Phylogenetic Relationships of the Bumblebee Subgenus Pyrobombus (Hymenoptera: Apidae) Inferred from Mitochondrial Cytochrome B and Cytochrome Oxidase I Sequences
Total Page:16
File Type:pdf, Size:1020Kb
Research Collection Journal Article Phylogenetic Relationships of the Bumblebee Subgenus Pyrobombus (Hymenoptera: Apidae) Inferred from Mitochondrial Cytochrome B and Cytochrome Oxidase I Sequences Author(s): Koulianos, Stella Publication Date: 1999 Permanent Link: https://doi.org/10.3929/ethz-b-000422490 Originally published in: Annals of the Entomological Society of America 92(3), http://doi.org/10.1093/aesa/92.3.355 Rights / License: In Copyright - Non-Commercial Use Permitted This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use. ETH Library SYSTEMATICS Phylogenetic Relationships of the Bumblebee Subgenus Pyrobombus (Hymenoptera: Apidae) Inferred from Mitochondrial Cytochrome B and Cytochrome Oxidase I Sequences STELLA KOULIANOS Experimental Ecology, ETH Zu¨ rich, ETH Zentrum NW, 8092 Zu¨ rich, Switzerland Ann. Entomol. Soc. Am. 92(3): 355Ð358 (1999) ABSTRACT Traditionally, the genus Bombus (Apidae: Apinae: Bombini) is divided into several subgenera. This study derives an approximation of the relationships between 10 species of the subgenus Pyrobombus (from both Europe and North America), B. lapidarius L. and B. sichelii Radoszkowski from the subgenus Melanobombus, and B. terrestris L. from the subgenus Bombus s. str., by comparing the mitochondrial cytochrome b and cytochrome oxidase I (COI) genes. Although bootstrap values for deep branches are mostly low, the sequences show signiÞcant phylogenetic signal, low homoplasy, and all trees share some patterns that are consistent with those from other phylogenetic studies on bumblebees. These results show that the subgeneric names Pyrobombus and Melanobombus do not accurately reßect phylogeny, and therefore it would seem wise to revise the existing system into monophyletic species-groups or even ignore the subgeneric names altogether. KEY WORDS Pyrobombus, bumblebees, cytochrome b, cytochrome oxidase I, mitochondrial DNA, phylogeny BUMBLEBEES OF THE subgenus Pyrobombus (Apidae: The current study provides a comparison of the Apinae: Bombini: Bombus) are Holarctic. Pyrobombus mitochondrial cytochrome b and cytochrome oxidase is the largest subgenus of Bombus, comprising of .40 I (COI) genes from 10 Pyrobombus species (from both species (Williams 1994), and accounts for almost half Europe and North America), B. lapidarius and B. sich- of the North American fauna. Like all bumblebees, elii Radoszkowski from the subgenus Melanobombus, species of this subgenus are morphologically very sim- and B. terrestris L. from the subgenus Bombus s. str., to ilar (Williams 1985, 1994), and this has resulted in derive an approximation of the relationships among much confusion about species deÞnitions because these species. Tetragona dorsalis Friese (Apidae: Api- they are also highly polymorphic with respect to coat nae: Meliponini), identiÞed as the sister group to the color (Scholl et al. 1995). Pyrobombus, however, has Bombini by several independent studies (Cameron been found to be heterogeneous with respect to en- 1991, 1993; Sheppard and McPheron 1991; Koulianos zymes (Pekkarinen et al. 1979, Pamilo et al. 1987, et al. 1999), and Apis mellifera L. (Apidae: Apinae: Scholl et al. 1995) and male marking pheromones Apini) (Crozier and Crozier 1993) were used as out- (Svensson 1979). groups. The close relationship of Pyrobombus to the subge- nus Melanobombus has been established by both mor- Materials and Methods phological (Plowright and Stephen 1973) and molec- Total DNA was extracted from part of the thorax or ular (Pamilo et al. 1987, Pedersen 1996) studies. from 2 legs of single individuals of 10 Pyrobombus Pyrobombus has been found to be paraphyletic by an species (Table 1), B. lapidarius, B. sichelii, and the extensive morphological study using wing venation by outgroup T. dorsalis using a modiÞcation of the pro- Plowright and Stephen (1973) and a molecular study, tocol of Edwards and Hoy (1993), as in Koulianos et which only included 2 Pyrobombus species, by Ped- al. (1999). ersen (1996). These 2 studies show a close association A 716-bp fragment of the cytochrome b gene was between B. lapidarius L. (Melanobombus) and the ampliÞed from these bees (excluding T. dorsalis, for Pyrobombus species B. bimaculatus Cresson and B. which the sequence was already available, see Table pratorum L. There is, however, evidence to the con- 1) using 2 oligonucleotide primers [CATTATT(G/ trary (i.e., no association between B. pratorum and B. A)TCC(T/A)AATATTGA(A/T)ATTGC and ATTA lapidarius) from the morphology of the penis valve CACCTCCTAATTTATTAGGAAT] (59 ends corre- (Williams 1985). Plowright and StephensÕs (1973) sponding to positions 11168 and 11884, respectively, in study also separated 6 species from the main Pyrobom- Crozier and Crozier [1993]), and Taq polymerase bus cluster and grouped these with species belonging (Promega, Wallisellen, Switzerland) in a