Abstract:Pomacea canaliculata (Lamarck) is an invasive aquatic gastropod endangering rice production in Southern China, with a feeding preference for rice seedlings in paddy fields. P. canaliculata is listed as one of the top 100 invasive species by the Invasive Species Group of the World Conservation Union. P. canaliculata mainly feeds on a diverse range of macrophyte species. We investigated the invasive effects of P. canaliculata on the aquatic macrophyte plant community in paddy fields. Our experimental macrophyte community consisted of Oryza sativa L., Pistia stratiotes L., Eichhornia crassipes, Myriophyllum aquaticum, Alternanthera philoxeroides, Monochoria vaginalis. The control and three treatments used different P. canaliculata densities: control (P. canaliculata-free); low (4 individuals/m2); medium (8 individuals/m2) and high (12 individuals/m2). We analyzed the effect that P. canaliculata had on the macrophyte community using the compound food preference test. Our results showed: 1) O. sativa densities in the three treatments decreased rapidly one week after P. canaliculata was introduced. In herbivore, in the P. canaliculata high-density treatment, O. sativa density declined approximately 70%. However, two weeks after P. canaliculata was introduced, the herbivore effect on O. sativa density decreased. 2) After eight weeks P. canaliculata reduced A. philoxeroides and M. vaginalis densities (30% and 25%, respectively) in the high-density herbivore treatment and inhibited the growth of A. philoxeroides and M. vaginalis. The plant densities of A. philoxeroides and M. vaginalis were higher in the low-density herbivore treatment compared with those in the medium and high-density herbivore treatments. However, P. canaliculata showed no significant effect on the plant densities of E. crassipes, P. stratiotes and M. aquaticum. 3) The daily density rate of increase was negative for O. sativa, and significantly different in the low, medium and high-density herbivore treatments. The daily density rate of increase for O. sativa showed a significant decrease compared with the P. canaliculata-free control. The daily density rate of increase for A. philoxeroides and M. vaginalis was positive, where the rate was significantly lower compared with the P. canaliculata-free control. Concurrently, the daily density rates of increase were not significantly influenced by P. canaliculata for P. stratiotes, E. crassipes and M. aquaticum. We found that P. canaliculata populations increased rapidly. The daily density rate of increase for P. canaliculata populations in the three different density treatments were all high, but not significantly different. The food preference frequencies of P. canaliculata on O. sativa, A. philoxeroides and M. vaginalis were all significantly higher compared with E. crassipes, P. stratiotes and M. aquaticum. 4) Eight weeks after the introduction of P. canaliculata the Simpson, Shannon and Homogeneity biodiversity indices showed a significant decline in the aquatic macrophyte community. The Simpson index values showed the decline in order of: the control > low- > medium-and high-densities. The Shannon and Homogeneity indices showed the decline in order of: the control > low- > medium- > high-densities. We conclude that P. canaliculata showed significantly different feeding preference on macrophyte plants species. The invasion of P. canaliculata would significantly endanger aquatic macrophyte community biodiversity in the paddy fields of China.