Fossil Information Research Articles

Practical guidelines for Bayesian phylogenetic inference using Markov chain Monte Carlo (MCMC)

Phylogenetic estimation is, and has always been, a complex endeavor. Estimating a phylogenetic tree involves evaluating many possible solutions and possible evolutionary histories that could explain a set of observed data, typically by using a model of evolution. Values for all model parameters need to be evaluated as well. Modern statistical methods involve not just the estimation of a tree, but also solutions to more complex models involving fossil record information and other data sources. Markov chain Monte Carlo (MCMC) is a leading method for approximating the posterior distribution of parameters in a mathematical model. It is deployed in all Bayesian phylogenetic tree estimation software. While many researchers use MCMC in phylogenetic analyses, interpreting results and diagnosing problems with MCMC remain vexing issues to many biologists. In this manuscript, we will offer an overview of how MCMC is used in Bayesian phylogenetic inference, with a particular emphasis on complex hierarchical models, such as the fossilized birth-death (FBD) model. We will discuss strategies to diagnose common MCMC problems and troubleshoot difficult analyses, in particular convergence issues. We will show how the study design, the choice of models and priors, but also technical features of the inference tools themselves can all be adjusted to obtain the best results. Finally, we will also discuss the unique challenges created by the incorporation of fossil information in phylogenetic inference, and present tips to address them.

Practical guidelines for Bayesian phylogenetic inference using Markov Chain Monte Carlo (MCMC)

Phylogenetic estimation is, and has always been, a complex endeavor. Estimating a phylogenetic tree involves evaluating many possible solutions and possible evolutionary histories that could explain a set of observed data, typically by using a model of evolution. Modern statistical methods involve not just the estimation of a tree, but also solutions to more complex models involving fossil record information and other data sources. Markov Chain Monte Carlo (MCMC) is a leading method for approximating the posterior distribution of parameters in a mathematical model. It is deployed in all Bayesian phylogenetic tree estimation software. While many researchers use MCMC in phylogenetic analyses, interpreting results and diagnosing problems with MCMC remain vexing issues to many biologists. In this manuscript, we will offer an overview of how MCMC is used in Bayesian phylogenetic inference, with a particular emphasis on complex hierarchical models, such as the fossilized birth-death (FBD) model. We will discuss strategies to diagnose common MCMC problems and troubleshoot difficult analyses, in particular convergence issues. We will show how the study design, the choice of models and priors, but also technical features of the inference tools themselves can all be adjusted to obtain the best results. Finally, we will also discuss the unique challenges created by the incorporation of fossil information in phylogenetic inference, and present tips to address them.

Imbalanced speciation pulses sustain the radiation of mammals.

The evolutionary histories of major clades, including mammals, often comprise changes in their diversification dynamics, but how these changes occur remains debated. We combined comprehensive phylogenetic and fossil information in a new "birth-death diffusion" model that provides a detailed characterization of variation in diversification rates in mammals. We found an early rising and sustained diversification scenario, wherein speciation rates increased before and during the Cretaceous-Paleogene (K-Pg) boundary. The K-Pg mass extinction event filtered out more slowly speciating lineages and was followed by a subsequent slowing in speciation rates rather than rebounds. These dynamics arose from an imbalanced speciation process, with separate lineages giving rise to many, less speciation-prone descendants. Diversity seems to have been brought about by these isolated, fast-speciating lineages, rather than by a few punctuated innovations.

