Bombus separandus Vogt, 1909

Williams, Paul H., Altanchimeg, Dorjsuren, Byvaltsev, Alexandr, Jonghe, Roland De, Jaffar, Saleem, Japoshvili, George, Kahono, Sih, Liang, Huan, Mei, Maurizio, Monfared, Alireza, Nidup, Tshering, Raina, Rifat, Ren, Zongxin, Thanoosing, Chawatat, Zhao, Yanhui & Orr, Michael C., 2020, Widespread polytypic species or complexes of local species? Revising bumblebees of the subgenus Melanobombus world-wide (Hymenoptera, Apidae, Bombus), European Journal of Taxonomy 719, pp. 1-120 : 93-96

Bombus separandus Vogt, 1909 stat. rev.

Figs 16 , 173–180 , 204 , 210 , 212

Bombus lapidarius Form [subsp.] separandus Vogt, 1909: 61 .

Bombus lapidarius Form [subsp.] kohli Vogt, 1909: 61 (non Cockerell, 1906: 75 = B. morio (Swederus)) . syn. nov.

Bombus kozlovi Skorikov, 1910a: 413 , replacement name for kohli Vogt, 1909: 61. syn. nov.

Bombus lapidarius var. [subsp.] tenellus Friese, 1913: 86 . syn. nov.

Bombus tenellus var. [not subsp.] alpivagus Richards, 1930: 639, but not infrasubspecific after Reinig 1935: 333 ( ICZN 1999: Art. 45.6.4.1). syn. nov.

Lapidariobombus alagesianus subsp. pamirus Skorikov, 1931: 226 (non Skorikov, 1931: 232 = B. oberti Morawitz , not B. difficillimus Skorikov as given by Skorikov). syn. nov.

Lapidariobombus alagesianus subsp. mongolicus Skorikov, 1931: 226 (non Friese, 1916: 110?= B. diversus Smith ). syn. nov.

Bombus keriensis f.g. [subsp.] bucharicus Reinig, 1935: 340. syn. nov.

Bombus keriensis f.g. [subsp.] richardsi Reinig, 1935: 341 (non Frison, 1930: 6 = B. rufipes Lepeletier ).

Bombus tenellus subsp. tibetensis S.-F. Wang, 1982: 439, replacement name for richardsi Reinig, 1935: 341. syn. nov.

Bombus kozlowi – Bischoff, 1936: 9, unjustified emendation.

Bombus keriensis (part) – Reinig 1935: 341. — Williams 1998: 134 (non Morawitz, 1887: 199).

Vogt (1909: 46) stated that his “Formen” are subpopulations, which can be considered equivalent to subspecies. Reinig (1935: 333) stated explicitly that his “f.g.” (forma geographica) are equivalent to subspecies. Vogt (1909: 61) described his taxon separandus as having the hair of the face and top of the head (“Gesicht und Stirn”) black and the underside of the thorax dark haired. He described his taxon kohli (later renamed kozlovi ) as having the face with a few yellow hairs. However, Vogt (1911: 58) then characterised the taxon separandus as differing from the taxon keriensis only in having an absence of pale hairs on the face (he did not include the taxon kohli / kozlovi as part of the taxon separandus but did include the taxon incertoides as part of separandus ). The taxon separandus was treated as a part of the species B. keriensis s. lat. by Reinig (1935) and by Williams (1998).

Our PTP analysis ( Fig. 10 ) of coalescents in the COI gene within the keriensis -complex supports six species including B. separandus , corroborated by differences in morphology. These species are also supported by the absence of a positive divergence-with-distance relationship among them ( Fig. 20 ) (see Divergence and geographical distance, page 12).

From morphology, after examining>1000 specimens of the keriensis -group from across Asia and after comparing the distribution of states of several characters as part of this study (PW), Vogt’s character state of exclusively black hair on the face does appear to remain one of the most consistently diagnostic characters (in comparison with COI sequences) for B. separandus in the QTP and Central Asia, at least for queens (but see the comments on B. keriensis ). Nonetheless, there are some individuals of the yellow-banded taxon kozlovi from Mongolia (part of B. separandus s. lat. according our COI-coalescent results, Fig. 10 ) that do have a few yellow hairs on the face, so that exclusively black hair on the face is not diagnostic in all parts of the range. Several aspects of the colour pattern vary strongly within the species: (1) B. separandus includes both yellow-banded and white-banded colour patterns; (2) the extent of the black hair between the wing bases varies; (3) the extent of the pale hair on the side and ventral area (leg bases) of the thorax varies; (4) the extent of the pale hair posteriorly on T3 varies (the ‘ciliation’ of authors); (5) the contrast of the pale posterior fringes on T4–5 varies. Many of these characters might appear to show a continuum of variation that is also continuous with the variation within B. keriensis s. str. ( Williams 1991), which might then appear consistent with the two taxa being parts of the same species ( Reinig 1935). However, as noted above, evidence from COI barcodes shows that the two genecoalescent groups differ most consistently in the absence of pale hairs on the face for B. separandus , supporting Vogt’s characterisation of the species, providing integrated morphological support for its status as a separate species. The overlapping colour variation of B. separandus but only in Mongolia (where B. keriensis s. str. does not occur) does not negate the fact that B. separandus is recognisable by morphological characters throughout its geographical range.

