- Species abundance distributions of moth and beetle assemblages, in addition to beetle feeding guilds, were examined using data collected by light traps in a cool-temperate deciduous forest at Hokkaido, Japan.
- Four types of species abundance distributions with a continuous shift in shape (broken stick, log-normal, power function, and dispersal-limited multinomial) were fitted and compared, and then the deviation of observed data from that expected for the neutral model (dispersal-limited multinomial distribution) was tested using species evenness and richness.
- Significantly better fits were obtained for the neutral model than for the other models for all assemblages and all beetle guilds. This result implies that the better fits of the neutral model might be characteristic of transient assemblages having an excess of rare species.
- The evenness of species abundance distributions for assemblages and guilds, except for moths associated with the understorey and predatory beetles, significantly deviated from that expected by the neutral model. The departure of beetle guilds (detritivores, herbivores, and xylophages) from the neutral model suggests their microhabitat dependence.
- The neutral model may prove to be a viable null hypothesis for examining the species abundance distributions of insect assemblages. In this study, we found that although stochastic dynamics appeared to have an increasing influence on insect community assembly, there are also complex biological processes still likely to be present.
- The fits of the neutral model suggest that habitat connectivity and microhabitat diversity are potentially important for conserving moth and beetle diversity in a temperate forest.