J Insect Conserv DOI 10.1007/s10841-016-9853-2

ORIGINAL PAPER

Biogeography and designatable units of Bombus occidentalis Greene and B. terricola Kirby (Hymenoptera: Apidae) with implications for conservation status assessments Cory S. Sheffield1 • Leif Richardson2 • Syd Cannings3 • Hien Ngo4 Jennifer Heron5 • Paul H. Williams6

•

Received: 12 August 2015 / Accepted: 8 February 2016  Springer International Publishing Switzerland 2016

Abstract Conservation action for species of concern requires that ‘‘designatable units’’ (e.g., species, subspecies, geographic races, genetically distinct forms) are clearly defined, or that the species complex is treated as a whole. Several species of bumble bee are currently threatened, and some of these have cryptic colouration (resembling other species), or form complexes that vary considerably in colour patterning. Here we address the taxonomy and distribution of Bombus occidentalis Greene and B. terricola Kirby, both of which are currently of conservation concern in North America. Bombus occidentalis includes two apparently monophyletic groups of COI barcode haplotypes (recently considered as subspecies) with ranges mostly separated by that of their sister species, B. terricola. The southern B. o. occidentalis ranges throughout the western United States and into western Canada Electronic supplementary material The online version of this article (doi:10.1007/s10841-016-9853-2) contains supplementary material, which is available to authorized users. & Cory S. Sheffield [email protected]

from southern Saskatchewan and Alberta, and throughout British Columbia north to ca. 55N; the northern B. o. mckayi Ashmead, is restricted to north of this in British Columbia, westernmost Northwest Territories, Yukon Territory and Alaska. Bombus o. mckayi exists, as far as is known, only with a ‘‘banded’’ colour pattern. By contrast, B. o. occidentalis occurs in both banded and nonbanded colour patterns, although the southern banded colour pattern is geographically isolated from the northern subspecies. Bombus o. occidentalis has declined throughout its range, perhaps due in part to exposure to novel parasites. Despite having similar levels of parasitism (ca. 40 %) as the southern subspecies, B. o. mckayi appears to have stable populations at present. There is therefore compelling evidence that the two subspecies should be distinguished for conservation and management purposes. We present the evidence for their distinction and provide tools for subspecies recognition. Keywords Bombus  Cryptic species  Species at risk  Conservation  DNA barcode  Distribution

1

Royal Saskatchewan Museum, 2340 Albert Street, Regina, SK S4P 2V7, Canada

Introduction

2

Gund Institute for Ecological Economics, University of Vermont, 617 Main St., Burlington, VT 05405, USA

3

Canadian Wildlife Service, Environment Canada, 91780 Alaska Highway, Whitehorse, YT Y1A 5X7, Canada

4

Department of Biology, York University, 4700 Keele St., Toronto, ON M3J 1P3, Canada

5

Species Conservation Science Unit, British Columbia Ministry of Environment, Vancouver, BC V6T 1Z1, Canada

6

Department of Life Sciences, The Natural History Museum, Cromwell Road, London SW7 5BD, UK

Bumble bees (Bombus, Hymenoptera: Apidae) are large, robust bees common throughout temperate regions of the world, ranging well into the Arctic Circle (Milliron and Oliver 1966; Milliron 1971, 1973; Richards 1973; Williams 1998; Michener 2007; Williams et al. 2014). Primarily primitively eusocial species (excluding subgenus Psithyrus Lepeletier, the social parasites or cuckoo bumble bees), bumble bees play a critical role as pollinators in most plant communities, including for several important crops (Free 1993; Kearns and Thomson 2001; Koch and

123

J Insect Conserv

Strange 2009; Button and Elle 2014; Zhang et al. 2015). In the past decade, several North American bumble bee species have undergone rapid declines in abundance and/or distribution (Thorp 2005; Colla and Packer 2008; Grixti et al. 2009; Koch and Strange 2009; Colla and Ratti 2010; Cameron et al. 2011; Colla et al. 2012), with several potential causative agents offered in explanation, including climatic and/or dietary specialization, climate change, queen emergence time, land use changes and loss of floral hosts, and increased levels of pathogens and pesticides (e.g., Goulson and Darvill 2004; Goulson et al. 2005, 2006, 2008, 2015; Colla et al. 2006; Williams and Osborne 2009; Williams et al. 2009; Connop et al. 2010; Cameron et al. 2011; Meeus et al. 2011; Kerr et al. 2015; Sachman-Ruiz et al. 2015). Interestingly, in North America these declines have been most prominent in the subgenus Bombus Latreille (Evans et al. 2008), which is represented there by five currently recognized species (Bertsch et al. 2010; Williams et al. 2012a, 2014): Bombus affinis Cresson, B. cryptarum (Fabricius), B. franklini (Frison), B. occidentalis Greene, and B. terricola Kirby, four of which are species of concern in North America (Koch and Strange 2009). Bombus franklini, the geographically most restricted species in the genus in North America (southern Oregon and northern California) is also the rarest, now thought extinct (Thorp 2005). Bombus affinis has also become very rare, and has received much recent attention due to its rapid decline in northeastern North America (Colla and Packer 2008; Grixti et al. 2009; Cameron et al. 2011); only a handful of specimens have been observed in the last decade. This species was once one of the most common bumble bees in northeastern United States, and in southern Ontario and Que´bec in Canada (with one specimen known from New Brunswick), but now has ‘‘Endangered’’ species status in Canada, as recognized by the Committee on the Status of Wildlife in Canada (COSEWIC 2010) and the federal Species at Risk Act (SARA). In addition, the U.S. Environmental Protection Agency has been petitioned (by the Xerces Society in 2013) to list B. affinis as Endangered. Bombus terricola, the most widely distributed member of the subgenus in North America (Williams et al. 2014), and its sister species B. occidentalis (Williams et al. 2012a; Owen and Whidden 2013) also appear to be declining (Evans et al. 2008). Bombus terricola is typically best known in eastern North America (Mitchell 1962; Laverty and Harder 1988; Colla et al. 2011; Williams et al. 2014), ranging west into the prairies (Frison 1923; Hobbs 1968; Koch et al. 2012), south to Tennessee and North Carolina, and north into northern Alberta and southeastern areas of the Northwest Territories. Though B. terricola was not recorded in the western coastal United States by Stephen (1957), including California (Thorp et al. 1983) (and see range maps of Milliron (1971) and Koch et al. (2012)),