There are some queens of the yellow-banded taxon richardsi (later renamed tibetensis ) from the northwestern QTP that have no pale hairs on the face, very few or no yellow hairs on the leg bases, and very few black hairs within the yellow bands of the thoracic dorsum. Many of the available specimens have failed to sequence, so the species’ diagnosis remains tentative. Particularly influential in the interpretation of these individuals is one yellow-banded queen with a reduced extent of pale hair on the side of the thorax from the Zanskar range (PW: ML311, Figs 177 , 210 ), which has produced a short sequence (382 bp) that nonetheless shows bases diagnostic for B. separandus compared to B. keriensis (base positions 267A, 286C, 301C, 306T, 339T, 363C, 423T, 498C, 529G, 540T, 541C, 556C). This is one of the specimens that most closely resemble some yellow-banded individuals of the candidate species keriensis s. str. also from Kashmir ( Figs 170 , 209 ) (the difference in the breadth of the black band between the wings is not diagnostic for these species).

Bombus separandus co-occurs locally with the similar B. incertoides in the Altai and Khangai mountains of Mongolia. Bombus separandus also co-occurs locally with the similar B. keriensis s. str. in the Zanskar mountains of Kashmir. Examination of specimens in the IOZ collection labelled B. incertus (S.-F. Wang 1985; S.-F. Wang & Yao 1996; Niu et al. 2018) shows that these are mostly misidentified B. separandus (with a few B. keriensis ) from Xinjiang.

Diagnosis

Females

Queens medium-sized body length 17–20 mm, workers 9–14 mm. Can be distinguished in Central Asia and Mongolia by the combination of hair of the face usually without pale hairs (except in Mongolia where B. keriensis does not occur), the leg bases usually without pale hairs, T2 posteriorly entirely pale without black hairs, and T3 laterally usually without pale hairs (cf. B. sichelii , B. keriensis ).

Males

Body length 10–15 mm. Can be distinguished reliably at present only by their COI sequence, although in the western Himalaya they may be distinguished by entirely black hair between the wing bases and hair of T3 entirely black. Genitalia ( Fig. 204 ) with the gonostylus shorter than broad, its inner basal projection reduced to a short stub (cf. rufipes- group, festivus- group, rufofasciatus -group); volsella with the inner distal corner broadly produced but without a narrow hook (cf. rufipes- group, festivus- group, rufofasciatus -group); eye unenlarged relative to female eye.

Material examined

Lectotype

CHINA • ♀ (queen), lectotype of Bombus lapidarius Form separandus Vogt, 1909 by designation of Williams (1991: 97); Xinjiang, Borohoro Shan, Tisilikau; RMNH (examined PW).

Material sequenced (25 specimens)

MONGOLIA • 1 ♀ (queen); Bulgan; 47.4257° N, 103.6940° E; 22 Jun. 2013; NHMUK seq: NHMSR1; PW: ML147 • 1 ♀ (worker); Bayanolgiy , Bulgan; 46.9333° N, 91.1667° E; 26 Jul. 1992; Culverwell leg.; BOLD seq: 6877B11; PW: ML151 • 1 ♀ (queen); Khenteyn Nuruu, Terelj ; 47.9905° N, 107.3719° E; 16 Jun. 2015; R. DeJonghe leg.; BOLD seq: 6877A08; RJ: ML333 • 1 ♀ (queen); Khenteyn Nuruu, Terelj ; 47.9956° N, 107.3303° E; 17 Jun. 2015; R. DeJonghe leg.; BOLD seq: 6880B05; RDJ: ML489 • 1 ♀ (worker); Arkhangai, Chuluut Khujirt Mts ; 47.5750° N, 101.1092° E; 11 Aug. 2018; R. DeJonghe leg.; BOLD seq: 6880B12; RDJ: ML497 .

