There is controversy about the means of environmental risk assessment and management of transgenic crops because of the
speed and quantity of these cr ops being planted, particularly those containing genes to produce the insecticid al toxin from Bacillus thuringiensis (Bt). Most research on
Bt crops has foc used on invasiveness, gene flow to indigenous organisms, development of resistance i n target pests, and direct or indirect effects on non-target organisms and ecos y stems. However, after the commercially usable portion of Bt crops has been harve sted, the remainder of plant biomass containing the toxins is usually incorporat ed into soil. The toxin is also introduced into soil in root exudates and from p ollen during tasseling. This is raising concern about the effects of residual Bt toxins on the
soil ecosystem. This review summarizes current research in six important areas related to the en vironmental fate of Bt toxin from Bt crops and its ecological effects in soil. ( 1) Bt toxins absorb rapidly onto mined clay minerals and the clay-size fraction of soil. Rapid adsorption and binding of Bt from root exudates and Bt corn biom ass onto surface-active particles in soil has been demonstrated in both in vitr o and in situ studies. (2) Binding of the toxin on surface-active particles can r educe their availability to microbes and hence biodegradation, which is probably responsible for their persistence in soil. Free Bt toxin is readily utilized as sole sources of carbon and nitrogen by pure and mixed cultures of microbes, whe reas the bound toxins are not utilized as a source of carbon and only slightly a s a source of nitrogen. The literature estimates the length of toxin persistence based on half-life' values range from 8～17 d for purified toxin and from 2～ 41 d for transgenic corn, cotton, and potato biomass. (3) The toxin released from b iomass or in root exudates of Bt corn, however, remains larvicidal for at least 180 d. Insect bioassays indicate that the Cry 1Ab protein of Bt corn tissue has an estimated DT 50 of 1.6 d and a DT 90 of 15 d. (4) The toxi n from Bt corn remai ning in soil is not taken up by subsequent non-Bt crops, such as corn, carrot a n d radish. The toxin has been found in the guts and casts of earthworms in soil p lanted with Bt corn or amended with Bt corn biomass, but it is cleared from the guts within 2～3 d after transfer into fresh soil. (5) Accumulated and persisten t Bt toxin in soil can subsequently be leached into groundwater or move horizonta lly into surface waters by rain, irrigation, snow melts, etc., as has been obser ved with heavy metals. (6) The Bt corn toxin has no apparent affect on total num bers of culturable bacteria, fungi, protozoa and nematodes. Bt cotton lines 247 and 249 did produce a transient but significant increase in number of culturable aerobic bacteria and fungi. The populations of culturable, aerobic bacteria an d fungi, and the species of fungi on Bt potato plants differed minimally from no n-Bt potato. The nematode populations in the soil surrounding Bt tobacco litter b ags were greater and had a different trophic group composition than in soil surr ounding the parental tobacco. There are few studies that emphasize soil ecologic al processes. Binding of Bt toxin onto surface-active particles can result in its accumulatio n in the environment to concentration levels that may enhance the control of targ et pests but may also constitute a hazard to non-target organisms. There are ma n y interactions that can occur between organisms in soil including, predation, co mpetition, antagonism, and mutualism. Therefore it is important that risk assess ment procedures should be targeted not only to specific organisms, but also orga nismal functions and soil processes. We suggest that risk assessment studies sho uld begin by evaluating the effects of Bt crops on both microbial communities an d processes. The studies should be of sufficient duratio