The Rusty-patched Bumblebee (Bombus affinis) was at one time among the more common and widespread bumblebees in eastern North America. However, this is no longer the case. The first decade of the 21st century saw growing concern about declining pollinator populations in general in several regions of the world, with particular attention focused on bees, and much of the available data on declining bee populations has focused on bumblebees. Although populations of some bumblebee species appear to be robust, many others have apparently gone extinct in recent years or suffered dramatic declines. Bombus affinis is one of several North American bumblebee species that have experienced clear declines. Colla and Packer (2008) documented an impoverishment of the bumblebee community in general in southern Ontario (Canada) between the early 1970s and the first decade of the 21st century and found that B. affinis, in particular, declined dramatically in abundance not only in southern Ontario but throughout its native range. There is evidence of declines in three other North American bumblebees as well (all four belong to the subgenus Bombus): B. franklini and B. occidentalis in the west and B. terricola in the east. Bombus franklini, which had a historically small geographic distribution, is thought to be at the brink of extinction (or possibly extinct). Bombus affinis, B. terricola, and B. occidentalis have much larger historical ranges, but have disappeared from numerous sites where they were previously common. (Colla and Packer 2008 and references therein)

The Xerces Society, an organization dedicated to invertebrate conservation, is an excellent resource for more information about Bombus affinis and about broader bumblebee conservation issues.

The rusty patched bumble bee (Bombus affinis) is named after the small rust-colored patch visible on worker bees' abdomens. Worker bees also have yellow on the first and rear half of the second abdominal segment. The remaining abdominal segments are black. Queens and males resemble workers except that they lack the rust-colored patch. Queens are larger in size than workers and have a small central bare patch on the thorax.

This bee was historically found throughout the eastern and upper midwestern United States. However, surveys between 2003 and the present have found only a small number of this species in Illinois and Wisconsin.

Global Range: (20,000-200,000 square km (about 8000-80,000 square miles)) The historic range was at least from central Maine across southernmost parts of Quebec and Ontario to most of Minnesota and a small part of the Dakotas extending south to New Jersey, and south mostly (but not entirely) in the mountains into Georgia eastward, but only to northern Illinois westward (Evans et al., 2008; Colla and Packer, 2008). Colla and Packer (2008) found it at one site in Ontario in 2004-2006. There reportedly was also a collection near St. Louis, Missouri in the 1990s (Michael Arduser pers. comm. to D. Schweitzer, September 2008) but that should clearly be treated as historic now. This bumblebee was not documented from the coastal plain south of New Jersey. In Quebec, this species has been confirmed from Gatineau and Montréal (COSEWIC 2010).
There is little basis for defining any current range, but since 2003 records have been from extreme southern Ontario to the Indianapolis area, northern Illinois and Wisconsin with records in Ontario (S. Colla) and Indiana (E. Day) in 2009 according to Leif Richardson. The Xerces Society website (http://www.xerces.org/rusty-patched-bumble-bee/) as of December (2008) indicates that there has been a record in Wisconsin and a few in Illinois since 2003, but some of these occurrences could easily have died out since. Sarina Jepsen of the Xerces Society (email to Nicole Capuano, September 2009) reports records near Peoria and Rockford, Illinois in 2009 (also see Berenbaum et al., 2010). Other recent records compiled by Evans et. al. (2008) one in Maryland in 2002 but no more in 2003-2007(sample size >1000). In Indiana they report B. affinis was 12% of a 2001 sample, 0.004% in 2002 and none in 2003, and no records for that state are known since then. Similarly in Iowa the last collections were in 2000 and 2001. One turned up in northern Illinois in 2003-2004 and in Wisconsin in 2006. More recent efforts in Illinois found the species locally common at one site in McHenry County in 2007 and singletons were collected in three other counties that year and one individual in 2008. Surveys in Virginia in 2002-2005 and western North Carolina 2002-2007 found none, and it had been regular through 2001 at the North Carolina site. The last record in Great Smoky Mountains National Park was in 1997, and a 2002 bee survey produced none. Richardson (2008) noted only the Ontario population as "known extant" but there would appear to have still been at least one or two in Illinois at that time. There is no basis to determine any plausible current range extent. Considering the rapidity of decline since 2002, for this subgenus records before 2006 probably should be considered historic, but nevertheless this species seems to be turning up repeatedly, but not commonly, from 2003 to 2009 in a limited area of Illinois, Wisconsin, Minnesota and Ontario, which suggests a core area, perhaps two areas, of persistence. It is probably localized even within this core area.

Sarina Jepson in an email in June 2011 to Dale Schweitzer indicated that Xerces has "recent" photo-documented citizen scientist records from Dane Co., Wisconsin; Hennepin and Washington Cos., Minnesota; Rockford in Winnebago Co., Illinois, which are more or less in or near the recent core area; but also two much farther east in Barnstable Co., Massachusetts {=Cape Cod] and at Ridley Creek State Park in Delaware Co., Pennsylvania (2006). The latter is just outside Philadelphia in a very urban setting (D. Schweitzer).

occurs (regularly, as a native taxon) in multiple nations

Canada

Origin: Native

Regularity: Regularly occurring

Currently: Present

Confidence: Confident

Type of Residency: Year-round

United States

Origin: Native

Regularity: Regularly occurring

Currently: Present

Confidence: Confident

Type of Residency: Year-round

The historical range of Bombus affinis extends from southern Ontario and southwestern Quebec (but apparently not New Brunswick, contrary to apparently erroneous literature references) in Canada south to Georgia and west to the Dakotas in the United States, with occurrences in the southern portion of the range limited mainly to higher elevations (Thorp and Shepherd 2005; Colla 2010).

