introduction of diseases such as citrus canker than the interstate movement of regulated fruit from a quarantined area. Allowing citrus fruit to be moved interstate from quarantined areas into California will not decrease the efficacy of the biosecurity in place at these facilities.

It should also be noted that, under this final rule, packinghouses will be inspected to ensure that they are complying with the requirements to treat regulated fruit with an APHISapproved disinfectant and to ensure that the fruit is free of leaves, twigs, and other plant parts, except for stems that are less than 1 inch long and attached to the fruit. With regard to the other commercial fruit production practices described in the November 2007 RMA, we assume that commercial growers and packinghouses will continue to employ procedures that reduce the incidence of citrus canker in their fruit, as citrus canker lesions reduce the market value of infected fruit.
New Evidence We Considered in the Updated PRA and Supplemental RMA

Several commenters generally addressed the Gottwald et al. (2009) and Shiotani et al. (2009) publications. We address these comments below.

One commenter stated that the premise of both publications was to prove that citrus canker cannot be transmitted by infected or contaminated citrus fruit. The commenter stated that, scientifically, a negative premise cannot be proven, and the commenter cited this as one major flaw of these studies. Another commenter stated that Shiotani et al. (2009) did not demonstrate that Xcc cannot be transmitted from fruit to susceptible tissue, as it did not adequately resolve the ability of Xcc to spread from asymptomatic fruit.

One commenter, responding to the first commenter, stated that the two publications never set out to prove that something cannot happen because, philosophically and scientifically, this is impossible. However, the commenter stated, both publications soundly proclaim that risks can very effectively, very simply, and very reliably be reduced below any reasonable and measurable risk of transmitting citrus canker disease.

As the last commenter states, neither of the publications concluded that citrus canker cannot be spread by fruit. Gottwald et al. (2009) concluded that ``harvested and packinghousedisinfested citrus fruit are extremely unlikely to be a pathway for Xcc to reach and infect susceptible citrus and become established in cankerfree areas.'' Shiotani et al. (2009) concluded that ``there is a low risk [of] transmission'' of Xcc from fruit. These conclusions are consistent with the conclusions of the updated PRA and supplemental RMA, as described earlier.

Two commenters stated that the research in the Gottwald et al. (2009) and Shiotani et al. (2009) publications should be tested and retested by others who were not involved in the original research before changing the conditions under which fruit is allowed to move from an area quarantined for citrus canker. Three commenters stated that a national task force consisting of scientists from citrus producing areas other than Florida (and besides ARS personnel) should be assembled to address any change in current quarantine regulations that might result in the introduction of known destructive pathogens from known infected areas to noninfected areas (i.e., California, Arizona, Texas, etc.).

The Gottwald et al. (2009) and Shiotani et al. (2009) publications were produced independently, published in a peerreviewed journal, and came to similar conclusions regarding the epidemiological significance of fruit as a pathway for the spread of citrus canker. Among other topics they address, these publications provide valuable evidence regarding the potential for Xcc to spread from infected fruit to host plants in the field; this evidence is what prompted us to prepare the updated PRA and supplemental RMA.

However, the updated PRA and supplemental RMA considered all the available evidence regarding the potential of fruit to serve as an epidemiologically significant pathway for the introduction and spread of citrus canker, not just the evidence in those publications. The weight of all the available evidence is what led us to the conclusion that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc. We have determined that the evidence provides adequate certainty regarding this conclusion to remove some restrictions on the interstate movement of commercially packed and disinfected fresh citrus fruit from an area quarantined for citrus canker.

The November 2007 PRA and RMA and the supplemental RMA prepared for this rulemaking were all submitted for peer review in accordance with the Office of Management and Budget's bulletin on peer review. The peer reviewers for the November 2007 PRA and RMA and the supplemental RMA were experts in plant pathology, phytobacteriology, and risk assessment. The comments we received from these peer reviewers indicated that our analysis of the available evidence regarding the risk associated with the movement of fruit from an area quarantined for citrus canker was sound.

It should also be noted that the authors of the Shiotani et al. (2009) publication were not affiliated with the State of Florida in any way, and the experiments in the Gottwald et al. (2009) publication were conducted by an international consortium of scientists working cooperatively and reaching the same conclusion after conducting similar experiments in two different countries, with participants from Argentina as well as Florida.

