Introduction
Pollinating insects are inarguably important for humans (see Potts et al. Reference Potts, Imperatriz-Fonseca, Ngo, Aizen, Biesmeijer and Breeze2016). Bees (Hymenoptera: Apidae), in general, are considered the most important pollinating insects (Klein et al. Reference Klein, Vaissière, Cane, Steffan-Dewenter, Cunningham, Kremen and Tscharntke2007) and the abundance of wild bees, specifically, is more correlated to crop yields than the abundance of honey bees (Breeze et al. Reference Breeze, Bailey, Balcombe and Potts2011; Garibaldi et al. Reference Garibaldi, Steffan-Dewenter, Winfree, Aizen, Bommarco and Cunningham2013; Mallinger and Gratton Reference Mallinger and Gratton2015). However, the abundance and diversity of wild pollinators has decreased worldwide due to a number of factors (Biesmeijer et al. Reference Biesmeijer, Roberts, Reemer, Ohlemüller, Edwards and Peeters2006; Colla and Packer Reference Colla and Packer2008; Grixti et al. Reference Grixti, Wong, Cameron and Favret2009; Committee on the Status of Endangered Wildlife in Canada 2010; Potts et al. Reference Potts, Biesmeijer, Kremen, Neumann, Schweiger and Kunin2010; Bommarco et al. Reference Bommarco, Lundin, Smith and Rundlöf2012; Szabo et al. Reference Szabo, Colla, Wagner, Gall and Kerr2012; Burkle et al. Reference Burkle, Marlin and Knight2013; Carvalheiro et al. Reference Carvalheiro, Kunin, Keil, Aguirre-Gutiérrez, Ellis and Fox2013; Committee on the Status of Endangered Wildlife in Canada 2014a, 2014b; Committee on the Status of Endangered Wildlife in Canada 2015; Goulson and Hughes Reference Goulson and Hughes2015; Koh et al. Reference Koh, Lonsdorf, Williams, Brittain, Isaacs, Gibbs and Ricketts2016), resulting in the ongoing search for and development of additional commercially managed pollinators. The global transportation of bees for commercial pollination has been, and continues to be a mechanism contributing to widespread introduction and establishment of non-native species and bee diseases. These diseases threaten native bee diversity and the vital ecosystem service that these pollinators provide to crops and wildflowers worldwide (Goulson and Hughes Reference Goulson and Hughes2015; Potts et al. Reference Potts, Imperatriz-Fonseca, Ngo, Aizen, Biesmeijer and Breeze2016).
The unique climate and isolation from mainland areas experienced by the island portion of the Canadian province of Newfoundland and Labrador have resulted in a bee fauna that is much less diverse than other areas of North America (Sheffield et al. Reference Sheffield, Heron, Gibbs, Onuferko, Oram and Best2017). Some growers in Newfoundland import commercially available colonies of the non-native bumble bee Bombus impatiens Cresson (Hymenoptera: Apidae) to supplement pollination of their crops, even though native Newfoundland bee species provide important pollination services to small fruit producers (Hicks Reference Hicks2011). Hicks and Sircom (Reference Hicks and Sircom2016) determined that such importations do not necessarily increase pollination on Newfoundland cranberry farms and this practice should be reconsidered. Furthermore, as some crops flower significantly later on the island than on the mainland of eastern Canada, some farmers purchase pre-used commercial B. impatiens colonies from different Maritime Provinces to supplement pollination, a practice that is not known from other areas.
