Gregory W. Stull

A Targeted Enrichment Strategy for Massively Parallel Sequencing of Angiosperm Plastid Genomes

Premise of the study: We explored a targeted enrichment strategy to facilitate rapid and low-cost next-generation sequencing (NGS) of numerous complete plastid genomes from across the phylogenetic breadth of angiosperms. Methods and Results: A custom RNA probe set including the complete sequences of 22 previously sequenced eudicot plastomes was designed to facilitate hybridization-based targeted enrichment of eudicot plastid genomes. Using this probe set and an Agilent SureSelect targeted enrichment kit, we conducted an enrichment experiment including 24 angiosperms (22 eudicots, two monocots), which were subsequently sequenced on a single lane of the Illumina GAIIx with single-end, 100-bp reads. This approach yielded nearly complete to complete plastid genomes with exceptionally high coverage (mean coverage: 717×), even for the two monocots. Conclusions: Our enrichment experiment was highly successful even though many aspects of the capture process employed were suboptimal. Hence, significant improvements to this methodology are feasible. With this general approach and probe set, it should be possible to sequence more than 300 essentially complete plastid genomes in a single Illumina GAIIx lane (achieving ∼50× mean coverage). However, given the complications of pooling numerous samples for multiplex sequencing and the limited number of barcodes (e.g., 96) available in commercial kits, we recommend 96 samples as a current practical maximum for multiplex plastome sequencing. This high-throughput approach should facilitate large-scale plastid genome sequencing at any level of phylogenetic diversity in angiosperms.

Resolving basal lamiid phylogeny and the circumscription of Icacinaceae with a plastome-scale data set

PREMISE OF THE STUDY Major relationships within Lamiidae, an asterid clade with ∼40000 species, have largely eluded resolution despite two decades of intensive study. The phylogenetic positions of Icacinaceae and other early-diverging lamiid clades (Garryales, Metteniusaceae, and Oncothecaceae) have been particularly problematic, hindering classification and impeding our understanding of early lamiid (and euasterid) character evolution. METHODS To resolve basal lamiid phylogeny, we sequenced 50 plastid genomes using the Illumina sequencing platform and combined these with available asterid plastome sequence data for more comprehensive phylogenetic analyses. KEY RESULTS Our analyses resolved basal lamiid relationships with strong support, including the circumscription and phylogenetic position of the enigmatic Icacinaceae. This greatly improved basal lamiid phylogeny offers insight into character evolution and facilitates an updated classification for this clade, which we present here, including phylogenetic definitions for 10 new or converted clade names. We also offer recommendations for applying this classification to the Angiosperm Phylogeny Group (APG) system, including the recognition of a reduced Icacinaceae, an expanded Metteniusaceae, and two orders new to APG: Icacinales (Icacinaceae + Oncothecaceae) and Metteniusales (Metteniusaceae). CONCLUSIONS The lamiids possibly radiated from an ancestry of tropical trees with inconspicuous flowers and large, drupaceous fruits, given that these morphological characters are distributed across a grade of lineages (Icacinaceae, Oncothecaceae, Metteniusaceae) subtending the core lamiid clade (Boraginales, Gentianales, Lamiales, Solanales, Vahlia). Furthermore, the presence of similar morphological features among members of Aquifoliales suggests these characters might be ancestral for the Gentianidae (euasterids) as a whole.

Fruits of an “Old World” tribe (Phytocreneae; Icacinaceae) from the Paleogene of North and South America

Abstract The Phytocreneae (Icacinaceae) are a tribe of scrambling shrubs and lianas presently distributed in tropical Africa, Madagascar, and Indo-Malesia. We describe the oldest known fossils of this tribe and provide the first recognition of this group in the Neotropical fossil record based on distinctive fruit remains. Palaeophytocrene piggae sp. nov., from the late Paleocene of western North America, and Palaeophytocrene hammenii sp. nov. and cf. Phytocrene sp., from the middle-late Paleocene of Colombia, constitute the oldest confirmed records of this tribe. Pyrenacantha austroamericana sp. nov., from the Oligocene of Peru, represents an extant Old World genus known also from the Eocene fossil record of North America and Europe. Collectively, these fossils indicate that the Phytocreneae were previously established in the Neotropics, despite their current absence from the region, and may provide evidence for Paleogene floristic exchange between North and South America.

