Predicting genetic biodiversity in salamanders using geographic, climatic, and life history traits

RESEARCH ARTICLE Predicting genetic biodiversity in salamanders using geographic, climatic, and life history traits Danielle J. Parsons1,2, Abigail E. Green3, Bryan C. CarstensID1,2*, Tara A. Pelletier3 1 Museum of Biological Diversity, The Ohio State University, Columbus, Ohio, United States of America, 2 Department of Evolution, Ecology and Organismal Biology, The Ohio State University, Columbus, Ohio, United States of America, 3 Department of Biology, Radford University, Radford, Virginia, United States of America * carstens.12@osu.edu Abstract The geographic distribution of genetic variation within a species reveals information about its evolutionary history, including responses to historical climate change and dispersal ability across various habitat types. We combine genetic data from salamander species with geo- graphic, climatic, and life history data collected from open-source online repositories to develop a machine learning model designed to identify the traits that are most predictive of unrecognized genetic lineages. We find evidence of hidden diversity distributed throughout the clade Caudata that is largely the result of variation in climatic variables. We highlight some of the difficulties in using machine-learning models on open-source data that are often messy and potentially taxonomically and geographically biased. Introduction Documenting biodiversity is an important first step in understanding both ecological and evo- lutionary processes [1], particularly the functional roles that act to connect processes function-

ing at both shallow and deep time scales [2]. Notably, any such documentation of biodiversity implicitly assumes that the units (e.g., species) are comparable across different geographic regions. Given that a Linnean shortfall (i.e., the ratio of recognized to unrecognized species [3]) exists in most clades and may be substantial across Eukaryota [4], it is not clear that this assumption is reasonable. An alternative approach is to utilize evolutionary significant units [5], or genetic lineages, in place of species in broad analyses of biodiversity (e.g., [6]). This may be particularly useful in clades with relatively high degrees of morphological and ecological conservatism. One such clade is Caudata (i.e., salamanders and newts), which exhibits high frequencies of cryptic species (e.g., [7–9]). Identifying hidden genetic lineages in Caudata can have important conservation implica- tions. For example, Mead et al. [10] discovered a new species of western Plethodon salamander that was originally thought to be either P. elongatus or P. stormi [10]. All three of these species are listed on the IUCN Red List as either near threatened (P. elongatus), vulnerable (P. asupak), PLOS ONE PLOS ONE | https://doi.org/10.1371/journal.pone.0310932 October 18, 2024 1 / 20 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Parsons DJ, Green AE, Carstens BC, Pelletier TA (2024) Predicting genetic biodiversity in salamanders using geographic, climatic, and life history traits. PLoS ONE 19(10): e0310932. https:// doi.org/10.1371/journal.pone.0310932

Editor: Christopher Nice, Texas State University, UNITED STATES OF AMERICA Received: February 22, 2024 Accepted: September 9, 2024 Published: October 18, 2024 Copyright: © 2024 Parsons et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Data are uploaded to Dryad and can be accessed using the following link: https://doi.org/10.5061/dryad.m63xsj474. Code related to this manuscript, including data cleaning, imputation, predictive modeling, and significance testing has been deposited in GitHub (https://github.com/parsons463/ HiddenSalamanders). All remaining data are available in the manuscript and/or Supporting information.