It has been documented that commercially supplied bumble bees carry diseases that are transmitted to native bumble bees (Niwa et al. Reference Niwa, Iwano, Asada, Matsuura and Goka2004; Otterstatter and Thomson Reference Otterstatter and Thomson2008; Arbetman et al. Reference Arbetman, Meeus, Morales, Aizen and Smagghe2013; Murray et al. Reference Murray, Coffey, Kehoe and Horgan2013; Graystock et al. Reference Graystock, Yates, Evison, Darvill, Goulson and Hughes2013b; Graystock et al. Reference Graystock, Goulson and Hughes2015). In North America, B. impatiens is commercially supplied and is also known to transmit several diseases to native bumble bees (Colla et al. Reference Colla, Otterstatter, Gegear and Thomson2006; Sachman-Ruiz et al. Reference Sachman-Ruiz, Narváez-Padilla and Reynaud2015; Cameron et al. Reference Cameron, Lim, Lozier, Duennes and Thorp2016). The main mechanism of transmission is contact with infected nest mates, nest material, or flowers (Schmid-Hempel and Tognazzo Reference Schmid-Hempel and Tognazzo2010; Graystock et al. Reference Graystock, Yates, Darvill, Goulson and Hughes2013a, Reference Graystock, Goulson and Hughes2015).
Bees that enter non-natal colonies are known as drifters (Free Reference Free1958). The term “drifting” is believed to be attributed to honey robbing or the result of orientation errors (Free Reference Free1958; Jay Reference Jay1966; Pfeiffer and Crailsheim Reference Pfeiffer and Crailsheim1998; Neumann et al. Reference Neumann, Moritz and Mautz2000). Supersedure is different than drifting; here, one queen enters a nest and, if successful, will kill the resident queen and take over (Alford Reference Alford1975). Drifting in bumble bees is likely a temporary phenomenon where foraging worker bees enter non-natal nests to rob honey (Alford Reference Alford1975; Genersch et al. Reference Genersch, Yue, Fries and de Miranda2006), or in some cases to lay eggs (Lopez-Vaamonde et al. Reference Lopez-Vaamonde, Koning, Brown, Jordan and Bourke2004; O’Connor et al. Reference O’Connor, Park and Goulson2013). Where commercially reared bumble bees are used in agricultural settings and greenhouses, intraspecific drifting commonly occurs (Birmingham and Winston Reference Birmingham and Winston2004; Birmingham et al. Reference Birmingham, Hoover, Winston and Ydenberg2004; Lefebvre and Pierre Reference Lefebvre and Pierre2007). In addition, Hobbs (Reference Hobbs1966, Reference Hobbs1967) observed both intraspecific and interspecific supersedure of queen bumble bees using artificial domiciles placed in natural habitats. The only evidence of intraspecific drifting in natural colonies comes from Takahashi et al. (Reference Takahashi, Martin, Ono and Shimizu2010) where they used molecular analysis to show non-natal males of Bombus deuteronymus Schulz reared by unrelated workers.
Here we follow up on questions we had following our previous study (Hicks and Sircom Reference Hicks and Sircom2016) where B. impatiens colonies were initially purchased to look at the efficiency of these bees as pollinators in Newfoundland cranberry farms. As Hicks and Sircom (Reference Hicks and Sircom2016) determined that commercial colonies are likely not needed in Newfoundland, their continued importation (either new or pre-used) puts the native bumble bee species of the island at an undue risk. The questions we address include: (1) Will bumble bees in Newfoundland enter commercial Bombus Latreille colonies? If so, (2) what is the pathogen profile of selected bee diseases in the introduced bees and native bees that may enter and/or leave these colonies? (3) What is the significance of potential pathogen spillover to Newfoundland native bees?
Methods
To help answer the questions, we examined commercial B. impatiens colonies for the presence of native bumble bee species after the colonies were removed from the field. In addition, we used the polymerase chain reaction with taxon-specific oligonucleotides to screen native and commercial bees for pathogens and address the potential of pathogen spillover from commercial bumble bee colonies to the native bees of Newfoundland. As per Graystock et al. (Reference Graystock, Yates, Evison, Darvill, Goulson and Hughes2013b), bees were tested for seven pathogens: Crithidia bombi (Lipa and Triggiani) (Kinetoplastea: Trypanosomatidae), Apicystis bombi (Liu, Macfarlane, and Pengelly) (Neogregarinorida: Lipotrophidae), Nosema bombi Fantham and Porter (Dissociodihaplophasida: Nosematidae), Nosema apis (Zander) (Dissociodihaplophasida: Nosematidae), Nosema ceranae Fries et al. (Dissociodihaplophasida: Nosematidae), Melissococcus plutonius (White) (Lactobacillales: Enterococcaceae), and Ascosphaera Olive and Spiltoir (Ascomycota: Onygenales: Ascosphaeraceae) species.