123

some specimens have been observed in Canada as far west as central British Columbia (Williams et al. 2012a, 2014). Bombus occidentalis is a western species, occurring along the Pacific coast (Stephen 1957; Milliron 1971; Thorp et al. 1983; Rao and Stephen 2007) as far east as southeastern Saskatchewan (Regina; C. S. Sheffield, pers. obs.), north into the western-most Northwest Territories, the Yukon and Alaska, and south to New Mexico (Cockerell and Porter 1899; Milliron 1971). In the subgenus Bombus s. str., only the Holarctic B. cryptarum (=B. moderatus Cresson), found in Alaska, Yukon, Northwest Territories, Nunavut (excluding the arctic islands), south to the James Bay area of Ontario, west to Alberta and British Columbia (Scholl et al. 1990; Williams et al. 2012a, 2014), seems to have stable populations in North America (Koch and Strange 2012), although this species may have undergone range expansion in the last few years (Owen et al. 2012). One of the most important steps in conserving bumble bees is to know the distribution and specific life-history requirements (e.g., habitat and food plant preferences, nesting requirements, climatic niche, etc.) of the species of concern. However, being able to assess the status of species and develop conservation recovery strategies is dependent on knowing which species or ‘‘designatable units’’ are (is) involved, especially for cryptic species (Murray et al. 2008; Williams et al. 2011, 2012b, 2013; Carolan et al. 2012; Lecocq et al. 2014; Scriven et al. 2015). Bumble bees, despite being well studied taxonomically in North America (e.g., Franklin 1912; Frison 1926; Stephen 1957; Milliron 1971, 1973; Thorp et al. 1983; Laverty and Harder 1988; Colla et al. 2011; Koch et al. 2012; Williams et al. 2014) often form cryptic groups with closely similar colour patterns shared by many species (Stephen 1957; Owen and Plowright 1980, 1988; Williams 2007; Murray et al. 2008; Lecocq et al. 2011; Carolan et al. 2012; Williams et al. 2014; Huang et al. 2015). As such, it is no surprise that of the 250? species of bumble bee currently recognized globally, over 2800 names have been proposed (Williams 1998). Of the members of the subgenus Bombus in North America this is especially true for B. occidentalis, once considered a subspecies of B. terricola (Handlirsch 1888; Milliron 1971; Williams et al. 2012a; Owen and Whidden 2013); Milliron (1971) even considered B. franklini a synonym of B. terricola occidentalis. In the Old World, the closely related B. terrestris also shows high levels of variation, and nine subspecies are recognized (Rasmont et al. 2008). Though Bertsch et al. (2010) indicated that B. terricola and B. occidentalis are easily distinguished from each other by colour alone, in areas where their ranges overlap in western North America, the colour patterns can be closely similar (Franklin 1912; Milliron 1971; Koch et al. 2012) and some of the specimens from the west identified as B.

J Insect Conserv

terricola may actually be B. occidentalis (and vice versa) so that confirmation with COI barcodes may be advisable (Williams et al. 2012a). For instance, although Titus (1902) reported B. terricola from Colorado (specimen identified by W. Ashmead), Lutz and Cockerell (1920) expressed doubts. Frison (1926) indicated that this record might be valid, and also described a new variety, B. terricola var. severini from South Dakota (which is now considered to be B. occidentalis). Hobbs (1968) studied both species in Alberta, and although he did not clearly state how the species were distinguished for his studies, he indicated that B. occidentalis was the more common of the two species in the southern part of the province, not occurring north of Banff (51N); in contrast B. terricola was more northern in distribution. More recently, and based on a molecular phylogeny of the subgenus Bombus globally, Williams et al. (2012a) recognized two subspecies of B. occidentalis in North America. As both B. occidentalis and B. terricola are being considered for conservation status assessment in Canada, and both are of concern in the United States (Evans et al. 2008; Cameron et al. 2011), our objective is to recognize clearly designatable units, including geographic considerations, for use in the assessment of B. occidentalis. We also review the colour patterns of B. occidentalis, and use combined evidence of morphology (colour-pattern distribution), and COI barcodes to clarify its taxonomy.

Materials and methods Colour patterns and biogeography We examined the colour patterns and geography of 1062 specimens of B. occidentalis and 642 specimens of B. terricola in nine collections in North America, including the Canadian National Collection of Insects, Arachnids and Nematodes (Ottawa, ON), Royal Saskatchewan Museum (Regina, SK), the Packer Collection at York University (Toronto, ON), Wallis/Roughly Collection at the University of Manitoba (Winnipeg, MB), Manitoba Provincial Museum (Winnipeg, MB), Royal Alberta Museum (Edmonton, AB), Royal British Columbia Museum (Victoria, BC), Royal Ontario Museum (Toronto, ON), and the American Museum of Natural History (New York, NY). Additionally, in order to clarify/delimit the distribution of both species of concern, in 2013 we conducted over 350 h of field work in Saskatchewan, Alberta, British Columbia and the Yukon Territory, with over 9000 bumble bee captures, including an additional 130 specimens of B. occidentalis and 209 of B. terricola. All specimens of B. occidentalis examined were assigned to two broad colour groups based on original descriptions: ‘‘non-banded’’ refers

