Context. Camera traps are an essential tool for biodiversity monitoring, yet variations in the definition of time-to-independence can influence species records and data comparability across studies. Aims. Our study examines how different time-to-independence filters (30 min, 60 min, 12 h, and 24 h) impact the detection, and estimated richness, abundance, and composition of medium- and large-bodied mammals in the Tapajós National Forest, Brazilian Amazon. We expected that (1) estimated species richness would be similar between filters and (2) apparent species composition would vary more for rarely captured species than for commonly captured species ones across time-to-independence filters. Methods. Using standardized camera trap data from 38 plots, we applied the four thresholds to generate species accumulation curves, rank-abundance curves, and apparent species composition matrices to assess how filtering alters species records and inferred community structure. Key results. Our findings indicate that estimated species richness remained consistent across filters, suggesting that studies based on species richness can integrate datasets with varying time-to-independence filters without introducing significant biases. Conversely, species abundance ranks and apparent composition varied notably across filters, particularly for rarely captured species, which exhibited greater sensitivity to temporal aggregation. While commonly captured species were consistently detected across all filters, certain species that were relatively infrequently recorded using short time-to-independence filters (i.e., 30 min and 60 min) became relatively more frequently detected with longer filters (i.e., 24h), altering their classification from 'rare' to 'common'. Conclusions. The stability of species richness across filters supports the use of datasets with different time frames in macroecological studies, enabling large-scale biodiversity assessments. However, the temporal sensitivity of apparent species composition, especially for rarely captured species, underscores the necessity of caution in community-level analyses. Implications. Studies investigating species interactions or community structure should account for the effects of time-to-independence filters when comparing datasets. Standardizing camera trap protocols or implementing methodological adjustments for different time-to-independence filters will be crucial for enhancing data comparability in ecological studies and ensuring robust conservation assessments.
Silva et al. (Sun,) studied this question.