Process‐based species delimitation leads to identification of more biologically relevant species*

Abstract

Most approaches to species delimitation to date have considered divergence‐only models. Although these models are appropriate for allopatric speciation, their failure to incorporate many of the population‐level processes that drive speciation, such as gene flow (e.g., in sympatric speciation), places an unnecessary limit on our collective understanding of the processes that produce biodiversity. To consider these processes while inferring species boundaries, we introduce the R‐package delimitR and apply it to identify species boundaries in the reticulate taildropper slug (Prophysaon andersoni). Results suggest that secondary contact is an important mechanism driving speciation in this system. By considering process, we both avoid erroneous inferences that can be made when population‐level processes such as secondary contact drive speciation but only divergence is considered, and gain insight into the process of speciation in terrestrial slugs. Further, we apply delimitR to three published empirical datasets and find results corroborating previous findings. Finally, we evaluate the performance of delimitR using simulation studies, and find that error rates are near zero when comparing models that include lineage divergence and gene flow for three populations with a modest number of Single Nucleotide Polymorphisms (SNPs; 1500) and moderate divergence times (<100,000 generations). When we apply delimitR to a complex model set (i.e., including divergence, gene flow, and population size changes), error rates are moderate (∼0.15; 10,000 SNPs), and, when present, misclassifications occur among highly similar models.

Phylogeographic model selection using convolutional neural networks

“…The overall error was 1.5% (0.75-3%; Figure S2) and the best model had a probability of 87%. In a more recent study, Smith & Carstens (2020) applied Random Forest to the reticulate taildropper slug (Prophysaon andersoni) and found an average error of 5.2% when comparing 208 demographic models. These results show that CNN has an accuracy comparable to the best results reported for other methods (i.e., ABC with Random Forest).…”

“…One example of a SML approach to phylogeographic inferences is implemented in the R package delimitR (Smith & Carstens, 2020), which uses a Random Forest classifier to create hundreds of individual decision trees (a forest) from SNP data, summarized using SFS, to train the model. Next, the set of decision trees are combined via a consensus tree, and this tree is used to classify a new dataset.…”

The genomic revolution and species delimitation in birds (and other organisms): gene flow matters, but what about other evolutionary forces?

“…By explicitly considering the role that forces like gene flow, selection or changes in population size influencing genetic drift may play in speciation, process-based approaches to species delimitation using genomic data appear especially promising to connect the above theory with taxonomic practice (Smith and Carstens 2020). While such approaches continue to be developed and used, however, we point out that researchers already have at their disposal rich data and tools allowing for meaningful species delimitation given that (1) gene flow is not the only force setting the limits of evolutionary lineages and (2) we often lack information on loci underlying adaptations that alter demographic exchangeability.…”

Machine‐learning‐based detection of adaptive divergence of the stream mayfly Ephemera strigata populations

“…RF is an ensemble tree-based method that constructs multiple decision trees from a data set and combines results from all the trees to create a final predictive model. In ecological studies, RF has been applied to community-level studies to predict the distributions of species and identify constrained environmental factors (Evans, Murphy, Holden, & Cushman, 2011;Pelletier, Carstens, Tank, Sullivan, & Espíndola, 2018;Smith & Carstens, 2019;Wedger, Topp, & Olsen, 2019). In most of these studies, environmental data have been used as independent variables to predict the presence or absence of species (dependent variables).…”