Analysis of biodiversity data suggests that mammal species are hidden in predictable places

Abstract

Significance Only an estimated 1 to 10% of Earth’s species have been formally described. This discrepancy between the number of species with a formal taxonomic description and actual number of species (i.e., the Linnean shortfall) hampers research across the biological sciences. To explore whether the Linnean shortfall results from poor taxonomic practice or not enough taxonomic effort, we applied machine-learning techniques to build a predictive model to identify named species that are likely to contain hidden diversity. Results indicate that small-bodied species with large, climatically variable ranges are most likely to contain hidden species. These attributes generally match those identified in the taxonomic literature, indicating that the Linnean shortfall is caused by societal underinvestment in taxonomy rather than poor taxonomic practice.

phylogatR: Phylogeographic data aggregation and repurposing

“…Studies using data from our initial data aggregation pipelines have further demonstrated the utility of single‐locus large‐scale studies that also utilize data layers from other sources. Parsons et al (2022) explore cryptic diversity in mammals using molecular species delimitation methods for single‐locus data in conjunction with natural history and environmental data for over 4000 species. They found that hundreds of mammal species are still probably undescribed and that these are mostly small‐bodied taxa with large ranges (scripts for this project can be found at https://github.com/parsons463/HiddenDiversity).…”

“…We recommend individual users to capitalize on available tools for checking taxonomy when appropriate for their needs. For example, the R package taxize (Chamberlain & Szöcs, 2013) accesses many data sources to update species names, or standardized databases can be used directly to update species names such as the Mammal Diversity Database published by the American Society of Mammalogists (as in Parsons et al, 2022).…”

“…First, the default MAFFT version 7 parameters were used. Sequence alignments were checked by eye for 10 families (117 species‐level alignments) that were previously determined to require post‐alignment adjustments (Parsons et al, 2022). Several alignments were found to have large sequence gaps at the ends of the alignment, while others contained unwanted sequences (e.g., parasitic sequences that have been named as the host species).…”

“…Two plant groups have relatively high values for π (e.g., Lycopodiophyta, Pinophyta). This suggests these groups are worth further exploration, as either they may be in need of database updates to reflect taxonomic revisions and misidentifications, or these groups may harbour a high number of cryptic species (Parsons et al, 2022). Additionally, though still highlighting the need for further work, this could be a sampling issue as these groups had lower species representation in the database and we might be misrepresenting the average.…”

Geographic and taxonomic biases in the vertebrate tree of life

“…It is important to note that I analysed only described species for which distributional data are available. The character of the study prevented to include unrecognized, cryptic, unnamed, or unmapped species despite that they may represent a substantial component of the world's biodiversity (Parsons et al, 2022). For example, given the frequency of uncovering cryptic species through genetic analysis (Bickford et al, 2007) it is likely that many of the unsampled species will be found to harbour more taxa.…”