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

SignificanceOnly 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.

Artificial intelligence enables unified analysis of historical and landscape influences on genetic diversity.

While genetic variation in any species is potentially shaped by a range of processes, phylogeography and landscape genetics are largely concerned with inferring how environmental conditions and landscape features impact neutral intraspecific diversity. However, even as both disciplines have come to utilize SNP data over the last decades, analytical approaches have remained for the most part focused on either broad-scale inferences of historical processes (phylogeography) or on more localized inferences about environmental and/or landscape features (landscape genetics). Here we demonstrate that an artificial intelligence model-based analytical framework can consider both deeper historical factors and landscape-level processes in an integrated analysis. We implement this framework using data collected from two Brazilian anurans, the Brazilian sibilator frog (Leptodactylus troglodytes) and granular toad (Rhinella granulosa). Our results indicate that historical demographic processes shape most the genetic variation in the sibulator frog, while landscape processes primarily influence variation in the granular toad. The machine learning framework used here allows both historical and landscape processes to be considered equally, rather than requiring researchers to make an a priori decision about which factors are important.

Genetic structure and landscape effects on gene flow in the Neotropical lizard Norops brasiliensis (Squamata: Dactyloidae).

One key research goal of evolutionary biology is to understand the origin and maintenance of genetic variation. In the Cerrado, the South American savanna located primarily in the Central Brazilian Plateau, many hypotheses have been proposed to explain how landscape features (e.g., geographic distance, river barriers, topographic compartmentalization, and historical climatic fluctuations) have promoted genetic structure by mediating gene flow. Here, we asked whether these landscape features have influenced the genetic structure and differentiation in the lizard species Norops brasiliensis (Squamata: Dactyloidae). To achieve our goal, we used a genetic clustering analysis and estimate an effective migration surface to assess genetic structure in the focal species. Optimized isolation-by-resistance models and a simulation-based approach combined with machine learning (convolutional neural network; CNN) were then used to infer current and historical effects on population genetic structure through 12 unique landscape models. We recovered five geographically distributed populations that are separated by regions of lower-than-expected gene flow. The results of the CNN showed that geographic distance is the sole predictor of genetic variation in N. brasiliensis, and that slope, rivers, and historical climate had no discernible influence on gene flow. Our novel CNN approach was accurate (89.5%) in differentiating each landscape model. CNN and other machine learning approaches are still largely unexplored in landscape genetics studies, representing promising avenues for future research with increasingly accessible genomic datasets.

Predictors of genomic diversity within North American squamates.

Comparisons of intraspecific genetic diversity across species can reveal the roles of geography, ecology, and life history in shaping biodiversity. The wide availability of mitochondrial DNA (mtDNA) sequences in open-access databases makes this marker practical for conducting analyses across several species in a common framework, but patterns may not be representative of overall species diversity. Here, we gather new and existing mtDNA sequences and genome-wide nuclear data (genotyping-by-sequencing; GBS) for 30 North American squamate species sampled in the Southeastern and Southwestern United States. We estimated mtDNA nucleotide diversity for 2 mtDNA genes, COI (22 species alignments; average 16 sequences) and cytb (22 species; average 58 sequences), as well as nuclear heterozygosity and nucleotide diversity from GBS data for 118 individuals (30 species; 4 individuals and 6,820 to 44,309 loci per species). We showed that nuclear genomic diversity estimates were highly consistent across individuals for some species, while other species showed large differences depending on the locality sampled. Range size was positively correlated with both cytb diversity (phylogenetically independent contrasts: R2 = 0.31, P = 0.007) and GBS diversity (R2 = 0.21; P = 0.006), while other predictors differed across the top models for each dataset. Mitochondrial and nuclear diversity estimates were not correlated within species, although sampling differences in the data available made these datasets difficult to compare. Further study of mtDNA and nuclear diversity sampled across species' ranges is needed to evaluate the roles of geography and life history in structuring diversity across a variety of taxonomic groups.

Geographic isolation and long-distance gene flow influence the genetic structure of the blue fan palm Brahea armata (Arecaceae).