Gottwald et al. (2009)

We received several comments specifically addressing Gottwald et al. (2009).

Some of the experiments included in Gottwald et al. (2009) examined the
[[Page 54435]]
effectiveness of treatment with a disinfectant at reducing Xcc populations on citrus fruit. One commenter stated that the disinfection procedures significantly reduced pathogen survival but did not completely eliminate it. The commenter stated that, considering the large amount of fruit being shipped, even a low survival rate of the pathogen poses a high risk for the introduction of Xcc to a disease free area.

This commenter also stated that the limitation of treatments in disinfecting fruit with lesions or fruit wounds contaminated with inoculum of the pathogen is well known. Oxidizing agents cannot effectively remove or reduce inoculum to acceptable levels in wounded tissue because of the natural reducing agents that occur in fruit tissue. Furthermore, these treatments would have little or no effect on established fruit lesions that act as reservoirs of inoculum. Thus, the commenter stated, without any inspections, even a few lesions on fruit would pose a high risk because the pathogen could not be eliminated using existing disinfection practices.

Another commenter stated that one cannot in a practical sense sterilize the surface of fruit; it would do more harm than good, and there is no biological reason to do so. The commenter stated that there is an inoculum threshold necessary to naturally establish citrus canker under even the most conducive conditions (10\5\ colonyforming units (cfu)/milliliter (ml) for intact tissue infection, 10\3\ cfu/ml for wounded) and that fruit disinfection easily achieves the low levels of inoculum necessary to avoid the risk of disease transmission. The commenter stated that the concern that inoculum in wounds on fruit could not be completely eliminated overlooks the fact that the bacteria do not even cause an infection at the wound site, let alone become liberated to possibly induce a lesion elsewhere.

The November 2007 RMA and the supplemental RMA both acknowledge the fact that disinfection treatments are not completely effective against Xcc bacteria in lesions. However, as the November 2007 RMA stated, there is abundant evidence that shows that packinghouse disinfection treatments destroy surface bacteria and reduce the viability of all bacteria on fruit. We did not rely solely on the Gottwald et al. (2009) publication in making our determination that treatment with an APHIS approved disinfectant is an effective mitigation against the risk of spread of citrus canker; rather, we considered all the available evidence regarding the effectiveness of disinfectant treatments.

In addition, other evidence indicates that bacteria that remain in lesions after disinfection are not epidemiologically significant. For example, Gottwald et al. (2009) provided additional evidence supporting the conclusion that the viability of bacteria on fruit and in lesions and wounds diminishes after the fruit is harvested and that the viability of Xcc bacteria which survive the packing process will further diminish during shipping.

We disagree with the first commenter that the effectiveness of disinfectant treatment on bacteria in wounds is a concern. The second commenter is correct to note that Xcc bacteria in wounds do not cause infections at the wound site. As discussed in the supplemental RMA, evidence indicates that wounds on harvested fruit containing Xcc inoculum do not lead to citrus canker lesion development, and Xcc populations generally decline, although wounds might occasionally retain Xcc populations that decline more slowly.

Finally, with respect to the first commenter's concern about elimination of bacteria, we acknowledge that the surface disinfectant treatments approved by APHIS reduce numbers of Xcc cells to low or undetectable levels, but do not necessarily provide complete eradication. As the second commenter notes, complete eradication would be impractical. In any case, it is not necessary to completely eradicate Xcc in order to ensure that disinfected fruit is not an epidemiologically significant pathway. While the updated PRA and supplemental RMA conclude specifically that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc, it is not just the disinfection process that makes fruit not an epidemiologically significant pathway for Xcc, but also the biology of Xcc and the conditions that must be fulfilled in order for Xcc transmission from infected fruit to a host plant to occur, among other factors.

Some commenters addressed experiments in the Gottwald et al. (2009) publication that were designed to investigate the likelihood that citrus fruit disposed of by consumers may serve as a source of inoculum for nearby host material. Gottwald et al. (2009) studied the transmission of Xcc from unprocessed, infected `Ruby Red' grapefruit and `Lisbon' lemon and packinghouseprocessed `Ruby Red' grapefruit in cull piles to `Duncan' grapefruit seedlings during natural weather events. During the course of the experiments, citrus canker lesions did not develop on the grapefruit seedlings (488 seedlings total) surrounding the diseased fruit, in spite of extensive leafminer damage present on some of the seedlings. Xcc bacteria were not detected in assays of the foliage.

