Spatially structured statistical network models for landscape genetics

• DOI: 10.1002/ecm.1355
• Publication Date: 2019-01-10T15:24:36Z
• Authors: Erin E. Peterson; Ephraim M. Hanks; Mevin B. Hooten; Jay M. Ver Hoef; Marie‐Josée Fortin
• Keywords: Landscape connectivity

Abstract A basic understanding of how the landscape impedes, or creates resistance to, dispersal organisms and hence gene flow is paramount for successful conservation science management. Spatially structured ecological networks are often used to represent spatial landscape‐genetic relationships, where nodes individuals populations movement represented using non‐binary edge weights. Weights typically assigned estimated by user, rather than observed, validating such weights challenging. We provide a synthesis current methods estimate an overview common model types, stressing advantages disadvantages each approach their ability data. further explore set spatial‐statistical that ecologists with alternative approaches modeling spatially explicit processes may affect genetic structure. This includes autoregressive models, particular focus on correlation partial neighborhood structure inverse covariance matrix (i.e., precision matrix). then demonstrate specifying appropriate statistical nodes, conditioned weights, through matrix. integration network ecology statistics provides practical analytical framework studies. The results can be make inferences about relative importance individual characteristics, as vegetative cover, hillslope, presence roads rivers, flow. In addition, R code we include allows readers in own datasets, which will potentially new insights into evolutionary generated networks, well valuable information optimal characteristics corridors.