Assessing the environmental factors that influence the ability of a threatened species to move through the landscape can be used to identify conservation actions that connect isolated populations. However, direct observations of species’ movement are often limited making the development of alternate approaches necessary. Here we use landscape genetic analyses to assess the impact of landscape features on the movement of individuals between local populations of a threatened snake, the Eastern Massasauga Rattlesnake (Sistrurus catenatus). We linked connectivity data with habitat information from two landscapes of similar size: a large region of unfragmented habitat and a previously studied fragmented landscape consisting of isolated patches of habitat. We used this analysis to identify features of the landscape where modification or acquisition would enhance population connectivity in the fragmented region. We found evidence that current connectivity is impacted by both contemporary landcover features, especially roads, and inherent landscape features such as elevation. Next, we derived estimates of expected movement ability using a recently developed pedigree-based approach and Least Cost Paths through the unfragmented landscape. We then used our pedigree and resistance map to estimate resistance polygons of the potential extent for S. catenatus movement in the fragmented landscape. These polygons identify possible sites for future corridors connecting currently isolated populations in this landscape by linking the impact of future habitat modification or land acquisition to dispersal ability in this species. Overall, our study shows how modeling landscape resistance across differently fragmentated landscapes can identify habitat features that affect contemporary movement in threatened species in fragmented landscapes and how this information can be used to guide mitigation actions whose goal is to connect isolated populations.

Dataset is ddRADseq from 99 individual Sistrurus catenatus (Eastern Massasauga Rattlesnake) captured at Killdeer Plains Wildlife Area in Central Ohio, USA from 2006-2019. Individuals were sequenced on Illumina HiSeq2500 and HiSeq4000 platforms, using single-end reads size selected from 300-600bp using EcoR1 and Pst1 for our restriction enzymes. Reads were aligned against a reference genome using ipyrad, and filtering done in PLINK 2.0, with 1481 loci called in every individual.

No missing values should be present in the data. All loci are bi-allelic, and anonymized survey sites linked to sample names have been uploaded in a separate document. Due to concerns over the sensitive nature of these sites, no GPS coordinates or morphometric data are reported. All S. catenatus in this dataset came from Wyandot County, Ohio, USA. Rscripts have been uploaded as examples to follow, but due to removed location and morphology data, cannot be run.

R is required to access the scripts, and we recommend using RStudio. The package ResistanceGA also requires Julia to be installed with the current setting. All packages used were accessed either via cran repository or GitHub.