Deren A. R. Eaton

Inferring Phylogeny and Introgression using RADseq Data: An Example from Flowering Plants (Pedicularis: Orobanchaceae)

Phylogenetic relationships among recently diverged species are often difficult to resolve due to insufficient phylogenetic signal in available markers and/or conflict among gene trees. Here we explore the use of reduced-representation genome sequencing, specifically in the form of restriction-site associated DNA (RAD), for phylogenetic inference and the detection of ancestral hybridization in non-model organisms. As a case study, we investigate Pedicularis section Cyathophora, a systematically recalcitrant clade of flowering plants in the broomrape family (Orobanchaceae). Two methods of phylogenetic inference, maximum likelihood and Bayesian concordance, were applied to data sets that included as many as 40,000 RAD loci. Both methods yielded similar topologies that included two major clades: a “rex-thamnophila” clade, composed of two species and several subspecies with relatively low floral diversity, and geographically widespread distributions at lower elevations, and a “superba” clade, composed of three species characterized by relatively high floral diversity and isolated geographic distributions at higher elevations. Levels of molecular divergence between subspecies in the rex-thamnophila clade are similar to those between species in the superba clade. Using Patterson’s D-statistic test, including a novel extension of the method that enables finer-grained resolution of introgression among multiple candidate taxa by removing the effect of their shared ancestry, we detect significant introgression among nearly all taxa in the rex-thamnophila clade, but not between clades or among taxa within the superba clade. These results suggest an important role for geographic isolation in the emergence of species barriers, by facilitating local adaptation and differentiation in the absence of homogenizing gene flow. [Concordance factors; genotyping-by-sequencing; hybridization; partitioned D-statistic test; Pedicularis; restriction-site associated DNA.]

A Framework Phylogeny of the American Oak Clade Based on Sequenced RAD Data

Previous phylogenetic studies in oaks (Quercus, Fagaceae) have failed to resolve the backbone topology of the genus with strong support. Here, we utilize next-generation sequencing of restriction-site associated DNA (RAD-Seq) to resolve a framework phylogeny of a predominantly American clade of oaks whose crown age is estimated at 23–33 million years old. Using a recently developed analytical pipeline for RAD-Seq phylogenetics, we created a concatenated matrix of 1.40 E06 aligned nucleotides, constituting 27,727 sequence clusters. RAD-Seq data were readily combined across runs, with no difference in phylogenetic placement between technical replicates, which overlapped by only 43–64% in locus coverage. 17% (4,715) of the loci we analyzed could be mapped with high confidence to one or more expressed sequence tags in NCBI Genbank. A concatenated matrix of the loci that BLAST to at least one EST sequence provides approximately half as many variable or parsimony-informative characters as equal-sized datasets from the non-EST loci. The EST-associated matrix is more complete (fewer missing loci) and has slightly lower homoplasy than non-EST subsampled matrices of the same size, but there is no difference in phylogenetic support or relative attribution of base substitutions to internal versus terminal branches of the phylogeny. We introduce a partitioned RAD visualization method (implemented in the R package RADami; http://cran.r-project.org/web/packages/RADami) to investigate the possibility that suboptimal topologies supported by large numbers of loci—due, for example, to reticulate evolution or lineage sorting—are masked by the globally optimal tree. We find no evidence for strongly-supported alternative topologies in our study, suggesting that the phylogeny we recover is a robust estimate of large-scale phylogenetic patterns in the American oak clade. Our study is one of the first to demonstrate the utility of RAD-Seq data for inferring phylogeny in a 23–33 million year-old clade.

Floral diversity and community structure in Pedicularis (Orobanchaceae)

A pervasive hypothesis at the interface of ecology and evolution is that biotic interactions contribute to regional biodiversity by accelerating adaptation and speciation. We investigated this question in the context of closely related, bumble bee-pollinated plants (Pedicularis spp.) in the Hengduan Mountains of south-central China, where they exhibit spectacular levels of richness, endemism, and floral diversity. Because these species co-occur frequently, flower synchronously, and share pollinators during the brief reproductive season, we predict that pollinator-mediated interactions may influence their community assembly and evolutionary diversification. If disparity in floral traits reduces competitive interactions between species, as would happen if floral isolation mitigates reproductive interference caused by heterospecific pollen flow, then species with dissimilar flowers should co-occur more often, yielding greater floral diversity at local scales than expected by chance. Moreover, if such interactions have repeatedly driven character displacement, then floral traits should exhibit homoplasy, the phylogenetic signature of labile evolution. We present evidence supporting these predictions, and find that local species richness is best explained by a model including both floral diversity and phylogenetic distance. Our results suggest that a dynamic mosaic of pollinator-mediated interactions among Pedicularis in the Hengduan region promotes ecological sorting through recurrent selection against reproductive interference, causing rapid species turnover at local scales, and accelerating the rate of floral divergence among species. Together these processes may have contributed to the remarkable accumulation of florally diverse species of Pedicularis endemic to the Hengduan Mountains biodiversity hotspot.

