Abstract: To determine the effects of different sampling methods on the structural characteristics of insect communities, the wet
evergreen broad-leaved forest on the eastern of Cangshan mountain was taken as the study site, and four sampling methods, which are the Malaise trap method （MS）, the ground trap method （XJ）, the light trap method （DY）, and the net method （SW）, were used to conduct the survey of insect communities for one year, and compare the structural characteristics of insect communities obtained by the different sampling methods, analyze the differences in results, and clarify the advantages, disadvantages, and substitutability of different sampling methods. The results showed that among the 59 families and 13 340 insects collected, 41 families and 6 896 heads were captured by MS, 6 201 specimens belonging to 31 families were captured by XJ, 141 specimens belonging to 24 families were captured by DY and 102 specimens belonging to 22 families were captured by SW. Comparing the number of families, species, and individuals collected by different sampling methods, the MS is the best, followed by the XJ and DY, and the WB is the worst. There are differences in the advantageous groups of different sampling methods. There are significant differences in community similarity among the four methods （Cs≤0.60）. Except for the combinations of SW+XJ, DY+XJ and SW+XJ+DY, other combinations have similar sampling effects and can be replaced.

Key words: sampling method, community survey of insect, similarity, diversity, method combination