Comparing biodiversity found with insect samplers and camera traps
In total, iDNA and camera trapping methods revealed 90 vertebrate species across the sampled landscape (Table 2). iDNA alone described 82 species with carrion fly data describing the most diversity of any sampler with 62 species while sandfly data revealed 49 species and mosquito data revealed 16 species. Camera trap data revealed 28 vertebrate species and added 8 species to the diversity already described with iDNA.
Species accumulation curves show a statistically significant difference in species richness gleaned from each iDNA source and camera trapping (Fig. 2). Carrion fly data and mosquito data are extrapolated to match the number of sandfly samples, and consequently the extrapolated curves provide less certainty than the interpolated curves. However, the confidence intervals show significant separation in the total species richness from each iDNA source across both the interpolated and extrapolated curves. Using carrion flies as a measure of biodiversity reveals a significantly greater number of vertebrate species across the landscape than both sandflies and mosquitos. Additionally, the total number of carrion flies analyzed (n = 1,759) was lower than both the total number of sandflies (n = 48,686) and mosquitos (n = 4,776) analyzed. The species accumulation curve for camera trap data shows that species richness plateaus after approximately 1,500 camera days meaning that few to no new species are revealed by camera traps even with continued camera days (Fig. 2). The species richness from camera traps was most similar to the species richness from the extrapolated curve for the pooled mosquito samples, which was the least efficient iDNA sampler in describing vertebrate diversity (Fig. 2; Table 1).
Comparisons of species’ RAI from iDNA samplers and camera traps show differences in diversity profiles across taxa groups (Fig. 3). As expected, camera traps showed a high diversity of carnivore and ungulate species and a low diversity of arboreal species or birds. Camera traps also did not capture the occurrence or diversity of domestic species and rodent species particularly when compared to the diversity of these groups found using iDNA methods. Carrion fly data consistently showed a high diversity of species in each taxa group. Sandfly data also showed high diversity of species in most taxa groups with an especially high diversity and relative abundance of armadillo species (Fig. 3; Table 2). Lastly, mosquito data showed the lowest species diversity for many of the taxa groups and only showed a high diversity of domestic species when comparing mosquito data to the other datasets (although carrion fly data still displayed greater diversity and relative abundance from domestic species). When species incidence data was parsed at the site level, camera traps showed the highest median species richness at sites (Fig. 4) and a consistent diversity profile across sites (Appendix S1: Fig. S3) indicating that camera trapping consistently samples the same species across sites.