Insect. Soc. 52 (2005) 45 –54 0020-1812/05/010045-10 DOI 10.1007/s00040-004-0761-1 © Birkhäuser Verlag, Basel, 2005

Insectes Sociaux

Research article

Male labial gland secretions and mitochondrial DNA markers support species status of Bombus cryptarum and B. magnus (Hymenoptera, Apidae) A. Bertsch 1, *, H. Schweer 2, A. Titze 1 and H. Tanaka 3 1

2 3

Department of Biology, Philipps-University Marburg, Karl-von-Frisch-Strasse 8, 35032 Marburg, Germany, e-mail: [email protected], [email protected] Department of Pediatrics, Philipps-University Marburg, Deutschhausstr. 12, 35033 Marburg, Germany, e-mail: [email protected] Primate Research Institute, Kyoto University, Inuyama, Aichi 484-8506, Japan, e-mail: [email protected]

Received 25 March 2004; revised 13 June 2004; accepted 15 June 2004.

Summary. Spring queens of Bombus cryptarum and B. magnus from 2 localities in Brandenburg/Germany and Scotland/ United Kingdom respectively were determined by morphological characteristics. The lateral border of the collare at the border of the pronotallobus or at the episternum proved to be an especially useful character. Artificial colonies were reared from safely determined spring queens and the cephalic part of the labial glands of males from these colonies were investigated by GC/MS. The investigation identified approximately 50 compounds, as a mixture of straight chain fatty acid derivatives (alcohols, esters and hydrocarbons). The labial secretions of B. cryptarum and B. magnus are significantly different. Mitochondrial cytochrome oxidase 1 (CO1) of two queens from each locality and species were sequenced. Each species from the different localities formed a cluster. Sequence divergence between B. cryptarum and B. magnus was about 30 base substitutions and approximately 0.04 in Tamura-Nei genetic distance. Bombus cryptarum and B. magnus were closer to each other than to B. lucorum and made the sister group in the topology of the tree. Both the CO1 sequences and the labial gland secretions of males of B. cryptarum from Brandenburg and of males from artificial colonies reared from safely determined spring queens from Scotland are identical. B. cryptarum has thus, for the first time, been identified as part of the British bumble bee fauna. The differences of both the labial gland secretions, used as species recognition signals, and the genetic differences established by sequencing CO1 confirm the morphological findings that B. cryptarum and B. magnus are distinct taxa which should be treated as distinct species.

* Author for correspondence.

Key words: Bumblebees, Bombus cryptarum, Bombus magnus, labial gland secretion, mtDNA, cytochrome oxidase 1.

Introduction Though the species of the Subgenus Bombus s. str. (syn. Terrestribombus) are abundant everywhere in Central Europe, the question of how many species are involved is still unresolved. In spite of their size and conspicuous coloration, identification of species is often difficult because most species share a similar general appearance in colour and morphology. Small differences in morphology have often been used as diagnostic characteristics to distinguish the species. The subgenus Bombus s. str. is a group in which classification of species is especially complicated, partly due to considerable intraspecific variation. In Europe, five species in the subgenus Bombus s. str., B. (Bombus) terrestris (Linnaeus), B. (B.) lucorum (Linnaeus), B. (B.) magnus Vogt, B. (B.) cryptarum (Fabricius) and B. (B.) sporadicus Nylander, are known. Their taxonomical status has been extensively examined based on morphology (Krüger, 1939, 1951, 1954, 1956, 1958; Løken, 1973; Pekkarinen, 1979; Rasmont, 1984; Rasmont et al., 1986), enzyme electrophoretic data (Pamilo et al., 1984; Scholl and Obrecht, 1983; Scholl et al., 1992) and analyses of the compounds of the male labial glands (Pamilo et al., 1997; Bertsch, 1997; Urbanová et al., 2001, 2002). The species status of B. sporadicus, B. terrestris and B. lucorum is generally accepted, however, the taxonomic status of B. magnus and B. cryptarum is still in dispute. Whereas Rasmont (1984) treated B. magnus and B. cryptarum as separate species, Williams (1991, 1998) lumped them together with B. lucorum. Having investigated enzyme electrophoretic data from a broad spectrum of sites

46

A. Bertsch et al.

Species discrimination by labial gland secretion and mtDNA

Scholl et al. (1992) could show that B. magnus and B. cryptarum are genetically distant from B. lucorum. This view could be confirmed for B. cryptarum by investigation of the male labial glands (Bertsch, 1997). As the species recognition signals of B. cryptarum are significantly different from those of B. lucorum, both taxa cannot belong to the same species. Recent publications (Pamilo et al., 1997; Williams, 2000; Pedersen, 2002) could not resolve these problems and did not clarify the status of B. cryptarum and B. magnus. In the present study, we use morphological characteristics of the collare of queens, the compounds of the male cephalic labial glands and mitochondrial DNA cytochrome oxidase 1 sequences to elucidate the taxonomic status of B. cryptarum and B. magnus. For this purpose, the specimens examined were taken from localities where these taxa occur sympatrically with B. lucorum.

Materials and methods Figure 1. REM of anterior part of thorax (partly shaved). Not = Notum, Pro = Pronotum, Lob = Pronotallobus, + anterior border of tegula.