Gottwald et al. (2009) repeated the cull pile experiment to see if transmission of Xcc from infected, unprocessed `Ruby Red' grapefruit fruit is possible under simulated extreme wind and rain conditions. Infected fruit were either placed in a cull pile or suspended by vertical strings. One seedling 0 meters (m) downwind from the cull pile became infected when subjected to the highest wind speed (25 m per second (m/s)) and simulated rain, developing 1 lesion on a single leaf injured by the action of the highspeed fan. The other 191 plants in the study did not develop Xcc lesions. No Xcc lesions developed on the 192 plants placed at the same distance and subjected to the same wind speed (0, 10, and 25 m/s with water) from Xccinfected grapefruit suspended from string. Xcc was recovered from 1 collection screen set up 2 m from suspended fruit, but no Xcc was recovered from the other 144 collection screens set up at various distances (0 to 10 m) from cull piles or suspended fruit. Gottwald et al. (2009) stated that this cull pile experiment was ``a highly contrived situation designed to provide every possible opportunity for dispersal of Xcc and would be unlikely to occur in most areas, except those locations where hurricanes or tropical storms are common occurrences.''

One commenter noted that one plant surrounding infected fruit in cull piles did develop the disease in one of the simulated wind and rain experiments, indicating that this pathway of transmission is possible. The commenter stated that one might think that this level of transmission from an infected fruit to a healthy plant is very low, but this can be interpreted as very high under the set of conditions established for the experiments. The commenter stated that conducting these studies in regions where other environmental conditions exist and with a different group of scientists may lead to a different conclusion.

A second commenter stated that both Gottwald et al. (2009) and Shiotani et al. (2009) demonstrate that transmission of the bacterium is a difficult process to replicate and expressed a view that the natural spread of the bacterium from infected fruit to host plants remains poorly understood. The commenter stated that the cull pile transmission experiments conducted by Gottwald et al. (2009) do not provide conclusive
[[Page 54436]]
evidence that the risk of fruittotree transmission is insignificant. The commenter stated that these trials were conducted with little replication and did not adequately represent weather events that are conducive to the transmission of the bacterium, that the authors did not demonstrate that Xcc could initiate infections under the experimental conditions in positive controls, and that the employed diagnostic methods were not tested in positive controls.

This commenter also noted that transmission of Xcc from infected fruit to host plants did occur, despite each wind speed treatment being applied for only 5 minutes. While APHIS concluded that the experimental conditions that produced this result were ``highly contrived,'' the commenter stated, due to the smallscale nature of this trial, small sample sizes, short exposure times, and lack of adequate controls, the risk of transmission under natural conditions remains feasible and significant. The commenter concluded that the experiments by Gottwald et al. (2009) demonstrated the ability of Xcc to be spread from symptomatic citrus fruit.

A third commenter stated that the transmission of Xcc from infected fruit to host plants in the simulated extreme wind and rain conditions was probably because of mechanical contact and injury, not from anything most people would consider as a natural transmission event. This commenter also noted that the cull pile in that experiment was composed of freshly picked and heavily infected fruit, not fruit that had been graded and disinfected according to packinghouse protocol. The commenter stated that the value of this experiment is that it demonstrates the ``tipping point'' for canker infection from fruit. The commenter stated that if the other commenters envision a pile of freshly picked cankerinfected grapefruit suddenly arriving in a grapefruit orchard in Australia, Arizona, or California immediately adjacent to susceptible plants and experiencing 25 m/s winds accompanied by rain, the scenario is excessively imaginary. The ``tipping point,'' in this commenter's view, identifies the dangerous conditions for shipping fresh fruit from a canker endemic area so they can be completely avoided.

We agree with the first two commenters that it would have been optimal to have additional replications of the experiment in which Xcc was transmitted from infected fruit to host plants, to better determine the rate at which transmission occurs in these conditions. However, as noted, the conditions in the experiment in which Xcc was successfully transmitted from infected fruit to host plants were extreme conditions, designed (as the third commenter states) to establish whether transmission of Xcc from infected fruit to host plants is possible, not whether it is likely. (As the third commenter notes, Gottwald et al. (2009) concluded that the lesion that resulted from the simulated wind and rain cull pile experiment ``was the result of a leaf wound.'')

