PS 13-116 - The role of habitat composition and configuration on species turnover among temperate seagrass communities

Background/Question/Methods

Seagrass beds are among the most productive ecosystems worldwide, forming patchworks of shallow-water habitats that vary in spatial arrangement across which species are distributed. While niche filtering and spatial arrangement of habitats are predicted to affect compositional differences across landscapes, attributing patterns of composition to mechanisms such as environmental affinities and dispersal remains a challenge. We hypothesize that (1) compositional differences increase with environmental differences between sample locations as well as their surrounding areas, and (2) compositional differences are greater across areas separated by unsuitable habitat (e.g., deep water, unvegetated) due to decreased movement of individuals.

We tested these hypotheses using beach seining samples of temperate nearshore fish communities on the central coast of British Columbia, Canada. We fit generalized dissimilarity models to fish composition to test for non-linear responses of compositional difference across environmental gradients and geographic distances. We used model selection to test the importance of local environment (e.g., vegetation cover, exposure, temperature) and proximity to potentially important landscape features (e.g., kelp, deep water, freshwater inputs). Additionally, we created and tested several biologically-guided metrics of pairwise connectivity using GIS generated resistance landscapes.

Results/Conclusions

We found significant positive relationships between fish community compositional difference and vegetation cover, ocean exposure, distance to freshwater and geographic connectivity. Vegetation cover accounted for almost half of the explained difference in composition (45.8%), but only when vegetation cover was low (< 50%). Distance to freshwater affected composition, but only when freshwater sources were nearby (< 1500m). Ocean exposure affected composition, with the greatest compositional differences between higher exposure values. A connectivity metric with increased resistance across deep water was selected in preference to over-water distance accounting for 16.6% of the explained difference in composition, where the effect may become weaker at greater distances.

The non-linear response of community composition to environmental variables highlights the importance of using such approaches in modeling beta-diversity, which can help with conservation prioritization. For example, the influence of freshwater sources on turnover may not extend much beyond 1500 m. The importance of deep water in limiting connectivity suggests that islands or shorelines separated by deep channels have greater compositional differences that may mitigate homogenization effects. Using biologically-guided connectivity metrics can refine our understanding of the spatial forces that shape fish distribution and can help with regional management and stock identification.