Bumblebee samples Queens of B. lucorum, B. cryptarum and B. magnus were collected in Brandenburg, Germany, about 5 km northwest of Menz, Lkr.Oberhavel (13° 02¢ 59¢¢ O, 53° 06¢ 08¢¢ N) during spring. All three species were abundant in a pine forest with sandy soil and a closed layer of mosses (Leucobryo-Pinetum sylvestris Matuszkiewicz 1962). In the beginning of May, luxuriant vegetation of Vaccinium myrtillus is a good food source for spring queens, which show strong flight activity in the early morning. Specimens were also taken from Teupitz, Lkr. Dahme-Spreewald (13° 36¢ 33¢¢ O, 52° 08¢ 08¢¢ N) feeding on planted Cotoneaster. In Scotland B. magnus is abundant (Alford 1975, maps ITE 1980). Specimens were taken from the North coast of Scotland in the Sand Dunes of Dunnet, Caithness (3° 20¢ 30¢¢ W, 58° 36¢ 30¢¢ N) feeding on Anthyllis vulneraria and near the light house at Duncansby Head, Caithness (3° 01¢ 40¢¢ W, 58° 38¢ 40¢¢ N) feeding on Cirsium palustre. At both localities individuals determined morphologically as B. cryptarum were also taken for investigation. The queens collected in the field were used in morphological and molecular genetic examinations. For the molecular analysis, we examined two individuals from each locality for each species. Table 1 shows bumblebee samples used for DNA sequencing, the localities of their origin, the number of haplotypes found in this study and their Genbank accession numbers. After collection, bees were kept alive in a cool-box. During transportation, the characteristics essential for identification, such as the tufts of hair at the thorax and abdomen, were sometimes

Table 1. List of Bombus samples used in the present mitochondrial DNA analysis Species

Localities

Haplotypes

Genbank Acc.. No.

B.lucorum

Menz, Brandenburg, Germany Menz, Brandenburg, Germany Duncansby Head, Scotland, United Kingdom Menz, Brandenburg, Germany Duncansby Head, Scotland, United Kingdom

2

1

AY530009, AY530010 AY530012, AY530013 AY530011

1

AY530015

1

AY530014

B.cryptarum

B.magnus

2

soaked and stuck together, especially during wet weather. In such cases, we put the bees in small flight cages with some honey-water. After feeding, they started to clean and brush their hair by themselves, resulting in the restoration of the morphological characteristics. For GC/MS investigations, males obtained from the colonies artificially developed in a greenhouse were used. Voucher specimens of the bees examined are stored at the Primate Research Institute, Kyoto University and at the Entomological Collection of the Zoological Museum, Humboldt-University, Berlin. Morphology The details of morphology were studied with a stereomicroscope (Wild M8, Planar 1.0, Oculars 10¥/21) using only fresh specimens with undisturbed colour patterns. As described by E. Krüger, details in hair was best studied in diffuse light (use of diffuse filter in combination with Novoflex Macrolight Plus) with high magnification by stroking the hair with a pinpoint artists paintbrush. Distribution of hair on different parts of the body were thus carefully investigated. To avoid minor equipment vibrations, photographs were taken on a table with a shock absorbing granite plate, using extension bellows and a macro-lens (Olympus Zuiko 1:1 Macro 1:4/80 mm). The thorax was mounted on stubs with conductive carbon cement, coated with gold and viewed with a Hitachi S-530 scanning electron microscope (Fig. 1). Queens of B. lucorum have been carefully examined by Krüger (1939). According to his investigations, the yellow hair of the collare always ended at the border of the pronotallobus, with only a few yellow hairs found at the dorsal border of the episternum. Queens of B. cryptarum show a characteristic darkened S-shaped band of hair, which follows the border of the pronotallobus and separates a yellow collare from two patches of yellow hair at the upper border of the episternum (Fig. 2a–c). The broad, bright collare of B. magnus queens is never melanised and reaches far down below the tegula at both sides of the thorax (Fig. 3a–c). The lower and anterior border is often characteristically diffused by long yellow hair running in parallel to the lateral parts of the thorax.

Insect. Soc.

Vol. 52, 2005

Research article

47

Figure 2. Bombus cryptarum (Fabricius) queen, anterior part of thorax with yellow hair pattern of collare. (a) Menz, Brandenburg, Germany. (b) Duncansby Head, Scotland, United Kingdom. (c) Dunnet, Scotland, United Kingdom.

Figure 3. Bombus magnus (Vogt) queen, anterior part of thorax with yellow hair pattern of collare. (a) Menz, Brandenburg, Germany. (b) Duncansby Head, Scotland, United Kingdom. (c) Dunnet, Scotland, United Kingdom.

Gland preparation and GC/MS The cephalic parts of the labial glands were dissected from the heads of males (reared in artificial colonies) under freezing conditions and placed in vials (glands from 5 males per vial) containing 0.2 ml pentane. A Finnigan MAT TSQ700 gas chromatograph/tandem mass spectrometer was employed. Gas chromatography was carried out on a Hewlett Packard Ultra 1 column (50 m, 0.2 mm i. d., 0.11 mm film thickness) in the splitless mode with helium as carrier gas at an inlet pressure of 300 kPa. Initial temperature of 120°C was held for 1 min, then

increased at 8°/min to 280°C, at 3°/min to 310°C and at 1°/min to 320°C. This temperature was held for 10 min. Mass spectrometer conditions were: interface temperature 300°C, source temperature 130 °C, electron energy 70 eV, emission current 0.2 mA, and electron multiplier 1400 V. In the positive ion chemical ionization mode ammonia CI gas pressure was 70 Pa. Dimethyl disulfide adducts were prepared as described by Buser et al. (1983). Compounds were identified by comparing their mass spectra with those of the NIST ¢02 Library (National Institute of Standards and Technology, USA) and by coinjection with commercially available standards.

