Julien Clavel

Multivariate phylogenetic comparative methods for modeling high-dimensional traits such as 3D shapes or gene expression profiles have been recently developed. However, these approaches are impractical and almost impossible to use when the number of traits exceeds a few thousands, as they become computationally prohibitive. We overcome these limitations by proposing a new maximum likelihood approach based on the Empirical Bayes framework. This approach takes into account the information of the complete covariances (among species and traits) to infer parameters and compare trait evolution models for high-dimensional data sets. Through simulations, we demonstrate that the proposed approach can accurately estimate parameters of various trait evolution models, even when the number of traits is 10 times larger than the number of lineages; it requires less memory and is at least 10 times faster than currently available approaches. This efficient framework enabled us to extend the high-dimensional multivariate phylogenetic comparative toolkit by including an Ornstein–Uhlenbeck process with multiple optima to study adaptation to various selective regimes. Applying our approach to the evolution of jaw morphology in relation to dietary adaptation in mammals, we demonstrate morphological convergence in carnivorous and herbivorous lineages. The proposed Empirical Bayes approach, implemented in the R package mvMORPH, enables phylogenetic comparative methods to efficiently handle high-dimensional data sets and complex models of trait evolution.

Meiotic recombination, the exchange of genetic material between parental chromosomes during gamete production, is critical for fertility, genome stability, and evolutionary adaptation. In eukaryotes, meiotic recombination is carried out by a deeply conserved molecular machinery. Despite this conservation, the sequences of proteins involved in meiotic recombination evolve at a remarkably high rate in mammalian species. Several biological processes have been proposed to explain this rapid evolution, but none have been quantitatively tested. In this study, we analyzed the variation in evolutionary rates of 100 recombination proteins across approximately 400 placental mammals. Our results show that this rapid evolution is primarily driven by lower levels of purifying selection compared to the rest of the proteome. We show that, although selective pressures exerted on recombination proteins are generally shared across mammals, a few recombination proteins exhibit strong variation in selective pressures. Contrary to previous hypotheses, we demonstrate that chromosome number and genome-wide recombination rates do not account for much of this variation. Instead, these variations are primarily associated with chromosome pairing and synapsis proteins, which tend to experience increased selective pressures throughout mammalian evolution, especially in large and long-lived species. This pattern probably reflects more intense selection for stability of chromosome pairing proteins in the oocytes of long-lived species, in which chromosomes sometimes need to stay paired for decades.

Given the ever-increasing availability of highly detailed phenotypes, modelling trait evolution in a multivariate framework is becoming a challenging task. Current phylogenetic comparative methods often struggle with high-dimensional datasets because they suffer computational limitations and interpretability. Here, we propose a maximum likelihood-based approach called Probabilistic and Phylogenetic Principal Components Analysis (P3CA) to circumvent current limitations. This approach is based on a continuous latent variable model, whereby observed traits are explained by a smaller number of unobserved variables that evolve according to a given evolutionary model. We implement the approach under Pagel’s lambda model using an Expectation-Maximisation algorithm that makes it computationally efficient and allows missing values. Using simulations, we demonstrate that evolutionary parameters are accurately estimated, regardless of phylogenetic signal, the number of traits or the proportion of missing values. The reconstruction of the reduced space is more accurate than the one obtained using other dimensionality reduction approaches, such as phylogenetic and conventional PCA. Likewise, the estimated values for missing data are more accurate than those obtained using current phylogenetic data imputation approaches. We illustrate the approach on a 3D geometric morphometric dataset describing Crocodyliformes skull shapes and containing around 4% of missing data. Our P3CA method unlocks the possibility to analyse and more easily interpret the large-scale multivariate datasets generated in recent decades within a phylogenetic comparative framework.

