Resolving the deep phylogeny: Implications for early adaptive radiation, cryptic, and present-day ecological diversity of Papuan microhylid frogs.

The microhylid frogs of the New Guinea region are the largest and most ecologically diverse subfamily (Asterophryinae) of one of the largest anuran families in the world and can live in communities of up to 20 species. While there has been recent progress in resolving the phylogenetic relationships of Asterophryinae, significant uncertainties remain, impeding further progress in understanding the evolution of microhabitat use, parental care, and life history variation in this group. In particular, the early divergences at the base of the tree remain unclear; as does the monophyly of some genera; and recent studies have discovered that species with wide geographic distribution are instead cryptic species complexes. In this study, we fortified geographic sampling of the largest previous phylogenetic effort by sequencing an additional 62 taxa and increased data quality and quantity by adding new layers of data vetting and by filling in previously incomplete loci to the five gene dataset (2 mitochondrial, 3 nuclear protein-coding genes) to obtain a dataset that is now 99% complete in over 2400 characters for 233 samples (205 taxa) of Asterophryinae and 3 outgroup taxa, and analyzed microhabitat use data for these taxa from field data and data collected from the literature. Importantly, our sampling includes complete community complements at 19 sites as well as representatives at over 80 sites across New Guinea and its offshore islands. We present a highly resolved molecular phylogeny which, for the first time, has over 95% of nodes supported (84% highly supported) whether using Maximum Likelihood or Bayesian Inference, allowing clarification of all genera (whether monophyletic or clearly not), their sister genera relationships, as well as an age estimate for the Asterophryinae at approximately 20MYA. Early generic diversification occurring between 17 and 12 MYA gave rise to a surprising diversity of about 18 genera as well as the 5 putative microhabitat types. Our tree reveals extensive cryptic diversity calling any widespread taxa into doubt, and clearly demonstrates that complex multispecies communities of Asterophryinae are ecologically diverse, are numerous, and of ancient origin across New Guinea. We discuss the implications of our phylogeny for explaining the explosive diversification of Asterophryinae as the result of adaptive radiation, niche conservatism, and non-adaptive radiation.

Testing Geology with Biology: Plate Tectonics and the Diversification of Microhylid Frogs in the Papuan Region.

Studies of the Papuan region have provided fundamental insights into the evolutionary processes generating its exceptional biodiversity, but the influence of geological processes merits further study. Lying at the junction of five tectonic plates, this region has experienced a turbulent geological history that has not only produced towering mountains allowing elevational specialization and island archipelagos with varying degrees of isolation promoting vicariance, but also active margins where land masses have collided and been subsequently rifted apart creating a mosaic of intermixed terranes with vastly different geological histories. Asterophryine frogs are a hyperdiverse clade representing half the world's microhylid diversity (over 360 species) centered on New Guinea and its satellite islands. We show that vicariance facilitated by geological history explains this far and wide distribution of a clade that should have poor dispersal abilities. We recovered a mainland tectonic unit, the East Papua Composite Terrane (EPCT), as the center of origin for Asterophryinae and no fewer than 71 instances of what appear to be long-distance dispersal events, 29 of which are between mainland regions, with 42 from the mainland to the islands, some presently as far as 200 km away from source populations over open ocean. Furthermore, we find strong support for a "Slow and Steady" hypothesis for the formation of the northern margin of New Guinea by many separate accretion events during the Miocene, over other major geological alternatives, consistent with the 20 M year age of the clade and arrival via the EPCT. In addition, the historical biogeography of our frogs strongly supports an affiliation of the Louisiade Archipelago and Woodlark Island with the Owen Stanley Range on the EPCT, and the recent proximity of the large New Britain Island. Our results show that Asterophryinae did not have to repeatedly and independently disperse across large ocean barriers to the offshore islands, against the predictions of island biogeography theory, but that the current distribution can be explained through vicariance and short-distance oceanic dispersal as historical land connections disappeared and islands slowly became separated from each other. We show that islands have a life history, changing in distance from other land masses, with consequent opportunities for dispersal, isolation, and cladogenesis of their biotas. More broadly, we can begin to see how the geological history of the Papuan region can result in the rapid accumulation and staggering number of extant species.

Molecular and phylogenetic datasets for the Asterophryinae frogs of New Guinea with additional data on lifestyle, geography, and elevation.

The data provided here are related to the article "Resolving the Deep Phylogeny: Implications for Early Adaptive Radiation, Cryptic, and Present-day Ecological Diversity of Papuan Microhylid Frogs" [1]. The dataset is based on 233 tissue samples of the subfamily Asteroprhyinae, with representatives from all recognized genera, in addition to three outgroup taxa. The sequence dataset contains over 2400 characters per sample for five genes: three nuclear (Seventh in Absentia (SIA), Brain Derived Neurotrophic Factor (BDNF), Sodium Calcium Exchange subunit-1 (NXC-1)), and two mitochondrial loci (Cytochrome oxidase b (CYTB), and NADH dehydrogenase subunit 4 (ND4)); and is 99% complete. New primers were designed for all loci and accession numbers for the raw sequence data are provided. The sequences are used with geological time calibrations to produce time-calibrated Bayesian inference (BI) and Maximum Likelihood (ML) phylogenetic reconstructions using BEAST2 and IQ-TREE. Lifestyle data (arboreal, scansorial, terrestrial, fossorial, semi-aquatic) were collected from the literature and field notes and used to infer ancestral character states for each lineage. Collection location and elevation data were used to verify sites where multiple species or candidate species co-occur. All sequence data, alignments, and associated metadata (voucher specimen number, species identification, type locality status, global positioning system [GPS] coordinates, elevation, site with species list, and lifestyle) as well as the code to produce all analyses and figures are provided.