to the occidentalis pattern as described by Greene (1858) (Figs. 1, 5f, k), ‘‘banded’’ refers to all patterns with a partial or complete band of yellow pubescence on tergum 3, including the more poorly defined banded patterns of the taxa ‘‘coloradensis’’ (Titus 1902) and ‘‘perixanthus’’ (Cockerell and Porter (1899); Figs. 1, 5g–j, l). For each specimen examined, we recorded and georeferenced the collection locality. We used Maxent species distribution modeling software (version 3.3.3) to model climate suitability for both subspecies of B. occidentalis (as per Williams et al. 2012a) and B. terricola (Phillips et al. 2004, 2006; Phillips and Dudı´k 2008). We supplied the model with collection locations of the specimens described above, as well as those of specimens with credible species determinations from museums around North America (B. o. mckayi, n = 200; B. o. occidentalis, n = 11,339; B. terricola, n = 14,499; see Online Resources 1 and 2). For environmental data we used 19 global bioclimatic variables (2.5 arc-minute resolution; see Online Resource 3) derived from meteorological temperature and precipitation recordings and clipped to North America (Hijmans et al. 2005). For each taxon, we partitioned the data into 10 replicates and constructed iterative models using the replicated run ‘subsample’ method, each time reserving 20 % of samples for model testing and jackknifing to measure importance of each variable to the prediction. We imported mean model outputs for each taxon to ArcGIS 10.1 and converted to raster format. For comparative purposes, we arbitrarily excluded areas with low prediction probability (p B 0.25) then binned the remaining raster as areas of low (0.25 \ p B 0.37), medium (0.37 \ p B 0.47) and high (0.47 \ p B 0.64) suitability for each species. We then overlaid specimen point records (by colour pattern) onto the maps to compare areas of known and predicted occurrence. COI-barcode data We used DNA barcoding to clarify the range extent and verify colour patterns of the two subspecies of B. occidentalis and B. terricola, with specimens from Williams et al. (2012a) serving as DNA Barcode reference material within the Barcodes of Life Data System (BOLD). Sample specimens (total N = 149; B. occidentalis N = 95, B. terricola N = 54) were sequenced for the short 50 barcode region of the mitochondrial cytochrome c oxidase, subunit 1 (COI) gene. Protocol for bee specimen preparation follows Sheffield et al. (2009); COI was extracted, amplified and sequenced at the Biodiversity Institute of Ontario, University of Guelph, using the standard protocols described by Hebert et al. (2003a, b). Universal primers for the COI barcode sequence for insects were used (variants LepF1 and LepR1; Hebert et al. 2004). Data for the

123

J Insect Conserv

Fig. 1 Summary of colour patterns of Bombus occidentalis Greene (‘‘non-banded’’ colour patterns outlined in red; ‘‘banded’’ colour patterns outlined in black) and B. terricola Kirby (outlined in blue).

‘‘Banded’’ colour patterns of B. occidentalis are partitioned into named colour patterns; grey outline for colour patterns ‘‘mckayi’’. Bee diagrams from Williams et al. (2014). (Color figure online)

specimens processed at Guelph have been uploaded to the BOLD online database (Ratnasingham and Hebert 2007) in project BOCTR, with sequences submitted to Genbank; Genbank accession numbers are included in Online Resource 4.

(United States) showing mean intra-species variation of 0.24 % (maximum 1.07 %), and forming a group (i.e., the subspecies mckayi of Williams et al. 2012a) separate from southern populations (see Williams et al. 2012a, and Online Resource 5). Colour patterns within the northern haplotype group are consistent in that most (i.e., 85 %) of the specimens examined collected between 55 and 60N, and virtually all above 60N show the banded pattern of tergum 3 (Fig. 2). In contrast, specimens from the southern barcode group (i.e., \55) show great variation in colour pattern throughout the range (Fig. 2), even from within the same local population (e.g., captured in the same trap in southeastern Alberta), although variation in COI remains low (mean intra-specific variation 0.14 %; maximum 1.96 %). Some of these specimens, though superficially matching B. terricola in colour pattern (Figs. 1, 5j, l) have patterns of COI consistent with the subspecies occidentalis of Williams et al. (2012a) (see Online Resource 5). Such individuals have been found in several localities throughout the range of B. occidentalis.

Results COI-barcodes and colour pattern Throughout its range, mean intra-species variance in COI of 0.11 % was observed for B. terricola, with low maximum intra-species variation at 1.39 %. These patterns correspond to relatively consistent patterns of colour variation for B. terricola, a species which maintains high levels of morphological consistency in pubescence colour in diagnostic areas (i.e., tergum 2 and 3 are normally entirely yellow, though few specimens have dark hairs apicomedially on tergum 2), varying only in degree of yellow pubescence on the scutellum, pleura, and apical terga (Figs. 1, 5c–e). Bombus occidentalis s. l. shows similar intra-specific variation in COI (mean intra-species variance 0.28 %), but with higher maximum variation of 2.03 %. The higher maximum variation for B. occidentalis s.l. corresponds with the larger range size; specimens from North America north of ca 57, including northern British Columbia, the Northwest Territories and Yukon (Canada) and Alaska

123

Biogeography The climatic suitability models derived from our data accurately predicted most test data (B. o. mckayi: AUC = 0.950 ± 0.038; B. o. occidentalis: AUC = 0.905 ± 0.003; B. terricola: AUC = 0.895 ± 0.005; Online Resource 1). Nearly all predicted habitat for the putative subspecies of B. occidentalis (mckayi and occidentalis) is geographically

J Insect Conserv Fig. 2 The proportion of ‘‘banded’’ (i.e., tergum 3 yellow; shaded bars) and ‘‘nonbanded’’ (i.e., tergum 3 black; open bars) specimens (N = 1062) of Bombus occidentalis Greene examined for 5 latitude intervals. Sample size for each 5 increment is indicated in [ ] on each bar

separated (Figs. 3, 4). Even where collections of the two taxa have been made in close proximity to each other (i.e., north-central British Columbia), the models predict a fine scale allopatry (Fig. 3). The predicted range of B. o. mckayi was best explained by temperature variables, including

mean annual temperature, diurnal temperature range and mean temperature of driest quarter; precipitation during the driest month was also explanatory (Online Resource 3). For the subspecies B. o. occidentalis, the most important variable contributing to the model was temperature

Fig. 3 Probability of occurrence of Bombus terricola (blue), B. occidentalis occidentalis (green) and B. occidentalis mckayi (red), based on habitat suitability models produced with Maxent software. Dots represent actual data points on which the models are based, and are coloured in blue, green, and red to match their models. (Color figure online)

123

J Insect Conserv Fig. 4 Probability of occurrence for Bombus terricola (blue), B. occidentalis occidentalis (green) and B. occidentalis mckayi (red) in North America, based on habitat suitability modeled with Maxent software. (Color figure online)

evenness (isothermality, a comparison of daily and yearly temperature fluctuations; Online Resource 3). Also important were mean winter temperature and mean summer precipitation and temperature. From our models we expect that B. terricola habitat may be found in close proximity to that of the two subspecies of B. occidentalis where they meet in British Columbia (Figs. 3, 4). Our surveys included sites where both B. terricola and the southern subspecies of B. occidentalis (i.e., non-banded colour patters) were found together. The predicted range of B. terricola is best explained by mean and maximum summer temperatures, as well as annual mean temperature and summer precipitation (Online Resource 3).