Perkin Elmer to Bombus s. str. (Rotkreuz, Switzerland) Cetus thermal cycler for 40 0013-8746/99/0355Ð0358$02.00/0 q 1999 Entomological Society of America 356 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 92, no. 3 Table 1. List of taxa used Table 2. Pairwise distances calculated according to Tajima and Nei (1984) for the combined cytochrome b and COI sequences Pyrobombus Bombus lapponicus (F.) B. jonellus (Kirby) 2345678910 B. hypnorum (L.) B. pratorum (L.) 1 0.14 0.12 0.11 0.12 0.11 0.14 0.12 0.12 0.13 B. monticola Smith (COI only) 2 Ñ 0.11 0.12 0.11 0.10 0.11 0.11 0.12 0.13 B. frigidus (Smith) (cyt b only) 3 Ñ 0.07 0.08 0.08 0.10 0.09 0.09 0.12 B. bifarius (Cresson) 4 Ñ 0.07 0.07 0.09 0.07 0.09 0.11 B. ternarius (Say) 5 Ñ 0.07 0.11 0.08 0.09 0.12 B. flavifrons (Cresson) 6 Ñ 0.09 0.06 0.08 0.10 B. huntii (Greene) (COI only) 7 Ñ 0.09 0.10 0.13 Melanobombus B. lapidarius (L.) 8 Ñ 0.09 0.12 B. sichelii Radoszkowski 9 Ñ 0.12 Bombus s. str. B. terrestris (L.) Outgroups Tetragona dorsalis ziegleri (Friese) 1, B. terrestris; 2, B. lapidarius; 3, B. sichelii; 4, B. jonellus; 5, B. Apis mellifera (L.) hypnorum; 6, B. lapponicus; 7, B. pratorum; 8, B. ternarius; 9, B. flavi- frons; 10, B. bifarius. Apis mellifera sequences are from Crozier and Crozier (1993) (GenBank accession #L06178); B. hypnorum, B. terrestris, and B. lapidarius COI sequences are from Pedersen (1996) (GenBank ac- combined sequences were compared with estimate cession #L26569-70, 73); and B. terrestris and T. dorsalis cytochrome the phylogenetic relationships among the species in- b sequences are from Koulianos et al. (in press) (GenBank accession cluded in this study and to assess the phylogenetic #AF002721, 25). informativeness of these sequences. All new se- quences have been deposited in GenBank and are cycles under the following conditions: denaturation at accessible by the numbers AF066969Ð72, 84Ð85, 89, 928C for 1 min, annealing at 408C for 1 min, and ex- 91, AF077918Ð19, and AF084908Ð17. tension at 728C for 1 min. A 521-bp fragment of the Pairwise distances from the combined sequences COI gene was also ampliÞed from these bees (exclud- ranged from 0.06 to 0.14 (Table 2). The distance be- ing those already sequenced by Pedersen [1996]; see tween B. sichelii and the Pyrobombus species B. jonel- Table 1 again) using primers AP-L-2176 and AP-H- lus Kirby, B. hypnorum L., and B. lapponicus F. was as 2650, and conditions in Pedersen (1996). All products low or lower than comparisons within the Pyrobombus were puriÞed using the Wizard polymerase chain re- subgenus. In general, the largest distances were ob- action (PCR) Preps DNA PuriÞcation System (Pro- tained by comparing B. terrestris, B. lapidarius, and B. mega). bifarius Cresson to each other and any other species. One individual per species was sequenced in both The distance between B. lapponicus and B. monticola directions from the puriÞed double-stranded PCR Smith calculated from the COI sequences alone was product, by Microsynth GmbH (Bulgach, Switzer- 0.006, .10-fold smaller than all other comparisons. land). This suggests that B. monticola and B. lapponicus are Sequence information was aligned by eye. Diver- not 2 separate species, contrary to evidence from mor- gence between bumblebee species (d) was estimated phology, behavior, and male marking pheromones by the method of Tajima and Nei (1984). Parsimony (Svensson 1979). analyses were carried out separately on the cyto- Parsimony analysis of the cytochrome b gene chrome b and COI sequences, as well as on the com- yielded 133 informative characters, 2 most-parsimo- bined sequences using PAUP 4.0 d64 (Swofford, per- nious trees, a skewness value of g1 521.26 (P , 0.01), sonal communication) and a branch and bound and a homoplasy index of 0.08. Analysis of the COI search. Support for the nodes was estimated by 1,000 gene yielded 51 informative characters, 12 most-par- bootstrap replications using Seqboot, Dnapars, and simonious trees, a skewness value of g1 520.44 (P , Consense in PHYLIP 3.57c (Felsenstein 1995). Max- 0.01), and a homoplasy index of 0.06. Analysis of the imum likelihood analyses were performed using PAUP combined sequences yielded 175 informative charac- 4.0 d64 and corrected with an estimated shape param- ters, 1 most-parsimonious tree, a skewness value of g1 eter of the gamma distribution. 521.32 (P , 0.01), and a homoplasy index of 0.05. For each of the data partitions, PAUP 4.0 d64 was The skewness values show that the cytochrome b, used to evaluate the strength of the phylogenetic sig- COI, and combined sequences all contain highly sig- nal by performing skewness tests (Hillis and Huelsen- niÞcant nonrandom structure which is likely to reßect beck 1992) using 10,000 randomly selected trees, and phylogenetic signal. The topology based on the anal- to assess the impact of character homoplasy by re- ysis of the COI region showed the lowest average weighting characters based on their rescaled consis- bootstrap values and the lowest homplasy index, tency indices and reanalyzing them with parsimony.