RUSSIA • 1 ♀ (worker); Altai, Ulagansky , 6 km NE of Aktash; 50.3158° N, 87.7036° E; 6–7 Jul. 2017; S. Knyazev leg.; BOLD seq: 1555F07; PW: ML264 • 1 ♀ (worker); Altai, Ulagansky , 8 km NE of Aktash; 50.3160° N, 87.7253° E; 5–6 Jul. 2017; S. Knyazev leg.; BOLD seq: 1555F08; PW: ML265 • 1 ♀ (worker); Tuva, 27 km S of Erzin ; 50.0049° N, 95.1719° E; 11–12 Jul. 2014; A. Lelej leg.; BOLD seq: 1555F09; PW: ML266 • 1 ♀ (worker); Tuva, 25 km of Erzin ; 50.1611° N, 95.4778° E; 14–15 Jul. 2014; A. Lelej leg.; BOLD seq: 1555F10; PW: ML267 • 1 ♀ (worker); Tuva, 31 km of Erzin ; 50.0366° N, 95.4348° E; 18 Jul. 2014; A. Lelej leg.; BOLD seq: 1555F12; PW: ML269 • 1 ♀ (worker); same collection data as for preceding; 16 Jul. 2014; A. Lelej leg.; BOLD seq: 1555G01; PW: ML270 .

KYRGYZSTAN • 1 ♀ (queen); Jalalabad , Jylamysh; 42.6522° N, 74.4812° E; 27 May 2009; L. Best leg.; BOLD seq: 1552H03; PW: ML156 • 1 ♀ (queen); Oshakaya , Kyzyl-Suu; 39.4761° N, 71.8933° E; 18 Jul. 1998; H. Rausch leg.; BOLD seq: 1552A08; OLML: ML191 • 1 ♀ (worker); Chong Aksuu Valley nr Karakol lake; 42.8386° N, 77.3889° E; 26 Jun. 2016; R. DeJonghe leg.; BOLD seq: 1555F04; RJ: ML261 • 1 ♀ (worker); same collection data as for preceding; BOLD seq: 1555F05; RJ: ML262 • 1 ♀ (worker); same collection data as for preceding; BOLD seq: 1555F06; RJ: ML263 • 1 ♀ (queen); Kungey Ala-Too , Chong Aksuu nr Karakol lake; 42.8397° N, 77.37° E; 25 Jun. 2016; R. DeJonghe leg.; BOLD seq: 6877A09; RJ: ML336 • 1 ♀ (queen); Terskey Ala-Too , Turgon-Aksuu river; 42.4183° N, 78.9467° E; 2 Jul. 2016; R. DeJonghe leg.; BOLD seq: 6877A11; RJ: ML341 • 1 ♀ (queen); same collection data as for preceding; BOLD seq: 6877A10; RJ: ML342 • 1 ♀ (queen); Terskey Ala-Too , Turgon-Aksuu river; 42.4019° N, 79.0486° E; 3 Jul. 2016; R. DeJonghe leg.; BOLD seq: 6877A12; RJ: ML343 • 1 ♀ (queen); Terskey Ala-Too , Saruu, Juktuu river; 42.0947° N, 77.9617° E; 6 Jul. 2016; R. DeJonghe leg.; BOLD seq: 6877B01; RJ: ML348 • 1 ♀ (queen); Santash ; 42.7389° N, 79° E; 29 Jun. 2016; R. DeJonghe leg.; BOLD seq: 6877B02; RJ: ML356 • 1 ♀ (queen); same collection data as for preceding; BOLD seq: 6877B03; RJ: ML357 .

PAKISTAN • 1 ♀ (worker); Chitral, Khyber-Pakhtunkhwa, Tirich Valley ; 36.3150° N, 71.9824° E; 8 Aug. 1984; W. Budenburg leg.; BOLD seq: 6877C02; PW: ML383 .

INDIA • 1 ♀ (queen); Kashmir , Zanskar, Nimaling Plain; 33.7927° N, 77.5881° E; 15 Jul. 1980; P. Williams leg.; BOLD seq: 1555H10; PW: ML311 .

Global distribution

(Central Asian and Mongolian mountain species extending into the northwestern corner of the QTP) North Asia: MONGOLIA, RUSSIA: Tuva, Altai Mountains. – Central Asia: KAZAKHSTAN, KYRGYZSTAN, TAJIKISTAN. – East Asia: CHINA: Xinjiang. – Himalaya: PAKISTAN, INDIA: Kashmir. (IAR, IOZ, NHMUK, OLML, PW, RDJ, RMNH, ZIN, ZMHB.). This species is widespread and can be common ( Fig. 212 ).

Behaviour

Very few food-plant records but expected to be generalists ( Caragana versicolor in Williams 1991 ). The male mate-searching behaviour is expected to resemble the patrolling of B. keriensis s. str.

Figures

Figs 14‒16. Maps of sequenced samples as in Figs 12–13. 14. The tanguticus-group and the lapidarius- group. 15. The sichelii-group. 16. The keriensis-group.

