Abstract

Anthropogenic habitat alteration has long been neglected as a factor in the analysis of predictability patterns in biological communities. We tested this factor by investigating anuran leaf litter assemblages in primary and secondary forests of Taï National Park, Ivory Coast, during two years. We measured predictability of assemblage composition by analyzing correlations between the off-diagonal elements of distance matrices based on (1) species distribution, (2) environmental characteristics, and (3) geographic distance. Pairwise correlations between matrices were significant in all cases when considering data pooled across time and habitats. A different pattern emerged when data were split according to season and disturbance level (i.e., primary vs. secondary habitats). Assemblage composition in primary habitats was correlated with geographic proximity of sites exclusively, indicating otherwise stochastic recruitment from a regional species pool at the local community level. In contrast, assemblage composition in secondary habitats was predictable based on environmental parameters, not geographical proximity. This can be inferred to be the result of a strong local site filter effect (i.e., physiologically more-restrictive conditions within secondary forest habitats, especially due to an altered microclimate). Results were consistent throughout seasons. The observed transition in predictability patterns indicates that anthropogenic disturbance not only affects system descriptors, such as species richness, abundance, and diversity, but may also alter the system's dynamics.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call