Robustness of biodiversity surrogates to temporal variation and imperfect detection

Common species can predict overall species richness and composition, serving as a shortcut to biodiversity and supporting well-informed management of areas. However, the impacts of temporal variation in underlying species occurrence and imperfect detection on surrogacy effectiveness are unknown. We hypothesized that common species will represent total species composition both across sites and along environmental gradients and that this surrogacy effectiveness is maintained over time and robust to imperfect detection.

We sampled ground-dwelling ants using pitfall traps over 25 km² on Amazon. We measured the congruence between a complete dataset and subsets containing a smaller number of species based on their commonness and compared the results among three sampling years. Further, we simulated virtual assemblages to determine the effect of species detection probability on surrogacy effectiveness and compared the costs associated with using all species vs. common species only. The common species composition responded to environmental gradients similarly to the overall species composition, regardless of year. The simulated data showed similar trends, although including imperfect detection increased the minimum number of common species required to recover overall patterns reasonably.

This study represents the first report on the effectiveness of common species in predicting ant assemblage patterns over time. We propose that imperfect detection should be considered in biodiversity surrogacy studies, especially regarding its impact on surrogate effectiveness over time.