Bee sampling
On 15 July 2013, six quads (i.e., each quad is four externally connected colonies, each with a queen and its own foragers) of B. impatiens that were previously used for lowbush blueberry pollination in New Brunswick, Canada (hereafter, called “pre-used quads”) were obtained privately by a local cranberry farmer. An additional four quads (hereafter, called “new quads”) of B. impatiens were purchased new on 31 July 2013 from Biobest Canada (Leamington, Ontario, Canada) and placed on the same cranberry (Vaccinium Linnaeus; Ericaceae) field. The cranberry farm was located near Stephenville, Newfoundland and Labrador (48°27'13"N, 58°24'25"W). At the end of the pollination season (14 August 2013) the new quads and one pre-used quad were taken off the field and frozen, their contents later examined. All specimens of B. impatiens located inside the colonies of the quads were collected and frozen (−20°C) and specimens of native species found inside the colonies were frozen separately. To determine the disease presence in free-living native bumble bees, we sampled bees that were pan-trapped on the field where the commercial bees were located, and from one other area away from the field. That area had bees pan-trapped from a commercial cranberry farm located 17 km northwest of the study field (48°34'17"N, 58°31'27"W) for a different study (but during the same time). These native specimens had been pinned and air dried for the other study (Hicks and Sircom Reference Hicks and Sircom2016). Identification of native bumble bee species were done using the key of Laverty and Harder (Reference Laverty and Harder1988). Voucher specimens from this study were deposited in the general collection of the College of the North Atlantic, Carbonear, Newfoundland and Labrador, Canada.
Molecular analysis
Table 1 summarises the 440 bee specimens that were used in the molecular analysis. DNA was extracted from individual bees with the exception of the B. impatiens specimens sampled at the end of the pollination season from the new and pre-used quads for which five to six specimens were pooled before extraction. DNA was extracted from frozen or dried bee abdomens, using the Qiagen DNeasy blood and tissue kit (Qiagen, Toronto, Ontario, Canada) following the tissue protocol. Abdomens were minced before extraction, and lysed overnight. Due to the increased amount of starting material for the pooled B. impatiens, specimens were lysed in twice the volume of buffer ATL and proteinase K, mixed with twice the volume of 95% ethanol and buffer AL, and added to the spin column in two separate volumes. DNA was eluted with two consecutive 75 μL volumes of AE buffer.
Table 1 The number of specimens (n) used in in the molecular analysis with species identification, stage, provenance, and sample date.
Note: The Bombus impatiens numbers were pooled in groups of either five or six specimens for the analysis.
DNA extractions were tested for seven pathogens by taxon-specific polymerase chain reactions, as detailed in Graystock et al. (Reference Graystock, Yates, Evison, Darvill, Goulson and Hughes2013b). The Apidae 18S rRNA gene was also amplified from each DNA extraction to confirm that the quality of DNA was sufficient for polymerase chain reaction. Reactions contained 1× Qiagen Type-it Master Mix, 0.2–0.4 μM each forward and reverse primer, and 3 μL of DNA. Polymerase chain reaction profiles consisted of an initial denaturation at 95°C for five minutes, 35 cycles at 95°C for 30 seconds, annealing temperature for 45 seconds, and 72°C for 30–45 seconds, with a final elongation at 72°C for 10 minutes (see Table 2 for pathogen-specific profile parameters). Polymerase chain reaction products were stained with EZ-vision three DNA dye (Amresco LLC, Cleveland, Ohio, United States of America) and visualised on a 1.5% agarose gel by electrophoresis. Samples that produced a single amplicon of expected size (Table 2) were identified as positive for the pathogen. Samples with no amplicons were considered to be free from the pathogen. Some samples resulted in either multiple amplicons or an amplicon of unexpected size; the result for these samples was identified as unknown. Positive (known infected specimens) and negative (no-template) controls were included for each set of reactions.