48

A. Bertsch et al.

Polymerase chain reaction (PCR) and DNA sequencing of mitochondrial CO1 We examined two individuals from each locality for each species. Total DNA was extracted from a mid-leg of each individual using the conventional phenol-chloroform method. The PCR was carried out to amplify the 1064 base-pair region in the mitochondrial cytochrome oxidase subunit 1 (CO1) according to Tanaka et al. (2001a). Primers used in PCR were forward: 5¢-ATAATTTTTTTTATAGTTATA-3¢ and reverse: 5¢-GATATTAATCCTAAAAAATGTTGAGG-3¢. The reaction mixture consisted of 2 ml template DNA (approx. 200 ng/ml), 3 ml 10 ¥ buffer (100 mM Tris-HCl, pH 8.3, 500 mM KCl and 15 mM MgCl2), 2.4 ml dNTP mixture (each 25 mM), 0.6 ml each primer, 0.15 ml polymerase (TAKARA) and 21.25 ml distilled water. The PCR conditions were: initial denaturing at 93°C for 1 min, 30 cycles of denaturing at 93 °C for 45 sec, annealing at 45°C for 1 min and extension at 60°C for 3 min, and final extension at 60°C for 4 min. The PCR products were purified using QIAquick PCR purification Kit (QIAGEN). The sequencing reaction was done using the purified DNA and a BigDye Terminator Ver. 3.1 Cycle Sequencing Kit (Applied Biosystems) with the primers mentioned above and the internal primers, forward: 5¢ATTTGATCGAAATTTTAATAC-3¢ and reverse: 5¢-CCAGAAGTATATATTTTAAT-3¢. DNA sequencing was performed using the ABI PRISM 3100 Genetic Analyzer (Applied Biosystems). DNA sequences determined from both directions were confirmed using a software, Sequence Navigator (Applied Biosystems). Analysis of sequence divergence of mitochondrial CO1 The absolute number of substitutions were counted based on pairwise comparison of CO1 haplotypes. The analysis for the specimens investigated was performed using the maximum likelihood method of the TREE-PUZZLE program distributed by K. Strimmer and A. von Haeseler and Tamura-Nei genetic distance was calculated. Because the distribution of nucleotides in CO1 genes of Hymenoptera is known to be heterogeneous, with a strong A + T bias, we selected the Tamura-Nei model of base substitution (Tamura and Nei, 1993), which corrects this bias in its assumption of sequence evolution. The nucleotide frequencies and the parameters necessary for this model were estimated from the sequence data. The tree topology was inferred by the quartet puzzling method (Strimmer and von Haeseler, 1996) and support indices for internal branches were calculated with 1000 puzzling steps. The CO1 sequence data of B. sporadicus (GenBank accession No. AF279500, Tanaka et al., in preparation) were used as outgroup. To extract more information for the branching patterns and genetic distances between species a maximum likelihood tree was also calculated by Bayesian analysis (Huelsenbeck et al., 2001) using MRBAYES, a program distributed by Huelsenbeck and Ronquist.

Results The diagnostic character ‘coloration of the collare’ and ‘border of collare at pronotallobus/episternum’ of queens Queens of B. cryptarum showed a characteristic S-shaped band of dark hair, which followed the border of the pronotallobus and separates a yellow collare from two patches of yellow hair at the upper border of the episternum. In northern Germany many specimen of B. cryptarum are strongly melanised, and the collare is not bright yellow but often dark-brown due to the mixture of yellow and black hair. Only the two yellow patches at the upper border of the episternum remain unchanged by melanisation, this distinct characteristics made a safe determination of queens

Species discrimination by labial gland secretion and mtDNA

especially easy (Fig. 2 a). The broad, bright collare of the B. magnus queen was never melanised. It reached far down below the tegula to both sides of the thorax, with both the lower and anterior border often characteristically diffused by long yellow hair running in parallel to the lateral parts of the thorax (Fig. 3 a). The combination of broad bright collare and the broad bright band on the abdomen made the queen of B. magnus both especially brilliant and unmistakable. In Scotland B. lucorum, B. magnus and B. cryptarum queens are large and bright, melanisation nearly never occurs. The characteristic dark S-shaped band of B. cryptarum, which follows the border of the pronotallobus was very faint (Fig. 2b, c), sometimes it was completely absent. Nevertheless the characteristic curved patch of yellow hair at the upper border of the episternum was always present, and helped to discriminate B. cryptarum from B. magnus, where the yellow hair of the collare extend much lower on both sides of the thorax (Fig. 3b, c). However discrimination was not easy and it was helpful to sharpen the eye for the characteristics by first inspecting specimens of B. cryptarum with the characteristic S-shaped band at the border of the pronotallobus (Fig. 2b). It was also very helpful to do field work early in spring when only queens of B. cryptarum were in flight. GC/MS of the cephalic labial glands of males Typical gas chromatograms (GC) for the cephalic labial gland secretions of males of B. cryptarum and B. magnus from Teupitz (LKr. Dahme-Spreewald, Brandenburg/Germany) and Dunnet (Caithness, Scotland/United Kingdom) are given in Fig. 4 and the compounds are summarised in Table 2. The gland secretions contained the usual mixture of straight-chain fatty acid derivatives (alcohols, esters, aldehydes and hydrocarbons) normally detected in the GC of the subgenus Bombus s. str. (Calam, 1969; Bergström et al., 1981; Bertsch, 1997; Valterová and Urbanová, 1997; Urbanová et al., 2001, 2002). The major compound in both species was ethyl dodecanoate (peak 4), considerable amounts of ethyl tetradecanoate (peak 8) and ethyl 9-octadecenoate (peak 18, B. magnus) and minor amounts of ethyl 9-hexadecenoate (peak 10) and ethyl hexadecanoate (peak 11) were also detected. Methyl dodecanoate (peak 2) was present in both species, whereas methyl tetradecanoate (peak 6), methyl hexadecanoate (peak 9) and methyl octadecenoate (peak 14) could only be detected in B. magnus. Substantial amounts of dodecanoic acid (peak 3) and 9-octadecenoic acid (peak 16, B. magnus) were also identified. 9,12-Octadecadienol (peak 12) and 9,12,15-octadecatrienol (peak 13) were characteristic for the labial glands of B. magnus. Both substances were absent in B. cryptarum. Icosenol (peak 20 & 21), docosenol (peak 26 & 27), tetracosenol (peak 33 & 34) and hexacosenol (peak 39 & 40) were also identified. Whereas for B. magnus the prominent peak of icosenol (peak 21) was characteristic, in B. cryptarum docosenol (peak 27) was the dominant alkenol besides two

