The fig and pollinator wasp obligate mutualism is diverse (~750 described species), ecologically important, and ancient (~80-90 Ma), providing model systems for generating and testing many questions in evolution and ecology. Once thought to be a prime example of strict one-to-one cospeciation, current thinking suggests that genera of pollinator wasps coevolve with corresponding subsections of figs, but the degree to which cospeciation or other processes contributes to the association at finer scales is unclear. Here we use genome-wide sequence data from a community of Panamanian strangler figs (Ficus subgenus Urostigma, section Americana) and associated fig wasp pollinators (Pegoscapus spp.) to infer the process of coevolutionary diversification in this obligate mutualism. Using a model-based approach adapted from the study of gene family evolution, our results indicate pervasive and ongoing host switching of pollinator wasps at this fine phylogenetic and regional scale. Although the model estimates a modest amount of cospeciation, simulations reveal this signal to be consistent with levels of co-association expected under a model of free host switching. Our findings provide an outline for testing how ecological and evolutionary processes can be modeled to evaluate the history of association of interacting lineages in a phylogenetic framework.

The total-evidence approach to divergence time dating uses molecular and morphological data from extant and fossil species to infer phylogenetic relationships, species divergence times, and macroevolutionary parameters in a single coherent framework. Current model-based implementations of this approach lack an appropriate model for the tree describing the diversification and fossilization process and can produce estimates that lead to erroneous conclusions. We address this shortcoming by providing a total-evidence method implemented in a Bayesian framework. This approach uses a mechanistic tree prior to describe the underlying diversification process that generated the tree of extant and fossil taxa. Previous attempts to apply the total-evidence approach have used tree priors that do not account for the possibility that fossil samples may be direct ancestors of other samples, that is, ancestors of fossil or extant species or of clades. The fossilized birth–death (FBD) process explicitly models the diversification, fossilization, and sampling processes and naturally allows for sampled ancestors. This model was recently applied to estimate divergence times based on molecular data and fossil occurrence dates. We incorporate the FBD model and a model of morphological trait evolution into a Bayesian total-evidence approach to dating species phylogenies. We apply this method to extant and fossil penguins and show that the modern penguins radiated much more recently than has been previously estimated, with the basal divergence in the crown clade occurring at ∼12.7 Ma and most splits leading to extant species occurring in the last 2 myr. Our results demonstrate that including stem-fossil diversity can greatly improve the estimates of the divergence times of crown taxa. The method is available in BEAST2 (version 2.4) software www.beast2.org with packages SA (version at least 1.1.4) and morph-models (version at least 1.0.4) installed.

Paederinae is one of the most diverse subfamilies among rove beetles, yet their evolutionary history remains poorly understood. This is attributed to the limited number of phylogenetic studies, which either sought answers at a shallower taxonomic level or included limited taxon sampling. Especially problematic is the position of the rare Neotropical tribe Cylindroxystini, morphologically one of the most puzzling groups of Paederinae. The phylogenetic position of this group within Paederinae was never understood, though its rank in the classification has already been shifted twice. We assembled molecular and morphological data matrices sampled from all currently recognized Paederinae subtribes, including both genera of Cylindroxystini, and used these data to estimate phylogenetic relationships using Bayesian inference. A total of 123 morphological characters and 4,631 bp of nuclear (28S, TP, Wg, CADA, CADC, ArgK) and mitochondrial (COI) sequences were analyzed for 76 taxa. The current tribe Cylindroxystini was resolved as a monophylum within the tribe Lathrobiini as sister to the genus Pseudolathra, and together they are sister to the so-called ‘Medonina and allied taxa’ clade. Based on these results, we downgraded Cylindroxystini back to the subtribal level, Cylindroxystina status reinstated, now with a known sister group. The resulting phylogeny is the largest of the subfamily Paederinae to date and lays the foundation for establishing a natural classification of the group.