Table 2 Primers and polymerase chain reaction conditions used to screen seven pathogens in Newfoundland bees.
TA refers to annealing temperature; A refers to Apidae; Cb refers to Crithidia bombi; Ex refers to external; In refers to internal.
We were unable to secure bee specimens with known N. bombi infections or N. bombi DNA to test as positive controls, so instead we sequenced the amplicons to confirm their provenance. Polymerase chain reaction products were purified for cycle sequencing using a 100 K Pall AcroPrep 96 Multi-Well filter plate (Pall Life Sciences, Port Washington, New York, United States of America) according to the instructions of the manufacturer. Sequencing was carried out using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, California, United States of America) following the protocol of the manufacturer, and purified via ethanol precipitation. Sequencing products were electrophoresed in an Applied Biosystems 3730 DNA analyser using Sequencing Analysis v5.2 Software. Sequences were edited, aligned, and compared to those available in GenBank (www.ncbi.nlm.nih.gov/genbank) using the BLAST sequence alignment software (www.blast.ncbi.nlm.nih.gov; Altschul et al. Reference Altschul, Gish, Miller, Myers and Lipman1990, Reference Altschul, Madden, Schäffer, Zhang, Zhang, Miller and Lipman1997), to confirm that amplicons resulted from the targeted pathogen.
Statistical analysis
Minitab version 15 was use to perform post-hoc χ 2 tests (after a Bonferroni correction was applied) for sample independence to determine differences in presence and absence of the specific pathogens between the native bumble bees found inside the new colonies and pre-used colonies as well as natives collected away from the field and ones collected on the field. Pairwise comparisons were done using Fisher’s exact tests and statistical significance was determined after a Bonferroni correction was applied.
Results
A total of 23 specimens of native bumble bee were collected from inside the four new B. impatiens quads (14 B. ternarius Say; four B. terricola Kirby; two B. vagans bolsteri Franklin; three B. frigidus Smith) and 18 native specimens were observed in the one pre-used B. impatiens quad (10 B. ternarius; one B. terricola; two B. vagans bolsteri; five B. frigidus).
Five of the seven pathogens screened were detected among the native and imported bumble bees tested (Table 3) Nosema apis and Melisococcus plutonis were not detected. Bombus impatiens workers sampled from the new quads at the end of the pollination period tested positive for three out of the seven pathogens screened. While there was an observation of dysentery outside the entrances of some of the colonies, we did not observe any outward signs of disease in the bees.
Table 3 Percentage (prevalence) of seven pathogens screened from native and commercial bees in Newfoundland.
* The native species located in the quads and sampled freely include: B. ternarius, B. terricola, B. vagans bolsteri, and B. frigidus.
≅ Values of particular disease prevalence for native Bombus followed by the same letter signifies significant statistical difference at P=0.05.
Crithidia bombi was detected most frequently in native Newfoundland bumble bee species. In native bees collected away from the study area (Table 3), C. bombi was present in 75.0% (21/28) of the bees sampled while only 53.3% (16/30) of bees sampled on the field-tested positive. By contrast, native bumble bees found inside the new quads tested positive with C. bombi at a rate of 81% (17/21), with all (18/18) native species inside the pre-used colonies were positive. A post-hoc χ 2 test showed a significant difference of the C. bombi rates among all of the native bumble bees collected (new quads, pre-used quads, away from field, and on field) (χ 2 (3, 97)=13.60, P=0.004). Only the bees on the field had a lower detection of pathogens than expected. Further pairwise comparisons using Fisher’s exact test and a Bonferroni correction showed only the native bumble bees found inside the pre-used colonies and the bumble bees collected on the field were significantly different (Table 3).