Insect. Soc.

Vol. 52, 2005

Research article

49

Table 2. Compounds of the labial glands of Bombus cryptarum (CRY) and B. magnus (MAG). a RT: Retention time. b RI: Retention index. c MI: Molecular ion. d CN: Carbon number = chain length. Major component = xxx, medium components = xx, minor & trace components = x. No

RT a

RI b

IUPAC-Name

MI c

CRY

MAG

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 Wax ester 45 46 47 48 49 50 51 52 53 54 55 56

6:02 7:41 8:31 8:41 9:55 10:22 11:12 11:25 13:09 13:32 13:46 14:46 14:49 15:15 15:20 15:46 15:58 16:08 16:16 17:04 17:08 17:17 17.38 18:29 18:36 19:13 19:19 19:24 19:28 19:42 20:23 20:34 21:15 21:20 21:30 21:34 21:37 21:49 23:18 23:25 23:32 23:37 23:51 25:49

1378 1507 1562 1579 1677 1709 1764 1778 1909 1964 1978 2035 2040 2086 2100 2125 2146 2155 2178 2248 2258 2270 2300 2378 2382 2453 2464 2471 2491 2500 2548 2600 2658 2266 2269 2672 2691 2700 2862 2874 2880 2882 2900 3081 CN d 20 20 22 22 22 24 24 26 26 26 28 28

Ethyl decanoate Methyl dodecanoate Dodecanoic acid Ethyl dodecanoate Propyl dodecanoate Methyl tetradecanoate Ethyl 9-tetradecenoate Ethyl tetradecanoate Methyl hexadecanoate Ethyl 7-hexadecenoate Ethyl hexadecanoate 9,12-Octadecadien-1-ol 9,12,15-Octadecatrien-1-ol Methyl octadecenoate Henicosane 9-Octadecenoic acid Ethyl octadecadienoate Ethyl octadecenoate Ethyl octadecanoate Icosen-1-ol Icosen-1-ol Tricosene Tricosane Ethyl icosenoate Ethyl icosanoate Docosen-1-ol Docosen-1-ol Pentacosene Pentacosene Pentacosane Dodecyl dodecanoate Hexacosene Tetracosen-1-ol Tetracosen-1-ol Heptacosadiene Heptacosene Heptacosene Heptacosane Hexacosen-1-ol Hexacosen-1-ol Nonacosadiene Nonacosene Nonacosane Hentriacontene

200 214 200 228 242 242 254 256 270 282 284 266 264 296 296 282 308 310 312 296 296 322 324 338 340 324 324 350 350 352 368 364 334 334 376 378 378 380 380 380 404 406 408 434

xx x xx xxx x – x xx – x x – – – xx x – x – x xx x xx – – x xx x x xx – – xx xx – x x x x x – x x x

xx x xx xxx x x x xx x x x xx xx xx xx xx x xx x x xx x xx x x – xx x x xx x x x x x x x x x x x x x x

Octadecadienyl dodecanoate Octadecatrienyl dodecanoate Icosenyl dodecanoate Icosenyl dodecanoate Icosyl dodecanoate Docosenyl dodecanoate Docosyl dodecanoate Octadecadienyl octadecenoate Octadecatrienyl octadecenoate Tetracosenyl dodecanoate Hexacosenyl dodecanoate Hexacosenyl dodecanoate

448 446 478 478 480 506 508 532 530 534 562 562

– – x x x x x – – x x x

xx xx x x x x x x x x x x

26:27 26:32 29:13 29:17 29:23 31:30 31:41 34:22 34:26 34:42 38:04 38:16

50

A. Bertsch et al.

Species discrimination by labial gland secretion and mtDNA

Figure 4. Typical gas chromatogram (GC) of cephalic part of labial glands for males of B. cryptarum and B. magnus. Teupitz, Brandenburg, Germany. (a) B. cryptarum. (c) B. magnus. Dunnet, Scotland, United Kindom. (b) B. cryptarum. (d) B. magnus. Major diagnostic peaks for GC B. magnus absent in B. cryptarum: 12*/13* octadecadienol & octadecatrienol, 18* ethyl-9-octadecenoate, 45*/46* octadecadienol dodecanoate & octadecatrienol dodecanoate. For details see Table 2.

Insect. Soc.

