Phylogenetic analysis of the corbiculate bee tribes based on 12 nuclear protein-coding genes (Hymenoptera: Apoidea: Apidae) Atsushi Kawakita, John S. Ascher, Teiji Sota, Makoto Kato, David W. Roubik To cite this version: Atsushi Kawakita, John S. Ascher, Teiji Sota, Makoto Kato, David W. Roubik. Phylogenetic anal- ysis of the corbiculate bee tribes based on 12 nuclear protein-coding genes (Hymenoptera: Apoidea: Apidae). Apidologie, Springer Verlag, 2008, 39 (1), pp.163-175. hal-00891935 HAL Id: hal-00891935 https://hal.archives-ouvertes.fr/hal-00891935 Submitted on 1 Jan 2008 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Apidologie 39 (2008) 163–175 Available online at: c INRA/DIB-AGIB/ EDP Sciences, 2008 www.apidologie.org DOI: 10.1051/apido:2007046 Original article Phylogenetic analysis of the corbiculate bee tribes based on 12 nuclear protein-coding genes (Hymenoptera: Apoidea: Apidae)* Atsushi Kawakita1, John S. Ascher2, Teiji Sota3,MakotoKato 1, David W. Roubik4 1 Graduate School of Human and Environmental Studies, Kyoto University, Kyoto, Japan 2 Division of Invertebrate Zoology, American Museum of Natural History, New York, USA 3 Department of Zoology, Graduate School of Science, Kyoto University, Kyoto, Japan 4 Smithsonian Tropical Research Institute, Balboa, Ancon, Panama Received 2 July 2007 – Revised 3 October 2007 – Accepted 3 October 2007 Abstract – The corbiculate bees comprise four tribes, the advanced eusocial Apini and Meliponini, the primitively eusocial Bombini, and the solitary or communal Euglossini. Recovering a robust phylogeny for the four tribes is of considerable importance for understanding the evolution of eusociality, yet previous morphological and molecular studies reached strikingly different conclusions. We study an expanded data set consisting of 12 nuclear genes to explore lines of support for the molecular hypothesis. Results corrob- orate previous molecular studies; support increases as more genes are added. Across genes, support for the molecular hypothesis is positively correlated with the number of informative sites and the relative substitu- tion rate. Phylogenetic signals supporting the molecular tree rest almost entirely upon synonymous changes at the first and third codon positions. We discuss possible future approaches for resolving the frustratingly persistent corbiculate bee controversy. advanced eusociality / bumble bee / corbiculate bee / honey bee / nuclear genes / orchid bee / phylogeny / stingless bee 1. INTRODUCTION and maximize their foraging efficiency by us- ing sophisticated communication and rapid The advanced eusocial colonies of the recruitment of nestmates to food sources. A honey bees and stingless bees are among the myriad of fascinating biological features of most elaborate of the insect societies (Wilson, the bees and their economic importance as 1971; Michener, 1974; Winston, 1987). In pollinators and honey producers have sparked these colonies, the morphologically differ- numerous ecological, behavioral, neural, ge- entiated queens are devoted to producing netic, and genomic studies, and led the western offspring, and the non-reproductive workers honey bee, Apis mellifera L., to become one gather and process food, feed larvae, and build of the first insect species to have its genome and defend nests. Workers specialize through sequenced (The Honeybee Genome Sequenc- division of labor by age, size, and genetics, ing Consortium, 2006). Despite the widely ac- cepted paradigms of social evolution repre- Corresponding author: A. Kawakita, [email protected] sented by such bees (e.g., division of labor, * Manuscript editor: Eduardo A.B. Almeida sterile workers, overlap of generations), and Online material is available at: several recent phylogenetic studies of tribal http://www.apidologie.org relationships, the origin of their advanced Article published by EDP Sciences and available at http://www.apidologie.org or http://dx.doi.org/10.1051/apido:2007046 164 A. Kawakita et al. societies remain unexplained (Schultz et al., 1993; Koulianos et al., 1999; Mardulyn and 1999, 2001; Ascher et al., 2001; Cameron and Cameron, 1999; Cameron and Mardulyn, Mardulyn, 2001; Engel, 2001a; Lockhart and 2001). Most notably, Bombini and Meliponini Cameron, 2001). are nearly always recovered as sisters, a The first study that rigorously explored the clade not supported by known morphological phylogenetic relationships among the corbic- data (Figs. 1d–f). Sheppard and McPheron ulate tribes was by Michener (1944), who (1991) and