Using social networks to evaluate the potential of resource distribution as a predictor of population structure in a metapopulation of the Eastern collared lizard in the Missouri Ozarks

Background/Question/Methods

Preservation of genetic variation is a vital aspect of conservation management. While the impact of environmental factors, such as resource distribution, on maintenance of population size is well understood for many species, how environmental factors influence maintenance of genetic variation is less clearly understood.  One way to better predict whether habitat will be a good steward of genetic variation is to examine the impact of environmental features, like the distribution of resources, on population structure. We examined six populations of the Eastern collared lizard, Crotaphytus collaris collaris; three of which existed on glades where the resource of perching/hiding habitat was homogeneously distributed across the glade, and three in which the resources were clustered together, separated by grassy, less optimal habitat. We used home range analysis to create social networks to evaluate social structure, and behavioral observations to quantify levels of territorial behavior.  We conducted a genetic analysis to test if the patterns observed were reflected in reproductive output. We assigned offspring to potential parents using data from 6 microsatellite markers and the program MasterBayes to assign the statistically most likely set of parents.  We used the inferred parents to measure reproductive skew. 

Results/Conclusions

Egocentric degree centrality is a network measure that indicates the degree to which a single individual could monopolize a network.  Individuals in habitats with clustered resource distribution displayed social networks with higher levels of centrality than networks on homogenous habitat(N=54, p<0.05) .  This suggests that reproduction in populations on habitat with clustered resources will be more easily monopolized by fewer individuals than populations on habitats with homogenous resource distribution.  One would also expect to see a higher degree of territorial behavior when a male was defending a more valuable resource.  We found males displayed aggressive and territorial behavior more frequently among populations where resources were clustered (N=44,p<0.05). To examine if the patterns inferred from home range distribution and behavioral analysis are reflected in the reproductive patterns of populations, we inferred parentage from yearlings captured in 2011 and hatchlings captured in early fall 2012.  Our early analysis revealed that reproductive skew, as measured using the B index , tends to be higher in populations on habitats with clustered resources, but lacked the power needed to test significance.  We hope to augment the number of testable offspring in future work.   Our results support a link between resource distribution and social structure useful to planning for long-term conservation success.