Comprendre comment évoluent les écosystèmes lors de perturbations majeures de l'environnement nécessite de prendre en compte l'histoire évolutive des espèces, c'est-à-dire leur phylogénie. Dans ce Mémoire de thèse, j'utilise une approche modélisatrice afin d'étudier l'évolution phénotypique en lien avec l'environnement (écomorphologie), approche incluant des données actuelles et fossiles et visant in fine à comprendre comment les écosystèmes se diversifient et s'organisent structurellement et fonctionnellement à l'échelle des temps géologiques. Le cadre historique, fourni par la phylogénie, définit une trame analytique commune à l'étude de taxons actuels et fossiles. Différents outils dédiés, d'une part au traitement de données morphométriques souvent incomplètes, et d'autre part aux études écomorphologiques évolutives dans un contexte phylogénétique, sont développés et discutés. Cette thèse s'articule autour de deux ateliers consacrés à l'étude macroévolutive et macroécologique de grands mammifères « ongulés » : Bovidae, Cervidae, et Equidae. Le premier atelier concerne les ruminants actuels, clade très diversifié et à l'écologie des espèces constitutives connue. Les analyses révèlent des modes évolutifs contrastés entre Cervidae et Bovidae actuels, en fonction des niches écologiques et des caractères écomorphologiques considérés. La diversification moi pliocène des Bovidae africains apparaît corrélée à des évènements globaux, tandis que leur évolution phénotypique révèle des modes de diversification différents selon les habitats. Le second atelier est focalisé sur l'impact d'un événement majeur ayant affecté l'ensemble du pourtour méditerranéen à la limite Miocène- Pliocène, il y a 5,3 millions d'années : la Crise de Salinité Messinienne. Il repose sur l'étude d' « ongulés » fossiles et apporte des précisions sur l'aspect temporel de l'évolution de ces mammifères avant et après cet événement. Les principaux résultats illustrent des variations rapides de la structure phylogénétique des assemblages d'ongulés qu'il est possible de relier au contexte biogéographique et aux variations climatiques locales et régionales. Les renouvellements fauniques (acteurs) et fonctionnels (rôles), initiés dès la fin du Miocène moyen, apparaissent progressifs et non soudains. Les résultats obtenus offrent une meilleure caractérisation et compréhension des réponses évolutives de ces mammifères, grands consommateurs primaires souvent parmi les premiers menacés lors de perturbations climatiques et environnementales majeures

Multivariate phylogenetic comparative methods for modelling high-dimensional traits such as 3D shapes or gene expression proBiles have been recently developed. However, these approaches are impractical and almost impossible to use when the number of traits exceeds a few thousands, as they become computationally prohibitive. We overcome these limitations by proposing a new maximum likelihood approach based on the Empirical Bayes framework. This approach takes into account the information of the complete covariances (among species and traits) to infer parameters and compare models of trait evolution for high-dimensional datasets. Through simulations, we demonstrate that the proposed approach can accurately estimate parameters of various trait evolution models, even when the number of traits is ten times larger than the number of lineages; it requires less memory and is at least 10 times faster than currently available approaches. This fast, efBicient framework enabled us to extend the high-dimensional multivariate phylogenetic comparative toolkit by including an Ornstein-Uhlenbeck process with multiple optima to study adaptation to various selective regimes. Applying our approach to the evolution of jaw morphology in relation to dietary adaptation in mammals, we demonstrate morphological convergence in carnivorous and herbivorous lineages. The proposed Empirical Bayes framework, implemented in the R package mvMORPH, enables phylogenetic comparative methods to efBiciently handle high-dimensional datasets and complex models of trait evolution.

Morphological change occurs over macroevolutionary timescales under the action of natural selection and genetic drift combined with developmental processes shaping organogenesis. Although determining their relative weight is made difficult by discrepancies between paleontological and neontological data, mammalian tooth morphology may bridge the gap between fossil record and laboratory observations. Fossils indicate that mammals have frequently diversified after evolving molars bearing more cusps, while developmental biology shows these emerge through the iterative signaling of enamel knots. However, this theoretical evo-devo model of mammalian tooth evolution has not been tested with empirical data from both fossils and laboratory experiments. In doing so, we identify a shared developmental basis for the convergent, ratcheted evolution of increasingly complex molars in arvicoline rodents (voles, lemmings, muskrats). Longer, narrower molars lead to more cusps throughout development and deep time, suggesting that tooth development directed morphological evolution. Both the arvicoline fossil record and vole tooth development show slower transitions toward the highest cusp counts. This pattern suggests that the developmental processes fueling the evolution of increasingly complex molars may also limit the potential for further complexity increases. Integrating paleontological and developmental data shows that long-term evolutionary trends can be accurately and mostly explained by the simple tinkering of developmental pathways.