Historical climate change and vicariance events contributed to the intercontinental disjunct distribution pattern of ash species (Fraxinus, Oleaceae)

The Northern Hemisphere temperate forests exhibit a disjunct distributional pattern in Europe, North America, and East Asia. Here, to reveal the promoter of intercontinental disjunct distribution, Fraxinus was used as a model organism to integrate abundant fossil evidence with high-resolution phylogenies in a phytogeographic analysis. We constructed a robust phylogenetic tree using genomic data, reconstructed the geographic ancestral areas, and evaluated the effect of incorporating fossil information on the reconstructed biogeographic history. The phylogenetic relationships of Fraxinus were highly resolved and divided into seven clades. Fraxinus originated in western North America during Eocene, and six intercontinental dispersal events and five intercontinental vicariance events were occured. Results suggest that climate change and vicariance contributed to the intercontinental disjunct distribution pattern of Fraxinus. Moreover, results highlight the necessity of integrating phylogenetic relationship and fossil to improve the reliability of inferred biogeographic events and our understanding of the processes underlying disjunct distributions.

Looking for the oldest diatoms

Paleontological observations of ancient flora and fauna provide powerful insights into past diversity and relationship dynamics between organisms and their environments. Diatoms are globally distributed protists that influence major biogeochemical cycles and sustain oceanic food webs. The fossil diatom record extends 120 million years back to the Early Cretaceous where rare deposits were discovered worldwide and are occasionally represented by diverse communities. However scarce, the taxonomic richness and geographical spread of these diatom communities suggest prior evolutionary events and therefore earlier deposits. To complement the existing fossil information and to discover diatom deposits predating 120 Ma, we examined 33 study sites from cores and outcrops across oceans and continents. These efforts did not generate new fossil discoveries, however. Our assessment suggests biogenic silica that comprises the cell wall of diatoms was likely dissolved from Mesozoic sediments through diagenetic processes. Altogether, the search for the oldest diatoms must continue but should target sediments that experienced shallow burial and concretions.

Validating a molecular clock for nudibranchs-No fossils to the rescue.

Time calibrated phylogenies are typically reconstructed with fossil information but for soft-bodied marine invertebrates that lack hard parts, a fossil record is lacking. In these cases, biogeographic calibrations or the rates of divergence for related taxa are often used. Although nudibranch phylogenies have advanced with the input of molecular data, no study has derived a divergence rate for this diverse group of invertebrates. Here, we use an updated closure date for the Isthmus of Panama (2.8 Ma) to derive the first divergence rates for chromodorid nudibranchs using multigene data from a geminate pair with broad phylogeographic sampling. Examining the species Chromolaichma sedna (Marcus & Marcus, 1967), we uncover deep divergences among eastern Pacific and western Atlantic clades and we erect a new species designation for the latter (Chromolaichma hemera sp. nov.). Next, we discover extensive phylogeographic structure within C. hemera sp. nov. sensu lato, thereby refuting the hypothesis of a recent introduction. Lastly, we derive divergence rates for mitochondrial and nuclear loci that exceed known rates for other gastropods and we highlight significant rate heterogeneity both among markers and taxa. Together, these findings improve understanding of nudibranch systematics and provide rates useful to apply to divergence scenarios in this diverse group.

Practical guidelines for Bayesian phylogenetic inference using Markov chain Monte Carlo (MCMC).

Phylogenetic estimation is, and has always been, a complex endeavor. Estimating a phylogenetic tree involves evaluating many possible solutions and possible evolutionary histories that could explain a set of observed data, typically by using a model of evolution. Values for all model parameters need to be evaluated as well. Modern statistical methods involve not just the estimation of a tree, but also solutions to more complex models involving fossil record information and other data sources. Markov chain Monte Carlo (MCMC) is a leading method for approximating the posterior distribution of parameters in a mathematical model. It is deployed in all Bayesian phylogenetic tree estimation software. While many researchers use MCMC in phylogenetic analyses, interpreting results and diagnosing problems with MCMC remain vexing issues to many biologists. In this manuscript, we will offer an overview of how MCMC is used in Bayesian phylogenetic inference, with a particular emphasis on complex hierarchical models, such as the fossilized birth-death (FBD) model. We will discuss strategies to diagnose common MCMC problems and troubleshoot difficult analyses, in particular convergence issues. We will show how the study design, the choice of models and priors, but also technical features of the inference tools themselves can all be adjusted to obtain the best results. Finally, we will also discuss the unique challenges created by the incorporation of fossil information in phylogenetic inference, and present tips to address them.