The formation of the Baja California Peninsula (BCP) has impacted the microevolutionary dynamics of different species in ways that depend on biological traits such as dispersal capacity. Plants with relatively low levels of vagility have exhibited high genetic divergence between the BCP and Continental mainland. Brahea armata (Arecaceae) is a palm species inhabiting the northern part of the BCP and Sonora; its distribution occurs in isolated oases of vegetation. We aimed to evaluate the influence of the formation of the BCP on the genetic structure of B. armata using nuclear microsatellites and chloroplast markers (cpDNA) to compare patterns of genetic diversity and structure with previous published studies. Because gene flow through seeds is usually more limited compared to pollen flow, we expect to find stronger genetic structure at (cpDNA) than at nuclear markers. Moreover, larger genetic structure might also be explained by the smaller effective population size of cpDNA. We analyzed six microsatellite markers and two cpDNA regions. The main results indicated high levels of genetic differentiation among isolated populations located in the BCP, while low genetic differentiation was found between southern populations of the BCP and Sonora, suggesting long distance gene flow. In contrast, chloroplast markers indicated high levels of genetic structure between BCP and Sonora populations, suggesting asymmetrical gene flow between pollen (measured by nuclear microsatellites) and seed (cpDNA markers). This study provides valuable information on genetic diversity of B. armata that can be relevant for conservation and management; and develops microsatellites markers that can be transferred to other Brahea species.

The role of multiple Pleistocene refugia in promoting diversification in the Pacific Northwest.

Pleistocene glacial cycles drastically changed the distributions of taxa endemic to temperate rainforests in the Pacific Northwest, with many experiencing reduced habitat suitability during glacial periods. In this study, we investigate whether glacial cycles promoted intraspecific divergence and whether subsequent range changes led to secondary contact and gene flow. For seven invertebrate species endemic to the PNW, we estimated species distribution models (SDMs) and projected them onto current and historical climate conditions to assess how habitat suitability changed during glacial cycles. Using single nucleotide polymorphism (SNP) data from these species, we assessed population genetic structure and used a machine-learning approach to compare models with and without gene flow between populations upon secondary contact after the last glacial maximum (LGM). Finally, we estimated divergence times and rates of gene flow between populations. SDMs suggest that there was less suitable habitat in the North Cascades and Northern Rocky Mountains during glacial compared to interglacial periods, resulting in reduced habitat suitability and increased habitat fragmentation during the LGM. Our genomic data identify population structure in all taxa, and support gene flow upon secondary contact in five of the seven taxa. Parameter estimates suggest that population divergences date to the later Pleistocene for most populations. Our results support a role of refugial dynamics in driving intraspecific divergence in the Cascades Range. In these invertebrates, population structure often does not correspond to current biogeographic or environmental barriers. Rather, population structure may reflect refugial lineages that have since expanded their ranges, often leading to secondary contact between once isolated lineages.

Comparative and predictive phylogeography in the South American diagonal of open formations: Unravelling the biological and environmental influences on multitaxon demography.

Phylogeography investigates historical drivers of the geographical distribution of intraspecific lineages. Particular attention has been given to ecological, climatic and geological processes in the diversification of the Neotropical biota. Several species sampled across the South American diagonal of open formations (DOF), comprising the Caatinga, Cerrado and Chaco biomes, experienced range shifts coincident with Quaternary climatic changes. However, comparative studies across different spatial, temporal and biological scales on DOF species are still meagre. Here, we combine phylogeographical model selection and machine learning predictive frameworks to investigate the influence of Pleistocene climatic changes on several plant and animal species from the DOF. We assembled mitochondrial/chloroplastic DNA sequences in public repositories and inferred the demographic responses of 44 species, comprising 70 intraspecific lineages of plants, lizards, frogs, spiders and insects. We then built a random forest model using biotic and abiotic information to identify the best predictors of demographic responses in the Pleistocene. Finally, we assessed the temporal synchrony of species demographic responses with hierarchical approximate Bayesian computation. Biotic variables related to population connectivity, gene flow and habitat preferences largely predicted how species responded to Pleistocene climatic changes, and demographic changes were synchronous primarily during the Middle Pleistocene. Although 22 (~31%) lineages underwent demographic expansion, presumably associated with the spread of aridity during the glacial Pleistocene periods, our findings suggest that nine lineages (~13%) exhibited the opposite response due to taxon-specific attributes.