Discussion For assessing the conservation status of biological diversity, it is often desirable to recognize taxonomic units below the species level, though sometimes delimiting taxa to even the species level can be difficult (De Queiroz 2007). According to COSEWIC, for status assessment of species under the Species at Risk Act (SARA) in Canada, designatable units can include taxa currently recognized as subspecies, and/or as taxa with geographically or genetically distinct populations (Green 2005). The majority of the North American species of Bombus s. str. are of concern; one is thought extinct (i.e., B. franklini), another virtually

123

extinct (B. affinis), and the two focal species of this study are of much concern (Evans et al. 2008; Cameron et al. 2011). As discussed by Plowright and Stephen (1980) and Williams et al. (2012a), Bombus franklini and B. affinis (the two species of most concern) pose little problem in terms of morphological and molecular distinctiveness within the subgenus in North America. Although Handlirsch (1888) and Dalla Torre (1896) considered Cresson’s (1863) [not Smith’s (1861)] ‘‘modestus’’ ([= moderatus Cresson, 1863] = B. cryptarum) (Fig. 5a, b) a subspecies of B. terricola, it is morphologically distinct and the only Holarctic member of the subgenus. At the present time, B. cryptarum is apparently not declining (Koch and Strange 2012, as B. moderatus), and appears to be becoming more common and widespread (Owen et al. 2012). Similarly, Milliron (1971) considered B. occidentalis as a subspecies of B. terricola and indicated that in ‘‘certain northwestern allopatric [he presumably meant ‘sympatric’, not allopatric] locations’’ it was often difficult to distinguish individuals of the two species, especially those of worker and male castes. In western North America very similar colour patterns exist making recognizing species difficult; some of the described colour patterns of B. occidentalis (e.g., B. occidentalis nigroscutatus, B. terricola var. severini) bear very close resemblance to B. terricola, and it is quite possible that some of the specimens identified as B. terricola from western British Columbia, and the United States from Colorado-west, are actually

J Insect Conserv Fig. 5 a Bombus cryptarum female and b male, c, d B. terricola females and e male, f– j B. occidentalis females and k– l males. Bombus occidentalis exhibits the greatest variation in colour, including ‘‘non-banded’’ (f, k) and ‘‘banded’’ (g–j, l) colour patterns, ‘‘banded’’ referring to partial or complete yellow pubescence on metasomal tergum 3. Only ‘‘banded’’ specimens are known to occur north of 60

B. occidentalis. Without pubescence, distinguishing B. terricola from B. occidentalis would be very difficult based on morphology alone (Bertsch et al. 2010; but see Owen and Whidden 2013). Stephen (1957), Milliron (1971) and Thorp et al. (1983) did not recognize B. terricola from costal western North America, though some specimens of B. occidentalis from San Francisco examined here are of a very similar colour pattern (i.e., nigroscutatus and severini) to B. terricola. Frison (1926) wrote that his variety of B. occidentalis (i.e., severini) is structurally very similar to B.

terricola, though having extensive yellow pubescence on the sides and dorsal surface of the mesosoma, some yellowish pubescence on the apicolateral edges of tergum 1, and tergum 5 with extensive white pubescence. Though the status of B. occidentalis as a distinct taxon has been unclear in the past (i.e., Franklin 1912; Milliron 1971; Cameron et al. 2007), recent molecular-based analyses (Bertsch et al. 2010; Williams et al. 2012a; Owen and Whidden 2013) have agreed on its relationship with B. terricola. Williams et al. (2012a) indicated that the concept of B. terricola as a

123

J Insect Conserv

species distinct from B. occidentalis is supported by several diagnostic changes in COI. In western North America their data appeared to show B. terricola extending its distribution between the two disjunct populations (i.e., northern and southern haplotype groups or subspecies) of B. occidentalis, which is supported by our analysis (Figs. 3, 4). Bombus terricola can be distinguished from most other members of the subgenus Bombus in North America by the characteristic metasomal colour pattern: tergum 1 black (with some exceptions noted by Milliron (1971)), terga 2–3 yellow, tergum 4 black (Figs. 1, 5c–e), with slight variations existing primarily in the extent of yellow pubescence on the dorsal and lateral mesosomal surfaces (Milliron 1971; Owen and Whidden 2013; Williams et al. 2014) and the apex of the metasoma. In contrast, most specimens of B. occidentalis (see Fig. 1) (and B. cryptarum) have extensive white pubescence on the apical metasomal segments (Rao and Stephen 2007). As mentioned above, some specimens of B. occidentalis do approach the colour patterning observed in B. terricola (Figs. 1, 5h, j, l). Milliron (1971) commented on specimens of B. terricola in which the anterior portion of tergum 2 would be interfused with black pubescence (e.g., Fig. 1; and see Williams et al. 2014), noting the similarity to Franklin’s (1912) nigroscutatus; unfortunately, geographic information was not provided for these specimens. For the specimens available to us, this intermixture of black hairs is very limited. Milliron (1971) also stated that there are ‘‘no evident constant or reliable morphological features by which specimens of these two overlapping allopatric [he presumably meant ‘sympatric’, not allopatric] populations can be positively differentiated in these areas of overlap. In such…locations, obviously the two subspecies (terricola and occidentalis) interbreed and produce numerous perplexing subspecific hybrids’’. Although this is unlikely, the pubescence colour of B. occidentalis is highly variable (Bertsch et al. 2010; Williams et al. 2014), typically much more so than in B. terricola, leading to the recognition of several forms previously recognized as species, subspecies, and taxa of lower rank (Fig. 1; summarized in Online Resource 6). Franklin (1912), Lutz and Cockerell (1920), and Rao and Stephen (2007) recognized two subspecies, occidentalis and nigroscutatus, each having several colour patterns which overlap, making them difficult to distinguish. Burks (1951) considered three taxa: nigroscutatus, occidentalis, and occidentalis var. proximus, though again with so much variation shared among and within these colour patterns (based on published descriptions) that these taxa again were not clearly delimited. The most extensive treatments of colour patterns were given by Franklin (1912), Stephen (1957), and more recently, Williams et al. (2014). The typical non-banded (‘‘occidentalis’’) colour pattern (Figs. 1, 5f, k) and the southern banded colour