The Gaia-ESO Survey: Chemical tagging in the thin disk

Context. The chemical makeup of a star provides the fossil information of the environment where it formed. Under this premise, it should be possible to use chemical abundances to tag stars that formed within the same stellar association. This idea – known as chemical tagging – has not produced the expected results, especially within the thin disk where open stellar clusters have chemical patterns that are difficult to disentangle. Aims. The ultimate goal of this study is to probe the feasibility of chemical tagging within the thin disk population using high-quality data from a controlled sample of stars. We also aim at improving the existing techniques of chemical tagging and giving some kind of guidance on different strategies of clustering analysis in the elemental abundance space. Methods. Here we develop the first blind search of open clusters’ members through clustering analysis in the elemental abundance space using the OPTICS algorithm applied to data from the Gaia-ESO survey. First, we evaluate different strategies of analysis (e.g., choice of the algorithm, data preprocessing techniques, metric, space of data clustering), determining which ones are more performing. Second, we apply these methods to a data set including both field stars and open clusters attempting a blind recover of as many open clusters as possible. Results. We show how specific strategies of data analysis can improve the final results. Specifically, we demonstrate that open clusters can be more efficaciously recovered with the Manhattan metric and on a space whose dimensions are carefully selected. Using these (and other) prescriptions we are able to recover open clusters hidden in our data set and find new members of these stellar associations (i.e., escapers, binaries). Conclusions. Our results indicate that there are chances of recovering open clusters’ members via clustering analysis in the elemental abundance space, albeit in a data set that has a very high fraction of cluster members compared to an average field star sample. Presumably, the performances of chemical tagging will further increase with higher quality data and more sophisticated clustering algorithms, which will likely became available in the near future.

The origin of modern patterns of continental diversity in Mauritiinae palms: the Neotropical museum and the Afrotropical graveyard.

While the latitudinal diversity gradient has received much attention, biodiversity and species richness also vary between continents across similar latitudes. Fossil information can be used to understand the evolutionary mechanisms that generated such variation between continents of similar latitudes. We integrated fossil data into a phylogenetic analysis of the Mauritiinae palms, whose extant diversity is restricted to the Neotropics, but extended across Africa and India during most of the Cenozoic. Mauritiinae diverged from its sister lineage Raphiinae ca 106 Ma. Using ancestral state estimation and a lineage through time analysis, we found that diversity arose globally during the late Cretaceous and Palaeocene across South America, Africa and India. The Palaeocene-Eocene transition (ca 56 Ma) marked the end of global Mauritiinae expansion, and the beginning of their decline in both Africa and India. Mauritiinae disappeared from the Indian subcontinent and Africa at the end of the Eocene and the Miocene, respectively. By contrast, Neotropical diversity steadily increased over the last 80 Myr. Taken together, our results suggest that the Neotropics functioned as a continental-scale refuge for Mauritiinae palms, where lineages survived and diversified while global climatic changes that drastically reduced rainforests led to their demise on other continents.

Hydrodynamics and sliding posture analysis of the Cambrian arthropod Ercaicunia multinodosa

Dynamic mechanical analysis offers the opportunity to explore the motility and feeding strategies of extinct organisms, but the prerequisite is to have an accurate recovery engineering model. As shown with micro-computed tomography (CT) scanning, fossils of the bivalved arthropod Ercaicunia multinodosa from the Cambrian (Series 2, Stage 3) Chengjiang biota of China have well-preserved three-dimensional (3D) morphological details with a certain degree of compression. Here, we propose a palaeontological restoration method using computational fluid dynamics (CFD) to analyse multiple hypothetical models based on the fossil information to obtain a reasonable restoration model. Furthermore, we carry out hydrodynamic experiments to verify the palaeontological restoration results. Our simulation and experimental results suggest that, a dorsally convex and ventrally straight body shape with the valves opening at an angle of approximately 120° works best for E. multinodosa to overcome resistance, and in the meantime obtain most lift while sliding in the water column. The combination of three-dimensional reconstruction, CFD simulation, and hydrodynamic experiments provides a useful method for restoring the morphologies of extinct animals and exploring their palaeoecology.