A role of asynchrony of seasons in explaining genetic differentiation in a Neotropical toad.

The process of diversification can be studied at the phylogeographic level by attempting to identify the environmental features that promote and maintain population divergence. Here we investigate diversification in Rhinella granulosa, a Neotropical toad from northeastern Brazil, by testing a range of hypotheses that encompass different putative mechanisms reducing gene flow among populations. We sequenced single nucleotide polymorphisms and examined individual predictions related to the role of geographic barriers (rivers), ecological gradients, historical habitat stability, and spatial variation in climate seasonality, also known as the asynchrony of seasons hypothesis. This hypothesis postulates that temporal asynchrony of wet and dry seasons over short distances causes parapatric populations to become isolated by time. After determining genetic structure, inferring past distributions, ranking demographic models, and estimating the power of monthly climatic variables, our results identified two populations that are not associated with geographic barriers, biome gradients, or historical refugia. Instead, they are predicted by spatial variation in monthly rainfall and minimum temperature, consistent with the asynchrony of seasons hypothesis, supported also by our comparative framework using multiple matrix regression and linear mixed effects modeling. Due to the toad's life history, climate likely mediates gene flow directly, with genetic differentiation being provoked by neutral mechanisms related to climate driven population isolation, and/or by natural selection against migrants from populations with different breeding times. The asynchrony of seasons hypothesis is seldom considered in phylogeographic studies, but our results indicate that it should be tested in systems where breeding is tightly coupled with climate.

Predicting plant conservation priorities on a global scale.

The conservation status of most plant species is currently unknown, despite the fundamental role of plants in ecosystem health. To facilitate the costly process of conservation assessment, we developed a predictive protocol using a machine-learning approach to predict conservation status of over 150,000 land plant species. Our study uses open-source geographic, environmental, and morphological trait data, making this the largest assessment of conservation risk to date and the only global assessment for plants. Our results indicate that a large number of unassessed species are likely at risk and identify several geographic regions with the highest need of conservation efforts, many of which are not currently recognized as regions of global concern. By providing conservation-relevant predictions at multiple spatial and taxonomic scales, predictive frameworks such as the one developed here fill a pressing need for biodiversity science.

Population genetic structure and demographic history of the lone star tick, Amblyomma americanum (Ixodida: Ixodidae): New evidence supporting old records.

Range expansions are a potential outcome of changes in habitat suitability, which commonly result as a consequence of climate change. Hypotheses on such changes in the geographic distribution of a certain species can be evaluated using population genetic structure and demography. In this study we explore the population genetic structure, genetic variability, demographic history of, and habitat suitability for Amblyomma americanum, a North American tick species that is a known vector of several pathogenic microorganisms. We used a double digestion restriction site-associated DNA sequencing technique (dd-RAD seq) and discovered 8,181 independent single nucleotide polymorphisms (SNPs) in 189 ticks from across the geographic range of the species. Genetic diversity was low, particularly when considering the broad geographic range of this species. The edge populations were less diverse than populations belonging to the historic range, possibly indicative of a range expansion, but this hypothesis was not statistically supported by a test based on genetic data. Nonetheless, moderate levels of population structure and substructure were detected between geographic regions. For New England, demographic and species distribution models support a scenario where A. americanum was present in more northern locations in the past, underwent a bottleneck, and subsequently recovered. These results are consistent with a hypothesis that this species is re-establishing in this area, rather than one focused on range expansion from the south. This hypothesis is consistent with old records describing the presence of A. americanum in the northeastern US in the early colonial period.

Evidence that Myotis lucifugus "Subspecies" are Five Nonsister Species, Despite Gene Flow.

