Abstract:It is important to evaluate the ecological risks of transgenic Bt rice on soil non-target organisms before it is widely released in the market. Rhizosphere and detritusphere formed by plant roots and organic residues that contact soil directly, and represent two distinct microsites that are highly affected by plants. Pot experiments were established to investigate the effects of rice, a variety genetically modified to express the Bt protein, KMD, and its parental line, XSD, on soil chemical, microbial properties and nematode assemblages of the two microsites. Results indicated that effects of transgenic Bt rice on soil properties varied with plant growth stage and microsite type. Compared with XSD, KMD increased the concentrations of soil dissolved organic nitrogen in detritusphere (D) at seedling and elongation stages, whereas the concentrations of ammonium in the detritusphere were decreased significantly by KMD at seedling stage (P<0.05). At maturing stage, KMD increased the concentrations of ammonium in the detritusphere and overlapped microsite (DR, i.e. the area that affected by both rhizosphere and detritusphere) significantly (P<0.05). With regard to soil microbial biomass, KMD significantly increased the concentrations of soil microbial biomass carbon (MBC) in detritusphere at seedling and maturing stages (P<0.05), whereas significantly decreased microbial biomass nitrogen (MBN) in rhizosphere at elongation and maturing stages (P<0.05). Integrated effects of the two microsites had a significant effect on concentrations of microbial biomass carbon and nitrogen as well as dissolved organic matter, total nematode numbers and nematode numbers of different feeding groups (P<0.05). Generally, KMD decreased the total nematode numbers, especially there were significantly fewer nematodes in the detritusphere of KMD than XSD at seedling and elongation stages (P<0.05). Repeated-measure ANOVA indicated that transgenic Bt rice KMD had significant effects on numbers of bacterial feeders (P<0.01) and fungal feeders (P<0.05), nematode channel ratio (NCR) and enrichment index (EI). In addition, the genetically modified KMD significantly decreased (P<0.05) the proportions of bacterial feeders and fungal feeders in detritusphere at seedling and elongation stages, and such trends were consistent to the responses of ammonium concentrations of the same microsite and growth stage. In light of soil Bt protein concentrations under KMD treatments after maturing stage, Bt rice showed no significant effect, though the content of Bt protein in straw of KMD was significantly higher than that of XSD, indicating that significant differences of effects of KMD from XSD might not result from Bt protein itself, but from the difference of other traits such as rice growth and residue chemical composition. Together, we found that microsite studies combining the rhizosphere and detritusphere would facilitate to demonstrate the responses of soil microbial properties and nematode assemblages to explore the ecological impacts and mechanisms of growing genetically modified rice.