Bombus affinis queens and workers differ slightly in coloration (an uncommon feature in bumble bees). Other than size, the primary difference between queens and workers is the presence of a medial rusty patch on the second abdominal segment of the worker. The hairs of workers are entirely black on the head, the bottom of the thorax, and in large part on the legs. The rest of the thorax has mostly yellow hair, with a black area in the middle of the thorax. Hairs are entirely yellow on the first two abdominal segments and black on the rest of the abdomen. On workers, there is more black intermixed with yellow near the base of the wings, forming something of a band between the wings, and black hairs extend posteriorly in a narrow "V" that partially bisects the yellow on the scutellum. The second abdominal segment has a rusty reddish patch centrally, with yellow hairs around the edges of the segment. Bombus affinis males have hairs largely black on the head, but with a few pale hairs intermixed near the top of the head. Black hairs sometimes form an obscure band across the middle of the thorax, but the hair on the thorax is otherwise largely pale yellowish. The first two abdominal segments have pale yellow hair. The hair on the rest of the abdominal segments is black. (Evans et al. 2008)

Lectotype for Bombus affinis Cresson, 1863
Collection: Smithsonian Institution, National Museum of Natural History, Department of Entomology
Sex/Stage: Male; Worker;
Preparation: Pinned
Locality: Canada, Unknown

Bombus affinis has been collected in a wide variety of habitats including mixed farmland, sand dunes, marshes, and both urban and wooded areas (Colla 2010).

Macfarlane (1974, cited in Colla and Packer 2008) observed B. affinis visiting at least 65 plant genera. These bees have been observed biting holes (i.e. nectar-robbing) in flowers with long corolla tubes such as Jewelweed (Impatiens capensis), Yellow Toadflax (Linaria vulgaris) (R. Gegear pers. comm., cited in Colla and Packer 2008), and Cow Vetch (Vicia cracca) (Harder 1983, cited in Colla and Packer 2008).

Bombus affinis is one of two host species for the socially parasitic Bombus (Psithyrus) ashtoni (the other host being the also declining B. terricola) (Fisher 1983). In 2010, Colla reported that B. ashtoni had not been observed in a decade. Bombus species in the subgenus Psithyrus lack pollen-collecting corbiculae and rely on the host workers for the rearing of reproductives. Sladen (1912, cited in Fisher 1983a) reported that the host queen was always killed or displaced by the invading Psithyrus, while Plath (1934, cited in Fisher 1983a) found that in the case of the two North American Psithyrus species he studied (B. ashtoni and B. citrinus) the queen was seldom killed. Subsequent investigations showed that B. ashtoni never kills queens of either of its host species, B. affinis and B. terricola (Fisher 1983a). Fisher found that B. ashtoni is incapable itself of suppressing ovarian development in queenless workers of B. affinis. An alternative strategy to physiological suppression of oogenesis is behavioral domination. Fisher reported that B. ashtoni females were often seen mauling workers, i.e., grasping and pulling them underneath the body as if to sting, but not actually doing so. This behavior only occurred in colonies which had no queen or had a queen that had lost her dominance (ovarian development of workers is normally supressed by the queen). Physical dominance by B. ashtoni females would not prevent oogenesis, but could eliminate or minimize egg laying by fecund workers. Fisher presented data suggesting that females of B. ashtoni benefit by the supression of host worker reproduction by the host queen. At the same time, B. ashtoni females prevent Bombus affinis males and queens from developing by selectively eating eggs and ejecting larvae. The non-aggressive invasion strategy used by B. ashtoni is quite distinct from that of other Psithyrus species, such as B. citrinus, which according to Fisher always kills or displaces its B. impatiens host queen. These other species presumably possess themselves the physiological or behavioral means of eliminating or reducing the frequency of egg laying by host workers in the nests they invade and thus do not need to rely on the host queen to control the workers. (Fisher 1983a) Fisher (1983b) investigated host nest finding by B. ashtoni. These bees search for nests within one to two weeks of host queen emergence at a time prior to emergence of the first worker brood and therefore cannot use odor trails of workers to recognize nests. In laboratory experiments, Fisher showed that B. ashtoni females can recognize host nest odor without actual contact with the nest or with worker-laid trails, successfully distinguishing nests of B. affinis and B. terricola from those of B. bimaculatus and from controls consisting only nest material.

Martin et al. (2010) investigated the cuticular hydrocarbon cues of 14 European Bombus species, including 5 socially parasitic species ("cuckoo bees" in the subgenus Psithyrus) (B. affinis, B. ashtoni, or any other North American species, were not among the species studied, but the general conclusions of the analysis by Martin et al. likely apply more broadly). They found that found that bumblebees possess species-specific alkene positional isomer profiles that are stable over large geographic regions and are mimicked by three host-specific Psithyrus parasites. In three host-cuckoo associations where mimicry is poor, possibly as a result of recent host shifts, these cuckoos produce dodecyl acetate a known chemical repellent that allows the cuckoos to invade their host colonies. Thus, various Psithyrus species may use both mimicry and repellents to invade host colonies.

