Testing the heterospecific attraction hypothesis with time-series data on species co-occurrence.

The distributional patterns of actively moving animals are influenced by the cues that the individuals use for choosing sites into which they settle. Individuals may gather information about habitat quality using two types of strategies, either directly assessing the relevant environmental factors, or using the presence of conspecifics or heterospecifics as an indirect measure of habitat quality. We examined patterns of heterospecific attraction with observational time-series data on a community of seven waterbird species breeding in artificial irrigation ponds. We fitted to the data a multivariate logistic regression model, which attributes the presence-absence of each species to a set of environmental and spatial covariates, to the presence of con- and heterospecifics in the previous year and to the presence of heterospecifics in the same year. All species showed a clear tendency to continue breeding in the same sites where they were observed in the previous year. Additionally, the presence of heterospecifics, both in the previous year and in the same year, generally increased the probability that the focal species was found breeding on a given pond. Our data thus give support for the heterospecific attraction hypothesis, though causal inference should be confirmed with manipulative experiments.