MOLECULAR PHYLOGENY OF THE SHINING LEAF BEETLES (COLEOPTERA: CHRYSOMELIDAE: CRIOCERINAE)

Abstract

Shining leaf beetles (Coleoptera, Chrysomelidae, Criocerinae; ~1500 spp) are considered amongst the earliest diverging leaf beetle lineage to attack early angiosperms. Although they are distributed worldwide, little is known about their biology and evolutionary relationships. Schmitt (1988) generated the first morphology-based phylogeny using four genera: ((Lilioceris + Crioceris) + (Lema + Oulema)). Teo (1999) is the second phylogenetic hypothesis (unpublished), this morphology-based phylogeny shows ((Pseudocriocerii + Criocerini (Lemini). Vencl et al (2004) proposed a phylogenetic hypothesis for the Central American genera ((Crioceris + Metopoceris) + (Lema + (Neolema + Oulema)). These three studies sampled a subset of the recognized genera and lacked large outgroup representation. Nevertheless, they provide a general understanding about phylogenetic relationships in Criocerinae. In this research I tested: 1) the systematic position of Criocerinae, 2) the monophyly of the subfamily, and 3) the intrageneric relationships by generating a molecular dataset and developing a new phylogentic hypothesis of evolutionary relationships. I sampled 76 species in 7 genera of Criocerinae and 9 outgroups from other chrysomelid subfamilies, to generate a molecular data set of three molecular markers (COI, 18S, and 28S). Phylogenetic analyses using parsimony, maximum likelihood, and posterior probabilities show strong support ( 0.90 posterior probabilities/ 1 - 0.75 bootstrap) for placing Criocerinae within the Sagrinae clade of Chrysomelidae, as either sister group to Donaciinae or Sagrinae. The monophyly of Criocerinae has been supported by several morphological characters--stridulatory apparatus and frontal grooves in adults, and ambulatory warts, dorsal anus, and fecal shield in larvae). Yet, this phylogenetic analyses showed no support for the monophyly of this subfamily. My phylogenetic analyses do not clarify the pattern of evolution in Criocerinae because systematic relationships within Criocerinae, at tribal or generic levels, were not recovered from our tree topologies (individual genes and combined data analyses). Our most resolved phylogeny was recovered using posterior probabilities and these results were consistent with Teo's (1999) strict consensus topology. Both phylogenies are not fully resolved and show that Lema Fabricius and Lilioceris Reitter are not monophyletic. Additionally, parametric bootstrapping was performed to test the monophyly of each genus and tribe. The only significant improvement was constraining Neolema Monrós as monophyletic (better ML scores, but not MP score). Using the most resolved topology, I examined the geographic pattern of species distributions. I found that species clusters are more related to their geographical distribution (i.e., the existence of Nearctic-Neotropical and Oriental species clusters). Similarly, I examined host plant record patterns with this tree topology, and observed that most of my sampled criocerinae species are monocot feeders. However, some species in certain genera (e.g., Lema, Lilioceris, Neolema, Metopoceris Heinze) are also feeding on eudicot plants. Future research in Criocerinae needs to focus on developing stronger diagnostic characters for the subfamily since traditional characters supporting it monophyly also occur in other chrysomelid subfamilies (e.g., Sagrinae & Hispinae). Future research should also sample members of Pseudocriocerini and Criocerini, which have been thought to be basal in the evolution of the subfamily. These will ultimately contribute to resolving the evolutionary patterns in Criocerinae.