Microscopic endoparasites recorded infecting B. affinis include Sphaeruluria bombi (a nematode infecting 10% of overwintered queens) and the apicomplexan protozoan Apicystis bombi (Neogregarinida: Ophrocystidae) (Macfarlane et al. 1995, cited in Colla 2010). Other parasites that are known to infect sympatric Bombus species are Nosema bombi (Microsporidia: Nosematidae) and the trypanosome protozoan Crithidia bombi (Kinetoplastea: Trypanosomatidae), both of which may be acquired at flowers via fecal transmission (Colla et al. 2006). Nosema bombi has recently been found infecting B. affinis (Cameron et al. 2011), but infection of B. affinis by C. bombi has apparently not yet been documented (possibly because of the relative recency of the presumed introduction of this parasite from Europe in combinatuon with the rarity of B. affinis in recent years) (Colla 2010). Several species of parasitoid conopid flies (Diptera: Conopidae) attack foraging bumblebees on the wing and lay their eggs inside the bee’s abdomen. (Gillespie 2010 and references therein)

Conopids and C. bombi can affect colony reproduction and worker foraging behavior. Nosema bombi may reduce colony fitness and worker survival. These parasites could affect local abundance of bumblebee populations and C. bombi and N. bombi have been tentatively implicated in the overall decline of bumblebees. In a study in Massachusetts, Gillespie (2010) found a high level of parasitism of bumblebees by C. bombi, N. bombi, and conopid flies (although no B. affinis were encountered in this study). (Gillespie 2010 and references therein)

Bombus affinis has been shown to be an excellent pollinator of cranberry and also to pollinate other important crops such as plum, apple, alfalfa, and onion for seed production. Evans et al. provide a long list of wild plants known to be visited by B. affinis. (Evans et al. 2008 and references therein).

Bombus affinis Cresson: Apidae (Bombini), Hymenoptera
(observations are from Reed, Graenicher, Betz et al., Williams, Arnold, Clinebell, Schoen, Ne'eman et al., Larson & Barrett, Macior, Conger, Lindsey, Cane et al., Reader, Cane & Schiffhauser, and Costelloe)

Apiaceae: Thaspium barbinode (Lnd); Asclepiadaceae: Asclepias hirtella [plup sn] (Btz), Asclepias meadii [plup sn] (Btz); Asteraceae: Arctium lappa sn cp (Gr), Aster drummondii sn cp (Gr), Aster lanceolatus (Re), Aster ontarionis (Re), Aster oolentangiensis (Re), Aster puniceus sn cp (Gr), Aster sericeus (Re), Cirsium altissimum sn (Gr), Crepis tectorum (Re), Echinacea purpurea (Cl), Eupatoriadelphus purpureus sn (Gr), Eupatorium perfoliatum sn cp (Gr), Helianthus giganteus sn cp (Gr), Liatris pycnostachya (Cl), Oligoneuron rigidum (Re), Solidago canadensis sn cp (Gr, Re), Solidago nemoralis (Re), Solidago speciosa (Re); Boraginaceae: Onosmodium molle (Wm); Campanulaceae: Lobelia siphilitica fq (Mc); Caprifoliaceae: Diervilla lonicera (Sch); Ericaceae: Andromeda glaucophylla (Rd), Chamaedaphne calyculata (Rd), Vaccinium macrocarpon sn (CS), Vaccinium stamineum (Cn); Fabaceae: Amorpha canescens (Re), Dalea purpurea (Re), Dalea villosa (Re), Melilotus alba (Re); Fumariaceae: Dicentra cucullaria sn prf sn@prf fq (Mc); Grossulariaceae: Ribes hirtellum sn (Gr); Lamiaceae: Agastache foeniculum (Re), Monarda fistulosa (Re), Nepeta cataria (Re), Pycnanthemum virginianum (Re), Stachys aspera sn/cp (Cng), Teucrium canadense sn/cp (Cng); Melastomataceae: Rhexia virginica cp (LBt); Parnassiaceae: Parnassia glauca sn (Gr); Primulaceae: Dodecatheon meadia cp (Mc); Ranunculaceae: Aquilegia canadensis sn prf sn@prf (Mc), Delphinium tricorne prf sn@prf fq np (Mc); Sarraceniaceae: Sarracenia purpurea cp (NNE); Scrophulariaceae: Linaria vulgaris fq (Arn, Mc), Pedicularis canadensis sn (Mc), Pedicularis lanceolata cp (Mc, Cst), Penstemon grandiflorus (Re); Solanaceae: Solanum dulcamara cp (Mc); Verbenaceae: Verbena stricta (Re)

Note: For many non-migratory species, occurrences are roughly equivalent to populations.

Estimated Number of Occurrences: 1 - 300

Comments: About 10-15 years ago there were probably at least thousands of occurrences with millions of hives, now by far most are gone. A recent survey from Ontario to Georgia in 2004-2006 (Colla and Packer, 2008) produced one specimen in Ontario. Except that there has been a huge decline in abundance and probably range in recent decades, nothing else can be stated with any confidence about a current range. The species has become undetectable and may be extirpated from most of its range. There is no basis to suggest the current number of occurrences. Since 2005 about half a dozen occurrences have been documented, but some, especially the Maryland one, probably have died out.

Unknown

The flight season of Bombus affinis is longer than that of most other North American bumblebees and it visits numerous plant genera in diverse habitats (Colla 2010).