123

pattern(s) show very little variation in COI across the range of the species; both colour patterns have even been collected in the same locations in southeastern Alberta, and no variation in COI was detected within these populations. As indicated by Pongprasert (2000) and Williams et al. (2012a), the northern banded colour pattern of B. occidentalis represents a distinct group of haplotypes, and in the present study, appears to be the dominant colour pattern of B. occidentalis occurring north of 55 (Fig. 2). Bombus mckayi was described in 1902 from Alaska, and is the only taxon now considered to be part of B. occidentalis with the type locality north of 60. Williams et al. (2012a) indicate that the northern ‘‘mckayi’’ is also distinct morphologically, possessing longer pile, and noted that it is apparently disjunct from the southern forms of B. occidentalis, which is supported here (Figs. 3, 4). Interestingly, Franklin (1912) considered B. mckayi a synonym of B. occidentalis occidentalis, though listed several of the variant banded female colour patterns (i.e., his colour ‘‘variants’’ 2, 4, 5) from Alaska; his colour pattern 4 he considered to be the same as B. proximus Cresson (described from Utah). Similarly, his (Franklin 1912) concept of the taxon B. occidentalis nigroscutatus (described from California—see Online Resource 6) ranged into Alaska and the Yukon Territory [this name was originally published by Franklin in 1908 (in Fletcher and Gibson 1908) for material from the ‘‘Skagway District’’ of northern British Columbia]. Also, as discussed above, a colour pattern referred to as ‘‘Bombus occidentalis nigroscutatus Alaska type’’ (Pongprasert 2000) was found to be genetically distinct (based on COII) from southern colour-pattern individuals. Ashmead (1902) also recorded what he called B. proximus from Alaska in the same publication in which he described B. mckayi. Clearly, much variation in colour pattern exists in both northern and southern subspecies of B. occidentalis, although in the north the non-banded colour pattern either does not occur, or is very rare and is restricted to the southern part of Alaska and adjacent British Columbia (i.e., below 60N; the Chilkoot Trail/White Pass region). The present study supports this, as we also observed variation in pubescence colour in specimens from north of 60, although all had yellow pubescence on the third tergum and consistent COI. However, increased sampling for colour variation and COI in the critical latitudes (55–60N) would further help refine our understanding of the ranges of these subspecies. Milliron (1971) did not comment on the distribution of the colour patterns of occidentalis, and the problems inherent in using such shaded distribution maps (i.e., Milliron 1971), specifically for conservation purposes, have been discussed by Koch and Strange (2009). Koch and Strange (2009) used Maxent models to infer historic range maps for Bombus s. str. and the results, specifically with respect to B. occidentalis

J Insect Conserv

(and B. terricola) are worth consideration here, as they are typically much constricted in terms of areas where the species are likely to be encountered, especially along the extremes, when compared to the shaded-distribution maps of Milliron (1971). Figure 3a (particularly) and 3c (see Koch and Strange 2009) show the estimated probability of encountering B. occidentalis above 60N is low in comparison to more southern parts of the range, in contrast to the probabilities of occurrence estimated here (Figs. 3, 4). However, they later report B. occidentalis as very common in Alaska, above 60N (Koch and Strange 2012; distribution data used in present study). From the combined evidence of an ‘‘isolated’’ northern distribution for the B. o. mckayi haplotype group, which is geographically separated from the similar ‘‘banded’’ colour patterns of the other B. o. occidentalis haplotype group in the south, we can consider each subspecies as a distinct designatable unit. In fact, it seems that the southern designatable unit is the one of most concern, considered in decline by several accounts (Evans et al. 2008; Colla and Ratti 2010; Cameron et al. 2011). In contrast, B. o. mckayi, like the more northern species B. cryptarum, appears to have relatively stable populations, and both of these taxa are regularly observed in the Yukon Territory and in northern British Columbia (S. Cannings, pers. obs.), as well as in Alaska (Koch and Strange 2012). In fact, Koch and Strange (2012) indicate that B. occidentalis was the most abundant bumble bee species collected in Alaska, accounting for almost 40 % of the individuals surveyed, though other summaries of that fauna report this species at just over 10 % (Pampell et al. 2015). Koch and Strange (2012) report a high parasite load (44 % infection of Nosema bombi Fantham and Porter) in B. occidentalis, higher even than that reported previously for populations in the continental United States (Cameron et al. 2011: 37 %), although seemingly in Alaska the species is not responding as southern populations are in terms of declines. They suggest that populations of B. occidentalis in Alaska are healthy despite the presence of N. bombi infections, and that Alaskan bumble bee populations may thus provide important insights on the role of pathogens in bumble bee decline in the south (Koch and Strange 2012). The Alaskan specimens they collected are certainly of the B. o. mckayi haplotype group, so it is also worth considering that the northern subspecies may respond differently to N. bombi infections compared to the southern subspecies, or that the N. bombi strain is different and more virulent in the south—perhaps introduced. Based on recent collection efforts, the northern taxa of Bombus s. str. (i.e., Bombus o. mckayi, B. cryptarum, and seemingly B. terricola populations in the Northwest Territories, Saskatchewan, and the island of Newfoundland in Canada do not appear to be declining in number as are the more southern taxa (Colla

