The genetic network of greater sage‐grouse: Range‐wide identification of keystone hubs of connectivity

Abstract

Genetic networks can characterize complex genetic relationships among groups of individuals, which can be used to rank nodes most important to the overall connectivity of the system. Ranking allows scarce resources to be guided toward nodes integral to connectivity. The greater sage‐grouse (Centrocercus urophasianus) is a species of conservation concern that breeds on spatially discrete leks that must remain connected by genetic exchange for population persistence. We genotyped 5,950 individuals from 1,200 greater sage‐grouse leks distributed across the entire species’ geographic range. We found a small‐world network composed of 458 nodes connected by 14,481 edges. This network was composed of hubs—that is, nodes facilitating gene flow across the network—and spokes—that is, nodes where connectivity is served by hubs. It is within these hubs that the greatest genetic diversity was housed. Using indices of network centrality, we identified hub nodes of greatest conservation importance. We also identified keystone nodes with elevated centrality despite low local population size. Hub and keystone nodes were found across the entire species’ contiguous range, although nodes with elevated importance to network‐wide connectivity were found more central: especially in northeastern, central, and southwestern Wyoming and eastern Idaho. Nodes among which genes are most readily exchanged were mostly located in Montana and northern Wyoming, as well as Utah and eastern Nevada. The loss of hub or keystone nodes could lead to the disintegration of the network into smaller, isolated subnetworks. Protecting both hub nodes and keystone nodes will conserve genetic diversity and should maintain network connections to ensure a resilient and viable population over time. Our analysis shows that network models can be used to model gene flow, offering insights into its pattern and process, with application to prioritizing landscapes for conservation.

Spatial differences in genetic diversity and northward migration suggest genetic erosion along the boreal caribou southern range limit and continued range retraction

“…Telemetry records have shown that eastern migratory animals overlap seasonally with boreal animals in Ontario (Berglund et al, 2014;Pond, Brown, Wilson, & Schaefer, 2016). Our network analysis allowed for the calculation of node metrics, which can be used to identify hubs of connectivity and prioritize areas for management (Cross et al, 2018;Garroway et al, 2008;Koen et al, 2016). Our network analysis allowed for the calculation of node metrics, which can be used to identify hubs of connectivity and prioritize areas for management (Cross et al, 2018;Garroway et al, 2008;Koen et al, 2016).…”

“…Consequently, that ecotype potentially interbreeds with boreal animals and their greater mobility likely increases gene flow within and among boreal caribou sampled areas (Thompson, ), which is similar to findings in Quebec (Boulet, Couturier, Côté, Otto, & Bernatchez, ). Our network analysis allowed for the calculation of node metrics, which can be used to identify hubs of connectivity and prioritize areas for management (Cross et al, ; Garroway et al, ; Koen et al, ). We found that most node metrics based on F ST suggested that sampled areas in northern regions of the Ontario boreal range may be important for supplementing southernmost populations (see Appendix , Table ).…”

A landscape experiment of spatial network robustness and space‐use reorganization following habitat fragmentation

“…We classified each experimental site in one of two categories based on the type of disturbed nodes in the pre‐disturbance network: if at least one hub was disturbed, the site was categorized as hub disturbed; otherwise, it was categorized as non‐hub disturbed. A node was identified as a hub when its degree was larger or equal to the 90th percentile of the degree distribution obtained from predicted degree of all visited patches in the 10 experimental sites (Bishop‐Taylor, Tulbure, & Broich, , ; Cross et al, ). In total, four sites out of 10 were hub disturbed.…”

graph4lg: A package for constructing and analysing graphs for landscape genetics in R

“…Once genetic graphs have been created, the compute_node_metric function computes graph‐theoretic metrics such as the degree, closeness and betweenness centrality indices, which identify keystone hubs of genetic connectivity (Cross et al., 2018). It also computes the average and sum of the inverse genetic distance weighting the links.…”