Plant diversity darkspots for global collection priorities.

More than 15% of all vascular plant species may remain scientifically undescribed, and many of the > 350 000 described species have no or few geographic records documenting their distribution. Identifying and understanding taxonomic and geographic knowledge shortfalls is key to prioritising future collection and conservation efforts. Using extensive data for 343 523 vascular plant species and time-to-event analyses, we conducted multiple tests related to plant taxonomic and geographic data shortfalls, and identified 33 global diversity darkspots (those 'botanical countries' predicted to contain most undescribed and not yet recorded species).

Reliable biogeography requires fossils: insights from a new species-level phylogeny of extinct and living carnivores.

A central objective of historical biogeography is to understand where clades originated and how they moved across space and over time. However, given the dynamic history of ecosystem changes in response to climate change and geological events, the manifold long-distance dispersals over evolutionary timescales, and regional and global extinctions, it remains uncertain how reliable inferences based solely on extant taxa can be achieved. Using a novel species-level phylogeny of all known extant and extinct species of the mammalian order Carnivora and related extinct groups, we show that far more precise and accurate ancestral areas can be estimated by fully integrating extinct species into the analyses, rather than solely relying on extant species or identifying ancestral areas only based on the geography of the oldest fossils.

Trait-mediated speciation and human-driven extinctions in proboscideans revealed by unsupervised Bayesian neural networks.

Species life-history traits, paleoenvironment, and biotic interactions likely influence speciation and extinction rates, affecting species richness over time. Birth-death models inferring the impact of these factors typically assume monotonic relationships between single predictors and rates, limiting our ability to assess more complex effects and their relative importance and interaction. We introduce a Bayesian birth-death model using unsupervised neural networks to explore multifactorial and nonlinear effects on speciation and extinction rates using fossil data.

Exploring the Macroevolutionary Signature of Asymmetric Inheritance at Speciation.

Popular comparative phylogenetic models such as Brownian Motion, Ornstein-Ulhenbeck, and their extensions, assume that, at speciation, a trait value is inherited identically by two descendant species. This assumption contrasts with models of speciation at a micro-evolutionary scale where descendants' phenotypic distributions are sub-samples of the ancestral distribution. Different speciation mechanisms can lead to a displacement of the ancestral phenotypic mean among descendants and an asymmetric inheritance of the ancestral phenotypic variance.

Toward a Semi-Supervised Learning Approach to Phylogenetic Estimation.

Models have always been central to inferring molecular evolution and to reconstructing phylogenetic trees. Their use typically involves the development of a mechanistic framework reflecting our understanding of the underlying biological processes, such as nucleotide substitutions, and the estimation of model parameters by maximum likelihood or Bayesian inference. However, deriving and optimizing the likelihood of the data is not always possible under complex evolutionary scenarios or even tractable for large datasets, often leading to unrealistic simplifying assumptions in the fitted models.