Apicystis bombi was the next most prevalent pathogen detected. Among native bees, 50.0% (14/28) of the Bombus individuals away from the field and 20% (6/30) collected on the field-tested positive for A. bombi. Apicystis bombi was found in 38.1% (8/21) of native bees collected inside the new colonies and 72.2% (13/18) of native bees inside the pre-used colony tested positive for this pathogen (Table 3); these values were significantly different (χ 2 (3, 97)=13.55, P=0.004). In this case, the bees from new quads and bees on the field had A. bombi detected at levels lower than the expected rate but native bees in the pre-used colonies were much higher than the expected rate. Plus, the pairwise comparisons again only showed a significant difference in pathogen detection between pre-used colonies and native bumble bees collected away from the field (Table 3).
The third most prevalent pathogen, Nosema bombi, was detected in 3.6% (1/28) of natives away from the field and 6.7% (2/30) on the field, 34.8% (8/23) of natives inside new colonies and 16.7% (3/18) of native bumble bees inside the pre-used colony tested positive for this pathogen. While the χ 2 indicates these values were significantly different (χ 2 (3, 99)=12.12, P=0.007), caution should be exercised as some of the expected counts were less than five. Only the native bumble bees inside the new quads were significantly different from the native bees collected away from the field after the pairwise comparisons (Table 3).
The fourth most common pathogen detected was the fungus Ascosphaera species. While no native Bombus species living away from the study area or on the field-tested positive for this pathogen, native Bombus species located inside the new colonies were 18.2% (4/22) positive and native species located in the pre-used colony tested 50.0% (9/18) positive for this fungus (Table 3). While the χ 2 test showed that these values were significantly different (χ 2 (3, 98)=30.44, P<0.001), caution should be exercised as some expected counts had values lower than five. Pairwise comparisons of Ascosphaera species prevalence showed that the native bumble bees collected inside the pre-used colonies were significantly different than the native bumble bees collected away from the field and for bumble bees collected on the field (Table 3). There was no difference in pathogen detection between new and pre-used colonies. Nosema ceranae was only detected in one B. impatiens from a pre-used colony, with one native B. ternarius specimen testing positive from within the same colony box.
Discussion
This is the first documented report of interspecific drifting of native bumble bee species in commercial colonies of Bombus impatiens. Though the drifting of individuals of the same species (intraspecific) into foreign nests occurs frequently in bee species (Birmingham and Winston Reference Birmingham and Winston2004; Birmingham et al. Reference Birmingham, Hoover, Winston and Ydenberg2004; Lefebvre and Pierre Reference Lefebvre and Pierre2007), interspecific movement of native bumble bees into commercial nests has not been previously recorded. Drifting can be the result of disorientation, or provision robbing (Pfeiffer and Crailsheim Reference Pfeiffer and Crailsheim1998; Neumann et al. Reference Neumann, Moritz and Mautz2000; Birmingham et al. Reference Birmingham, Hoover, Winston and Ydenberg2004). However, it is unlikely that the observed drifting was caused by disorientation in the present study as the commercial nests were in very conspicuous boxes on cranberry fields. Therefore, it seems more likely that these drifters were attempting to steal provisions from the commercial colonies. We are unsure how long the native specimens were inside the B. impatiens nests or whether they may have returned repeatedly. Of more concern to us is the potential for disease transmission between the commercial bees and the native bees. Drifting bees can pick up pathogens from the host nest (spillover) or they may bring diseases to the nest where the host species can become infected. The diseases could spread rapidly because of close proximity of individuals and then the disease can be transmitted back to other drifting bees (spillback) which may go on to infect other native species. As we see from the analysis, in some instances the native bees away from the field had similar levels of disease detection as the bees in the new quads as no natural population will be diseases free. However, importing commercial bees into areas will exacerbate pathogen spillover and spillback with native bees. The mechanisms of spreading diseases among managed and wild bees include shared flower use, drifting, and honey robbing (see Goulson et al. Reference Goulson, Whitehorn and Fowley2012; Graystock et al. Reference Graystock, Goulson and Hughes2015). O’Connor et al. (Reference O’Connor, Park and Goulson2013) suggested that drifting by worker Bombus is important among intraspecific disease transmission, however, we think it is fair to speculate that interspecific disease transmission by drifting is very likely.