Phylogeny and biogeography of the American live oaks (Quercus subsection Virentes): a genomic and population genetics approach

The nature and timing of evolution of niche differentiation among closely related species remains an important question in ecology and evolution. The American live oak clade, Virentes, which spans the unglaciated temperate and tropical regions of North America and Mesoamerica, provides an instructive system in which to examine speciation and niche evolution. We generated a fossil‐calibrated phylogeny of Virentes using RADseq data to estimate divergence times and used nuclear microsatellites, chloroplast sequences and an intron region of nitrate reductase (NIA‐i3) to examine genetic diversity within species, rates of gene flow among species and ancestral population size of disjunct sister species. Transitions in functional and morphological traits associated with ecological and climatic niche axes were examined across the phylogeny. We found the Virentes to be monophyletic with three subclades, including a southwest clade, a southeastern US clade and a Central American/Cuban clade. Despite high leaf morphological variation within species and transpecific chloroplast haplotypes, RADseq and nuclear SSR data showed genetic coherence of species. We estimated a crown date for Virentes of 11 Ma and implicated the formation of the Sea of Cortés in a speciation event ~5 Ma. Tree height at maturity, associated with fire tolerance, differs among the sympatric species, while freezing tolerance appears to have diverged repeatedly across the tropical–temperate divide. Sympatric species thus show evidence of ecological niche differentiation but share climatic niches, while allopatric and parapatric species conserve ecological niches, but diverge in climatic niches. The mode of speciation and/or degree of co‐occurrence may thus influence which niche axis plants diverge along.

Historical introgression among the American live oaks and the comparative nature of tests for introgression

Introgressive hybridization challenges the concepts we use to define species and infer phylogenetic relationships. Methods for inferring historical introgression from the genomes of extant species, such as ABBA‐BABA tests, are widely used, however, their results can be easily misinterpreted. Because these tests are inherently comparative, they are sensitive to the effects of missing data (unsampled species) and nonindependence (hierarchical relationships among species). We demonstrate this using genomic RADseq data sampled from all extant species in the American live oaks (Quercus series Virentes), a group notorious for hybridization. By considering all species and their phylogenetic relationships, we were able to distinguish true hybridizing lineages from those that falsely appear admixed. Six of seven species show evidence of admixture, often with multiple other species, but which is explained by introgression among a few related lineages occurring in close proximity. We identify the Cuban oak as the most admixed lineage and test alternative scenarios for its origin. The live oaks form a continuous ring‐like distribution around the Gulf of Mexico, connected in Cuba, across which they could effectively exchange alleles. However, introgression appears highly localized, suggesting that oak species boundaries and their geographic ranges have remained relatively stable over evolutionary time.

Genotyping-by-sequencing as a tool to infer phylogeny and ancestral hybridization: A case study in Carex (Cyperaceae)

Determining phylogenetic relationships among very closely related species has remained a challenge for evolutionary biologists due to interlocus phylogenetic discordance and the difficulty of obtaining variable markers. Here, we used a Genotyping-by-Sequencing (GBS) approach to sample a reduced representation genomic data set and infer the phylogeny of seven closely related species in the genus Carex (Cyperaceae). Past attempts to reconstruct phylogenetic relationships among these species produced conflicting and poorly-supported results. We inferred a robust phylogeny based on >3000 GBS loci and >1300 SNPs (with a minimum sequence depth within individuals of 10) using maximum likelihood and Bayesian inference. We also tested for historical introgression using the D-statistic test. We compared these analyses with partitioned RAD analysis, which is designed to identify suboptimal trees reflecting secondary phylogenetic signal that may be obscured by the dominant signal in the data. Phylogenetic analyses yielded fully resolved trees with high support. We found two main clades, one grouping Carex scoparia populations and C. waponahkikensis, and a second clade grouping C. longii, C. vexans, C. suberecta and C. albolutescens. We detected marginally significant signals of introgression between C. scoparia and C. suberecta or C. albolutescens, and we rejected a hybrid origin hypothesis for C. waponahkikensis. Our results demonstrate the power of NGS data sets for resolving some of the most difficult phylogenetic challenges where traditional phylogenetic markers have failed.