Bombus affinis have relatively short tongues. On average, the tongues of workers are around 5 to 7 mm in length (some other bumblebee species have tongues as long as 10 mm). Their short tongues make them unable to access the nectar in flowers with deep tubes, although they sometimes use their mandibles to chew holes in the bottom of these flowers to access the nectar from the outside of the flower, thus cheating the flower of pollination. (Evans et al. 2008 and references therein)

Bombus affinis typically nests underground in abandoned rodent burrows located from six to eighteen inches below the surface. Occasionally nests are constructed on the surface in areas such as clumps of grass on the ground. Thus, nesting sites may be limited by the abundance of rodents and the presence of undisturbed grassland. (Evans et al. 2008 and references therein).

Brood cells and honey pots are made of wax produced by the queen and workers. Like other bumblebees, B. affinis have an annual life cycle (i.e., 1 year = 1 generation). Mated queens emerge from diapause in the spring to begin feeding and searching for potential nest sites to initiate new colonies. The queen collects nectar and pollen from flowers to support the production of her eggs (which are fertilized by sperm she has stored since mating the previous fall) and produces a brood of workers. In the early stages of colony development, the queen is responsible for all food collection and care of the young. As the colony grows, workers take over the duties of food collection, colony defense, and care of the young. The queen then remains within the nest and spends most of her time laying eggs. As the summer progresses, the colony reaches maximum worker production and begins producing males and potential queens (queen production is dependent on access to sufficient quantities of pollen). These reproductive individuals leave the colony and mate. After mating, young queens enter diapause and overwinter. The males and workers decline as fall approaches until they die in the winter. (Evans et al. 2008; Colla 2010).

The largest B. affinis colony on record produced 2,100 individuals in captivity (MacFarlane 1974, cited in Colla 2010), but in the wild colonies are much smaller (Colla 2010). More typically, B. affinis colonies consist of a queen and between 50 and 400 workers at their peak (Evans et al. 2008).

Bombus affinis is a "pollen-storer", meaning the larvae live in cells and are fed individually by adults opening the brood clump as the larvae develop. Pollen-storing adults emerge relatively equal in size compared to "pocket-making" bumble bee species, which tend to produce workers that vary greatly in size due to unequal food distribution within the brood clumps during development. (Colla 2010)

Bombus affinis was first described by Cresson in 1863. Although the taxonomy of some bumble bee species is controversial, the status of B. affinis as a distinct, valid species is not (Cameron et al. 2007).

Because of the particular haplodiploid mode of sex determination characteristic of bees, which normally yields diploid females and haploid males but results in the production of non-viable diploid males when allelic diversity is low (and hence homozygosity is high), effective population size relative to census size is much reduced and bee populations in general may therefore be especially vulnerable to extinction as population size shrinks (Zayed and Packer 2005).

Barcode of Life Data Systems (BOLDS) Stats
Public Records: 0
Species: 4
Species With Barcodes: 1

Rounded Global Status Rank: G1 - Critically Imperiled

Reasons: This was a very widespread bumblebee in eastern and central North America, but like three others of its subgenus it has suffered a severe to catastrophic decline starting in or just after the late 1990s. Williams et al. (2008) assign it the highest decline measure of any Canadian Bombus, also higher than any Chinese species, but lower than most British species. Williams and Osbourne (2009) conclude the species merits critically endangered by IUCN standards. While there are no actual historic population data, the decline in numbers and area of occupancy has almost certainly been over 99% and reduction in range extent apparently around 90-95% over a period of less than a decade. B. afffinis may be well on the way to extinction, perhaps within the current decade. Historical collection records and the literature indicate that Bombus affinis was still a common bumblebee in the mid or even late 1990s. It is also noteworthy that while the closely related B. terricola is still persisting fairly widely in Vermont, B. affinis has not turned up in recent efforts there.

If in fact pathogen spillover is the main cause, very widespread extirpations or even extinction of this and one or two other species of subgenus Bombus are likely. On the other hand the species is persisting in a limited core area near the middle of its range, and there could be scattered undetected refugia and possibly some resistance could evolve as in related European species. There is no conservation status rank that really captures this sort of extreme decline of recently common and widespread species and several important ranking factors are unknown. GU would be a very reasonable rank. Otherwise, the least misleading rank would be G1, or a combination rank including G1, but this rank and IUCN's critically endangered (not yet formally assigned) are not interchangeable. NatureServe's rank calculator produces a rank of G1G2 which is assigned for now with some reservartions.

Environmental Specificity: Broad. Generalist or community with all key requirements common.

Comments: This was a widespread rather common bumblebee that occurred in a variety of habitats including urban areas, which could possibly be refugia now.

Other Considerations: A serious negative factor could be the limited occurrence of this bumblebee from very cold regions which seem to be providing refugia for Bombus (Bombus) terricola and B. (B.) occidentalis. However B. affinis does not appear to be persisting in places like New England and in the Appalachians where they formerly overlapped and B. terricola persists, and in some places seems to be recovering. However only B. terricola occurred widely in Canada. B. (B.) affinis barely entered extreme southern Ontario and Quebec. Neither species apparently persists in what are among the warmest parts of the range of B. terricola in southernmost Ontario or Ohio. The persistence, so far anyway, of B. (B.) occidentalis in very cold mountain and northern regions of western North America, but not in warmer Pacific coastal and valley areas where it was also common, also suggests species of this subgenus might be more vulnerable in milder climates. Instead limited refugia for B. affinis are in and near northern Illinois, apparently to some extent in relatively urbanized areas.