Vol. 52, 2005

Research article

51

Figure 5. Cladogramm (branching order) and phylogram (branching order and distance information) of B. lucorum, B. magnus and B. cryptarum CO1 based on 891 characters of aligned sequences. Bombus sporadicus was used as outgroup. (a) Maximum likelihood tree obtained by quartet puzzling method using Tamura-Nei’s model of base substitution. Numbers above branches indicate support indices based on 5000 quartet puzzling steps. (b) Maximum likelihood tree obtained by using Bayesian MCMC (Markov Chain Monte Carlo method) analysis with the general time reversible model of base substitution, gamma distribution and 50000 generations.

characteristic hexacosenol peaks (peak 39 & 40). A complex mixture of 12 wax type esters with carbon chain lengths of between 20–28 were found in the chromatograms. The characteristic MS fragment ions of the alcohol and the acid part of these esters (Pepe et al., 1993) enabled the detection of small amounts of the respective alcohols or acids, which are otherwise difficult to detect and identify in gas chromatograms. The ions of m/z = 201 and 283 were the protonated dodecanoic and octadecanoic acids, respectively, the ions of m/z = 183 and 265 are attributed to the acylium ions of these acids. Characteristic components in the male labial gland secretions of many bumblebee species are primary alcohols. Therefore it is likely that they play a major role in communication. Both the main qualitative and quantitative differences between the labial gland secretions of B. magnus and B. cryptarum were differences in primary alcohols (9,12-octadecadien-1-ol, 9,12,15-octadecatrien-1-ol, icosen1-ol and docosen-1-ol), the species recognition-signals differ significantly, proving the specific status of both taxa. The chemistry of the labial glands show that B. lucorum is distinct from the species pair B. magnus and B. cryptarum, which share the same main component but differ in alcohols.

CO1 sequence divergence between and within species We used the 891 base-pair sequences of CO1 in analyses of sequence divergence and tree topology among the 3 species. The pairs of B. magnus individuals collected from Duncansby Head (United Kingdom), pairs of B. magnus from Menz (Germany), and pairs of B. cryptarum from Duncansby Head (United Kingdom) had identical CO1 sequences. Table 3 presents the matrix of evolutionary distance estimated by the TamuraNei model (Tamura and Nei 1993) and the number of base substitutions within and between the three taxa investigated. The intraspecific genetic variability was low for all the three taxa (1–3 base pairs, 0.001–0.003 in Tamura-Nei distance), even if the specimens of each taxon were collected in geographically distant localities. In contrast, the genetic distances between the species were approximately one order of magnitude larger (26–37 base pairs, 0.030–0.043 in Tamura-Nei distance). The cladogram constructed by the quartet-puzzling maximum likelihood analysis is shown in Figure 5a. Each species from the different localities formed a cluster. Bombus cryptarum and B. magnus were closer to each other than to B. lucorum and made the sister group in the topology of the tree. The molecular data clearly distinguished B. lucorum, B. cryptarum and B. magnus from each other.

52

A. Bertsch et al.

Species discrimination by labial gland secretion and mtDNA

Table 3. Pairwise genetic distance between and within the species of B. lucorum, B. cryptarum and B. magnus. Top diagonal: evolutionary distance estimated by the Tamura-Nei model, Bottom diagonal: number of substitutions in the 891 base-pair sequences of CO1 B. lucorum

1 luc-Menz-1 2 luc-Menz-2 3 cry-Duncansby 4 cry-Menz-1 5 cry-Menz-2 6 mag-Duncansby 7 mag-Menz

B. cryptarum

B. magnus

1

2

3

4

5

6

7

– 1 35 36 37 33 32

0.00112 – 34 35 36 32 31

0.04072 0.03953 – 1 2 28 26

0.04192 0.04072 0.00112 – 1 29 27

0.04315 0.04195 0.00225 0.00112 – 30 28

0.03840 0.03721 0.03239 0.03357 0.03478 – 3

0.03721 0.03602 0.03005 0.03123 0.03243 0.00338 –

Discussion Morphology As already described for B. cryptarum (Bertsch, 1997), contrary to many discussions in literature, the determination of spring queens in Germany is not difficult. Most individuals can easily be identified while feeding on plants. The strongly melanised dark collare, compared to the light yellow band of the abdomen characterises nearly all the females of B. cryptarum, whereas the broad, bright collare and a similar broad bright band at the abdomen makes the females of B. magnus especially brilliant against the dark-green background of Vaccinium myrtillus and the mosses of the forest floor. Inspection of the characteristic ‘border of the collare at the pronotallobus/episternum’ while the bee is in a glass vial usually confirms the first judgement. Only a few specimens need closer inspection in the laboratory with a stereomicroscope. The situation is more difficult in Scotland where spring queens of all three species are large and brightly coloured. Inspection of the specimen in a glass vial is therefore always required but, again, a closer inspection of the characteristic ‘border of the collare at the pronotallobus/episternum’ allows a safe determination of most females in the field. Only a few specimens must be transferred to the laboratory for closer inspection under a stereomicroscope. On wet and rainy days, removing the specimen from the net in most cases results in completely soaked bees and an inspection of the diagnostic characteristics is impossible. The bees have then to be transferred to flight cages and fed honey solution. There they warm-up, get dry and, after a short time, start cleaning and brushing their hair so that the diagnostic characteristics come out distinctively. Determination then becomes easy. Bombus magnus from Scotland was first described as a form of B. lucorum by Vogt (1911) and raised to rank of species by Krüger (1954) who carefully described the diagnostic characteristics of this species. Nevertheless the dispute surrounding the taxonomic status of B. magnus and its delimitation from B. lucorum is on going. Williams (2000) for instance investigated a series of 32 females from Scotland from which 6 had been morphologically determined as B. magnus by P. Rasmont. After measuring ‘how far the collare extends (dorso-ventrally) below the tegula’ and the

