Multiple ancient origins of neoteny in Lycidae(Coleoptera): consequences for ecology and macroevolution

Abstract

Neoteny, the maintenance of larval features in sexually mature adults, is a radical way of generating evolutionary novelty through shifts in relative timing of developmental programmes. While controlled by the environment in facultative neotenics, retention of larval features is obligatory in many species of Lycidae (net-winged beetles). They are studied here as an example of how developmental shifts and ecology interact to produce macroevolutionary impacts. We conducted a phylogenetic analysis of Lycidae based on DNA sequences from nuclear (18S and 28S rRNA) and mitochondrial (rrnL, cox1, cob and nad5) genes from a representative set of lineages (73 species), including 17 neotenic taxa. Major changes of basal relationships compared with those implied in the current classification generally supported three independent origins of neotenics in Lycidae. The southeast Asian Lyropaeinae and Ateliinae were in basal positions indicating evolutionary antiquity, also confirmed by molecular clock estimates, unlike the neotropical leptolycines nested within Calopterini and presumably much younger. neotenics exhibit typical K-selected traits including slow development, large body size, high investment in offspring and low dispersal. This correlated with low species richness and restricted ranges of neotenic lineages compared with their sisters. Yet, these factors did not impede the evolutionary persistence of affected lineages, even without reversals to fully metamorphosed forms, contradicting earlier suggestions of recent evolution from dispersive non-neotenics.

Figure 1

Female larva of Duliticola sp. from Mt Sinabung (Indonesia, Sumatra).

Figure 2

Phylogenetic hypothesis of 73 lycid taxa and six out-groups. One of five most parsimonious trees produced under default settings from the ClustalX alignment, analysed with indels considered as missing characters. Numbers above branches refer to bootstrap proportions (if more than 50%) and Bremer support values are given below selected branches. The tree topology is identical with a single tree produced by parsimony analysis of the same dataset when gaps were coded as fifth character. Vertical bars represent the biogeographic distribution of lineages, with species numbers given at the far right. The species numbers are estimations including not yet described species if these are present in collections. Significantly imbalanced nodes: ^*p values (Slowinski & Guyer 1993), corrected ^**p values (Davies et al. 2004).