While genetic exchange between nonsister species was traditionally considered to be rare in mammals, analyses of molecular data in multiple systems suggest that it may be common. Interspecific gene flow, if present, is problematic for phylogenetic inference, particularly for analyses near the species level. Here, we explore how to detect and account for gene flow during phylogeny estimation using data from a clade of North American Myotis bats where previous results have led researchers to suspect that gene flow among lineages is present. Initial estimates of phylogenetic networks and species trees indicate that subspecies described within Myotis lucifugus are paraphyletic. In order to explore the extent to which gene flow is likely to interfere with phylogeny estimation, we use posterior predictive simulation and a novel Approximate Bayesian Computation approach based on gene tree distances. The former indicates that the species tree model is a poor fit to the data, and the latter provides evidence that a species tree with gene flow is a better fit. Taken together, we present evidence that the currently recognized M. lucifugus subspecies are paraphyletic, exchange alleles with other Myotis species in regions of secondary contact, and should be considered independent evolutionary lineages despite their morphological similarity.

Phylogeographic model selection leads to insight into the evolutionary history of four-eyed frogs.

Phylogeographic research investigates biodiversity at the interface between populations and species, in a temporal and geographic context. Phylogeography has benefited from analytical approaches that allow empiricists to estimate parameters of interest from the genetic data (e.g., θ = 4Neµ, population divergence, gene flow), and the widespread availability of genomic data allow such parameters to be estimated with greater precision. However, the actual inferences made by phylogeographers remain dependent on qualitative interpretations derived from these parameters' values and as such may be subject to overinterpretation and confirmation bias. Here we argue in favor of using an objective approach to phylogeographic inference that proceeds by calculating the probability of multiple demographic models given the data and the subsequent ranking of these models using information theory. We illustrate this approach by investigating the diversification of two sister species of four-eyed frogs of northeastern Brazil using single nucleotide polymorphisms obtained via restriction-associated digest sequencing. We estimate the composite likelihood of the observed data given nine demographic models and then rank these models using Akaike information criterion. We demonstrate that estimating parameters under a model that is a poor fit to the data is likely to produce values that lead to spurious phylogeographic inferences. Our results strongly imply that identifying which parameters to estimate from a given system is a key step in the process of phylogeographic inference and is at least as important as being able to generate precise estimates of these parameters. They also illustrate that the incorporation of model uncertainty should be a component of phylogeographic hypothesis tests.

Combining allele frequency and tree-based approaches improves phylogeographic inference from natural history collections.

Model selection approaches in phylogeography have allowed researchers to evaluate the support for competing demographic histories, which provides a mode of inference and a measure of uncertainty in understanding climatic and spatial influences on intraspecific diversity. Here, to rank all models in the comparison set and determine what proportion of the total support the top-ranked model garners, we conduct model selection using two analytical approaches-allele frequency-based, implemented in fastsimcoal2, and gene tree-based, implemented in phrapl. We then expand this model selection framework by including an assessment of absolute fit of the models to the data. For this, we utilize DNA isolated from existing natural history collections that span the distribution of red alder (Alnus rubra) in the Pacific Northwest of North America to generate genomic data for the evaluation of 13 demographic scenarios. The quality of DNA recovered from herbarium specimen leaf tissue was assessed for its utility and effectiveness in demographic model selection, specifically in the two approaches mentioned. We present strong support for the use of herbarium tissue in the generation of genomic DNA, albeit with the inclusion of additional quality control checks prior to library preparation and analyses with multiple approaches that incorporate various data. Analyses with allele frequency spectra and gene trees predominantly support A. rubra having experienced an ancient vicariance event with intermittent and frequent gene flow between the disjunct populations. Additionally, the data consistently fit the most frequently selected model, corroborating the model selection techniques. Finally, these results suggest that the A. rubra disjunct populations do not represent separate species.

PHRAPL: Phylogeographic Inference Using Approximate Likelihoods.