In the context of the other experiments Gottwald et al. (2009) performed to assess the likelihood of fruittoplant transmission, and in the context of the conditions of the experiment, including not only the simulated extreme wind and rain conditions but also the fact that the fruit were unprocessed and untreated and the placement of those fruit directly adjacent to host plants, we have determined that this one successful transmission is consistent with a determination that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc, given all the available evidence about the potential for fruit to serve as a pathway.

Although the first commenter is correct that conducting the experiments in other environmental conditions and with another group of scientists might lead to a different conclusion, based on the available science regarding the transmission of citrus canker, the environmental conditions under which these experiments were conducted are extremely suitable to the potential transmission of citrus canker. Fruit that were specifically selected for their high level of infection and that were subjected to none of the packinghouse processes (including disinfection) that are known to reduce the viability of Xcc infection were used in attempts to infect highly susceptible grapefruit plants at the most susceptible stage of the plants'development. The one trap plant that was infected was placed immediately adjacent to the infected fruit and subjected to simulated extreme wind and rain conditions that are unlikely to occur in most areas. We have determined that it is unlikely that studies in other regions and under other environmental conditions would produce a greater level of transmission of the disease from infected fruit to host plants.

We have determined that the Gottwald et al. (2009) experiments adequately represented weather events that are conducive to the transmission of Xcc and represented a range of weather conditions as well. The trials were conducted both in field conditions that were not conducive to the transmission of Xcc, in Argentina, and that were conducive, in Florida.

It would be difficult to develop a positive control for the cull pile experiments, as a positive control would require the successful transmission of Xcc, which Gottwald et al. (2009) were only able to accomplish under conditions described in the publication as ``highly contrived.'' (It should be noted that this was not APHIS' description.) Nevertheless, it should be noted that the authors who performed the cull pile experiments have performed similar experiments using yard blowers, as documented in Bock et al. (2005) and Parker et al. (2005). These publications demonstrated that using a forced air source for wind and hose water for rain will elicit and spread Xcc from infected plants. In one experiment in Bock et al. (2005), the blower was run for 5 minutes, the same duration as in the 25m/s artificial wind and rain cull pile experiment, and bacteria were recovered from the water to which the infected plants were exposed. Different experiments in both papers using different durations produced the same results. We would presume that using similar techniques to elicit and spread Xcc from infected fruit would be effective, if fruit was an epidemiologically significant pathway.

The commenter correctly notes that the Gottwald et al. (2009) publication did not describe any positive controls for the immunostrips used in the cull pile experiments to determine whether Xcc was present. However, a personal communication with one of the authors of that publication indicates that the experimenters did use positive controls to confirm that the immunostrips were working properly and thus would have indicated that Xcc was present if it had been present.

We disagree with the second commenter that the exposure times in the cull pile experiments in Gottwald et al. (2009) were ``short.'' The 5minute exposure time in the 25m/s artificial wind and rain experiment was sufficient to infect 1 test plant. The commenter also ignores the field cull pile experiments, which each took place for several weeks, at different times of year.

Finally, it is important to note that our determination that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc does not rest solely upon the Gottwald et al. (2009) cull pile experiments, although they do provide
[[Page 54437]]
valuable evidence supporting that determination. Rather, that determination takes into account all the evidence considered in the November 2007 RMA, the updated PRA, and the supplemental RMA, including evidence about the biology of the disease, the effectiveness of disinfectant treatment, the conditions that must be fulfilled for disease transmission to occur, and the fact that the movement of commercial citrus fruit has not been associated with an outbreak of the disease anywhere in the world.

Shiotani et al. (2009)

We also received several comments specifically addressing Shiotani et al. (2009).

One commenter stated that, in Shiotani et al. (2009), proper positive controls proving that the polymerase chain reaction (PCR) detection technique is working were not included in one set of experiments. (We believe the commenter is referring to the examination of fruit collected from a diseased commercial orchard to investigate the survival of Xcc.)The commenter stated that the lack of controls casts doubts on the results of this research.