Canada

Rounded National Status Rank: N1 - Critically Imperiled

United States

Rounded National Status Rank: NNR - Unranked

Global Short Term Trend: Decline of 50-70%

Comments: The decline from 2002 to 2009 may be around 99% even in range extent (see Evans et al., 2008 and range comments), and there is very little doubt that number of individuals since 2002 has declined by well over 99%. More conservatively the data in Colla and Packer (2008) suggest about a 95% decline in range extent based on their survey of historic and expected sites from Maine and Ontario to Georgia, which produced only a single specimen from Ontario. Williams et al. (2009) assigned this one a substantially higher decline index than any other species in Ontario. Evans et al. (2008) reported it at a few other places in the upper Midwest in 2006-2008, and Sarina Jepsen of the Xerces Society (email to Nicole Capuano, September 2009) reports records near Peoria and Rockford, Illinois in 2009. Documented decline noted in New York by Committee on the Status of Pollinators (2007) book and the species is on the Xerces Society Red List of pollinators. This species was common until about 2001 and has had a decline in range and relative numbers found of substantially more than 90%, probably well over 99% in numbers, since then.

Global Long Term Trend: Decline of >90%

Comments: Same as short term. There is little evidence of decline prior to about 2002.

In the 1970s, Bombus affinis was among the more common and widespread bumblebee species in eastern North America. Dramatic declines were noted by the mid-1990s in both Canada and the United States. In Canada, extensive targeted searches from 2005 to 2009 detected just three individuals (one in 2005 and two in 2009). Similar population crashes have been observed in the Colla 2010 and references therein)

According to an analysis by Cameron et al. (2011), the relative abundances of four North American bumblebee species have declined by up to 96% and their geographic ranges have contracted by an estimated 23% to 87%, some within a span of two decades. In the recent large-scale bumblebee survey carried out by Cameron et al., which included the capture and identification of nearly 17,000 individuals (most of which were released), only 22 B. affinis individuals were found, and although this species was once found throughout the eastern United States and northern Midwest, individuals were detected at just three locations in Illinois and one in Indiana (yielding an estimated range reduction of 87%).

Grixti et al. (2009) found that bumblebee species richness in Illinois (U.S.A.) declined substantially during the middle of the century (1940 to 1960). Four species were locally extirpated: B. borealis, B., B. terricola and B. variabilis. The ranges of B. affinis, B. fraternus, B. pensylvanicus and B. vagans have also decreased dramatically in Illinois. The major decline in the Illinois bumblebee fauna coincided with large-scale agricultural intensification in Illinois, suggesting one likely factor driving bumblebee declines.

Comments: Like other severely declining bumblebees the main cause is thought to be pathogen spillover of an especially virulent strain of the imported microsporidian (Nosema bombi) and an imported protozoan parasite (Crithidia bombi) from domesticated bumblebees (Bombus impatiens, B. occidentalis) that were reared in Europe and returned to the USA for greenhouse pollination (e.g. Committee on Status of Pollinators, 2007, Colla and Packer, 2008, Evans et al., 2008 and references reviewed in all). While effects of N. bombi are sublethal and sometimes quite mild for B. impatiens and some other bumblebees, impacts to subgenus Bombus appear to be severe. The mite Locustacarum buckneri and the honeybee deformed wing virus may also be contributing to decline. There are probably other threats in some places such as land use changes and other forms of habitat loss, changes in nectar flora etc. A potentially serious threat might be novel pesticides especially new persistent neonicotinoids (Colla and Packer, 2008), but more evidence is needed, and this seems inconsistent with findings that some bumblebee species are stable or increasing. The consensus seems to be that the above pathogens and parasites, perhaps especially an introduced Nosema strain, are probably the main causes of the very recent drastic decline. However, Sokolova et al. (2010) suggest there may also be very closely related native North American Nosema. In a plausible worst case scenario some or all North American species of subgenus Bombus could be extinct within a decade or two, and one (B. franklini) may be already. Leif Richardson (pers. comm. to N.Capuano, January 2010) suggests this "bee is adapted to a fairly narrow range of climate conditions-seasonality, precipitation, temperature" so impact from climate change is more plausible for this species than most North American bumblebees (see Williams and Osborne, 2009). Its persistence near the middle of its overall range would be very consistent narrow tolerances, and the range is rather narrow westward. However, B. affinis ranged over a substantial range of climates east of the Appalachians, from cool central Maine, about 45° North, where the July mean is 21° C, and the mean is 20° C or higher for only about two months, to the the hot summers of southern New Jersey to Washington D.C. (38-39° North) where July means are 25-27°C, and the mean is 20°C or higher for over 3.5 months, and occurred to some extent in the North Carolina Piedmont around 36° North where the summer is yet a month longer. However most of the southern range was in the mountains with long, cool, wet summers.