Misconceptions on Missing Data in RAD-seq Phylogenetics with a Deep-scale Example from Flowering Plants

Abstract Restriction‐site associated DNA (RAD) sequencing and related methods rely on the conservation of enzyme recognition sites to isolate homologous DNA fragments for sequencing, with the consequence that mutations disrupting these sites lead to missing information. There is thus a clear expectation for how missing data should be distributed, with fewer loci recovered between more distantly related samples. This observation has led to a related expectation: that RAD‐seq data are insufficiently informative for resolving deeper scale phylogenetic relationships. Here we investigate the relationship between missing information among samples at the tips of a tree and information at edges within it. We re‐analyze and review the distribution of missing data across ten RAD‐seq data sets and carry out simulations to determine expected patterns of missing information. We also present new empirical results for the angiosperm clade Viburnum (Adoxaceae, with a crown age >50 Ma) for which we examine phylogenetic information at different depths in the tree and with varied sequencing effort. The total number of loci, the proportion that are shared, and phylogenetic informativeness varied dramatically across the examined RAD‐seq data sets. Insufficient or uneven sequencing coverage accounted for similar proportions of missing data as dropout from mutation‐disruption. Simulations reveal that mutation‐disruption, which results in phylogenetically distributed missing data, can be distinguished from the more stochastic patterns of missing data caused by low sequencing coverage. In Viburnum, doubling sequencing coverage nearly doubled the number of parsimony informative sites, and increased by >10X the number of loci with data shared across >40 taxa. Our analysis leads to a set of practical recommendations for maximizing phylogenetic information in RAD‐seq studies.

Introgression obscures and reveals historical relationships among the American live oaks

Introgressive hybridization challenges the concepts we use to define species and infer phylogenetic relationships. Methods for inferring historical introgression from the genomes of extant species are now widely used, however, few guidelines have been articulated for how best to interpret results. Because these tests are inherently comparative, they are sensitive to the effects of missing data (unsampled species) and non-independence (hierarchical relationships among species). We demonstrate this using genomic RADseq data sampled from all extant species in the American live oaks (Quercus series Virentes), a group notorious for hybridization. By considering all species, and their phylogenetic relationships, we were able to distinguish true hybridizing lineages from those that falsely appear admixed. Six of seven species show evidence of admixture, often with multiple other species, but which is explained by hybrid introgression among few related lineages occurring in close proximity. We identify the Cuban oak as the most admixed lineage and test alternative scenarios for its origin. The live oaks form a continuous ring-like distribution around the Gulf of Mexico, connected in Cuba, across which they could effectively exchange alleles. However, introgression appears highly localized, suggesting that oak species boundaries, and their geographic ranges have remained relatively stable over evolutionary time.

Reconciling species diversity in a tropical plant clade (Canarium, Burseraceae)

The challenges associated with sampling rare species or populations can limit our ability to make accurate and informed estimates of biodiversity for clades or ecosystems. This may be particularly true for tropical trees, which tend to be poorly sampled, and are thought to harbor extensive cryptic diversity. Here, we integrate genomics, morphology, and geography to estimate the number of species in a clade of dioecious tropical trees (Canarium L.; Burseraceae) endemic to Madagascar, for which previous taxonomic treatments have recognized between one and 33 species. By sampling genomic data from even a limited number of individuals per taxon, we were able to clearly reject both previous hypotheses, and support instead an intermediate number of taxa. We recognize at least six distinct clades based on genetic structure and species delimitation analyses that correspond clearly with geographic and discrete morphological differences. Two widespread clades co-occur broadly throughout eastern wet forests, one clade is endemic to western dry forests, and several slightly admixed clades are more narrowly distributed in mountainous regions in the north. Multiple previously described taxa were recovered as paraphyletic in our analyses, some of which were associated with admixed individuals, suggesting that hybridization contributes to taxonomic difficulties in Canarium. An improved understanding of Canarium species diversity has important implications for conservation efforts and understanding the origins of diversity in Madagascar. Our study shows that even limited genomic sampling, when combined with geography and morphology, can greatly improve estimates of species diversity for difficult tropical clades.

Concatenated alignment of RADseq data: virentes_ME_c85d6m4p3

Concatenated alignments of RADseq data for seven Virentes taxa and eight outgroup taxa (35 samples total) used to infer phylogenetic trees. The virentes_ME_c85d6m4p3.phy alignment is 6,252,421 bp in length. Filtered reads were clustered at an 85% threshold for within-sample clustering. The number of samples for which data needed to be recovered to include a RAD locus in the data set was 4.