and Packer 2008; Grixti et al. 2009; Colla and Ratti 2010; Cameron et al. 2011). It is also likely that other stresses, in addition to high parasite loads, may further make species of Bombus s. str. more susceptible to decline. Rates of habitat loss, agricultural intensity and pesticide use differ greatly between southern Canada (see Javorek and Grant 2011) and the adjacent United States, and areas of the continent north of 60. Acknowledgments We thank the following for access to bumble bee specimens for colour pattern determination: John Ascher and Jerry Rozen (American Museum of Natural History), Claudia Copley (Royal British Columbia Museum), Matthias Buck (Royal Alberta Museum), Barb Sharanowski (University of Manitoba), Andy Bennett and Sophie Cardinal (Canadian National Collection), and Laurence Packer (York University). In addition, we thank the data contributors from the museums and academic collections outlined in Online Resource 3. Thanks are also extended to those who assisted in collecting specimens in 2013; in BC: Dawn Marks, Brittany King, Joanne Neilson and Lynn Wescott; in AB: Gary Anweiler; in SK: Danae Frier, Graham Rothwell, Sam Jaques, Nick Cairns, Shelby Stecyk, Sarah MacDonald, Allie Gallon, Leanne Heisler, Ashley Fortney, Marla Anderson. Thanks also to Scott Tarof (Earth Rangers) for support and encouragement during this study. Funding in part for this study was provided by the Earth Rangers ‘‘Bring Back the Wild’’ campaign.

References Ashmead WH (1902) Hymenoptera. Proc Wash Acad Sci 4:117–274 Bertsch A, Hrabe´ de Angelis M, Przemeck GKH (2010) A phylogenetic framework for the North American bumblebee species of the subgenus Bombus sensu stricto (Bombus affinis, B. franklini, B. monderatus, B. occidentalis & B. terricola) based on mitochondrial DNA markers (Hymenoptera: Apidae: Bombus). Beitr Ent 60:229–242 Burks BD (1951) Tribe Bombini. In: Muesebeck CFW, Krombein KV, Townes HK (eds) Hymenoptera of America north of Mexico Synoptic Catalog. United States Department of Agriculture, Monograph Number 2, Washington, pp 1247–1255 Button L, Elle E (2014) Wild bumble bees reduce pollination deficits in a crop mostly visited by managed honey bees. Agric Ecosyst Environ 197:255–263. doi:10.1016/j.agee.2014.08.004 Cameron SA, Hines HM, Williams PH (2007) A comprehensive phylogeny of the bumble bees (Bombus). Biol J Linn Soc 91:161–188 Cameron SA, Lozier JD, Strange JP, Koch JB, Cordes N, Solter LF, Griswold TL (2011) Patterns of widespread decline in North American bumble bees. Proc Natl Acad Sci USA 108:662–667 Carolan JC, Murray TE, Fitzpatrick U, Crossley J, Schmidt H, Cederberg B, McNally L, Paxton RJ, Williams PH, Brown MJF (2012) Colour patterns do not diagnose species: quantitative evaluation of a DNA barcoded cryptic bumblebee complex. PLoS ONE 7(1):e29251. doi:10.1371/journal.pone.0029251 Cockerell TDA, Porter W (1899) The New Mexico bees of the genus Bombus. Ann Mag Nat Hist (7) 4:386–393 Colla SR, Packer L (2008) Evidence for decline in eastern North American bumblebees (Hymenoptera: Apidae), with special focus on Bombus affinis Cresson. Biodivers Conserv 17:1379–1391 Colla SR, Ratti CM (2010) Evidence for the decline of the western bumble bee (Bombus occidentalis Greene) in British Columbia. Pan-Pac Entomol 86:32–34

123

J Insect Conserv Colla SR, Otterstatter MC, Gegear RJ, Thomson JD (2006) Plight of the bumble bee: pathogen spillover from commercial to wild populations. Biol Conserv 129:461–467 Colla S, Richardson L, Williams P (2011) Guide to the bumble bees of the eastern United States. USDA Forest Service/Pollinator Partnership Colla SR, Gadallah F, Richardson L, Wagner D, Gall L (2012) Assessing declines of North American bumble bees (Bombus spp.) using museum specimens. Biol Conserv 21:3585–3595. doi:10.1007/s10531-012-0383-2 Connop S, Hill T, Steer J, Shaw P (2010) The role of dietary breadth in national bumblebee (Bombus) decline: simple correlation? Biol Conserv 143:2739–2746 COSEWIC (2010) COSEWIC assessment and status report on the Rusty-patched Bumble Bee Bombus affinis in Canada. Committee on the Status of Endangered Wildlife in Canada, Ottawa Cresson ET (1863) List of the North American species of Bombus and Apathus. Proc Entomol Soc Philos 2:83–116 Dalla Torre CG (1896) Catalogus Hymenopterorum hucusque descriptorum systematicus et synonymicus. Volume X: Apidae (Anthophila). Leipzig, Engelmann De Queiroz K (2007) Species concepts and species delimitation. Syst Biol 56:879–886 Evans E, Thorp R, Jepsen S, Black SH (2008) Status review of three formerly common species of bumble bee in the subgenus Bombus: Bombus affinis (the rusty patched bumble bee), B. terricola (the yellow banded bumble bee), and B. occidentalis (the western bumble bee). The Xerces Society of Invertebrate Conservation. Portland Fletcher J, Gibson A (1908 [1909]) Entomological Record 1908. Annu Rep Entomol Soc Ont 39:99–116 Franklin HJ (1912) The Bombidae of the new world. Trans Am Entomol Soc 38:177–486 Free JB (1993) Insect pollination of crops. Academic Press, London Frison TH (1923) Systematic and biological notes on bumblebees (Bremidae; Hymenoptera). Trans Am Entomol Soc 48:307–326 Frison TH (1926) Descriptions and records of North American Bremidae, together with notes on the synonymy of certain species (Hymenoptera). Trans Am Entomol Soc 52:129–145 Goulson D, Darvill B (2004) Niche overlap and diet breadth in bumblebees: are rare species more specialized in their choice of flowers? Apidologie 35:55–63 Goulson D, Hanley ME, Darvill B, Ellis JS, Knight ME (2005) Causes of rarity in bumblebees. Biol Conserv 122:1–8 Goulson D, Hanley ME, Darvill B, Ellis JS (2006) Biotope associations and the decline of bumblebees (Bombus spp.). J Insect Conserv 10:95–103 Goulson D, Lye GC, Darvill B (2008) Decline and conservation of bumble bees. Annu Rev Entomol 53:191–208 Goulson D, Nicholls E, Botı´as C, Rotheray EL (2015) Bee declines driven by combined stress from parasites, pesticides, and lack of flowers. Science. doi:10.1126/science.1255957 Green DM (2005) Designatable units for status assessment of endangered species. Conserv Biol 19:1813–1820. doi:10.1111/ j.1523-1739.2005.00284.x Greene JW (1858) Descriptions of several new hymenopterous insects from the north west coast of America. Ann Lyc Nat Hist NY 7:11–12 Grixti JC, Wong LT, Cameron SA, Favret C (2009) Decline of bumble bees (Bombus) in the North American Midwest. Biol Conserv 142:75–84 Handlirsch A (1888) Die Hummelsammlung des k. k. naturhistorischen Hofmuseums. Ann Naturhist Mus Wien 3:209–250 Hebert PDN, Cywinska A, Ball SL, deWaard JR (2003a) Biological identifications through DNA barcodes. Proc R Soc Lond B Biol Sci 270:313–322

