Weighting effective number of species measures by abundance weakens detection of diversity responses

The effective number of species (ENS) has been proposed as a robust measure of species diversity that overcomes several limitations in terms of both diversity indices and species richness (SR). However, it is not yet clear if ENS improves interpretation and comparison of biodiversity monitoring data, and ultimately resource management decisions. We used simulations of five stream macroinvertebrate assemblages and spatially extensive field data of stream fishes and mussels to show (a) how different ENS formulations respond to stress and (b) how diversity–environment relationships change with values of q, which weight ENS measures by species abundances. Values of ENS derived from whole simulated assemblages with all species weighted equally (true SR) steadily decreased as stress increased, and ENS-stress relationships became weaker and more different among assemblages with increased weighting. The amount of variation in ENS across the fish and mussel assemblages that was associated with environmental gradients decreased with increasing q. Synthesis and applications. Species diversity is valued by many human societies, which often have policies designed to protect and restore it. Natural resources managers and policy makers may use species richness and diversity indices to describe the status of ecological communities. However, these traditional diversity measures are known subject to limitations that hinder their interpretation and comparability. The effective number of species (ENS) was proposed to overcome the limitations. Unfortunately, our analyses show that ENS does not improve interpretability of how species diversity responds to either stress or natural environmental gradients. Moreover, incorporating the relative abundance of individuals in different species (evenness) into diversity measures as implemented in ENS can actually weaken detection of diversity responses. Natural resources managers and policy makers therefore need to be cautious when interpreting diversity measures, including ENS, whose values are jointly influenced by richness and evenness. We suggest that both researchers and practitioners measure and report three aspects of diversity (species richness, evenness, and composition) separately when assessing and monitoring the diversity of ecological communities.