CladeDate: Calibration information generator for divergence time estimation

Abstract Time‐scaled phylogenetic trees are essential tools in modern biology and node‐based calibrations have been the main approach to time‐tree estimation. But methods for generating the required calibration information are scarce and difficult to parameterize. Here, I present CladeDate, an R package for the generation of empirical calibration information from the fossil record. CladeDate uses simple mathematical models to estimate the age of a clade and its uncertainty based on fossil times. Using a Monte Carlo approach, CladeDate generates empirical densities representing the uncertainty associated with the age of the clade and fits standard probability density functions that can be used in time‐tree inference software such as BEAST2, MrBayes and MCMCtree. I show with simulations that the calibration information generated with CladeDate produces accurate time‐trees and compares favourably with more complex methods. I demonstrate the use CladeDate for the generation of calibration information, including point estimates, upper bounds and calibration densities, for passerine birds. CladeDate is particularly suited for groups with limited fossil information or when assumptions of more complex methods are not met, and provides a general and practical solution to the problem of generating calibration information for divergence time estimation.

Fossils Do Not Substantially Improve, and May Even Harm, Estimates of Diversification Rate Heterogeneity.

The fossilized birth-death (FBD) model is a naturally appealing way of directly incorporating fossil information when estimating diversification rates. However, an important yet often overlooked property of the original FBD derivation is that it distinguishes between two types of sampled lineages. Here, we first discuss and demonstrate the impact of severely undersampling, and even not including fossils that represent samples of lineages that also had sampled descendants. We then explore the benefits of including fossils, generally, by implementing and then testing two types of FBD models, including one that converts a fossil set into stratigraphic ranges, in more complex likelihood-based models that assume multiple rate classes across the tree. Under various simulation scenarios, including a scenario that exists far outside the set of models we evaluated, including fossils rarely outperform analyses that exclude them altogether. At best, the inclusion of fossils improves precision but does not influence bias. Similarly, we found that converting the fossil set to stratigraphic ranges, which is one way to remedy the effects of undercounting the number of k-type fossils, results in turnover rates and extinction fraction estimates that are generally underestimated. Although fossils remain essential for understanding diversification through time, in the specific case of understanding diversification given an existing, largely modern tree, they are not especially beneficial. [Fossilized birth-death; fossils; MiSSE; state speciation extinction; stratigraphic ranges; turnover rate.].

Dating Microbial Evolution with MCMCtree.

This protocol explains how to use the program MCMCtree to estimate divergence times in microbial phylogenies. The main advantage of MCMCtree is the implementation of an approximation to the molecular data likelihood that dramatically speeds up computation during Bayesian MCMC sampling of divergence times and evolutionary rates. The approximation allows the analysis of large phylogenies with hundreds of taxa and molecular alignments with thousands or millions of sites. Two examples are used to illustrate Bayesian clock dating with MCMCtree. The first is a phylogeny of (mostly) microbial eukaryotes and prokaryotes encompassing the major groups of life on Earth, and for which fossil information, to calibrate the nodes of the phylogeny, is available. The second is a phylogeny of influenza viruses with known sampling times. An overview of Bayesian MCMC sampling is given as well as practical advice on issues such as construction of the time and rate prior and assessment of convergence of MCMC chains. Strategies for estimating times in microbial phylogenies for which neither fossil information nor sampling times are known are discussed.

Biogeography of curimatid fishes reveals multiple lowland-upland river transitions and differential diversification in the Neotropics (Teleostei, Curimatidae).