During this study we screened native Newfoundland bees sampled freely and from inside pre-used and new B. impatiens colonies for seven pathogens associated with serious bee diseases. The trypanosome Crithidia bombi was the most prevalent pathogen detected among Newfoundland bees (Table 3). The parasite resides in the hindgut of its host where it attaches to the gut wall and multiplies with transmission stages passing out in the faeces of the host (Schmid-Hempel Reference Schmid-Hempel2001). This pathogen may be directly transmitted without any vector through contact with infected nest mates, nest material or via flowers (Schmid-Hempel and Tognazzo Reference Schmid-Hempel and Tognazzo2010). The prevalence of C. bombi in the native Newfoundland Bombus species no matter from where they were sampled was high and is consistent with other areas of North America (Colla et al. Reference Colla, Otterstatter, Gegear and Thomson2006; Otterstatter and Thomson Reference Otterstatter and Thomson2008; Gillespie Reference Gillespie2010).
The neogregarine, Apicystis bombi, was the second most prevalent pathogen in native Bombus. It infects the adipose tissue of bees (Lipa and Triggiani Reference Lipa and Triggiani1996) and causes mortality to bumble bees (Rutrecht and Brown Reference Rutrecht and Brown2008; Graystock et al. Reference Graystock, Meeus, Smagghe, Goulson and Hughes2016). The spillover of A. bombi from commercial B. terrestris in Argentina played a role in the decline of at least one bumble bee species there (Arbetman et al. Reference Arbetman, Meeus, Morales, Aizen and Smagghe2013). In Newfoundland, A. bombi prevalence was high in pre-used commercial colonies and there was a significant difference in prevalence in native bees located inside the pre-used colonies with those found outside.
The third most prevalent pathogen among Newfoundland bees was Nosema bombi, a fungal pathogen that has been implicated in the decline of bumble bees is North America (Gillespie Reference Gillespie2010; Cameron et al. Reference Cameron, Lozier, Strange, Koch, Cordes, Solter and Griswold2011; Bushmann et al. Reference Bushmann, Drummond, Beers and Groden2012; Malfi and Roulston Reference Malfi and Roulston2014; Sachman-Ruiz et al. Reference Sachman-Ruiz, Narváez-Padilla and Reynaud2015). Spores are released into the environment by the host feces and are the likely mechanism of transmission when bees share flowers (Graystock et al. Reference Graystock, Goulson and Hughes2015). Laboratory experiments have demonstrated that N. bombi can be transmitted from commercial B. terrestris colonies to the native species in both Japan and the United Kingdom (Niwa et al. Reference Niwa, Iwano, Asada, Matsuura and Goka2004; Murray et al. Reference Murray, Coffey, Kehoe and Horgan2013; Graystock et al. Reference Graystock, Yates, Evison, Darvill, Goulson and Hughes2013b). Otti and Schmid-Hempel (Reference Otti and Schmid-Hempel2007, Reference Otti and Schmid-Hempel2008) indicated that N. bombi can deform wings, decrease the survival of workers and males and prevent queens from mating. The incidence of N. bombi close to greenhouses supplied with commercial B. impatiens rose to 15% in Ontario, Canada (Colla et al. Reference Colla, Otterstatter, Gegear and Thomson2006). Recently, Cameron et al. (Reference Cameron, Lim, Lozier, Duennes and Thorp2016) showed that while N. bombi was historically present and widespread in North American native Bombus species, it was the spillover from heavily affected commercial colonies in the mid-1990s that may have greatly increased the prevalence of this pathogen in some declining native species.
