The arctic and alpine bumblebees of the subgenus Alpinobombus revised from integrative assessment of species’ gene coalescents and morphology (Hymenoptera, Apidae, Bombus) Author Williams, Paul H. Author Berezin, Mikhail V. Author Cannings, Sydney G. Author Cederberg, Björn Author Ødegaard, Frode Author Rasmussen, Claus Author Richardson, Leif L. Author Rykken, Jessica Author Sheffield, Cory S. Author Thanoosing, Chawatat Author Byvaltsev, Alexandr M. text Zootaxa 2019 2019-07-03 4625 1 1 68 journal article 26307 10.11646/zootaxa.4625.1.1 24213782-d4ec-4a95-9ca4-be0a424b06f9 1175-5326 3268175 D08DD464-F1AD-4253-888C-65A2A293F517 5. Bombus kirbiellus Curtis ( Figs. 5 , 77‒88 , 133 , 141 , 146 ) < Bombus > Kirbiellus Curtis 1835 :lxii and his plate A figure 2, type-locality citation ‘ Arctic’ (probably the Boothia Peninsula isthmus, Nunavut , Canada ). Lectotype queen (#294) by present designation, NHMUK examined PHW, no locality label. Note 1. Bombus Putnami Cresson 1878:185 , type-locality citation ‘ Colorado - Alpine’ ( USA ). Holotype female by monotypy, not seen (ANSP Type 2170, identity not in doubt). Synonymised with Bombus kirbyellus Curtis by Franklin (1913:290) . [< Bombus > kirbyellus Curtis ; von Dalla Torre 1896:527 , incorrect subsequent spelling.] < Bombus > putnamii Cresson; von Dalla Torre 1896:544 , unjustified emendation. PSITHYRUS KODIAKENSIS Ashmead 1902:130 , type-locality citation ‘Kodiak’ ( USA ). Syntype males, NMNH photos of specimen and genitalia examined PW, ‘Kadiak’ [sic] (Kodiak, Alaska , USA ). Synonymised with Bremus kirbyellus (Curtis) by Frison (1923:316) . Bombus atrifasciatus Morrill 1903:224 , type-locality citation ‘Gallatin Co., Montana’ ( USA ). Holotype female by monotypy, not seen (identity not in doubt). Synonymised with Bombus kirbyellus Curtis by Franklin (1913:291) . Bremus kirbyellus var. alexanderi Frison 1923:308 , type-locality citation ‘ Patagonia Mountains , Arizona’ ( USA ). Holotype queen (#5012) by original designation, INHS examined PW. Bremus kirbyellus var. arizonensis Frison 1923:309 , type-locality citation ‘ Patagonia Mountains , Arizona’ ( USA ). Holotype queen (#5013) by original designation, INHS examined PW. Note 1 ( kirbiellus ). The NHMUK is the most likely depository for the type material described by John Curtis from the British expedition led by John Ross to find the Northwest Passage. The type material has been believed to be lost ( Franklin 1913:291 ; Richards 1931:15 ; Milliron 1973:95 ) or to be ‘not traced’ ( Løken 1973:105 ). Fortunately, there is an illustration of a queen accompanying Curtis’s original description (his plate A figure 2). The NHMUK collection does contain one queen (#294) that could have been the model for Curtis’s illustration of B. kirbiellus . It is the only nineteenth century queen of this species in the NHMUK collection from the New World and it is also the only nineteenth century specimen of this species in the NHMUK collection that has been dried with the wings and legs extended, broadly as shown in the illustration in Curtis (which may have been rendered in a slightly more idealised, symmetrical pose). The lectotype queen of B. polaris (from the same expedition, see above) has been dried in a similar pose, but has a different colour pattern. The NHMUK queen (#294) matches the Curtis illustration of B. kirbiellus in colour pattern, in that: (1) the hair on the top of the head posterior to the ocelli is predominantly black, so that it appears black from a distance, even though there is some yellow hair intermixed in this area as is frequent for this species; (2) there is a narrow but clear indication of black hair anterior and ventral to the wing bases, indicating a patch of black hair on the side of the thorax posteriorly in addition to the yellow hair anteriorly on the side of the thorax that is most frequent and often most