The reasons for the sudden decline of Bombus affinis, a previously common species throughout its large range, are unknown. It has been suggested that, along with other vulnerable North American species in the subgenus Bombus (B. occidentalis, B. terricola, and the possibly extinct B. franklini), the species has suffered from introduced diseases transmitted from managed bumblebee colonies used for greenhouse pollination. In addition, habitat loss and the widespread use of certain pesticides likely represent significant threats. (Evans et al. 2008; Colla 2010)

A variety of circumstantial evidence supports the hypothesis that at least some recent bumblebee declines in North America have been driven or exacerbated by the spread of Nosema bombi, an obligate intracellular microsporidian fungal parasite found commonly in bumblebees throughout Europe, via commercial bumblebee rearing facilities that introduced this pathogen from Europe. Pathogenic effects of N. bombi may vary depending on the host species and reproductive caste, but they may include reductions in both colony growth and individual life span and fitness. Further research will be necessary to clarify the role of pathogens in bumblebee declines. (Cameron et al. 2011 and references therein) Based on a combination of modeling, laboratory experiments, and literature review, Otterstatter and Thomson (2008) found strong support for the hypothesis that spillover of Crithidia bombi (a destructive internal trypansomatid protozoan parasite) from bumblebees reared commercially to pollinate greenhouse crops has contributed to the ongoing decline of wild Bombus in North America. Wild bumblebees may also be negatively impacted by the spread of the bumble bee tracheal mite Locustacarus buchneri to wild populations from commercially reared colonies (Evans et al. 2005 and references therein).

Global Protection: None. No occurrences appropriately protected and managed

Comments: It is probably not possible to protect any occurrence from pathogen spillover.

Needs: Uncertain but in a plausible worst case scenario captive breeding might be needed. If stable populations are found somewhere, importation of bumblesbees from elsewhere should be prevented.

Stewardship Overview: The following is a summary of the management and conservation needs for the genus Bombus. There is a lot of literature on the decline of bumblebees and other pollinators (e.g. Goulson et al. 2005, Brown and Paxton 2009, Evans et al. 2008, and the Committee on the Status of Pollinators in North America 2007). Byrne and Fitzpatrick (2009) review pollinator conservation programs at national, regional, and global levels. DeVore (2009) offers many practical considerations for pollinator conservation in the U.S. Noordijk et al.'s (2009) discussion of mowing in Europe should be generally applicable in North America, especially northern portions. Goulson et al. (2005), among others, make more general recommendations.

It is generally agreed that declines, and in some regions extirpations, of bumblebees and other pollinators in Europe have been due primarily to habitat loss or alteration, including changes in forage plant availability (due especially to intensification of agriculture). In turn, some plants have declined due to loss of pollinators. Brown and Paxton (2009), based in the United Kingdom, suggest that future conservation strategies need to "prioritise (i) minimising habitat loss, (ii) making agricultural habitats bee-friendly, (iii) training scientists and the public in bee taxonomy and identification, (iv) basic autecological and population genetic studies to underpin conservation strategies, (v) assessing the value of DNA barcoding for bee conservation, (vi) determining the impact of invasive plants, animals, parasites and pathogens, and (vii) integrating this information to understand the potential impact of climate change on current bee diversity." Some needs may be different in the U.S. In particular, climate change is probably less of a concern, whereas parasites and diseases are of much greater immediate concern.

Williams et al. (2009) examine various hypotheses from the literature as related to the status of bumble bees in North America, Europe and China, including competition with congeners, climatic specialization, proximity to climatic range edge, food specialization, phenology, body size, and range size. Food specialization would be in part an index of habitat specialization, but possible special needs for overwintering or nesting sites are not addressed. Results of their meta-analysis of correlations showed support for the hypotheses that decline susceptibility is generally greater for species that have greater climatic specialization, in areas where species occur closest to the edges of their climatic ranges, and for species that have queens that become active relatively late in the year. At least on a multi-continent scale the other factors apparently do not widely explain bumblebee declines. Notably most North American bumble bees range through more than ten degrees of latitude and thus have adapted to a wide array of climates at least in terms of temperature, and some range from coast to coast, which requires adaptation to a wide array of precipitation regimens and habitat types as well. While climate change could impact bumble bees positively or negatively at the edges of their ranges, this is not a plausible explanation for range-wide declines of widespread species.

Williams et al. (2009) suggest that late queen phenology may render a species at a particular disadvantage when they have long colony cycles if there are losses of food plants in mid to late colony development. Among declining species in the Grixti et al. (2009) tabulation (see also Colla and Packer, 2008) two are early, one is intermediate, and two are late; the ratio is nearly the same for non-declining species six are early, one is intermediate and three are late, and late species are not over-represented among declining species (chi2=0.6, df=2, p=0.95). In contrast B. affinis, which is undergoing extreme decline and may be on the brink of extinction, has early queen phenology, as does B. terricola which is also in severe decline. Their phenologies should make them among the least at risk. Phenology is not driving the extremely rapid declines of these species. The meta-analysis of Williams et al. (2009) also does not support any relationship between declines and tongue length as some earlier studies suggested--long tongues generally indicate more specialized foragers. The data in Grixti et al. (2009) are consistent with this finding: among declining species two have short tongues, one is intermediate and two have long tongues while among non-declining species four have short tongues, four are intermediate and two have long tongues, again nearly identical (chi2=0.9, df=2, p=0.885). Some other factor, almost certainly diseases and parasites (Colla et al., 2006, Otterstatter and Thomson, 2008; Federman, 2009), is largely overriding phenology, tongue length (foraging ecology) and other life history traits that may be important determinants of risk in other countries leading to very rapid declines. Nevertheless managers should be aware of these ecological traits that may predispose species to future declines, or may be driving slower current declines. Practical implications include a need for legumes and other flowers favored by long-tongued species and for a reliable supply of flowers late in the season.

