3.1 Community diversity
A total of 1179 small mammal individuals were captured in our study, belonging to 30 species (Appendix), including 17 glire species (accounting for 56.7%) and 13 insectivore species (43.3%). The rarefaction curves showed that our survey results adequately represented the small mammal diversity in each forest type (Fig. 2). At the landscape scale, 28, 24 and 27 small mammal species were recorded in the primary, secondary and plantation forests, respectively. Two species (Niviventer andersoni and Berylmys bowersi ) were recorded only in the primary forests, Chodsigoa lamula was recorded only in the secondary forests, and Scaptonyx fusicauduswas the unique species sampled in the plantation forests (Appendix). All 17 glire species were recorded in the primary forests, accounting for 60.7% of the primary forest species. Thirteen glire species were recorded in the secondary forests, accounting for 54.1% of secondary forest species, while 15 glire species were recorded in the plantation forests, accounting for 55.6% of plantation forest species. There was no difference in the ratio of glire species among the three forest types (chi-squared test, χ2 =.259, df =2, p=0.879). In addition, FD and PD were highest in the primary forests, followed by the plantation forests and secondary forests (Fig. 3). At the site scale, the TD, FD and PD at the plantation forest sites were significantly lower than those at the primary forest sites (one-way ANOVA: df =2, p=0.003; Kruskal-Wallis Test: df =2, p=0.046; and one way ANOVA: df =2, p=0.043, respectively), and there was no significant difference in TD, FD or PD between the secondary and primary forest sites (Fig. 3). In addition, NMDS ordinations of species assemblage showed that community composition of small mammals was similar among the three forest types (Fig. 4).