A birth-death-sampling model gives rise to phylogenetic trees with samples from the past and the present. Interpreting “birth” as branching speciation, “death” as extinction, and “sampling” as fossil preservation and recovery, this model – also referred to as the fossilized birth-death (FBD) model – gives rise to phylogenetic trees on extant and fossil samples. The model has been mathematically analyzed and successfully applied to a range of datasets on different taxonomic levels, such as penguins, plants, and insects. However, the current mathematical treatment of this model does not allow for a group of temporally distinct fossil specimens to be assigned to the same species.

In this paper, we provide a general mathematical FBD modeling framework that explicitly takes “stratigraphic ranges” into account, with a stratigraphic range being defined as the lineage interval associated with a single species, ranging through time from the first to the last fossil appearance of the species. To assign a sequence of fossil samples in the phylogenetic tree to the same species, i.e., to specify a stratigraphic range, we need to define the mode of speciation. We provide expressions to account for three common speciation modes: budding (or asymmetric) speciation, bifurcating (or symmetric) speciation, and anagenetic speciation.

Our equations allow for flexible joint Bayesian analysis of paleontological and neontological data. Furthermore, our framework is directly applicable to epidemiology, where a stratigraphic range is the observed duration of infection of a single patient, “birth” via budding is transmission, “death” is recovery, and “sampling” is sequencing the pathogen of a patient. Thus, we present a model that allows for incorporation of multiple observations through time from a single patient.

Estimating speciation and extinction rates is essential for understanding past and present biodiversity, but is challenging given the incompleteness of the rock and fossil records. Interest in this topic has led to a divergent suite of independent methods—paleontological estimates based on sampled stratigraphic ranges and phylogenetic estimates based on the observed branching times in a given phylogeny of living species. The fossilized birth–death (FBD) process is a model that explicitly recognizes that the branching events in a phylogenetic tree and sampled fossils were generated by the same underlying diversification process. A crucial advantage of this model is that it incorporates the possibility that some species may never be sampled. Here, we present an FBD model that estimates tree-wide diversification rates from stratigraphic range data when the underlying phylogeny of the fossil taxa may be unknown. The model can be applied when only occurrence data for taxonomically identified fossils are available, but still accounts for the incomplete phylogenetic structure of the data. We tested this new model using simulations and focused on how inferences are impacted by incomplete fossil recovery. We compared our approach with a phylogenetic model that does not incorporate incomplete species sampling and to three fossil-based alternatives for estimating diversification rates, including the widely implemented boundary-crosser and three-timer methods. The results of our simulations demonstrate that estimates under the FBD model are robust and more accurate than the alternative methods, particularly when fossil data are sparse, as the FBD model incorporates incomplete species sampling explicitly.

Phylogenetics and phylodynamics are central topics in modern evolutionary biology. Phylogenetic methods reconstruct the evolutionary relationships among organisms, whereas phylodynamic approaches reveal the underlying diversification processes that lead to the observed relationships. These two fields have many practical applications in disciplines as diverse as epidemiology, developmental biology, palaeontology, ecology, and linguistics. The combination of increasingly large genetic data sets and increases in computing power is facilitating the development of more sophisticated phylogenetic and phylodynamic methods. Big data sets allow us to answer complex questions. However, since the required analyses are highly specific to the particular data set and question, a black-box method is not sufficient anymore. Instead, biologists are required to be actively involved with modeling decisions during data analysis. The modular design of the Bayesian phylogenetic software package BEAST 2 enables, and in fact enforces, this involvement. At the same time, the modular design enables computational biology groups to develop new methods at a rapid rate. A thorough understanding of the models and algorithms used by inference software is a critical prerequisite for successful hypothesis formulation and assessment. In particular, there is a need for more readily available resources aimed at helping interested scientists equip themselves with the skills to confidently use cutting-edge phylogenetic analysis software. These resources will also benefit researchers who do not have access to similar courses or training at their home institutions. Here, we introduce the “Taming the Beast” (https://taming-the-beast.github.io/) resource, which was developed as part of a workshop series bearing the same name, to facilitate the usage of the Bayesian phylogenetic software package BEAST 2.