Habitat fragmentation can also be important in bumblebee ecology (Hines and Hendrix 2005). A study by Bhattacharya et al. (2003) near Boston documented that foraging bumble bees (Bombus impatiens, B. affinis) have high site fidelity and flower constancy, and are reluctant to cross roads and railroads compared to more natural habitats. If the flower supply runs out they are more likely to locate another on the same side of a road rather than to cross it. Thus like many animals, bumblebees should benefit from reducing the number of roads, and the amount of other highly unnatural habitats, such as lawns, in and near natural areas.

Bumble bees have three critical sets of ecological needs: suitable overwintering places for the queens, suitable nesting microhabitats, and adequate flowers for foraging throughout the length of the colony cycle. This cycle is several months, typically mid spring to mid or late summer. Little information was found regarding hibernation sites, and information on nesting sites is usually rather general, except that Carvell (2002) discusses the nesting habits of the subgenus Thoracobombus. These are known in England as carder bees and they "nest on the ground surface and comb together material from around the nest as a covering" and "therefore require moss and dried grasses, often in the form of disused small mammal nests... hence the importance of undisturbed tall grassland with sufficient sunlight providing warmth to the surface nest." Most other bumblebees nest below ground in pre-existing cavities. Bombus impatiens will also use a variety of man-made situations such as under houses or old rodent nests in cardboard boxes, etc. (D. Schweitzer, pers. obs in New Jersey). Plath (1922, 1927) provides detailed observations of the nesting habits of North American bumble bees, including the now seriously imperiled B. (Bombus) affinis, the nests of which he observed to be solely subterranean. Prior to its decline, B. affinis had adapted well to urban areas and was observed nesting in the concrete rubble beside the foundation of buildings (Super and Moyer, 2003). Queen bumblebees probably usually hibernate in the leaf litter near the soil surface or perhaps underground.

Other than plowing (Hopwood, 2008, DeVore 2009), most common management activities should not directly affect underground nests. However bumblebees above ground in grasses would be vulnerable to fires, and to mowing if the blade is low enough to destroy them. Hibernating queens could be very vulnerable to prescribed burns if they are above ground in dry microhabitats. However, this might not affect the population in situations where nest sites are limiting such that many queens fail to establish colonies. Prescribed burning or any other management scheme potentially can have two sets of impacts, and the latter could be positive or negative: direct mortality to the pollinators and changes in vegetation composition and structure. Prescribed burning would likely render an area unsuitable for Bombus (Thoracobombus) for at least one season due to removal of nesting microhabitats (see Carvell, 2002). Much more information is needed regarding impacts of common management practices, perhaps especially fire, on bumblebees. In general maintaining healthy rodent populations in habitats where bumblebees nest should improve availability of nest sites. Besides providing habitat and cover for native rodents, elimination of free-roaming cats could be beneficial. In addition to a likely reduction in small rodents overall, high cat numbers may lead to an increased ratio of the non-native house mouse (Mus musculus) over native mice (Peromyscus spp.) (Hawkins, 1998, as cited by Longcore et al., 2009). It is not clear whether the species mix of small rodents has much affect on the availability of bumblebee nest sites. Besides old rodent holes and grass clumps, logs may provide useful nesting sites.

Most management activities involving bumblebees will be aimed at improving flower availability. Bumblebees depend on both nectar, mainly for carbohydrates, and pollen, for protein. A queen bumblebee needs nectar when she leaves hibernation, and for another month or more while she alone rears the first brood of workers. After that, workers need access to nectar and pollen for several months. Bumblebees will travel several kilometers. For example Schmidt and Jacobson (2005) note that B. pensylvanicus sonorus commonly flies to high elevations from desert nesting sites much lower, the vertical distance alone is often over a kilometer. Devore (2009) considers a mile (1.7 km) as about the typical distance over which bumblebees forage. Citing several studies (e.g. Dramstad, 1996), Hines and Hendrix (2005) state that bumble bees routinely forage up to 600-650 m from their nests. While some bumble bees are more generalized in their preferences than others, none are thought to be highly specialized in their foraging needs. Carvell (2002) found that the more common species in her study areas in England used between about nine and 15 plant species over the course of her study. Hines and Hendrix (2005) observed bumblebees foraging on 43 of 150 species of flowers monitored in Iowa prairie remnants. Bumblebee species with longer tongues tend to visit legumes and other plants with long corollas. Feeding is most efficient if the length of the tongue and corolla are similar. Thus habitats with a greater diversity of plants with varying corolla lengths can be expected to have a richer bumblebee fauna than less diverse communities.

Carvell (2002) in England found that "Numbers of both long- and short-tongued bumblebee species, abundance of all bumblebees and species richness per quadrat were significantly positively correlated with abundance of P[ilosella] officinarum and T[rifolium] pratense (red clover), total flower abundance, flowering plant species richness and continuity of bee-exploited species (the last excepting long-tongued bumblebee species)." Habitats most likely to have these features are generally open with simple vegetation structure and little moss or thatch cover. Gardens, even those in urban areas (Wojcik et al. 2008, Fetridge et al. 2008), can be useful foraging resources for bumblebees, as can croplands (e.g. Turnock et al. 2007) if they are not sprayed with insecticides during the flowering period. Hay fields with abundant red clover or alfalfa can be major foraging habitats. Roadsides with restored native prairie vegetation can also be very beneficial (Hopwood, 2008). Prairies, moist meadows, and restored roadsides with native flowers probably are among the most productive bumblebee foraging habitats. Tuell et al. (2008) provides many records of bees, including bumble bees, at native wild flowers in Michigan. Many other studies provide useful information about species of flowers visited by bumblebees (e.g. Hopwood, 2008, US fish and Wildlife Service, 1999, 2008, DeVore, 2009). Evans et al. (2008) provide summaries of plants known to be visited by three seriously declining or imperiled species of subgenus Bombus. Hopwood (2008) documents that in Kansas, roadsides restored to native prairie vegetation supported richer bee faunas than weedy roadsides or even local prairie remnants. Similarly Noordijk et al. (2009) discuss roadsides as important habitats for pollinators, including bumblebees, in northern Europe, and discuss the suitability of various mowing regimens. For Bombus species Hopwood had 37 observations of seven species on restored roadsides versus 10 representing two species on weedy controls. She concluded that infrequently mowed roadsides with diverse native flora actually make good native bee habitat because such places are not plowed, and thus are good nesting [and probably hibernating] habitats, and unlikely to be developed for other uses. She suggests roadsides could be important to pollinator management in much of the world.

