Objective To study spatial distribution patterns and sampling techniques for the larvae of rice water weevil, Lissorhoptrus oryzophilus, under seedling throwing transplanting and conventional seedling transplanting in Guangxi, and to provide a theoretical basis for insect survey and integrated control of rice water weevil.

Method The spatial distribution patterns of rice water weevil larvae were investigated by aggregation indices, Taylor's power principle and Iwao's regression methods. Different sampling techniques including five-point, linear parallel, double diagonal, checkerboard and zigzag sampling were compared.

Result The larvae of rice water weevil had an aggregated spatial pattern and fitted to negative binomial distribution with density dependence under two transplanting methods. The basic components of the distribution were colonies of individuals attracting each other. Different cultivation methods did not change the distribution patterns. The formula of field theoretical sampling based on Iwao's regression was N=t²(2.256/m+0.240)/D² under seedling throwing transplanting, and N=t²(4.878/m+0.126)/D² under conventional seedling transplanting. The coefficients of variation of rice water weevil larvae in rice field under both seedling throwing transplanting and conventional seedling transplanting were the lowest using double diagonal sampling, which were 63.048 8% and 84.796 0% respectively.

Conclusion Rice water weevil larvae show an aggregated distribution pattern due to their own characteristics as well as environmental factors. The double diagonal sampling was the optimum method to investigate rice water weevil larvae in paddy field under both transplanting methods.