‘maximum breadth (antero-posteriorly)’ and plotting the standardised values, Williams concluded that the material investigated showed a continuum and not, as he expected, a distinct gap separating the measurements for B. lucorum and B. cryptarum. Apart from the difficulties involved in measurements on hair patches with diffuse borders, attempts at quantifying subtle morphological differences often fails either because the database is insufficient or because B. cryptarum is not treated separately as for instance in the measurements of Løken (1973), Pekkarinen (1979) and Baker (1996). Gas chromatogram The components of the cephalic part of the male labial glands of bumblebees used for scent marking are species specific, however following the discussion of Peters (1998) on speciation it could be useful to see them more as a cohesion mechanism of the species than as an isolation mechanism. These compounds have been successfully used to discriminate or to recognise difficult bumblebee taxa, as for instance B. lapponicus and B. monticola (Svensson, 1979) and B. lucorum and B. cryptarum (Bertsch, 1997). In this investigation the species recognition signals of the labial glands are used to separate B. cryptarum from B. magnus. The labial gland secretions of B. cryptarum from Teupitz/Germany (Fig. 4a) and from Dunnet/United Kingdom (Fig. 4b) are identical, both show the same pattern of compounds as the GCs for B. cryptarum from different localities all over Germany (Bertsch, 1997, Table 1). The labial gland secretions of B. magnus from Teupitz/Germany (Fig. 4c) and Dunnet/United Kingdom (Fig. 4d) are also identical. Compared to B. cryptarum, the GC of B. magnus has qualitative and quantitative differences: the characteristic double peak of octadecadienol and octadecatrienol of B. magnus is absent in B. cryptarum, with the corresponding double peak octadecadienol dodecanoate/octadecatrienol dodecanoate also absent. A prominent peak of ethyl 9-octadecenoate is characteristic for the GC of B. magnus, which also contains both dodecanoic acid and 9-octadecenoic acid, whereas B. cryptarum contains only dodecanoic acid. B. magnus thus has a larger pattern of esters than B. cryptarum.

Insect. Soc.

Vol. 52, 2005

Males of B. lucorum in the broadest sense (B. lucorum s. lat., see Williams, 1991, 1998) show extensive variations in colour (Pekkarinen, 1979), a variation associated with the chemical composition of the marking pheromones produced by the labial glands (Bergström et al., 1973). Pheromonal studies revealed two distinct types of males; one representing the ‘blond’ and the other the ‘dark’ form. The ‘blond’ form was characterised by the component ethyl tetradecenoate, which was detected as the main component of the pheromone from B. lucorum (Calam, 1969). Later Bergström et al. (1981) regarded these forms as two sibling species on the basis of their pheromonal differences. In an attempt to relate the figures of the publication (Bergström et al., 1973, Figs 1–14) to taxa, Rasmont et al. (1986) came to the conclusion that most of the specimens illustrated belong to B. lucorum, some might belong to B. cryptarum and, probably, no B. magnus was included. A new attempt to discriminate between B. lucorum, B. magnus and B. cryptarum by GCs of the male labial glands from Finland (Pamilo et al., 1997) gives the same result. Most probably as already in the investigation of Bergström et al. (1973) only males of B. lucorum and B. cryptarum have been analysed. Our GC/MS results for males bees, reared in artificial colonies from unmistakable spring queens collected from Brandenburg/Germany and Scotland/United Kingdom clearly prove that carefully identified specimens result in distinctly different GCs. This variation in the compound distribution of male labial glands was confirmed by material from a range of localities in Germany (Nuremberg/Bavaria), Russia (St. Petersburg), France (Col de la Croix Morand/Puy-deDôme) and the United Kingdom (Porlock/England, Abergavenny/Wales and Glenmore Forest/Scotland), where all three species live sympatric (Bertsch et al., unpublished). B. cryptarum is not restricted to Central and Northern continental Europe, it is also a species of the bumblebee fauna of the British Isles. Genetic distance and Sequence divergence of mitochondrial CO1 Mitochrondrial cytochrome oxidase 1 (CO1), a conservative protein-coding gene, evolving under functional constraints, was used with particular success in analysing different Hymenoptera (Bombus: Pedersen, 1996, 2002; Tanaka, 2001, Lasius: Hasegawa, 1998; Lasioglossum: Danforth, 1999, Apis: Tanaka et al., 2001a,b). CO1 is also a good genetic marker for the taxa B. lucorum, B. cryptarum and B. magnus. Each taxon from the different localities formed a cluster (Fig. 5a). Bombus cryptarum and B. magnus were closer to each other than to B. lucorum and made the sister group in the topology of the tree. Since the molecular data clearly distinguished B. lucorum, B. cryptarum and B. magnus from each other they should be treated as distinct species. Even for specimens collected from geographically distant places such as Menz/Germany and Yakutsk/Russia for B. lucorum (linear distance about 7000 km), and Menz/Germany and Duncans-