The recommendations of DeVore (2009), which are specifically for heavily agricultural southern Minnesota and for bees in general, would apply to bumblebees almost anywhere in eastern and central North America. The following quote provides guidance for making farms and other landscape pollinator-friendly not just of bumblebees, but for bees in general: "Native wildflowers can provide excellent foraging for pollinators. So can cover crops that are allowed to flower. Reducing tillage causes less disruption of nesting habitat, because two-thirds of bees nest underground. Leaving logs, stumps, snags, and clumps of grass will provide nesting sites for the rest. Fencerows with willow, dogwood, or other flowering plants provide foraging habitat on working farmland without disrupting the agronomic productivity... Roadsides, ditches, and buffer strips can also serve as wild pollinator habitat." In some cases the pollination service will have significant economic benefits. Even narrow strips of buffer land can have important impacts on bees (Hopwood, 2008, DeVore, 2009). While suburban and urban gardens do provide nutrition (but sometimes also toxic pesticides) to wild bees (Frankie et al., 2009, McFrederick and LeBuhn, 2006), manicured lawns are among the least useful nesting and foraging habitats. DeVore also points out that protection from insecticides may be needed and that several studies find low levels of many pesticides in hives, brought in by workers on contaminated pollen. While levels are usually sublethal there may still be impacts to hive success. Recent articles (e.g. Colla and Packer, 2008) have suggested persistent neonicotinoid insecticides might be particularly hazardous to bees, but so far these have not been directly linked to major local declines.

On more natural lands, management would likely focus on maintaining diverse assemblages of primarily native flora, such that flowers would be constantly available throughout the active season, typically about April to September in many places. In some areas several different habitats would be needed to fulfill this need. To the extent practical, productive foraging sites should not be mowed during the flowering season, although in most contexts foraging workers will probably locate alternative flowers. Noordijk et al. (2009) found that while mowing virtually eliminates nectar for a period of days or longer afterward, summer mowing often stimulates re-flowering that benefits the bees later in the season. Also growing season mowing could be highly detrimental to other taxa of conservation concern, such as Lepidoptera larvae that could be killed directly--in some cases an entire population. Carvell (2002) found cattle grazing to be an effective management tool in England.

Regardless of the suitability of habitat management on a local scale, the surrounding landscape context will affect bumblebee communities. Hines and Hendrix (2005) found that variation in the availability of floral resources, especially as determined by the extent of land in grassland, in the surrounding 500-700 meter radius around prairie remnants explained most difference in the bumblebee community in prairie remnants, with the abundance of suitable flowers on site also being important. Landscape context is also likely to be important in terms of suitable nesting and hibernation sites, since these habitats may not be the same as foraging areas (e.g. Carvell, 2002). Thus small scraps of habitat may be poorly suited for bumblebee conservation even if the flora is relatively pristine. Ideally management aimed at conserving local bumblebees should be tailored to the needs of the local fauna, but in general unplowed, open, diverse, flowery habitats will be the goal. In most places such habitat will need active management--the exact nature of which may be determined by factors unrelated to bumblebees. Some degree of connectivity to other suitable habitats may be necessary, which is one reason roadside, power lines, and other long corridors may be quite suitable. See Russell et al. (2005) for a useful discussion of power line management as it relates to bees.

While habitat changes, especially a large-scale decline in foraging plants in some regions such as highly agricultural parts of the Midwest, have undoubtedly impacted North American bumblebees, these do not appear to be driving declines to anywhere near the extent they are in Great Britain. There is little that can be done to address what appears to be the greatest conservation need of the most imperiled North American bumble bees: protection from Nosema, Cirthidia, and other (mostly non-native) pathogens and parasites. For the most severely impacted species, Bombus affinis, B. franklini, and other species of subgenus Bombus, unless these can be mitigated in the near future or the species evolve some resistance, any conservation efforts are likely to be moot. Nosema could not be effectively managed in commercial hives of the closely related B. occidentalis and catastrophic outbreaks lead directly to the discontinuation of commercial use of that species. There is even less that can be done to protect wild bumble bees, and evolution of natural resistance may be needed for some species to persist in major portions, or all, of their ranges, although hypothetically the same result could be accomplished by genetic engineering. About the only suggestion now is to try to minimize contact between wild bumblebees and green house bees, especially of the same subgenus. Wild bees from near green house colonies tend to have higher infection rates than more distant ones (Colla et al., 2006). Several authors have pointed out the need for tight restrictions on the importation of bumblebees, whether of native or non-native species, that have been reared outside of North America.