extensive for this species; (3) metasomal tergum 2 posteriorly has black hair in a median triangle, which is infrequent for this species; (4) the hair of terga 4‒5 is a very pale pink-orange rather than the deep orange-red or yellow-white seen for many individuals. None of these colour-pattern characters is unique to this specimen, but no other candidate specimen is known with this combination of characters, so that this is the most likely specimen currently known as a candidate for the model for the illustration. There are nineteenth century New World queens of B. polaris in the NHMUK collection, but they all have a conspicuous band of yellow hair posteriorly on metasomal tergum 3, which is specifically excluded in the original description of B. kirbiellus . The NHMUK specimen (#294) carries the labels: (1) white, handwritten ‘ Kirbiellus [female].’; (2) white, round, handwritten ‘40 / 4 – 2 / 1719’; (3) green, printed ‘ Alpinobombus / AL# 294. det. PHW’; (4) red, printed ‘ LECTOTYPE [female] / Bombus / kirbiellus / Curtis, 1835 / det. PH Williams 2014 ’; (5) white, printed ‘ Bombus / ( Alpinobombus ) / kirbiellus / det. PH Williams 2015 ’. The accession label (2) ‘40 / 4 – 2 / 1719’ refers (according to the NHMUK accessions register) to a specimen ‘Bought at Mr Children’s sale’ in 1840. This is presumably a reference to John Children, who became Keeper of Zoology at the British Museum (BM), retiring in 1840. It is possible that Children may have acquired the specimen from Curtis through Children’s association with the BM. Smith (1854:397) indicated that this species was present in the BM collection in 1854 (the entry for ‘ BOMBUS KIRBIELLUS’ is marked ‘B.M.’), represented by material from ‘Boothia Felix’ (known to be visited at this time only by the Ross expedition), presumably therefore from Curtis. The hand-writing of the label (1) ‘ Kirbiellus [female].’ matches a sample of Curtis’s handwriting shown by Horn et al. ( 1935 ‒1937: plate XXXVIII no. 19) with a similar ‘r’, ‘b’, and ‘e’. The ‘s’ can be compared directly between the label and the sample published for Curtis, with a reduced upper section, and is more similar to Curtis’s handwriting than to the complete ‘s’ shown for Frederick Smith ( Horn et al. 1935 ‒1937:plate V no. 13), who worked on the bumblebee collection of the British Museum from 1849. However, the handwriting samples are too small to be conclusive. There is no single point of evidence that demonstrates beyond all doubt that this specimen (#294) is one of Curtis’s syntypes , but in combination the evidence from the specimen’s appearance and from the labels is considered on the balance of probabilities to support this conclusion. The specimen has been re-pinned, but it is complete except that it lacks the terminal joints of the left hind tarsus and of the mid and hind right tarsi. This queen is designated here as lectotype in order to reduce uncertainty in the identity and application of the name, especially in relation to the closely similar species B. polaris that also occurs on Boothia. A conspecific male (#293) in the NHMUK collection with an accession label ‘40 / 4 – 2 / 1718’ (also from the 1840 Children purchase) is also likely to be from the Ross expedition and is considered to be a paralectotype . Curtis (1835 :lxiii) wrote that B. kirbiellus ‘seemed to be the most abundant species’ [of bumblebee] seen on the Ross expedition, although it is likely that this included observations of the similar B. polaris . He gave the observation dates for most of the large female (queen) bumblebees (page lxiv) as June and July in 1830 and 1831 . The main text of the report (e.g. page 566) places the expedition on these dates in the region around Cape Isabella on the isthmus of the Boothia Peninsula. Taxonomy and variation. Bombus kirbiellus has often been considered part of B. balteatus but has recently been recognized as separate from evidence of a species coalescent in the COI gene ( Williams et al. 2015 ; Fig. 9 ). Bombus kirbiellus is vicariant with B. balteatus across the Bering Strait ( Figs. 59 , 77 ). FIGURES 77‒88 . 