A comparison of regression methods for model selection in individual‐based landscape genetic analysis

Abstract

Anthropogenic migration barriers fragment many populations and limit the ability of species to respond to climate-induced biome shifts. Conservation actions designed to conserve habitat connectivity and mitigate barriers are needed to unite fragmented populations into larger, more viable metapopulations, and to allow species to track their climate envelope over time. Landscape genetic analysis provides an empirical means to infer landscape factors influencing gene flow and thereby inform such conservation actions. However, there are currently many methods available for model selection in landscape genetics, and considerable uncertainty as to which provide the greatest accuracy in identifying the true landscape model influencing gene flow among competing alternative hypotheses. In this study, we used population genetic simulations to evaluate the performance of seven regression-based model selection methods on a broad array of landscapes that varied by the number and type of variables contributing to resistance, the magnitude and cohesion of resistance, as well as the functional relationship between variables and resistance. We also assessed the effect of transformations designed to linearize the relationship between genetic and landscape distances. We found that linear mixed effects models had the highest accuracy in every way we evaluated model performance; however, other methods also performed well in many circumstances, particularly when landscape resistance was high and the correlation among competing hypotheses was limited. Our results provide guidance for which regression-based model selection methods provide the most accurate inferences in landscape genetic analysis and thereby best inform connectivity conservation actions.

Beyond the landscape: Resistance modelling infers physical and behavioural gene flow barriers to a mobile carnivore across a metropolitan area

“…Mixed models were fitted using the maximum likelihood population effects (MPLE) parameterization (Clarke, Rothery, & Raybould, ) implemented in the R package lme4 (Bates, Mächler, Bolker, & Walker, ). A simulation study by Shirk, Landguth, and Cushman, (2018) has shown that this linear‐mixed‐effects‐model‐based method had a high accuracy in model selection.…”

Urbanization reduces genetic connectivity in bobcats (Lynx rufus) at both intra– and interpopulation spatial scales

“…Associations between landscape resistance matrices and pairwise genetic distances were tested using linear mixed effect models incorporating a maximum likelihood population effects (MLPE) approach (Clarke, Rothery, & Raybould, ; Van Strien, Keller, & Holderegger, ) using the lme4 package in R. This method incorporates a random effect structure that accounts for the nonindependence among pairwise data, and has been shown recently to outperform other model selection methods for landscape genetics (Shirk, Landguth, & Cushman, ). Prior to fitting models, matrices of D ps were log–transformed to satisfy normality assumptions and all dependent and independent variables were rescaled to units of standard deviation and a mean of zero.…”

ResistanceGA: An R package for the optimization of resistance surfaces using genetic algorithms

“…Pairwise genetic distance is the response and the scaled and centred pairwise effective distance is the predictor. The MLPE mixed effects parameterization accounts for non‐independence among the pairwise data (Clarke, Rothery, & Raybould, ), and Shirk, Landguth, and Cushman () recently found that MLPE models performed best in landscape genetic model selection among the seven regression methods assessed. An objective function, specified by the user, is obtained from the fitted MLPE model: log‐likelihood, AIC, or marginal R 2 (Nakagawa & Schielzeth, ). Steps 2–5 are repeated until the specified number of n individuals have been created. The genetic algorithm then conducts selection on the population, and the individuals with the best objective function values are carried over to the next generation to form the reproducing population (default = top 5% retained each generation).…”