The population consequences of spatial modularity: Insights from networks undergoing habitat destruction

Background/Question/Methods Hierarchical structuring occurs throughout many systems, including the internet, transportation infrastructures, and food webs. In such situations, systems are often composed of groups or modules that exhibit strong similarity or share function, but are only weakly linked to other groups. The concept of modularity is especially relevant to spatial ecology and could have wide-ranging implications for populations by reducing vulnerability to ongoing environmental change, altering synchrony in dynamics, and providing insight to long-standing questions of population stability. Here, I ask whether spatial modularity occurs in the movement dynamics of two species, and, if so, whether such dynamics can reduce the likelihood of extinction under habitat destruction. To do so, I couple models from social network analysis and statistical physics aimed at modeling modularity on networks with two mark-recapture datasets that vary in several orders of spatial extent: within-field movements of a cactus-feeding insect (Chelinidea vittiger) on a patchy resource, and breeding-season movements of a wide-ranging bird (Rostrhamus sociabilis) across a wetland network in peninsular Florida. Using modularity estimates from these models, I then parameterize a stochastic patch occupancy model to estimate the emergent metapopulation dynamics from this spatial structuring and contrast these dynamics to non-modular systems under habitat destruction

Results/Conclusions I find evidence for substantial spatial modularity in movements and network connectivity in both species. Spatial modularity in movement cannot be explained simply by geographic distance alone. Consistent with modeling from statistical physics, I find that spatial modularity reduces the likelihood of population extinction, increases population resilience, and alters the synchrony of populations relative to non-modular dynamics. These findings suggest that the concept of spatial modularity in population dynamics holds much promise for understanding the likelihood of population persistence under environmental change.