Skip to main content
Dryad logo

Origin and adaptive radiation of the exceptional and threatened bembidiine beetle fauna of St Helena (Coleoptera: Carabidae)

Citation

Maddison, David; Sproul, John; Mendel, Howard (2019), Origin and adaptive radiation of the exceptional and threatened bembidiine beetle fauna of St Helena (Coleoptera: Carabidae), Dryad, Dataset, https://doi.org/10.5061/dryad.n5tb2rbrh

Abstract

The central peaks of the isolated island of St Helena (in the south Atlantic Ocean) are home to an extraordinary set of ground beetles of the tribe Bembidiini, and belong to three endemic genus-group taxa. These beetles are strikingly different in overall body form from the many bembidiines found elsewhere in the world. At least some of the St Helena species are likely extinct, and all are threatened by habitat destruction and invasive species. Through next-generation sequencing of old museum specimens, we examine the phylogenetic relationships of the St Helena fauna. We find that, in spite of their morphological disparities, the endemic St Helena bembidiines form a clade of genetically very similar species, their sister group is Bembidion alsium from the island of La Réunion (east of Madagascar), and the sister group of this pair is the African subgenus Omotaphus.  We propose that the St Helena Peaks Bembidion are an adaptive radiation that arose from a single dispersal event to St Helena from a now-extinct African lineage (sister to Omotaphus), and that this extinct lineage also served as the ancestral source of Bembidion alsium.  As the St Helena Peaks Bembidion are deeply nested within the genus Bembidion, we move the three taxa back within that genus as subgenera, and provide a new name (Bembidion shepherdae) for the now-homonymous Bembidion wollastoni

Methods

Genome skimming using an Illumina sequencer from DNA extracted from pinned specimens that were 46-130 years old.  Sequences obtained by de novo and reference-based assembly.  Matrix of seven genes analyzed using maximum likelihood estimation using IQ-TREE.