Fruits of Icacinaceae from the Eocene of Southeastern North America and Their Biogeographic Implications

We describe three members of the pantropical family Icacinaceae from the Eocene Claiborne Group of Tennessee and Mississippi, providing the first recognition of this family from the Mississippi Embayment of southeastern North America. Natsiatum wilcoxiana (Berry) comb. nov. and Phytocrene densipunctata sp. nov. represent extant genera presently confined to Southeast Asia and Malesia. Croomiocarpon mississippiensis gen. et sp. nov. represents an extinct genus with affinities to the tribe Iodeae. These fossils increase the generic diversity of Icacinaceae’s fossil record and expand our understanding of the historical distribution of this family. Given the similarity of N. wilcoxiana and P. densipunctata with several fossils known from the Eocene of Europe, these records suggest a broader pattern of floristic connection between southeastern North America and Europe during the Eocene.

Plastid and Seed Morphology Data Support a Revised Infrageneric Classification and an African Origin of the Pantropical Genus Xylopia (Annonaceae)

Abstract The floristic treatment of Engler and Diels, published in 1901, provides the only infrageneric classification of the pantropical genus Xylopia (Annonaceae). Here we test and extend that classification using molecular and seed morphology characters. Phylogenetic relationships were reconstructed using data from four plastid regions obtained from 44 of the approximately 165 species in the genus, recovering four well-supported major clades. Seed characters were examined for these taxa, and six aril morphologies, three previously undocumented, were distinguished; we also document the presence of a sarcotesta on the seeds of many species. Molecular and seed data support recognition of five sections within the genus; one, Xylopia sect. Rugosperma, is proposed here as new. Our phylogenetic results suggest an African origin for the genus and reveal complex biogeographic patterns, likely facilitated by long-distance dispersal.

X-ray micro-computed tomography (Micro-CT) of pyrite-permineralized fruits and seeds from the London clay formation (Ypresian) conserved in silicone oil: A critical evaluation

Pyrite-permineralized fruits and seeds from the London Clay Formation (Ypresian; England) in the NHMUK are stored in silicone oil to retard decay processes. X-ray micro-computed tomography (micro-CT) has revealed internal morphology for multiple holotypes (including severely cracked and encrusted specimens) scanned in the protective fluid. Silicone oil alone has a similar X-ray attenuation to parts of the specimens, causing minor uncertainty for digitally rendered surfaces, but key systematic characters are readily visualized. Further work is needed to optimize visualization of fine-scale cellular detail. Labelling and segmentation to visualize important structures is achievable with these micro-CT datasets. However, manual labelling of individual slices is required, and defining boundaries between features can be difficult due to differential pyritization and silicone oil permeation. Digital sections through specimens can be made in any orientation and digital locule casts can be produced for studies in ...

