COS 5-2
Linking marine metacommunity patterns and processes across environmental gradients
Patterns of species diversity reflect the interplay between biotic and abiotic factors operating at multiple spatial scales. To unravel this complex relationship, ecologists are increasingly turning to statistical approaches that partition the variation in (meta)community structure into a spatial and an environmental component. The relative magnitude of these two components is then used to determine the strength of niche vs. neutral processes. If dispersal is not limiting, variation in community structure is expected to be dominated by the environmental component, as species are able to reach their environmental niche. Conversely, if dispersal is limiting, variation in community structure is expected to be dominated by the spatial component, since dispersal will act as a barrier preventing species from reaching their environmental niche. Dispersal limitation can thus erode the relationship between community structure and the environment. However, large-scale dispersal can also erode the relationship between community structure and the environment by allowing source-sink dynamics to operate across spatially-structured environmental gradients. We used a multi-year intertidal dataset to understand the drivers of (meta)community structure across an 1,800 km environmental gradient. To do so, we split our dataset into species groups characterized by different dispersal abilities and then applied variation partitioning.
Results/Conclusions
Our analyses show that the spatial and environmental components explain anywhere from 30-70% of the total variation in community structure across different species groups and years. Interestingly, we also find a significant positive relationship between the magnitude of the spatial component and dispersal ability across species groups. Hence, species characterized by greater dispersal ability had a larger proportion of their variation explained by the spatial component than species characterized by weaker dispersal ability. These results suggest that the ecological interpretation of the spatial component obtained from variation partitioning methods can be difficult. Indeed, a large spatial component can be interpreted as evidence of either (i) dispersal limitation acting as a physical barrier by keeping species from their environmental niche, or (ii) large-scale dispersal promoting source-sink dynamics that erode the relationship between community structure and environmental gradients. Hence, incorporating a priori information about species dispersal ability into variance partitioning analyses can be critical for correctly interpreting the relationship between ecological patterns and processes across scales.