Evaluating baselines for long‐term ecological monitoring of biodiversity trends: insights from the US National Ecological Observatory Network carabid data

Evidence is mounting that rapid environmental change threatens global insect biodiversity, underscoring the need for informed conservation strategies that protect both species and the ecosystem services they provide. Armed with accurate baseline community data, long‐term continental‐scale monitoring projects are invaluable for detecting and predicting responses to ecological change. However, high species diversity and temporal variability in population sizes can hinder our ability to establish baselines, and, thus, obscure, exaggerate, or reverse temporal trends in long‐term insect data. With its scale and consistent protocol, the US National Ecological Observatory Network (NEON) carabid pitfall trapping data provides an excellent case study for evaluating sampling effort. We use species incidence‐frequencies calculated from more than 200 000 identified carabids across up to 10 years of sampling and 46 field sites to extrapolate asymptotic richness and Shannon diversity. We find that the completeness of observed species richness and diversity is negatively related to year‐to‐year species turnover and diversity metrics themselves, but improves with increasing sampling duration. While observed Shannon diversity converges to asymptotic estimates within a few years, we find that NEON's intensive sampling is unlikely to capture all species, even if no biodiversity loss occurs over its 30‐year span. If the mechanisms driving these patterns can be understood, they hold important implications for optimizing sampling designs in studies focused on ecological change detection, particularly for diverse and temporally variable taxa. Our findings underscore the critical importance of long‐term monitoring and prompt reconsideration of how we interpret trends in existing biodiversity data, given the complexity of establishing robust baselines.