The Neotropics harbors a megadiverse ichthyofauna comprising over 6300 species with approximately 80% in just three taxonomic orders within the clade Characiphysi. This highly diverse group has evolved in tropical South America over tens to hundreds of millions of years influenced mostly by re‐arrangements of river drainages in lowland and upland systems. In this study, we investigate patterns of spatial diversification in Neotropical freshwater fishes in the family Curimatidae, a species‐rich clade of the order Characiformes. Specifically, we examined ancestral areas, dispersal events, and shifts in species richness using spatially explicit biogeographic and macroevolutionary models to determine whether lowlands–uplands serve as museums or cradles of diversification for curimatids. We used fossil information to estimate divergence times in BEAST, multiple time‐stratified models of geographic range evolution in BioGeoBEARS, and alternative models of geographic state‐dependent speciation and extinction in GeoHiSSE. Our results suggest that the most recent common ancestor of curimatids originated in the Late Cretaceous likely in lowland paleodrainages of northwestern South America. Dispersals from lowland to upland river basins of the Brazilian and Guiana shields occurred repeatedly across independently evolving lineages in the Cenozoic. Colonization of upland drainages was often coupled with increased rates of net diversification in species‐rich genera such as Cyphocharax and Steindachnerina. Our findings demonstrate that colonization of novel aquatic environments at higher elevations is associated with an increased rate of diversification, although this pattern is clade‐dependent and driven mostly by allopatric speciation. Curimatids reinforce an emerging perspective that Amazonian lowlands act as a museum by accumulating species along time, whereas the transitions to uplands stimulate higher net diversification rates and lineage diversification.

Chemical evidence for planetary ingestion in a quarter of Sun-like stars

Stellar members of binary systems are formed from the same material, therefore they should be chemically identical. However, recent high-precision studies have unveiled chemical differences between the two members of binary pairs composed by Sun-like stars. The very existence of these chemically inhomogeneous binaries represents one of the most contradictory examples in stellar astrophysics and source of tension between theory and observations. It is still unclear whether the abundance variations are the result of chemical inhomogeneities in the protostellar gas clouds or instead if they are due to planet engulfment events occurred after the stellar formation. While the former scenario would undermine the belief that the chemical makeup of a star provides the fossil information of the environment where it formed, a key assumption made by several studies of our Galaxy, the second scenario would shed light on the possible evolutionary paths of planetary systems. Here, we perform a statistical study on 107 binary systems composed by Sun-like stars to provide - for the first time - unambiguous evidence in favour of the planet engulfment scenario. We also establish that planet engulfment events occur in stars similar to our own Sun with a probability ranging between 20 and 35$\%$. This implies that a significant fraction of planetary systems undergo very dynamical evolutionary paths that can critically modify their architectures, unlike our Solar System which has preserved its planets on nearly circular orbits. This study also opens to the possibility of using chemical abundances of stars to identify which ones are the most likely to host analogues of the calm Solar System.

Of pandas, fossils, and bamboo forests: ecological niche modeling of the giant panda (Ailuropoda melanoleuca) during the Last Glacial Maximum

Abstract The giant panda (Ailuropoda melanoleuca) is the most basal living species in the phylogeny of the family Ursidae, with a specialized diet composed of a variety of bamboo species. The evolutionary history and past distribution patterns of the giant panda remain poorly understood. Our aim was to integratively apply distinct methods to evaluate the evolutionary history and distributional patterns of the giant panda; these included phylogeography, ecological niche modeling (ENM), and fossil data. To this end, we characterized the panda’s past and present ecological niches and the environmental conditions that define them. To estimate the panda’s phylogeographic patterns and the environmental conditions (precipitation and temperature) available across its historical geographic range, we evaluated its past distribution during the Last Glacial Maximum (LGM). Considering that modeling biotic interactions (e.g., foraging, predation) is still an enormous challenge, we propose a novel modeling strategy based on the panda’s specialized diet, using an ensemble of three bamboo genera with distribution across the panda’s historical geographic range. Finally, we tested the accuracy of our approach by evaluating its ability to predict the LGM fossils. Our results support that the panda’s diversification across its distribution happened ca. 2.7 million years (Mya), coinciding with the likely period when the panda changed from a carnivorous to a vegetarian diet (from the Pleistocene to the Pliocene), acquiring its exclusively bamboo-feeding habits until the mid-Holocene. Our findings provide evidence of a historical directional niche change along which the panda has currently reached the lower limits of temperature and precipitation conditions existing on the geography where its food is available. Our proposed ENM based on the panda’s food habits accurately predicted 85.7% of the LGM fossils, in stark contrast with the traditional approach of modeling the distribution of species by using exclusively its own occurrences. These results provide insights on how to include Eltonian components to undertake more robust ENM when only abiotic variables are available. We emphasize the importance of integrating fossil information, whenever available, into the niche modeling process to include the historical component when estimating species ecological niches.