The demographic history of most species is complex, with multiple evolutionary processes combining to shape the observed patterns of genetic diversity. To infer this history, the discipline of phylogeography has (to date) used models that simplify the historical demography of the focal organism, for example by assuming or ignoring ongoing gene flow between populations or by requiring a priori specification of divergence history. Since no single model incorporates every possible evolutionary process, researchers rely on intuition to choose the models that they use to analyze their data. Here, we describe an approximate likelihood approach that reduces this reliance on intuition. PHRAPL allows users to calculate the probability of a large number of complex demographic histories given a set of gene trees, enabling them to identify the most likely underlying model and estimate parameters for a given system. Available model parameters include coalescence time among populations or species, gene flow, and population size. We describe the method and test its performance in model selection and parameter estimation using simulated data. We also compare model probabilities estimated using our approximate likelihood method to those obtained using standard analytical likelihood. The method performs well under a wide range of scenarios, although this is sometimes contingent on sampling many loci. In most scenarios, as long as there are enough loci and if divergence among populations is sufficiently deep, PHRAPL can return the true model in nearly all simulated replicates. Parameter estimates from the method are also generally accurate in most cases. PHRAPL is a valuable new method for phylogeographic model selection and will be particularly useful as a tool to more extensively explore demographic model space than is typically done or to estimate parameters for complex models that are not readily implemented using current methods. Estimating relevant parameters using the most appropriate demographic model can help to sharpen our understanding of the evolutionary processes giving rise to phylogeographic patterns. [AIC; grid search; isolation-with-migration; migration rate; multispecies coalescent; parameter optimization; population genetics; tree topologies.].

Species Delimitation with Gene Flow.

Species are commonly thought to be evolutionarily independent in a way that populations within a species are not. In recent years, studies that seek to identify evolutionarily independent lineages (i.e., to delimit species) using genetic data have typically adopted multispecies coalescent approaches that assume that evolutionary independence is formed by the differential sorting of ancestral alleles due to genetic drift. However, gene flow appears to be common among populations and nascent species, and while this process may inhibit lineage divergence (and thus independence), it is usually not explicitly considered when delimiting species. In this article, we apply Phylogeographic Inference using Approximate Likelihoods (PHRAPL), a recently described method for phylogeographic model selection, to species delimitation. We describe an approach to delimiting species using PHRAPL that attempts to account for both genetic drift and gene flow, and we compare the method's performance to that of a popular delimitation approach (BPP) using both simulated and empirical datasets. PHRAPL generally infers the correct demographic-delimitation model when the generating model includes gene flow between taxa, given a sufficient amount of data. When the generating model includes only isolation in the recent past, PHRAPL will in some cases fail to differentiate between gene flow and divergence, leading to model misspecification. Nevertheless, the explicit consideration of gene flow by PHRAPL is an important complement to existing delimitation approaches, particularly in systems where gene flow is likely important. [approximate likelihoods; coalescent simulations; genealogical divergence index; Homo sapiens; isolation-with-migration; multispecies coalescent; Sarracenia; Scincella.].

Speciation with Gene Flow in North American Myotis Bats.

Growing evidence supports the idea that species can diverge in the presence of gene flow. However, most methods of phylogeny estimation do not consider this process, despite the fact that ignoring gene flow is known to bias phylogenetic inference. Furthermore, studies that do consider divergence-with-gene-flow typically do so by estimating rates of gene flow using a isolation-with-migration model (IM), rather than evaluating scenarios of gene flow (such as divergence-with-gene flow or secondary contact) that represent very different types of diversification. In this investigation, we aim to infer the recent phylogenetic history of a clade of western long-eared bats while evaluating a number of different models that parameterize gene flow in a variety of ways. We utilize PHRAPL, a new tool for phylogeographic model selection, to compare the fit of a broad set of demographic models that include divergence, migration, or both among Myotis evotis, $M$. thysanodes and M. keenii. A genomic data set consisting of 808 loci of ultraconserved elements was used to explore such models in three steps using an incremental design where each successive set was informed by, and thus more focused than, the previous set of models. Specifically, the three steps were to (i) assess whether gene flow should be modeled and identify the best topologies, (ii) infer directionality of migration using the best topologies, and (iii) estimate the timing of gene flow. The best model (AIC model weight ${\sim}0.98$) included two divergence events (($M$. evotis, $M$. thysanodes), M. keenii) accompanied by gene flow at the initial stages of divergence. These results provide a striking example of speciation-with-gene-flow in an evolutionary lineage. [Myotis bats; PHRAPL; P2C2M; phylogeographic model selection; speciation with gene flow.].

Geographical range size and latitude predict population genetic structure in a global survey.

