3. RESULTS
A total of 19,053 beetle individuals were recorded across all sampling sites during the study period, which represents 64 beetle families. The rarefaction and extrapolation curves, adjusting for the number of specimens collected in natural forests and plantation areas, clearly revealed that beetle community composition in natural forests had a higher overall cumulative diversity of families compared to plantations (Figure 2). We also determined that the highest diversity was sampled at site 8, 3, and 5 in natural forests respectively (Figure S1).
Beetle community composition did not appear to be dependent on any of the variables tested, according to the PERMANOVA analysis (Table 2). However, visual examination of NMDS indicated that the beetle families collected differed between trapping methods (Figure S2), and that beetle communities were largely overlapping between natural forests and plantation areas (Figure 3).
The richness of beetle communities (in terms of the estimated number of families per site) was found to be dependent on the trapping method (P = 0.006) (Table 3), but was not influenced by the other three variables of interest (landscape context, temperature, and precipitation). In addition, there was also no significant effect of any variable on beetle community diversity as estimated by Shannon and Simpson indices (Table S2, S3).
There was a strongly significant effect of all tested variables, except temperature, on the variation of beetle abundance across sites (Table 4). Specifically, the abundance of beetle communities was significantly higher (ca. ×3 on average) in the natural forest compared to plantations (P = 0.0014), differed between trapping methods (P = 0.0001), and was positively associated with precipitation (P = 0.0146) (Figure 4). When analyses were computed for each family separately, we highlighted a similar pattern of higher abundances in the natural forest compared to plantations; we also identified significant effects of temperature and precipitation as well as trapping methods (Table S1).