An integrated pathway for building regional phylogenies for ecological studies. Issue 12 (26th August 2019)

• Record Type: Journal Article
• Authors: Eme, David; Anderson, Marti J.; Struthers, Carl D.; Roberts, Clive D.; Liggins, Libby
• Editors: Davies, Jonathan
• Is Part Of: Global ecology & biogeography. Volume 28:Issue 12(2019)
• Page Start: 1899
• Page End: 1911
• Publication Date: 2019-08-26
• Subjects: Bayesian phylogenetics -- community phylogenetics -- DNA sequences -- geo‐referencing -- marine fishes -- metadata -- New Zealand -- regional species pool -- super‐matrix -- topological constraints; Ecology -- Periodicals; Biogeography -- Periodicals; Biodiversity -- Periodicals; Macroevolution -- Periodicals; 577
• DOI: 10.1111/geb.12986
• Languages: English
• ISSNs: 1466-822X

Abstract

Abstract: Aim: Phylogenies are increasingly used in community ecology, biogeography and macroecology. However, sourcing a phylogeny comprising the entire species pool for a focal region can be difficult. Typically, a bespoke phylogeny must be created requiring considerable data manipulation and the use of many standalone software packages. Here we present a suite of methodological tools within the popular R environment that help to build molecular phylogenies appropriate for ecological studies with a regional focus. Innovation: Our R package regPhylo provides a pipeline to construct a Bayesian posterior distribution of time‐calibrated trees suitable to address ecological questions. The novel contributions of regPhylo include options to: use prior phylogenetic knowledge through flexible topological constraints; include spatial metadata in sourcing DNA sequences; and include taxa without DNA sequences and then infer consequent phylogenetic uncertainty. Specifically, regPhylo helps researchers: retrieve DNA sequences; enhance available metadata; select DNA sequences based on their length or spatial proximity to the region of study; align sequences; and perform quality control. Output from the pipeline is a file ready to run in the Bayesian tree reconstruction software beast2, appropriate for estimating time‐calibrated trees and including phylogenetic uncertainty for downstream analyses. Main conclusions: Overall, regPhylo improves the integration of popular standalone phylogenetic software into the flexible R environment. It provides a novel approach to include topological constraints based on prior knowledge, include taxa without DNA sequences, and select spatially appropriate DNA sequences. When coupled with a Bayesian tree‐building process, our approach provides estimates of uncertainty in both topology and branch lengths. We demonstrate the utility of the package by constructing a posterior distribution of time‐calibrated phylogenies for the New Zealand marine ray‐finned fishes (Actinopterygii) providing the unprecedented opportunity to include phylogenetic information in downstream ecological analyses for marine fishes in this region.