Figs 139–180. Simplified diagrams for the colour patterns of the hair on the dorsum for the species from the integrative analysis. The dorsum is divided into regions, each of which shows only the predominant or most apparent colour for that region using a simplified colour palette, with olive indicating a mixture of black and yellow hair, and grey indicating a mixture of black and white hair. The keriensis-group. 139. Queen, China-Sichuan. 140. Worker, China-Sichuan. 141. Queen, China-Sichuan. 142. Queen, China-Sichuan. 143. Worker, China-Xizang. 144. Queen, India-Kashmir. 145. Male, China-Qinghai. 146. Male, India-Kashmir. 147. Male, China-Gansu. 148. Queen, Armenia. 149. Queen, Iran. 150. Male, Georgia. 151. Worker, China-Qinghai. 152. Male, China-Qinghai. 153. Queen, Mongolia. 154. Worker, Mongolia. 155. Male, Mongolia. 156. Male, Mongolia. 157. Queen, China-Sichuan. 158. Worker, China-Qinghai. 159. Worker, China-Qinghai. 160. Male, China-Qinghai. 161. Male, China-Qinghai. 162. Male, China-Qinghai. 163. Queen, Pakistan. 164. Queen, Pakistan. 165. Queen, Pakistan. 166. Queen, Pakistan. 167. Queen, Afghanistan. 168. Queen, Pakistan. 169. Queen, Pakistan. 170. Queen, India-Kashmir. 171. Male, India-Kashmir. 172. Male, India-Kashmir. 173. Queen, China- Xinjiang. 174. Queen, Kyrgyzstan. 175. Queen, Kyrgyzstan. 176. Queen, Mongolia. 177. Queen, India- Kashmir. 178. Male, Kazakhstan. 179. Male, China-Xinjiang. 180. Male, Mongolia.

Figs 199‒204. Morphology of the male genitalia for species of the subgenus Melanobombus von Dalla Torre, 1880 from the dorsal aspect, anterior at the bottom of the image, posterior at the top. 199. Bombus alagesianus Reinig, 1930 stat. rev., Georgia. 200. B. tibeticus sp. nov., China-Qinghai. 201. B. incertoides Vogt, 1911 stat. rev., Mongolia. 202. B. qilianensis sp. nov., China-Qinghai. 203. B. keriensis Morawitz, 1887, India-Kashmir. 204. B. separandus Vogt, 1909 stat. rev., Kyrgyzstan. Scale bars = 1 mm.

Figs 209–210. Images of queens of two cryptic species from around the western Qinghai-Tibetan plateau. 209. Bombus keriensis Morawitz, 1887 from Mt Apharwat (4000 m a.s.l.) in the Pir Panjal mountains (ML405). 210. B. separandus Vogt, 1909 stat. rev. from Nimaling plain (4800 m a.s.l.) in the Zanskar mountains (ML311). Viewed from the left lateral aspect.

Figs 211–212. Maps of sequenced samples showing the most frequent colour patterns locally. 211. The sichelii-complex, consisting of the single polytypic species B. sichelii Radoszkowski, 1859 (inset: B. sichelii worker, China-Neimenggu, photo PW). 212. The keriensis-complex, consisting of B. alagesianus Reinig, 1930 stat. rev. (blue spots), B. tibeticus sp. nov. (dark green spots), B. incertoides Vogt, 1911 stat. rev. (light green spots), B. qilianensis sp. nov. (yellow spots), B. keriensis Morawitz, 1887 (orange spots), and B. separandus Vogt, 1909 stat. rev. (red spots) (inset: B.qilianensis sp. nov. queen, China-Sichuan, photo PW). Maps projected onto a sphere in ArcGIS using World_Shaded_Relief basemap © 2014 ESRI and showing boundaries between countries as recognised by the UN.

Fig. 10. October 2019 (Fig. 8) Bayesian PTP analysis of MrBayes tree of unique COI barcodes (UHF- PTP). Symbols as in Fig. 9.

Figs 17–20. Plots of (y-axis) pairwise proportion of genetic divergence between COI barcode region sequences against (x-axis) pairwise Great Circle geographical distance in km between sample sites with linear trend lines (black). 17. Genetic divergences within the candidate species for the sichelii-complex identified using Bayesian UF-PTP (Mantel r = 0.52). 18. Genetic divergences among the candidate species for the sichelii-complex identified using Bayesian UF-PTP (Mantel r = 0.21). 19. Genetic divergences within the candidate species for the keriensis-complex identified using Bayesian UF-PTP (Mantel r = 0.16). 20. Genetic divergences among the candidate species for the keriensis-complex identified using Bayesian UF-PTP (relationship not positive). For details of the measurements see the