Character evolution and missing (morphological) data across Asteridae

PREMISE OF THE STUDY Our current understanding of flowering plant phylogeny provides an excellent framework for exploring various aspects of character evolution through comparative analyses. However, attempts to synthesize this phylogenetic framework with extensive morphological data sets have been surprisingly rare. Here, we explore character evolution in Asteridae (asterids), a major angiosperm clade, using an extensive morphological data set and a well-resolved phylogeny. METHODS We scored 15 phenotypic characters (spanning chemistry, vegetative anatomy, and floral, fruit, and seed features) across 248 species for ancestral state reconstruction using a phylogenetic framework based on 73 plastid genes and the same 248 species. KEY RESULTS Iridoid production, unitegmic ovules, and cellular endosperm were all reconstructed as synapomorphic for Asteridae. Sympetaly, long associated with asterids, shows complex patterns of evolution, suggesting it arose several times independently within the clade. Stamens equal in number to the petals is likely a synapomorphy for Gentianidae, a major asterid subclade. Members of Lamianae, a major gentianid subclade, are potentially diagnosed by adnate stamens, unilacunar nodes, and simple perforation plates. CONCLUSIONS The analyses presented here provide a greatly improved understanding of character evolution across Asteridae, highlighting multiple characters potentially synapomorphic for major clades. However, several important parts of the asterid tree are poorly known for several key phenotypic features (e.g., degree of petal fusion, integument number, nucellus type, endosperm type, iridoid production). Further morphological, anatomical, developmental, and chemical investigations of these poorly known asterids are critical for a more detailed understanding of early asterid evolution.

Missing the point (estimate): Bayesian and likelihood phylogenetic reconstructions of morphological characters produce generally concordant inferences. A comment on Puttick et al.

Puttick et al . [1] performed a simulation study to compare accuracy among methods of inferring phylogeny from discrete morphological characters. They report that a Bayesian implementation of the Mk model [2] was most accurate (but with low resolution), while a maximum likelihood (ML) implementation of the same model was least accurate. They conclude by strongly advocating that Bayesian implementations of the Mk model should be the default method of analysis for such data. While we appreciate the authors9 attempt to investigate the accuracy of alternative methods of analysis, their conclusion is based on an inappropriate comparison of the ML point estimate, which does not consider confidence, with the Bayesian consensus, which incorporates estimation credibility into the summary tree. Using simulation, we demonstrate that ML and Bayesian estimates are concordant when confidence and credibility are comparably reflected in summary trees, a result expected from statistical theory. We therefore disagree with the conclusions of PEA and consider their prescription of any default method to be poorly founded. Instead, we recommend caution and thoughtful consideration of the model or method being applied to a morphological dataset.Puttick et al. (2017) performed a simulation study to compare accuracy between methods inferring phylogeny from discrete morphological characters. They report that a Bayesian implementation of the Mk model (Lewis, 2001) was most accurate (but with low resolution), while a maximum likelihood (ML) implementation of the same model was least accurate. They conclude by strongly advocating that Bayesian implementations of the Mk model should be the default method of analysis for such data. While we applaud investigations into accuracy and alternative methods of analysis, this conclusion is based on an inappropriate comparison of the ML point estimate with the Bayesian consensus. We revisit these issues through simulation by considering uncertainty in ML reconstructions, and demonstrate that Bayesian and ML estimates are generally concordant when conventional edge support thresholds are considered. We therefore disagree with the conclusions of Puttick et al. (2017), and consider their prescription of any default method to be unfounded. Instead, we recommend caution and thoughtful consideration of the model or method being applied to a morphological dataset.Puttick et al . (hereafter, PEA) [1] performed a simulation study to compare accuracy among methods of inferring phylogeny from discrete morphological characters. They report that a Bayesian implementation of the Mk model [2] was most accurate (but with low resolution), while a maximum likelihood (ML) implementation of the same model was least accurate. They conclude by strongly advocating that Bayesian implementations of the Mk model should be the default method of analysis for such data. While we appreciate the authors9 attempt to investigate the accuracy of alternative methods of analysis, their conclusion is based on an inappropriate comparison of the ML point estimate, which does not consider confidence, with the Bayesian consensus, which incorporates estimation credibility into the summary tree. Using simulation, we demonstrate that ML and Bayesian estimates are concordant when confidence and credibility are comparably reflected in summary trees, a result expected from statistical theory. We therefore disagree with the conclusions of PEA and consider their prescription of any default method to be poorly founded. Instead, we recommend caution and thoughtful consideration of the model or method being applied to a morphological dataset.