The commenter correctly notes that there is no explicit discussion of controls in the ``Materials and Methods'' section of the paper. This does not mean that the proper controls were not used, but we cannot verify that they were. That said, the fact that isolations and bioassays made from the same material also yielded negative results supports the PCR results.

One commenter stated that the Shiotani et al. (2009) experiments used a laboratory strain of Xcc that has not been shown to be pathogenic but, the publication stated, ``is believed to be as robust as the wildtype.'' The commenter stated that this demonstrates critical flaws in the experimental design and that the conclusions of Shiotani et al. (2009) can thus not be accepted without reasonable doubts.

The commenter quotes from the ``Discussion'' section of the Shiotani et al. (2009) publication. In the ``Materials and Methods'' section, the authors discuss the laboratory strain in more detail: ``A marked strain of X. citri pv. citri (KC21Rif100) that is resistant to rifampicin was used as inoculum. This strain is a stable, spontaneously derived mutant from strain KC21 (Shiotani et al., 2008), which has been shown to be as pathogenic as other strains of X. citri pv. citri in infection studies.'' We believe this information addresses the commenter's concern.

The Shiotani et al. (2009) publication included experiments designed to assess the potential for spread of Xcc from mature Satsuma mandarin fruit inoculated with the marked strain of Xcc mentioned above and suspended in polypropylene net bags in navel orange trees. One commenter noted that, in one of the four experiments conducted, citrus canker was transmitted from culled mandarin fruit to leaves of navel orange trees in an orchard.

Another commenter, responding to the first commenter, noted that the infections in that experiment were not caused by the marked strain of Xcc but by the wild type. Citrus canker is endemic in the area where this study was done, so a tagged strain was used. That way, the commenter stated, the researchers have an idea where the inoculum is coming from. The commenter stated that the fact that wildtype canker bacteria occasionally are caught in traps or cause infection on plants in the experiment does not undermine the conclusion in any way; in fact, it demonstrates that conditions conducive to the transmission of canker existed, and the marked strain on and in fruit did not demonstrate any risks of disease transmission.

We agree with the second commenter.

One commenter stated that the Shiotani et al. (2009) publication does not provide a high degree of confidence that transmission of Xcc from contaminated fruit to host plants is not epidemiologically significant. Although no transmission of Xcc was observed, the commenter suggested that it is possible that this was due to unexplained variables. Rainfall data were provided but no information was provided on the growth stage of trap plants, insect presence in the orchard, potential wounds and insect damage, spray history within the orchard, or other significant wind and weather events. Because the experiments were conducted in a commercial orchard, the commenter stated, it would be expected that pest and disease management would have been practiced at some point prior to the study.

As noted earlier, the Shiotani et al. (2009) experiments used a marked strain of Xcc because Xcc is endemic in the area where the experiments took place. The wildtype strain of Xcc occurred in the orchard where the experiments took place, throughout the experiments. This indicates that at least some plants in the orchard were at a susceptible growth stage, and in general the transmission of Xcc between trees in the orchard indicates that whatever unexplained variables may have been present did not impede the normal transmission of Xcc.

In Shiotani et al. (2009), the authors state, for the initial assay of fruit from diseased orchards, ``No chemicals had been sprayed to control the disease,'' addressing the commenter's concern about the previous employment of disease control methods. Disease control is not addressed directly for the other experiments, including the experiments regarding the potential spread of Xcc from Satsuma mandarin fruits. However, other statements in the publication imply that no disease control techniques were employed in the orchard:
In September 2006, the Satsuma mandarin orchard in Saga was damaged by typhoon No. 0613. The typhoon brought rain with strong southerly winds with maximum speeds of 50 m/s to the orchard, which is located on a southfacing hillside. The severe meteorological conditions of this typhoon strongly facilitated spread of citrus canker, leading to the highest incidence of the disease in the orchard in the last decade. ... It is most likely that small populations of the wild strain of X. citri pv. citri survived in the orchard. Citrus canker infection caused by the wild strain indicated that conditions were also conducive for the establishment and spread of the introduced KC21Rif100 strain. The KC21Rif100 strain did not exude from lesions on Satsuma mandarin fruits after they were discarded in an orchard in October 2006, although conditions were conducive for the spread of X. citri pv. citri.

