WMV2; that information would have shown whether the viruses were present that year. Because of a severe drought, no FLCP plants were sampled in Texas. On the basis of anecdotal reports and this one-season survey of only nine locations, USDA-APHIS concluded (USDA 1994b):

Given the available knowledge, it is unlikely that resistance to ZYMV and WMV2 infection will confer a selective advantage or be maintained in the FLCP populations. Surveys of natural FLCP populations for the incidence and severity of ZYMV and WMV2 infections suggest that resistance to these viruses will confer little, if any, selective advantage, because disease caused by these viruses is apparently not among the factors important to the survival or reproductive success of FLCP.

The issue merits further empirical study, especially because selectively neutral crop genes are often maintained in the gene pools of wild and weedy plants (for example, Whitton et al. 1997). Also, the selective benefit of such genes could vary geographically and over time, and a small-scale survey like Asgrow's could easily miss infections that affect long-term population dynamics. No studies other than the one just mentioned have been conducted to determine whether viral diseases are important to the survival or reproductive success of FLCP.

Questions about the weediness of FLCP were addressed again when Asgrow requested deregulation of the CZW-3 squash in 1995. The CZW-3 squash is resistant to CMV (cucumber mosaic virus), as well as to ZYMV and WMV2. To the committee's knowledge, USDA-APHIS did not obtain any new, original data on the agroecological factors that regulate FLCP populations in their geographic range (USDA 1996b). Instead, the 1996 deregulation of the CZW-3 again relied on the 1993 Asgrow survey and anecdotal evidence from interviews with several weed scientists in Arkansas (USDA 1996b). Some of the reasoning in the permit document is not well supported. For example, USDA states that the arrival of ZYMV in the United States in the 1980s did not lead to decreases in FLCP populations, as would be expected if ZYMV suppressed FLCP populations. That statement is puzzling in light of the statement that FLCP populations apparently have become less of a weed problem in the 1990s, presumably because of changes in available herbicides (USDA 1996b). Without any studies of FLCP populations, one cannot rule out the possibility that viral diseases and other factors (such as frequency of floods) have also played a role in suppressing FLCP.

In summary, the committee concludes that

USDA's assumption that transgenic resistance to viruses will not affect the weediness of wild C. pepo might be correct, but longer-term empirical studies are needed to determine whether this is true.