Cameron (1991, 1993) were the used adult external morphology to investi- first to propose this alternative phylogeny, in gate the overall phylogeny of bees. He con- which sequences of nuclear and mitochondrial cluded that Euglossini were the first lineage ribosomal RNA were used to reconstruct the to diverge, followed by Bombini, and then corbiculate bee tribal relationships. Under this the advanced eusocial Meliponini and Apini hypothesis, advanced eusociality arose twice, (Fig. 1a), a view consistent with that of ear- independently, in Meliponini and Apini; lier authors such as Darwin (1859). Under reversals from advanced eusociality to other this hypothesis, general eusociality arose once stages, while equally parsimonious, were among corbiculate bees in the common ances- considered very unlikely to occur because the tor of Bombini, Meliponini, and Apini. Ad- queens of advanced eusocial colonies have vanced eusociality also arose only once among lost the ability to survive on their own without bees, in the common ancestor of Meliponini workers. Later studies utilized several other and Apini. Later studies explored a broad ar- genes, from both mitochondrial and nuclear ray of adult and larval morphological char- genomes (Koulianos et al., 1999; Mardulyn acters (Maa, 1953; Michener, 1974, 1990; and Cameron, 1999; Cameron and Mardulyn, Winston and Michener, 1977; Kimsey, 1984; 2001; Cameron, 2003). These studies con- Prentice, 1991; Roig-Alsina and Michener, sistently recovered Bombini and Meliponini 1993; Chavarría and Carpenter, 1994), and as a monophyletic group excluding Apini. some of them reached more or less differ- The most comprehensive study was that of ent conclusions (e.g., Fig. 1b, Michener, 1974; Cameron and Mardulyn (2001), who analyzed Fig. 1c, Winston and Michener, 1977). How- a data matrix of four genes (cytochrome b, ever, all recent analyses of extant corbiculate 16S rRNA, opsin, and 28S rRNA). Their morphology (Michener, 1990; Prentice, 1991; results gave Bombini and Meliponini strong Roig-Alsina and Michener, 1993; Chavarría support as sister groups, although the rela- and Carpenter, 1994; Schultz et al., 1999; tionship of this clade to the remaining two Ascher et al., 2001) are consistent with the hy- tribes remained unresolved. Cha et al. (2007) pothesis proposed by Michener (1944). More recently analyzed the complete mitochon- recently, Engel (2001a, b) significantly ex- drial genome of Bombini, Meliponini, and panded the morphological data set by includ- Apini, also finding that the first two tribes ing fossil corbiculate tribes; the combined data are unambiguously recovered as sisters (the set lent further support to the Michener (1944) mitochondrial genome of Euglossini is still phylogeny. Cardinal and Packer (2007) studied unavailable). a new set of morphological characters derived Resolving the controversy over tribal re- from the sting apparatus. Their results were lationships in corbiculate bees has proven perfectly congruent with the previous morpho- to be remarkably difficult (Schultz et al., logical studies. Noll (2002) analyzed a large 1999, 2001; Ascher et al., 2001; Lockhart and series of behavioral characters and arrived at Cameron, 2001), and additional data from new the same phylogenetic interpretation. sources are needed to broaden our perspective Despite the near unanimity of morpho- on the issue and to allow for more sophisti- logical, palaeontological, and behavioral cated analyses of molecular evolution. Recent data, phylogenetic analyses using DNA publication of the honey bee genome had fa- sequences suggest strikingly different rela- cilitated assembly of orthologous sequences tionships among the four tribes (Sheppard of previously unexplored nuclear genes in and McPheron, 1991; Cameron, 1991, many bees (Danforth et al., 2004). Because Phylogeny of the corbiculate bees 165 abc Euglossini Bombini Meliponini Apini Euglossini Bombini Meliponini Apini Meliponini Apini Euglossini Bombini def Figure 1. Previously proposed rooted phy- logenies of the four corbiculate tribes (a– Apini Euglossini Bombini Meliponini Euglossini Apini Bombini Meliponini Euglossini Apini Bombini Meliponini f) and the three theoretically possible un- rooted topologies for the four tribes (g–i). a. Phylogeny favored by the majority of mor- phological data sets (Michener, 1944, 1990; Maa, 1953; Prentice, 1991; Roig-Alsina and g h Michener, 1993; Chavarría and Carpenter, 1994; Schultz et al., 1999; Ascher et al., Euglossini Meliponini Euglossini Bombini