If disease control techniques had been employed in the orchard, we assume that the authors would not have described the conditions as conducive for the spread of Xcc.

These statements also indicate that information on significant wind and water events was provided, specifically with regard to typhoon No. 0613.

Shiotani et al. (2009) did not provide any information on insect presence or pest control in the orchard. The citrus leafminer is known to occur in Japan, but we do not know whether it occurs in the orchard. However, it is important to note that insects themselves are not known to be vectors for Xcc; the presence of the citrus leafminer or another insect in the orchard might increase the severity of canker in the orchard, but it would not enable transmission of Xcc from infected fruit to host plants.

The commenter stated it is likely that naturally infected tissues have a higher
[[Page 54438]]
ability to transmit the bacterium than artificially surfaceinoculated fruit, which were used in Shiotani et al. (2009).

Shiotani et al. (2009) determined that the bacteria in the lesions that resulted from the artificial inoculation were viable. We know of no evidence that suggests that bacteria in natural lesions are more effective than surfaceinoculated bacteria in spreading Xcc, and the commenter did not supply any.

The commenter stated that another limitation of the design of this experiment is that it did not include a control group to demonstrate treetotree transmission under a similar set of conditions.

Treetotree transmission was demonstrated through the incidence of the wildtype strain of Xcc, which the publication discussed. In this case, the wildtype strain acted as a control to show that transmission of Xcc within the orchard was possible and did occur.

The commenter also stated that the uncertainties cited by the commenter are acknowledged by the authors, who suggested that conditions may have been unfavorable for spread of the bacterium.

The statement in Shiotani et al. (2009) that conditions may have been unfavorable for disease spread referred to one replication of the experiment. The publication goes on to note that disease spread occurred at high levels in a subsequent replication:
In the experiments started in November 2005 and March 2006, no canker symptoms were observed on any branches beneath the discarded fruits. This may be because weather conditions were unfavourable for disease spread during this period. During the experiment started on May 2006, canker lesions were observed on leaves of navel oranges located beneath the discarded Satsuma mandarin fruits. [hellip]The severity of the disease was greater in 2006 than in 2005. The incidence of citrus canker in the orchard was 36.2 percent and severity was 18.0. The high incidence may be attributed to typhoon No. 0613 that occurred on September 17, 2006.

In addition, it should be noted that our determination that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc does not rest solely on the experiments in Shiotani et al. (2009), although they do provide valuable evidence supporting that determination. Rather, that determination reflected our analysis of all the evidence considered in the November 2007 RMA, the updated PRA, and the supplemental RMA, as discussed earlier.

Shiotani et al. (2009) also examined the survival of Xcc bacteria on the surface of artificially inoculated fruit that were retained for sampling. One commenter noted that viable Xcc was isolated from 3 canker lesions from 2 out of 6 Satsuma mandarin fruit (a cultivar resistant to citrus canker), 3 months after inoculation. Given these results, the commenter concluded that symptomatic citrus fruit (treated or untreated) remain a potential source of inoculum.

We agree with the commenter that some viable bacteria may remain in lesions of infected fruit. However, in those fruits, the strain KC21Rif100 was found in only 3 of 14 lesions andat a bacterial population lower than 3 x 10\3\ cfu per lesion. This is consistent with one of the findings of the November 2007 RMA and the supplemental RMA, which is that the viability of bacteria on fruit and in lesions and wounds diminishes after the fruit is harvested. Diminishing bacterial populations are less likely to provide adequate inoculum to incite infection.

It should also be remembered that the fruit that were sampled and found to have viable bacteria had been stored in protected conditions. The fruit that were artificially inoculated and used in the experiment regarding the potential of spread of citrus canker did not serve as sources of citrus canker transmission, even when the lesions had just been formed and presumably contained high levels of inoculum. The rinds of the artificially inoculated fruits retrieved after 3 days in the orchard did not have any viable bacteria. Finally, as noted earlier in the discussion of Gottwald et al. (2009), other evidence indicates that bacteria that remain on the fruit in lesions and wounds after disinfection are not epidemiologically significant.