Scolicia and its producer in shallow-marine deposits of the Miocene Chenque Formation (Patagonia, Argentina): functional morphology and implications for understanding burrowing behavior

Scolicia is one of the most conspicuous trace fossils in lower shoreface deposits of the Miocene Chenque Formation of Patagonia, Argentina. This ichnotaxon consists of horizontal, sinuous or meandering trace fossils with a laminated backfill and two parallel strings located at the base. Abundant body fossils attributed to Brisaster iheringi occur in close association to these trace fossils. The echinoids are very well preserved, and most of the specimens have some areas with their spines attached in life position. In particular, preservation of the subanal tufts of spines, associated to isopores (which are interpreted as being associated to funnel building tube feet), supports the interpretation that these organisms are the producers of Scolicia isp. A shaft connecting the burrow with the sediment-water interface was not observed in vertical sections of the excavations, although on bedding plane surfaces, some small circular tube outlets are recognized. However, the great diversity and abundance of ichnofossils co-occurring in these deposits may preclude adequate identification of vertical structures. These deposits provide an excellent opportunity to integrate trace fossil data, body fossil information and observations from modern analogues in order to perform a morpho-functional analysis of Scolicia.

THE IMPORTANCE OF THE MUSEUM IN ANTEBELLUM U.S. WESTERN TERRITORIAL EXPLORATION: UNDERSTANDING THE RELEVANCE OF COLLECTING FOSSILS AND THEIR CONSERVATION TO SOLVING LONG-STANDING GEOLOGIC AND PALEONTOLOGIC PROBLEMS—PART 1

The determination of pivotal moments in the history of a discipline of science can depend on the perspective of the observer. This narrative notes the importance of antebellum institutions in fostering research, research communication, and the potential for fossil conservation. The Smithsonian Institution (U.S. National Museum = National Museum of Natural History) provided a federal umbrella for fossil collection and curation when one was needed. However, along with other institutions, the success record of conserved fossil continental mollusks prior to 1855 is abysmal. Fossils from the first (Frémont in 1843), second (Harris–Audubon in 1843), and third (Evans–Shumard in 1853) expeditions to collect specimens are all now missing. As a clue to the general state of confusion one, terrestrial snail named by Hall and Meek (1855) was misplaced for over a century, but was recently found. Continental molluscan fossils should have served as temporal and environmental landmarks in the construction of geologic maps produced in the 1850s by Hitchcock, Marcou, Rogers, and Hall and Lesley. However, except for the Hall and Lesley map, they did not. Fossil information was published and available, but many fossils were not accessible. The Smithsonian was the recipient of Hayden's fossils and natural science specimens collected in 1854 and 1855. Hayden's fossils and observations resulted in numerous publications, not the least of which were those by Meek and Hayden in 1856 and 1857. For reasons that remain unknown, a number of type specimens (and associated material) used to describe species in 1856 were replaced by Meek in his 1876a monograph, when Meek and Hayden upper Missouri and Yellowstone River species were finally illustrated. Thus, undeclared neotypes have been masquerading as holotypes or members of syntypic (cotypic) series. Meek and Hayden entered the field of western territorial geological studies with only the preconceptions of geology not particularly relevant to what they were about to see. Their claim to fame was not subtle—they published based on observations and specimens. In almost all ways that were important, they were starting from scratch.

