Mapping climate adaptation corridors for biodiversity—A regional-scale case study in Central America

unprotected) that represent potential climate change refugia. Second, we created a conduc- tance surface for modeling least cost paths and third, we quantified the ecological characteris- tics and conservation status of potential climate adaptation corridors and their associated start and end nodes to develop a corridor priority index. We used Google Earth Engine [38] through the ‘rgee’ R package [39] to download data layers for connectivity assessments. Throughout our analysis, occasionally start and end nodes and often corridors spanned multi- ple countries, so we referred to these by their primary overlapping country when necessary to facilitate comparisons among countries. A permanent repository containing data and R code is available on Zenodo [40]. We used R version 4.3.0 for all analyses [41]. Start and end nodes We used the World Database on Protected Areas (WDPA) as the most up-to-date, compre- hensive source of protected areas globally [37]. We used the ‘wdpar‘R package [42] to filter out Fig 1. Study area, lowland protected areas (starting nodes), and highland target areas (end nodes). Country boundaries are public domain and were obtained from Natural Earth (50m-admin-0-countries-2) before clipping to our study area (https://www.naturalearthdata.com/). Lowland protected areas are polygons � 5 km2 and � 500 m elevation from the World Database on Protected Areas [37]. Highland target areas are patches (� 5 km2) of upper-montane (� 1500 m elevation) contiguous forest with medium or high Forest Landscape Integrity Index values [47].

https://doi.org/10.1371/journal.pone.0304756.g001 PLOS ONE Mapping climate adaptation corridors for biodiversity PLOS ONE | https://doi.org/10.1371/journal.pone.0304756 May 31, 2024 4 / 19