Research article

53

by/United Kingdom for B. cryptarum and B. magnus (linear distance about 1000 km), the genetic distance within species was less than 0.003 in Tamura-Nei distance. In an intensive molecular phylogenetic study of European Bombus, Pedersen (2002) recently analysed the CO1 sequences of B. lucorum, B. cryptarum and B. magnus. His results (Pedersen, 2002, Figs 4 and 5) show B. cryptarum as being genetically very close to B. lucorum (from the continent) and B. magnus as distinctly different but surprisingly close to B. lucorum from the United Kingdom. Pedersen (2002, page 382) concludes from his investigation ‘although the Bombus group of species seems to form a distinct monophyletic group, observed differences within the group indicate taxonomic problems so severe that likely only a closer study of morphology and molecular data from several localities in Europe will delimitate the species.’ We think, that will not be necessary. A maximum likelihood tree using Bayesian analysis (Huelsenbeck et al., 2001) was calculated to extract more information on the branching patterns and genetic distances between species. The CO1 sequence data of B. lucorum from Russia (GenBank accession No. AF279497, Tanaka et al., in prep.) and B. lucorum and B. magnus from Austria (Pedersen, 2001) were included in this analysis (Fig. 5b). The haplotype of B. magnus from Austria (AY181123) did not connect to B. magnus from Germany and Scotland but formed a cluster with B. cryptarum. The specimens of B. magnus investigated by Pedersen (2002) were collected from Switzerland and Austria, at localities of ± subalpine/alpine habitats (Sölkpass 1790 m, Austria and Julier-Pass 2284 m, Switzerland). B. magnus has until now never been reported from such subalpine localities. According to Amiet (1996) B. magnus is not a member of the fauna of Switzerland. However an exceptionally bright form of B. cryptarum (ssp. reinigianus Rasmont, 1984, Fig. 1b) can be found in the Alps of Austria and Switzerland (Rasmont, 1984, Map 3; Amiet, 1996). The specimens named B. magnus by Pedersen (2002) should be conspecific with B. cryptarum. Without morphological inspection the specimens named B. cryptarum by Pedersen (GenBank AY181101 Danmark and AY181100 Austria) cannot be properly categorised. Acknowledgement We would like to express our thanks to Dr. Dan Darley (at present Laboratory of Microbial Biochemistry, Fachbereich Biologie, Philipps-Universität Marburg) who carefully read the manuscript and suggested linguistic corrections.

References Alford, D.V., 1975. Bumblebees. London: Davis-Poynter. Amiet, F., 1996. Hymenoptera Apidae I. Teil Allgemeiner Teil, Gattungsschlüssel, die Gattungen Apis, Bombus und Psithyrus. Insecta Helvetica Fauna 12. Baker, D.B., 1996. On a collection of Bombus and Psithyrus principally from Sutherland, with notes on the nomenclature or status of three species ( Hymenoptera, Apoidea). Br. J. Ent. Nat. Hist. 9: 7–19.

54

A. Bertsch et al.

Bergström, G., B. Kullenberg and S. Ställberg-Stenhagen, 1973. Studies in natural odoriferous compounds. VII. Recognition of two forms of Bombus lucorum L. (Hymenoptera, Apidae) by analysis of the volatile marking secretion from individual males. Chemica Scripta 4: 174–182. Bergström, G., B.G. Svensson, M. Appelgren and I. Groth, 1981. Complexity of bumble bee marking pheromones: biochemical, ecological and systematical interpretations. In: Biosystematic of Social Insects (Howse P.E. and J.L. Clément, Eds), Academic Press, London, New York, Systematic Association Special Volume No.19, pp. 175–183. Bertsch, A., 1997. Abgrenzung der Hummel-Arten Bombus cryptarum und B. lucorum mittels männlicher Labialdrüsen-Sekrete und morphologischer Merkmale (Hymenoptera, Apidae). Entomol. Gen. 22: 129–145. Buser, H.R., H. Arn, P. Guerin and S. Rauscher, 1983. Determination of double bond position in mono-unsaturated acetates by mass spectrometry of dimethyl disulfide adducts. Anal. Chem. 55: 818– 822. Calam, D.H., 1969. Species and sex specific compounds from the heads of male bumblebees (Bombus spp.). Nature 221: 856–857. Danforth, B.N., 1999. Phylogeny of the genus Lasioglossum (Hymenoptera: Halictidae) based on mitochrondrial CO1 sequence data. Syst. Entomol. 24: 377–393. Hasegawa, E., 1998. Phylogeny and host-parasite relationships in social parasitism in Lasius Ants. Entomol. Sci. 1: 133–135. Huelsenbeck, J.P., F. Ronquist, R. Nielson and J.P. Bollback, 2001. Bayesian inference of phylogeny and its impact on evolutionary biology. Science 294: 2310–2314. ITE 1980: Atlas of the bumblebees of the British Isles. Published by the Institute of Terrestrial Ecology, Cambridge. Krüger, E., 1939. Die Hummeln und Schmarotzerhummeln von Sylt und dem benachbarten Festland. Schriften des Naturwissenschaftlichen Vereins für Schleswig-Holstein Band 23: 28–123. Krüger, E., 1951. Phaenoanalytische Studien an einigen Arten der Untergattung Terrestribombus O. Vogt (Hymenoptera, Bombidae). I. Teil. Tijdschr. Entomol. 93: 141–197. Krüger, E., 1954. Phaenoanalytische Studien an einigen Arten der Untergattung Terrestribombus O. Vogt (Hymenoptera, Bombidae). II. Teil. Tijdschr. Entomol. 97: 263–298. Krüger, E., 1956. Phaenoanalytische Studien an einigen Arten der Untergattung Terrestribombus O. Vogt (Hymenoptera, Bombidae). II. Teil. Tijdschr. Entomol. 99: 75–105. Krüger, E., 1958. Phaenoanalytische Studien an einigen Arten der Untergattung Terrestribombus O. Vogt (Hymenoptera, Bombidae). III. Teil. Tijdschr. Entomol. 101: 283–344. Løken, A., 1973. Studies on Scandinavian bumblebees (Hymenoptera, Apidae). Nor. Entomol. Tidsskr. 20: 1–218. Pamilo, P., S. Varvio-Aho and A. Pekkarinen, 1984. Genetic variation in bumblebees (Bombus, Psithyrus) and putative sibling species of Bombus lucorum. Hereditas 101: 245–251. Pamilo, P., J. Tengö, P. Rasmont, K. Pirhonen, A. Pekkarinen and E. Kaarnama, 1997. Pheromonal and enzyme genetic characteristics of the Bombus lucorum species complex in northern Europe. Entomol. Fenn. 7: 187–194. Pedersen, B.V., 1996. A phylogenetic analysis of cockoo bumblebees (Psithyrus, Lepelletier) and bumblebees (Bombus, Latreille) inferred from sequences of the mitochrondial gene cytochrome oxidase I. Mol. Phylogenet. Evol. 5: 289–297. Pedersen, B.V., 2002. European bumblebees (Hymenoptera: Bombini) – phylogenetic relationships inferred from DNA sequences. Insect Syst. Evol. 33: 361–386. Pekkarinen, A., 1979. Morphometric, colour and enzyme variation in bumblebees (Hymenoptera, Apidae, Bombus) in Fennoscandia and Denmark. Acta Zool. Fenn. 158: 1–60. Pepe, C., J. Dagaut, P. Scribe and A. Saliot, 1993. Double bond location in monosaturated wax esters by gas chromatography/mass spec-