123

Hebert PDN, Ratnasingham S, deWaard JR (2003b) Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc R Soc Lond B Biol Sci 270:S96– S99 Hebert PDN, Penton EH, Burns JM, Janzen DH, Hallwachs W (2004) Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proc Natl Acad Sci USA 101:14812–14817 Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. Int J Climatol 25:1965–1978 Hobbs GA (1968) Ecology of species of Bombus (Hymenoptera: Apidae) in southern Alberta. VII. Subgenus Bombus. Can Entomol 100:156–164 Huang J, Wu J, An J, Williams PH (2015) Newly discovered colourpattern polymorphism of Bombus koreanus females (Hymenoptera: Apidae) demonstrated by DNA barcoding. Apidologie 46:250–261. doi:10.1007/s13592-014-0319-9 Javorek SK, Grant MC (2011) Trends in wildlife habitat capacity on agricultural land in Canada, 1986–2006. Canadian Biodiversity: Ecosystem Status and Trends 2010, Technical Thematic Report No. 14. Canadian Councils of Resource Ministers. Ottawa http:// www.biodivcanada.ca/default.asp?lang=En&n=137E1147-1 Kearns CA, Thomson JD (2001) The natural history of bumblebees. A sourcebook for investigations. University Press of Colorado, Boulder Kerr JT, Pindar A, Galpern P, Packer L, Potts SG, Roberts SM, Rasmont P, Schweiger O, Colla SR, Richardson LL, Wagner DL, Gall LF, Sikes DS, Pantoja A (2015) Climate change impacts on bumblebees converge across continents. Science 349(6244):177–180 Koch JB, Strange JP (2009) Constructing a species database and historic range maps for North American bumblebees (Bombus sensu stricto Latreille) to inform conservation decisions. Uludag Bee J 9:97–108 Koch JB, Strange JP (2012) The status of Bombus occidentalis and B. moderatus in Alaska with special focus on Nosema bombi incidence. Northwest Sci 86:212–220 Koch J, Strange J, Williams P (2012) Guide to bumble bees of the western United States. USDA Forest Service/Pollinator Partnership Laverty TM, Harder LD (1988) The bumble bees of eastern Canada. Can Entomol 120:965–987 Lecocq T, Lhomme P, Michez D, Dellicour S, Valterova I, Rasmont P (2011) Molecular and chemical characters to evaluate species status of two cuckoo bumblebees: Bombus barbutellus and Bombus maxillosus (Hymenoptera, Apidae, Bombini). Syst Entomol 36:453–469 Lecocq T, Brasero N, De Meulemeester T, Michez D, Dellicour S, Lhomme P, de Jonghe R, Valterova´ I, Urbanova´ K, Rasmont P (2014) Integrative taxonomic approach to assess the taxonomic status of Corsican bumblebees: implication for conservation. Anim Conserv. doi:10.1111/acv.12164 Lutz FE, Cockerell TDA (1920) Notes on the distribution and bibliography of North American bees of the families Apidae, Meliponidae, Bombidae, Euglossidae, and Anthophoridae. Am Mus Nat Hist 42:491–641 Meeus I, Brown MJ, De Graaf DC, Smagghe GUY (2011) Effects of invasive parasites on bumble bee declines. Conserv Biol 25:662–671 Michener CD (2007) The bees of the world, 2nd edn. Johns Hopkins University Press, Baltimore Milliron HE (1971) A monograph of the western hemisphere bumblebees (Hymenoptera: Apidae; Bombinae) I. The genera Bombus and Megabombus subgenus Bombias. Mem Entomol Soc Can 82:1–80