The commenter is correct to note that Satsuma mandarin is a resistant variety of citrus. As noted in the supplemental RMA, the Gottwald et al. (2009) and Shiotani et al. (2009) publications used citrus cultivars that represented the extremes of susceptibility from highly susceptible (grapefruit) to less susceptible varieties (lemon, mandarins). APHIS assumes cultivars not specifically studied would fall within this range of susceptibility and the results are therefore applicable to all citrus cultivars. In any case, the supplemental RMA and November 2007 RMA consider many different sources of evidence in making the determination that the viability of bacteria on fruit and in lesions and wounds diminishes after the fruit is harvested, not just the Shiotani et al. (2009) publication.

One commenter noted that the authors of Shiotani et al. (2009) state: ``It is possible that bacterial cells of KC21Rif100 strain could not grow and colonize the surface of the contaminated fruits due to lack of nutrients.'' The commenter stated that, considering that at least a small percentage of fruit is always decaying during shipment and marketing, this decayed fruit can contaminate other fruit with nutrients that will make survival of the bacteria more likely.

The commenter provided no evidence suggesting that this would occur, and we are aware of none. The available evidence suggests that rotting fruit would not provide nutrients that would make survival of Xcc bacteria more likely. For example, Fulton and Bowman (1929) demonstrated that canker does not survive on rotting fruit. In addition, decaying fruit would be decaying due to the presence of other organisms, and Xcc does not compete well with other organisms, as described in Fulton and Bowman (1929) and Leite (1990).

One commenter stated that, at the end of the Shiotani et al. (2009) publication, the authors indicate that navel oranges are more susceptible to canker than mandarins. The commenter stated that this indicates that their pathogen survival studies on mandarins will not reflect the true risk of transmission of the pathogen/disease. Two other commenters echoed this concern and stated that, because California's growing situation is quite different than those in the research areas, there are serious issues about the extrapolation of data from study of only a few varieties. Another commenter, approaching this issue differently, suggested that restrictions on the interstate movement of different varieties of citrus fruit could vary based on the variety's resistance to citrus canker.

The Shiotani et al. (2009) publication does not actually state that Satsuma mandarins are more resistant to Xcc than navel oranges, although this is widely acknowledged to be true. In any case, as noted earlier, the Gottwald et al. (2009) and Shiotani et al. (2009) publications used citrus cultivars that represented the extremes of susceptibility from highly susceptible (grapefruit) to less susceptible varieties (lemon, mandarins). APHIS assumes cultivars not specifically studied would fall within this range of susceptibility and the results are therefore applicable to all citrus cultivars. The commenters did not provide any specific reasons to question this assumption. [[Page 54439]]

In general, although we recognize that there are limitations in extrapolating from results achieved with Satsuma mandarins, the Shiotani et al. (2009) provides valuable evidence supporting our determination that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc. We took this evidence into account along with the Gottwald et al. (2009) publication and the other evidence cited in the November 2007 RMA and the supplemental RMA in making this determination.

Other Issues in the Updated PRA and Supplemental RMA

One of the conclusions in the updated PRA is that standard packinghouse procedures and postharvest treatments will remove and/or devitalize epiphytic populations of Xcc. This conclusion is echoed in the supplemental RMA.

One commenter stated that the conclusion in the updated PRA that Xcc has a low survival potential is in contrast to earlier research by Golmohammadi et al. (2007), who reported that Xcc was frequently detected on fruit with cankerlike symptoms in commercial consignments of citrus from Uruguay and Argentina into Spain. These consignments were accompanied by phytosanitary certification stating that fruit had been treated with postharvest bactericides, including chlorine and sodium orthophenylphenate. The presence of Xcc on these samples was confirmed by molecular and pathogenicity testing. Pathogenicity assays on grapefruit leaves confirmed that Xcc cells remained viable and were able to produce symptoms despite the application of postharvest treatments and low temperature storage.

Both the updated PRA and the supplemental RMA addressed Golmohammadi et al. (2007). The updated PRA and supplemental RMA state that the results in Golmohammadi et al. (2007) indicate that disinfection protocols are not 100 percent effective. Some samples were only positive by PCR protocols. The authors concluded this was probably due to the disinfection treatments, which would reduce bacterial populations, and may induce the noncultivable state in the analyzed lesions. They further suggested that the bacterial cells in the lesions could be stressed after the fruit treatments (washing, disinfection, chemical treatments, transport, and storage at low temperatures for variable periods of time). Pathogenicity tests were successfully conducted only by artificial laboratory inoculations; the
epidemiological significance of these results was not evaluated.