Species discrimination by labial gland secretion and mtDNA trometry of their dimethyl disulphide derivatives. Org. Mass Spectrom. 28: 1365–1367. Peters, G., 1998. Der ‘point of no return’ im Speziationsprozess. Zoologische Abhandlungen, Staatliches Museum für Tierkunde Dresden. Band 50 Supplement Nr. 6: 100 Jahre Art-Konzept in der Zoologie: 69–76. Rasmont, P., 1984. Les bourdons du genre Bombus Latreille sensu stricto en Europe Occidentale et Centrale (Hymenoptera, Apidae). Spixiana 7: 135–160. Rasmont, P., A. Scholl, R. De Jonghe, E. Obrecht and A. Adamski, 1986. Identité et variabilité des mâles de bourdons du genre Bombus Latreille sensu stricto en Europe occidentale et centrale (Hymenoptera; Apidae, Bombinae). Rev. Suisse Zool. 93: 661–682. Scholl, A. and E. Obrecht, 1983. Enzymelektrophoretische Untersuchungen zur Artabgrenzung im Bombus lucorum-Komplex (Apidae, Bombini). Apidologie 14: 65–78. Scholl, A., E. Obrecht and R.E. Owen, 1990. The genetic relationship between Bombus moderatus Cresson and the Bombus lucorum auct species complex (Hymenoptera: Apidae). Can. J. Zool. 68: 2264– 2268. Scholl, A., R.W. Thorp and E. Obrecht, 1992. The genetic relationship between B. franklini (Frison) and other taxa of the Subgenus Bombus s. str. (Hymenoptera: Apidae). Pan-Pac. Entomol. 68: 46–51. Strimmer, K. and A. von Haeseler, 1996. Quartet puzzling: A quartet maximum likelihood method for reconstructing tree topologies. Molecular Biol. Evol. 13: 964–969. Svensson, B.G., 1979. Pyrobombus lapponicus auct., in Europe recognized as two species: P. lapponicus (Fabricius) and P. monticola (Smith) ( Hymenoptera, Apoidea, Bombinae). Entomol. Scand. 10: 275–296. Tamura. K. and M. Nei, 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzes. Mol. Biol. Evol. 10: 512–526. Tanaka, H., 2001. Phylogeographical study of East Asian bumble bees and Southeast Asian honey bees. PhD thesis, Kyoto University. (in Japanese). Tanaka, H., D.W. Roubik, M. Kato, F. Liew and G. Gunsalam, 2001a. Phylogenetic position of Apis nuluensis of northern Borneo and phylogeography of A. cerana as inferred from mitochondrial DNA sequences. Insect. Soc. 48: 44–51. Tanaka, H., T. Suka, D.W. Roubik and M. Mohamed, 2001b. Genetic differentiation among geographic groups of three honeybee species, Apis cerana, A. koschevnikovi and A. dorsata, in Borneo. Nature and Human Activities 6: 5–12. Urbanová, K., I. Valterová, O. Hovorka and J. Kindl, 2001. Chemotaxonomical characterisation of males of Bombus lucorum (Hymenoptera: Apidae) collected in the Czech Republic. Eur. J. Entomol. 98: 111–115. Urbanová, K., I. Valterová, P. Rasmont and M. Terzo, 2002 Znackovací feromony: slození sekretu labiální zlázy samcù Bombus magnus. [Marking pheromones of bumblebees: composition of the labial gland secretion of males of Bombus magnus]. Chem. Listy 96: 918– 919 (in Czech). Valterová, I. and K. Urbanová, 1997. Chemické signály Cmeláku [Chemical signals of bumble bees]. Chem. Listy 91: 846–857 (in Czech). Vogt, O., 1911. Studien über das Artproblem. 2. Mitteilung. Über das Variieren der Hummeln. 2. Teil (Schluss). Sitzungsberichte der Gesellschaft Naturforschender Freunde zu Berlin 1911: 31–74. Williams, P., 1991. The bumble bees of the Kashmir Himalaya (Hymenoptera: Apidae, Bombini). Bull. Br. Mus. (Nat. Hist.) Entomol. 60: 1–204. Williams, P., 1998. An annotated checklist of bumble bees with an analysis of patterns of description (Hymenoptera: Apidae, Bombini). Bull. Br. Mus. (Nat. Hist.) Entomol. 67: 79–152. Williams, P., 2000. Are Bombus lucorum and magnus separate species? BWARS Newsletter 2000(1): 15–17.