J Insect Conserv Milliron HE (1973) A monograph of the western hemisphere bumblebees (Hymenoptera: Apidae; Bombinae) II. The genus Megabombus subgenus Megabombus. Mem Entomol Soc Can 89:81–237 Milliron HE, Oliver DR (1966) Bumblebees from northern Ellesmere Island, with observations on usurpation by Megabombus hyperboreus (Scho¨nh.) (Hymenoptera: Apidae). Can Entomol 98:207–213 Mitchell TB (1962) Bees of the eastern United States. Volume II. N C Agric Exp Stn Tech Bull 152:1–557 Murray TE, Fitzpatrick U, Brown MJF, Paxton RJ (2008) Cryptic species diversity in a widespread bumble bee complex revealed using mitochondrial DNA RFLPs. Conserv Genet 9:653–666 Owen RE, Plowright RC (1980) Abdominal pile color dimorphism in the bumble bee Bombus melanopygus. J Hered 71:241–247 Owen RE, Plowright RC (1988) Inheritance of metasomal pile colour variation in the bumble bee Bombus rufocinctus Cresson (Hymenoptera: Apidae). Can J Zool 66:1172–1178 Owen RE, Whidden TL (2013) Discrimination of the bumble bee species Bombus occidentalis Greene and B. terricola Kirby by morphometric, colour and RAPD variation. Zootaxa 3608:328–344 Owen RE, Otterstatter MC, Cartar RV, Farmer R, Colla SR, O’Toole N (2012) Significant expansion of the distribution of the bumble bee Bombus moderatus (Hymenoptera: Apidae) in Alberta over twenty years. Can J Zool 90:133–138 Pampell R, Sikes D, Pantoja A, Holloway P, Knight C, Ranft R (2015) Bumble bees (Hymenoptera: Apidae: Bombus spp.) of interior Alaska: species composition, distribution, seasonal biology, and parasites. Biodivers Data J 3:5085. doi:10.3897/ BDJ.3.e5085 Phillips SJ, Dudı´k M (2008) Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography 31:161–175 Phillips SJ, Dudik M, Schapire RE (2004) A maximum entropy approach to species distribution modeling. In: Greiner R, Schuurmans D (eds) Proceedings of the 21st international conference on machine learning. ACM Press, New York, pp 655–662 Phillips SJ, Anderson RP, Schapire RE (2006) Maximum entropy modeling of species geographic distributions. Ecol Model 190:231–259 Plowright RC, Stephen WP (1980) The taxonomic status of Bombus franklini (Hymenoptera: Apidae). Can Entomol 112:475–479 Pongprasert W (2000) Phylogenetic relationships of the bumblebees (Apidae: Hymenoptera) in the pacific northwest of America inferred from mitochondrial cytochrome Oxidase gene sequences. Dissertation, Oregon State University Rao S, Stephen WP (2007) Bombus (Bombus) occidentalis (Hymenoptera: Apiformes): in decline or recovery? Pan-Pac Entomol 83:360–362 Rasmont P, Coppe´e A, Michez D, De Meulemeester T (2008) An overview of the Bombus terrestris (L. 1758) subspecies (Hymenoptera: Apidae). Ann Soc Entomol Fr 44:243–250 Ratnasingham S, Hebert PDN (2007) BOLD: the Barcode of Life Data System (www.barcodinglife.org). Mol Ecol Notes 7:355–364 Richards KW (1973) Biology of Bombus polaris Curtis and B. hyperboreus Schonherr at Lake Hazen, Northwest Territories (Hymenoptera: Bombini). Quaest Entomol 9:115–157 Sachman-Ruiz B, Narva´ez-Padilla V, Reynaud E (2015) Commercial Bombus impatiens as reservoirs of emerging infectious diseases

in central Me´xico. Biol Invasions 17:2043–2053. doi:10.1007/ s10530-015-0859-6 Scholl A, Obrecht E, Owen RE (1990) The genetic relationship between Bombus moderatus Cresson and the Bombus lucorum auct. species complex (Hymenoptera: Apidae). Can J Zool 68:2264–2268 Scriven JJ, Woodall LC, Tinsley MC, Knight ME, Williams PH, Carolan JC, Brown MJF, Goulson D (2015) Revealing the hidden niches of cryptic bumblebees in Great Britain: implications for conservation. Biol Conserv 182:126–133 Sheffield CS, Hebert PDN, Kevan PG, Packer L (2009) DNA barcoding a regional bee (Hymenoptera: Apoidea) fauna and its potential for ecological studies. Mol Ecol Resour 9(suppl 1):196–207 Smith F (1861) Descriptions of new genera and species of exotic Hymenoptera. J Entomol 1:146–155 Stephen WP (1957) Bumble bees of western North America (Hymenoptera: Apoidea). Agric Exp Stn Oregon State Coll Tech Bull 40:1–163 Thorp RW (2005) Species profile: Bombus franklini. In: Shepherd MD, Vaughan DM, Black SH (eds) Red list of pollinator insects of North America. CD-ROM Version 1. The Xerces Society for Invertebrate Conservation, Portland Thorp RW, Horning DS Jr, Dunning LL (1983) Bumble bees and cuckoo bumble bees of California. Bull Calif Insect Surv 23:1–79 Titus ESG (1902) Colorado Bombidae. Can Entomol 34:37–44 Williams PH (1998) An annotated checklist of bumble bees with an analysis of patterns of description (Hymenoptera: Apidae, Bombini). Bull Nat Hist Mus (Entomol) 67:79–152 Williams PH (2007) The distribution of bumblebee colour patterns world-wide: possible significance for thermoregulation, crypsis, and warning mimicry. Biol J Linn Soc 92:97–118 Williams PH, Osborne JL (2009) Bumblebee vulnerability and conservation world-wide. Apidologie 40:367–387 Williams P, Colla S, Xie Z (2009) Bumble vulnerability: common correlates of winners and losers across three continents. Conserv Biol 23:931–940 Williams PH, An J, Huang J (2011) The bumblebees of the subgenus Subterraneobombus: integrating evidence from morphology and DNA barcodes (Hymenoptera, Apidae, Bombus). Zool J Linn Soc Lond 163:813–862. doi:10.1111/j.1096-3642.2011.00729.x Williams PH, Brown MJF, Carolan JC, An J, Goulson D, Aytekin AM, Best LR, Byvaltsev AM, Cederberg B, Dawson R, Huang J, Ito M, Monfared A, Raina RH, Schmid-Hempel P, Sheffield CS, Sˇima P, Xie Z (2012a) Assessing cryptic species of the bumblebee subgenus Bombus s. str. world-wide with COI barcodes (Hymenoptera: Apidae). Syst Biodivers 10:21–56. doi:10.1080/14772000.2012.664574 Williams PH, An J, Brown MJF, Carolan JC, Goulson D, Huang J, Ito M (2012b) Cryptic bumblebee species: consequences for conservation and the trade in greenhouse pollinators. PLoS ONE 7(3):e32992 Williams PH, Byvaltsev A, Sheffield CS, Rasmont P (2013) Bombus cullumanus—an extinct European bumblebee species? Apidologie 44:121–132 Williams PH, Thorp R, Richardson L, Colla S (2014) Bumble bees of North America. Princeton University Press, Princeton Zhang H, Huang J, Williams PH, Vaissie`re BE, Zhou Z, Gai Q, Dong J, An J (2015) Managed bumblebees outperform honeybees in increasing peach fruit set in China: different limiting processes with different pollinators. PLoS ONE 10(3):e0121143

123