COS 194-3 - Determinism rather than dispersal limitation or priority effects governs community assembly across an artificial freshwater pond array

Abstract: The patterns governing community assembly have long been of interest to ecologists, as disturbances regularly perturb biotic communities, initiating new assembly processes that dictate where and when we find different species. The goal of the current project was to evaluate the roles of dispersal limitation and priority effects in influencing the determinism of aquatic community assembly. We constructed 36 artificial ponds at the Koffler Scientific Reserve, Ontario, Canada on a 3 x 3 grid, with four ponds at each of the nine nodes. The artificial ponds spanned three distances from existing source ponds and half were given an initial nutrient pulse to establish different priority effects. Over the course of 3 years, all ponds were surveyed 13 times using a combination of visual, dipnet, and stovepipe methods to determine abundances of all aquatic vertebrates and macroinvertebrates. 144,047 individuals from 40 taxa were identified and used to describe the community assembly process. Priority effects from the initial nutrient pulse were evident one month after pond construction, but by 1.5 months post-construction had disappeared. Similarly, the community that had established one month post-construction had minimal bearing on the final community composition at the end of 3 years (variance explained ranging from 9-18%). Priority effects thus appear to be minimal. Similarly, distance from source ponds was never a significant predictor of community composition, and also did not influence the variation in community composition between the four ponds at each node of the array. Moreover, taxa that one would expect to be dispersal limited (e.g., amphipods, leeches, snails) appeared at the most distant ponds at the same rate as in the artificial ponds closest to the source ponds. Thus, dispersal limitation appears to be minimal as well. Instead, community assembly appears to be very deterministic. All ponds, except one, first increased and then decreased on the first ordination axis, and all ponds increased on the second ordination axis through time. The first of these patterns is associated with a negative quadratic relationship between both abundance and richness with time. Both abundance and richness reached maximum values during the second year of community assembly. The second pattern is associated with a steady increase in predators. Abiotic variables were measured at 8 time points during the community assembly process. Among the abiotic variables, ORP, which is often considered a measure of pond health, mirrored the pattern in abundance and richness.