While genetic diversity within species is influenced by both geographical distance and environmental gradients, it is unclear what other factors are likely to promote population genetic structure. Using a machine learning framework and georeferenced DNA sequences from more than 8000 species, we demonstrate that geographical attributes of the species range, including total size, latitude and elevation, are the most important predictors of which species are likely to contain structured genetic variation. While latitude is well known as an important predictor of biodiversity, our work suggests that it also plays a key role in shaping diversity within species.

Giant tree frogs exemplify the promise of integrating multiple types of data in phylogeographic investigations.

Hugall et al. (2022) is one of the seminal publications from the single locus era of phylogeographic research. These authors were among the first to argue that genetic data are ideally suited to test hypotheses that are ultimately derived from other sources of information. While the testing of predictions from the fossil record has long been important to molecular systematics (e.g., Donoghue et al., 1989), phylogeographic investigations into the more recent evolutionary past lack a fossil record in most focal taxa. In lieu of fossils, which were not available for the small snails that served as the focal taxon, Hugall et al. (2002) applied the (then) new technique of environmental modelling to identify regions within the species range with habitat that was predicted to be stable throughout the Holocene. They then present data that suggests that these regions correspond to the areas with high genetic diversity. Apart from the inferences about snail evolutionary history, the core argument of Hugall et al. (2002) is that consilience (i.e., agreement between inferences drawn from different sources of data) is an important goal for phylogeographic investigation. Consilience in the inferences drawn from independent types of data has a multiplicative effect; when present the researcher is likely to have more confidence in their inference than would be possible to have in an inference from any one source of data. The manuscript by Jaynes et al. (2022) is a splendid illustration of this principle.

Do ecological communities disperse across biogeographic barriers as a unit?

Biogeographic barriers have long been implicated as drivers of biological diversification, but how these barriers influence co-occurring taxa can vary depending on factors intrinsic to the organism and in their relationships with other species. Due to the interdependence among taxa, ecological communities present a compelling opportunity to explore how interactions among species may lead to a shared response to historical events. Here we collect single nucleotide polymorphism data from five commensal arthropods associated with the Sarracenia alata carnivorous pitcher plant, and test for codiversification across the Mississippi River, a major biogeographic barrier in the southeastern United States. Population genetic structure in three of the ecologically dependent arthropods mirrors that of the host pitcher plant, with divergence time estimates suggesting two of the species (the pitcher plant moth Exyra semicrocea and a flesh fly Sarcophaga sarraceniae) dispersed synchronously across this barrier along with the pitcher plant. Patterns in population size and genetic diversity suggest the plant and ecologically dependent arthropods dispersed from east to west across the Mississippi River. In contrast, species less dependent on the plant ecologically show discordant phylogeographic patterns. This study demonstrates that ecological relationships may be an important predictor of codiversification, and supports recent suggestions that organismal trait data should be prominently featured in comparative phylogeographic investigations.

Demographic model selection using random forests and the site frequency spectrum.

Phylogeographic data sets have grown from tens to thousands of loci in recent years, but extant statistical methods do not take full advantage of these large data sets. For example, approximate Bayesian computation (ABC) is a commonly used method for the explicit comparison of alternate demographic histories, but it is limited by the "curse of dimensionality" and issues related to the simulation and summarization of data when applied to next-generation sequencing (NGS) data sets. We implement here several improvements to overcome these difficulties. We use a Random Forest (RF) classifier for model selection to circumvent the curse of dimensionality and apply a binned representation of the multidimensional site frequency spectrum (mSFS) to address issues related to the simulation and summarization of large SNP data sets. We evaluate the performance of these improvements using simulation and find low overall error rates (~7%). We then apply the approach to data from Haplotrema vancouverense, a land snail endemic to the Pacific Northwest of North America. Fifteen demographic models were compared, and our results support a model of recent dispersal from coastal to inland rainforests. Our results demonstrate that binning is an effective strategy for the construction of a mSFS and imply that the statistical power of RF when applied to demographic model selection is at least comparable to traditional ABC algorithms. Importantly, by combining these strategies, large sets of models with differing numbers of populations can be evaluated.