Estimating changes in species abundance from occupancy and aggregation
Predicting the change in abundance is pivotal for evaluating species' current conservation status and population viability. Empirical works have suggested that species with an increasing abundance have a more aggregated distribution than those with a declining abundance (namely, the change-aggregation hypothesis, CAH). Here we introduced an improved negative binomial distribution model of the occupancy-abundance relationship (OAR) to estimate the change in abundance from changes in occupancy or aggregation. Analysis of the model suggests that (i) in general the change in abundance is synchronized with the change in occupancy when the level of environmental heterogeneity remains constant, and (ii) there could exist a threshold of the population density above which the CAH is no longer valid. Tests using data of epigaeic ants in Fynbos of South Africa collected from different seasons and macro-invertebrates from different localities in streams of central Spain verified these model propositions and thus support the use of this model as a monitoring method for assessing species persistence. Results suggest that the change in abundance can be estimated from the change in occupancy often obtained from cost-efficient presence-absence records, and a revision of the traditional CAH is necessary to capture the threshold phenomenon in the change-aggregation relationship. This work thus signifies the use of the three distinct but related concepts of population structure (i.e. occupancy, abundance and aggregation) in conservation biology. © 2012 Gesellschaft für ökologie.