Does context-dependency among metacommunity processes require redefining the metacommunity framework?

Link To Share This Poster: https://cdmcd.co/abyvGJ

Jurek Kolasa, Matthew P. Hammond and Joyce Yan, Biology, McMaster University, Hamilton, ON, Canada

Background/Question/Methods
Metacommunity framework involves multiple dimensions. It develops predictions from states and process such as community separation, dispersal, habitat heterogeneity, and species interactions. Yet, a major complication emerges: the states and processes may modify one another reciprocally in response to impacts of metacommunity dynamics – show context-dependency, C-D. This can make predictions unpredictable or unreliable at best. Remedies for treating milder cases of C-D are readily available but severe cases, where some contributory factors cause non-linear changes in others are not. We need to understand the degree to which metacommunity framework may be vulnerable when its core notions (habitat structure, dispersal, species interactions) are being functionally entangled. To illustrate the significance of such interdependence, we test two hypotheses. One that changing just one parameter while holding others constant, can alter inference of a study and the second that the severity of context-dependency increases when core metacommunity dimensions interact and transform one another through a variety of mechanisms. The test aims to demonstrate that context-dependency may introduce a high degree of ambiguity into the meaning of results.
Results/Conclusions
We use an agent-based, spatially explicit model (NetLogo) of metacommunities, where some of the core dimensions are interdependent. We asked how different configurations of generalized dimensions (habitat heterogeneity, specialization, species interactions and dispersal) affected richness, S. Specifically, we (1) allowed species interactions to change across a range values in these factors; and (2) allowed species to affect habitat quality by affecting resource levels, which dynamically modifies habitat heterogeneity.
We found that any factor alone could not predict richness, that specialization of species had a great effect on their response to other factors, and that both richness and species specialization modified habitat suitability. This suggests that any specific hypothesis will have to confront a full range of potential outcomes because other, not controlled, dimensions may affect metric of interest such as species richness, functional groups, and many others.
We thus conclude that C-D predicts a dauntingly vast space of possible empirical outcomes and interpretations. However, we offer a hope in that these results reveal clear and more general patterns that will eventually allow a more organized, not less but more efficient, pursuit of understanding of metacommunity mechanisms.