Phylogeny and DNA barcoding of inquiline oak gallwasps (Hymenoptera: Cynipidae) of the Western Palaearctic

Abstract

We examine phylogenetic relationships within the Synergus complex of herbivorous inquiline gallwasps (Hymenoptera; Cynipidae; Synergini) associated with cynipid host galls on oak, a biologically diverse group whose genus-level morphological taxonomy has long been considered stable but whose species level taxonomy is problematic. We incorporate data for over 70% of recognised Western Palaearctic species in five morphology-based genera (Ceroptres, Saphonecrus, Synergus, Synophrus, Ufo), comprising sequence for two mitochondrial loci (coxI, cytb) and one nuclear locus (28S D2). In particular, we assess the evidence for monophyly of two long-established, morphology-defined sections within the genus Synergus that differ in a range of biological traits. To aid analyses of ecological interactions within oak cynipid communities, we also consider the utility of cytochrome oxidase I (coxI) DNA barcodes in the oak inquilines. In this assessment, we do not assume that species are delineated at a single threshold value of sequence divergence for a single gene, but examine concordance in the composition of molecular operational Taxonomic units (MOTUs) across a range of sequence divergences in each gene and across genes. We also assess the impact of sampling effort on MOTU stability. Phylogenetic reconstructions for all three loci support monophyly for Synergus and Synophrus, but reject monophyly for Saphonecrus and for the two sections within Synergus. The suites of traits associated with the two sections of the genus Synergus are thus homoplasious. All three loci also reject monophyly for three Synergus species (S. hayneanus, S. pallipes, S. umbraculus). Sequences for each locus identify robust MOTUs that are largely concordant across loci for a range of cut-off values. Though many MOTU's correspond to recognised Linnean species, there is significant, multigene disagreement between groupings supported by morphology and sequence data, with both allocation of different morphospecies to the same MOTU and allocation of the same morphospecies to multiple MOTUs, regardless of cut-off value. Our results imply that while DNA barcoding has considerable utility within this group, morphology-based identification needs major revision at both genus and species levels. Further, lifehistory traits currently attributed to single morphospecies probably confound attributes of multiple lineages. Revealing patterns of character state evolution in Synergus requires collection of new host association and life history data explicitly linked to DNA barcode data for the specimens concerned.