Pathogenicity tests of bacteria in the laboratory do not indicate whether the bacteria would actually be able to infect host plants in a field setting, where conditions are likely to be less favorable than in a laboratory. The fact that Golmohammadi et al. (2007) concluded that bacterial cells in the lesions could be stressed after the fruit treatments suggests that the bacteria would not have been able to do so, particularly given the results of the experiments Gottwald et al. (2009) and Shiotani et al. (2009) conducted that addressed the transmission of Xcc from infected fruit to host plants in the field. Since Gottwald et al. (2009) and Shiotani et al. (2009) both used untreated fruit in their experiments, and Golmohammadi et al. (2007) concluded that packinghouse processing and disinfection treatment further reduce the viability of the bacteria, we have determined that the results of Golmohammadi et al. (2007) are consistent with the determination that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc.

One commenter, specifically noting the detections of Xcc on fruit with cankerlike symptoms in commercial consignments of citrus from Uruguay and Argentina into Spain, stated that standard harvesting and packinghouse procedures may not effectively eliminate infected fruit from the export pathway.

Both the November 2007 RMA and the supplemental RMA acknowledge this. However, these procedures do reduce the prevalence of viable Xcc in commercial consignments of fruit, thus bolstering the conclusion that commercially packed and disinfected fresh citrus fruit is not an epidemiologically significant pathway for the introduction and spread of Xcc.

One commenter stated that the supplemental RMA claims that the ``uncertainties'' recognized in the November 2007 RMA are now answered, but the question of additional ``uncertainties'' is completely disregarded.

The supplemental RMA has an extensive discussion of remaining uncertainties in the discussion of options at the end of the document. The commenter did not identify any specific uncertainties that the supplemental RMA did not address.

One commenter stated that, in the supplemental RMA, there is not a single biological reference to fruit pests such as the peel miner and to the fact that there is no scientific work/information for its impact on diseases such as citrus canker. The supplemental RMAsimply disregards this classic epidemiological factor under the general assumption ``Vectors do not have a role in disease epidemiology and if they do, it is not subject to regulation.'' The commenter stated that this disregard of valid, researchable questions is highly disturbing.

The role of insects in citrus canker outbreaks was discussed in the November 2007 RMA. The supplemental RMA does not recreate or revise the entire body of evidence cited in the November 2007 RMA, but rather builds on that body of evidence and evaluates those areas of evidence addressed by the new research. Because none of the newer research cited in the supplemental RMA addressed the role of insects in citrus canker outbreaks, we did not update the discussion in the November 2007 RMA.

With regard to the issue of vectors, one commenter stated that canker is a local lesion disease that does not invade the vascular system and is not transmitted by sucking insects or mites, including citrus leafminer and peel miner. The commenter stated that citrus leafminer is not a vector for the canker bacterium.

The November 2007 RMA indicates that injuries caused by the Asian leafminer can produce wounds that serve as infection courts in leaves and, to a lesser extent, fruit, but the leafminer itself is not known to be a vector for the spread of citrus canker. In the November 2007 final rule, we discussed the peel miner, stating that injuries from the peel miner would be likely to increase the susceptibility of fruit to infection, and increase the severity of the infection if they became infected. In terms of overall spread of citrus canker, however, the peel miner would not likely be as epidemiologically significant as the Asian leafminer, since leaves of citrus trees and plants are more susceptible to citrus canker infection than the peels of citrus fruit.

We also note that there exists no evidence indicating that the peel miner is a vector for citrus canker, and we would presume that the peel miner is not a vector, for the reasons cited by the second commenter. Comments on the November 2007 RMA

The November 2007 RMA contained a discussion of the potential for introduction and establishment of Xcc in various climatic conditio

FOR FURTHER INFORMATION CONTACT

Mr. Stephen Poe, Senior Operations Officer, Emergency and Domestic Programs, Plant Protection and Quarantine, APHIS, 4700 River Road Unit 137, Riverdale, MD 207371231; (301) 7344387.