77: Bombus kirbiellus global distribution (as for Fig. 15). 78 ‒ 88 colour patterns (as for Figs. 16 ‒ 21): 78 (#556) Canada-Nunavut; 79 (#294) Canada-Nunavut; 80 (#100) Canada-Nunavut; 81 (#3791) USA-Colorado; 82 (#3790) USA-Colorado; 83 (#3765) Canada-Nunavut; 84 (#3792) USA-Colorado; 85 (#3797) USA-Colorado; 86 (#4915) USA Alaska; 87 (#3768) Canada-Nunavut; 88 (#3812) USA-Colorado. Variation in the colour pattern has been illustrated and analysed previously ( Williams et al. 2014 (under the name B. balteatus ); Williams et al. 2015 ). The colour pattern with yellow on T4‒5 ( Figs. 78‒79 ) is known to us only from Nunavut. Otherwise the principal variation in this species is in whether the pale hair of T4‒5 is predominantly orange ( Figs. 80‒83, 85‒87 ) or predominantly black ( Figs. 84, 88 ). The colour pattern from Lake Hazen on Elles- mere Island ( Fig. 83 ) has the side of the thorax predominantly black, resembling some B. balteatus . Individuals with both the orange-tailed and black-tailed colour patterns share some of the same alleles, although there may be some variation in the relative frequency of the two patterns with latitude ( Williams et al. 2015 ). Material examined. 257 queens 512 workers 307 males (plus 166 individuals with sex or caste undetermined), from USA and Canada ( Fig. 77 : AMNH BBSL CNC CSCA EMEC INHS JT LACM NHMUK PR PCYU PW PWRC RBCM RMBL ROM RSM RUAC SEMC UAM UCD UCMNH UOG WRMEYPM ), with 27 specimens barcoded.A record for Bartlett Bay plotted for Ellesmere Island by Williams et al . (2014) is corrected here to Alaska from revised information on the location of this site ( Mackovjak 2010 ). The records of B. balteatus in Milliron (1973) from Alaska and northern Canada are believed to be misidentifications of B. kirbiellus . Six dark specimens (#1618, 3763‒3766, 4250) from the far north of Canada at Lake Hazen (Ellesmere Island), from samples taken in 1961, 1967, and 2010, have been identified here as B. kirbiellus . The nearest known populations of this species, probably on the Boothia Peninsula of Nunavut ( Curtis 1835 ), appear to be distantly disjunct. This apparent gap is curious. Bombus kirbiellus was not recorded from Lake Hazen either by Milliron and Oliver (1966) or by Richards (1973) . The species may be rare in northern Nunavut. The colour pattern from Lake Hazen has a dark side to the thorax, similar to some Old World individuals of B. balteatus , and is less easy to distinguish from the locally more abundant B. polaris . This difference in colour pattern from their relatives further south in the New World makes a recent introduction from the southern population a less likely explanation of the apparently isolated Lake Hazen population. Alternatively these northern bees could be a relic from a darker population that was once more widespread during a previous warm interglacial period. Habitat and distribution. Flower-rich arctic/alpine tundra and subalpine meadows in the New World excluding Greenland , north to the Boothia Peninsula as well as disjunctly in Ellesmere Island and east to Newfoundland, extending southwards into the subarctic region in the Aleutian Islands and Hudson Bay, and in the alpine zone of the southern Rocky and Californian Sierra Nevada mountains. Regional distribution maps ( Milliron 1973 ; Thorp et al. 1983 ; Laverty & Harder 1988 ; Williams et al. 2014 ). Food plants. ( Milliron 1973 ; Thorp et al. 1983 ; Williams et al. 2014 ) Behaviour. ( Hobbs 1964 ) Conservation status. This species has not yet been fully assessed for Red List threat status using IUCN criteria (2001). Hatfield et al . (2016a) have listed B. kirbiellus as ‘Data deficient’.