Endothermy has independently evolved in several vertebrate lineages but remains rare among fishes. Using an integrated approach combining phylogenomic and ecomorphological data for 1051 ray-finned fishes, a time-dependent evolutionary model, and comparative genomic analyses of 205 marine vertebrates, we show that ecological interactions with modern cetaceans coincided with the evolution of endothermy in ray-finned fishes during the Eocene-Miocene. This result is supported by evidence of temporal and geographical overlap between cetaceans and endothermic fish lineages in the fossil record, as well as correlations between cetacean diversification and the origin of endothermy in fishes. Phylogenetic comparative analyses identified correlations between endothermy, large body sizes, and specialized swimming modes while challenging diet specialization and depth range expansion hypotheses. Comparative genomic analyses identified several genes under selection in endothermic lineages, including carnmt1 (involved in fatty acid metabolism) and dcaf6 (associated with development). Our findings advance the understanding of how ecological interactions and genomic factors shape key adaptations.

Metamorphosis, a common and spectacular developmental process, involves major phenotypic changes often linked to environmental transitions. Understanding how these changes affect morphological evolution is crucial for examining metamorphosis's evolutionary impacts. In this study, we analysed the role of metamorphosis and life cycle variation on the morphological evolution of the mandible in Caudata (salamanders and newts), a clade that exhibits a great diversity of life cycles and ecologies. We focused on the mandible, a key component of the masticatory apparatus, to investigate how life cycle variation may influence the diversification of feeding modes. Using 3D geometric morphometrics on 317 extant species of Caudata, we found that both life cycle and size influence mandible shape, with complex interactions between these factors. The morphology of the mandibular bones evolved more independently in species undergoing complete metamorphosis compared to those with partial metamorphosis, likely facilitating feeding diversification in terrestrial adults. On the other hand, high morphological integration in species undergoing partial metamorphosis could promote the rapid evolution of traits, resulting in the differentiation of more disparate morphologies, potentially aiding paedomorphic species in adapting to challenging environments. This study highlights the interplay between metamorphosis and the environment in shaping morphological evolution among species and provides new insights into the role of heterochrony in the morphological diversification of organisms. Read the free Plain Language Summary for this article on the Journal blog.

Morphological change occurs over macroevolutionary timescales under the action of natural selection and genetic drift combined with developmental processes shaping organogenesis ^1–3 . Although determining their relative weight is made difficult by discrepancies between palaeontological and neontological data ⁴ , mammalian tooth morphology may bridge the gap between fossil record and laboratory observations. Fossils indicate that placentals and marsupials diversified after evolving molars bearing more cusps ^5–7 , which emerge through the iterative signalling of enamel knots ^8,9 . However, this theoretical evo-devo model of mammalian tooth evolution has not been explicitly assessed on empirical data. Here, by combining the fossil record with laboratory experiments, we identify a shared developmental basis for the convergent, ratcheted evolution of increasingly complex molars in arvicoline rodents (voles, lemmings, muskrats). Longer, narrower molars lead to more cusps throughout development and deep time, suggesting a developmental driver for tooth evolution. Both the arvicoline fossil record and vole tooth development show slower transitions toward the highest cusp counts. This newly identified pattern suggests that the developmental processes fuelling the evolution of increasingly complex molars may also limit the potential for further complexity increases. Integrating palaeontological and developmental data shows that long-term evolutionary trends can be accurately explained by the simple tinkering of developmental pathways.

Organismal morphology has been at the core of study of biodiversity for millennia before the formalization of the concept of evolution. In the early to mid-twentieth century, a strong theoretical framework was developed for understanding both pattern and process of morphological evolution on a macroevolutionary scale. The past half century has been a transformational period for the study of evolutionary morphology, in both the quantification of morphology and novel analytical tools for estimating how and why morphological diversity changes through time, with a marked increase in studies apparent in the early 1990s. We are now at another inflection point in the study of morphological evolution, with the availability of vast amounts of high-resolution data sampling extant and extinct diversity allowing ‘omics’-scale analysis. Artificial intelligence is already increasing the pace of phenomic data collection even further. This new reality, where the ability to obtain data is quickly outpacing the ability to analyse it with robust, realistic evolutionary models, brings a new set of challenges, and we here present analyses demonstrating these challenges and discussing solutions. Fully transitioning the study of morphological evolution into the “Omics” era will involve the development of tools to automate the extraction of meaningful, comparable morphometric data from images, integrate fossil data into large phylogenetic trees and downstream evolutionary analyses, and generate robust models that both accurately reflect the complexity of evolutionary processes and are well-suited for high-dimensional data. Combined, these advancements will solidify the emerging field of evolutionary phenomics and appropriately center it around the analysis of unambiguously critical deep-time data.