Nosema ceranae was originally thought of as a pathogen of honey bees (Higes et al. Reference Higes, Martín-Hernández, Botías, Bailón, González-Porto and Barrios2008) but recent studies have indicated that Bombus species are susceptible as well (Graystock et al. Reference Graystock, Yates, Darvill, Goulson and Hughes2013a; Fürst et al. Reference Fürst, McMahon, Osborne, Paxton and Brown2014; Graystock et al. Reference Graystock, Goulson and Hughes2015). In the present study, we found one specimen of B. impatiens and one native bumble bee (B. ternarius) located in the same colony box infected with N. ceranae. Shutler et al. (Reference Shutler, Head, Burgher-MacLellan, Colwell, Levitt, Ostiguy and Williams2014), using molecular techniques, found N. ceranae in two of 55 colonies of Apis mellifera Linnaeus in one large beekeeping operation in Newfoundland. Moreover, several unpublished records recently indicate that this species is now common in Newfoundland honey bees. As Graystock et al. (Reference Graystock, Yates, Darvill, Goulson and Hughes2013a, Reference Graystock, Goulson and Hughes2015) showed that honey bees and bumble bees can acquire N. ceranae from flowers visits by other infected bees, there is possibility that Newfoundland bumble bees and honey bees are significantly at risk for this pathogen.
Ascosphera is known to infect the larvae of honey bees and other native bees (Stephen et al. Reference Stephen, Vandenberg and Fichter1981; Evison et al. Reference Evison, Roberts, Laurenson, Pietravalle, Hui and Biesmeijer2012; Wynns et al. Reference Wynns, Jensen and Eilenberg2013; Maxfield-Taylor et al. Reference Maxfield-Taylor, Mujic and Rao2015). Hedtke et al. (Reference Hedtke, Per Moestrup, Jensen and Genersch2011) and Evison et al. (Reference Evison, Roberts, Laurenson, Pietravalle, Hui and Biesmeijer2012) have suggested that non-host organisms may vector the fungal spores. None of this fungus was detected in native bumble bees sampled freely but was found in native species located inside both new and used B. impatiens colonies. Therefore, it appears that Ascosphera can readily move to native species from commercial bumble bees because of the close proximity of bees inside the nests. We can only speculate that drifting bees have the potential to transmit the fungus to other native species outside of the host nest.
The contact between native bumble bees and bees inside the B. impatiens colony boxes increases the likelihood that the native bees will acquire the diseases by pathogen spillover or help propagate diseases by pathogen spillback. These bees then can spread the diseases further when they return to their own colonies or share flowers with other insects. Pathogen spillover from commercial B. impatiens has been documented in other places (see Colla et al. Reference Colla, Otterstatter, Gegear and Thomson2006; Szabo et al. Reference Szabo, Colla, Wagner, Gall and Kerr2012; Murray et al. Reference Murray, Coffey, Kehoe and Horgan2013; Sachman-Ruiz et al. Reference Sachman-Ruiz, Narváez-Padilla and Reynaud2015) and it has been suggested that populations of native eastern North American bumble bee species have been negatively impacted by the commercial B. impatiens. In Canada, B. terricola is listed with “special concern” (Committee on the Status of Endangered Wildlife in Canada 2015) and its cuckoo B. (Psithyrus) bohemicus (Seidl) is listed as “endangered” by the Committee on the Status of Endangered Wildlife in Canada (2014b, 2015). These are two native bumble bee species in Newfoundland and a ban on the importation of new and especially pre-used commercial bumble bee species will be a major step in protecting native bee diversity.
Acknowledgements
The authors thank the following individuals for supplying positive control samples of the diseases: K. Burgher-MacLellan, Agriculture and Agrifood Canada, Kentville, Nova Scotia; V. Tomkies, Fera Science Limited, York, United Kingdom; E. Guzman, University of Guelph, Ontario; I. Meeus, University of Ghent, Belgium; E. Klinger, United States Department of Agriculture-Agricultural Research Service, Pollinating Insects Research Unit, Logan Utah, United States of America. The research was supported by a small research grant from College of the North Atlantic to B.H.
