Landscape context influences the abundance of amphibians and the strength of their food web interactions in small ponds

Trait variation in patchy landscapes: Morphology of spotted salamanders (Ambystoma maculatum) varies more within ponds than between ponds

The influence of intraspecific trait variation on species interactions makes trait-based approaches critical to understanding eco-evolutionary processes. Because species occupy habitats that are patchily distributed in space, species interactions are influenced not just by the degree of intraspecific trait variation but also the relative proportion of trait variation that occurs within- versus between-patches. Advancement in trait-based ecology hinges on understanding how trait variation is distributed within and between habitat patches across the landscape. We sampled larval spotted salamanders (Ambystoma maculatum) across six spatially discrete ponds to quantify within- and between-pond variation in mass, length, and various metrics associated with their relationship (scaling, body condition, shape). Across all traits, within-pond variation contributed more to total observed morphological variation than between-pond variation. Between-pond variation was not negligible, however, and explained 20-41% of total observed variation in measured traits. Between-pond variation was more pronounced in salamander tail morphology compared to head or body morphology, suggesting that pond-level factors more strongly influence tails than other body parts. We also observed differences in mass-length relationships across ponds, both in terms of scaling slopes and intercepts, though differences in the intercepts were much stronger. Preliminary evidence hinted that newly constructed ponds were a driver of the observed differences in mass-length relationships and morphometrics. General pond-level difference in salamander trait covariation suggest that allometric scaling of morphological traits is context dependent in patchy landscapes. Effects of pond age offer the hypothesis that habitat restoration through pond construction is a driver of variation in trait scaling, which managers may leverage to bolster trait diversity.

Does regional diversity recover after disturbance? A field experiment in constructed ponds

The effects of disturbance on local species diversity have been well documented, but less recognized is the possibility that disturbances can alter diversity at regional spatial scales. Since regional diversity can dictate which species are available for recolonization of degraded sites, the loss of diversity at regional scales may impede the recovery of biodiversity following a disturbance. To examine this we used a chemical disturbance of rotenone, a piscicide commonly used for fish removal in aquatic habitats, on small fishless freshwater ponds. We focused on the non-target effects of rotenone on aquatic invertebrates with the goal of assessing biodiversity loss and recovery at both local (within-pond) and regional (across ponds) spatial scales. We found that rotenone caused significant, large, but short-term losses of species at both local and regional spatial scales. Using a null model of random extinction, we determined that species were selectively removed from communities relative to what would be expected if species loss occurred randomly. Despite this selective loss of biodiversity, species diversity at both local and regional spatial scales recovered to reference levels one year after the addition of rotenone. The rapid recovery of local and regional diversity in this study was surprising considering the large loss of regional species diversity, however many aquatic invertebrates disperse readily or have resting stages that may persist through disturbances. We emphasize the importance of considering spatial scale when quantifying the impacts of a disturbance on an ecosystem, as well as considering how regional species loss can influence recovery from disturbance.

An extended patch-dynamic framework for food chains in fragmented landscapes

Habitat destruction, a key determinant of species loss, can be characterized by two components, patch loss and patch fragmentation, where the former refers to the reduction in patch availability, and the latter to the division of the remaining patches. Classical metacommunity models have recently explored how food web dynamics respond to patch loss, but the effects of patch fragmentation have largely been overlooked. Here we develop an extended patch-dynamic model that tracks the patch occupancy of the various trophic links subject to colonization-extinction-predation dynamics by incorporating species dispersal with patch connectivity. We found that, in a simple food chain, species at higher trophic level become extinct sooner with increasing patch loss and fragmentation due to the constraint in resource availability, confirming the trophic rank hypothesis. Yet, effects of fragmentation on species occupancy are largely determined by patch loss, with maximal fragmentation effects occurring at intermediate patch loss. Compared to the spatially explicit simulations that we also performed, the current model with pair approximation generates similar community patterns especially in spatially clustered landscapes. Overall, our extended framework can be applied to model more complex food webs in fragmented landscapes, broadening the scope of existing metacommunity theory.