A new, fast method to search for morphological convergence with shape data.

Morphological convergence is an intensely studied macroevolutionary phenomenon. It refers to the morphological resemblance between phylogenetically distant taxa. Currently available methods to explore evolutionary convergence either: rely on the analysis of the phenotypic resemblance between sister clades as compared to their ancestor, fit different evolutionary regimes to different parts of the tree to see whether the same regime explains phenotypic evolution in phylogenetically distant clades, or assess deviations from the congruence between phylogenetic and phenotypic distances. We introduce a new test for morphological convergence working directly with non-ultrametric (i.e. paleontological) as well as ultrametric phylogenies and multivariate data. The method (developed as the function search.conv within the R package RRphylo) tests whether unrelated clades are morphologically more similar to each other than expected by their phylogenetic distance. It additionally permits using known phenotypes as the most recent common ancestors of clades, taking full advantage of fossil information. We assessed the power of search.conv and the incidence of false positives by means of simulations, and then applied it to three well-known and long-discussed cases of (purported) morphological convergence: the evolution of grazing adaptation in the mandible of ungulates with high-crowned molars, the evolution of mandibular shape in sabertooth cats, and the evolution of discrete ecomorphs among anoles of Caribbean islands. The search.conv method was found to be powerful, correctly identifying simulated cases of convergent morphological evolution in 95% of the cases. Type I error rate is as low as 4–6%. We found search.conv is some three orders of magnitude faster than a competing method for testing convergence.

A tale of four bears: Environmental signal on the phylogeographical patterns within the extant Ursus species

AbstractAimAssessing the relevance of niche evolution in the diversification patterns and geographical distribution of species driven by climate remains a challenge. We apply an integrative approach to evaluate the role of the environment on the phylogeography of bear species, incorporating fossil data to characterize the changes in the ecological niche through time. We evaluate our approach with the four extant species of bears within Ursus, the best represented taxon in the fossil record of the family Ursidae.LocationEurasia and North America.TaxaAsian black bear, Ursus thibetanus; American black bear, U. americanus; Brown bear, U. arctos; and Polar bear, U. maritimus.MethodsWe built a genetic and a geographical database from all published mitochondrial DNA sequences and of species occurrence records. We defined the most significant climatic variables based on each species ecological realm using correlation matrices, and characterized the ecological niches and existing environmental conditions with ellipsoid models. We inferred their current and Last Glacial Maximum (LGM) ecological niche modellings (ENMs) and compared the results with the fossil record. We estimated the times of divergence (d‐loop sequences) of lineages and applied a phyloclimatespace approach to discern the phylogeographical patterns along each species’ ecological space.ResultsEcological niche modelling showed wider niches for U. thibetanus and U. americanus encompassing higher temperature and precipitation, while U. arctos and U. maritimus showed an opposite pattern. LGM models were consistent with the fossil record, predicting 55%–89% of the fossil occurrences (within their suitability areas). The phyloclimatespace revealed different degrees of environmental signal in the lineages’ phylogeographical patterns and ecological trajectories associated with LGM climatic conditions. Results indicated habitat tracking and ecological expansion since the LGM towards more extreme precipitation and temperature conditions for three species, except U. maritimus that showed ecological niche reduction.Main ConclusionsIncorporating fossil information from the LGM improved our characterization and interpretation of ecological models, by enabling definition of the limits of the climatic conditions explored by the species in the past. Our approach also provided insights about the existing set of environmental conditions shaping the ecological niche divergence of Ursus bears. We were able to depict key features of the lineages’ evolutionary history, ecology and distribution, revealing the dynamics of niche occupation and the environmental signal on the phylogeographical patterns of Ursus.