DNA-based taxonomy for associating adults and larvae in multi-species assemblages of chafers (Coleoptera: Scarabaeidae)

Abstract

DNA sequences provide a universal character system in taxonomy for associating all developmental stages of organisms, but ambiguity arises due to genetic variation within species. The problem is compounded where target groups are less well studied or incompletely represented in DNA databases. Here we investigate the utility of DNA for larval–adult species associations within chafer (Coleoptera: Scarabaeidae) communities from four sites in the tropical lowlands of Nepal. We sequenced ca. 1600 bp of mitochondrial cox1 and rrnL and 700 bp of nuclear 28S rRNA from 250 larval and adult specimens. Individuals were grouped into putative species using statistical parsimony analysis and population aggregation analysis (PAA), whereby specimens from each locality were grouped according to the presence of diagnostic nucleotides. In addition, species membership was determined based on shifts in branching rates on clock-constrained trees to detect the putative transition from speciation to population coalescence patterns. Using these two methods we delineated between 48 and 56 groups, of which 16–20 were composed of larval and adult individuals. Nuclear and mtDNA-based groups were highly congruent; variation of 28S rRNA within groups was very low, while one widespread 28S rRNA genotype was universally found in a paraphyletic group of five mtDNA clusters. Linnean names could be assigned to 19 groups, and hence between 86.1% and 92.7% of larval specimens could be associated to species by their membership in clearly delineated groups that contained fully identified adults. The remaining larvae were delineated as five species, four of which could be assigned to Anomala or Adoretus based on their phylogenetic position. We conclude that the sequence variation was highly structured in this complex assemblage of chafers and that any given individual (larva or adult) can be readily associated with a particular DNA group using the criterion of diagnosability. The association of different developmental stages therefore becomes a matter of determining the extent of the DNA-based groups, rather than matching of sequences from adult and larval individuals. This indicates the need for a purely sequence-based taxonomic system when associating different life stages via DNA.