Introduction
Honey bees, Apis mellifera Linnaeus (Hymenoptera: Apidae), are subject to injury from numerous diseases, pests, and parasites. A parasite of great concern is Nosema spp. (Nosematidae), a genus of globally distributed microsporidian parasites (Chen et al. Reference Chen, Evans, Smith and Pettis2008; Schwarz et al. Reference Schwarz, Huang and Evans2015; Holt and Grozinger Reference Holt and Grozinger2016). Based on recent molecular phylogenetic work, many species formerly classified as Nosema are being reassigned to the genus Vairimorpha (Tokarev et al. Reference Tokarev, Huang, Solter, Malysh, Becnel and Vossbrinck2020), but herein they are referred to as Nosema. Two species of Nosema that commonly infect European honey bees are N. apis and N. ceranae (Higes et al. Reference Higes, Meana, Bartolomé, Botías and Martín-Hernández2013; Schwarz et al. Reference Schwarz, Huang and Evans2015). Infection of honey bees by Nosema spp. (herein referred to as Nosema) causes nosemosis, a disease resulting in hives with lower number of bees (Botías et al. Reference Botías, Martín-Hernández, Barrios, Meana and Higes2013), impaired health (Mayack and Naug Reference Mayack and Naug2009), and poor colony performance (Higes et al. Reference Higes, Martín-Hernández, Botías, Bailón, González-Porto and Barrios2008; Goblirsch Reference Goblirsch2018). Nosemosis can be especially concerning during overwintering, where the additional stress can further increase the likelihood of colony loss via suppressing the honey bee immune system (Higes et al. Reference Higes, Martín-Hernández, Botías, Bailón, González-Porto and Barrios2008; Antúnez et al. Reference Antúnez, Martín-Hernández, Prieto, Meana, Zunino and Higes2009).
Stressful conditions such as transportation to the crop field and conditions experienced during pollination may exacerbate disease problems for honey bees (Zhu et al. Reference Zhu, Zhou and Huang2014). The relationship between stress and honey bee disease has been well documented, although the impacts of using and moving honey bees for crop pollination are not always consistent amongst and within studies (Zhu et al. Reference Zhu, Zhou and Huang2014; Alger et al. Reference Alger, Burnham, Lamas, Brody and Richardson2018; Dolezal and Toth Reference Dolezal and Toth2018). For example, Cavigli et al. (Reference Cavigli, Daughenbaugh, Martin, Lerch, Banner and Garcia2016) found that the prevalence of 16 pathogens was highest in honey bees immediately after they had been sent to almond pollination. In a second study, in which colonies were transported 8600 km across the contiguous United States of America, migratory colonies had approximately 20% fewer bees than did stationary colonies immediately after transport, with the effect persisting through at least one month following return (Alger et al. Reference Alger, Burnham, Lamas, Brody and Richardson2018). However, the same study revealed complexity amongst end points, finding no difference in the prevalence of deformed wing virus (Iflaviridae) between migratory and stationary bees and finding that varroa mite (Mesostigmata: Varroidae) loads were approximately 60% lower in migratory colonies compared to in control hives one month after return. A study examining hives transported for almond pollination in California, United States of America, found decreased lifespan in adult bees and increased oxidative stress compared to colonies that remained stationary (Simone-Finstrom et al. Reference Simone-Finstrom, Li-Byarley, Huang, Strand, Rueppell and Tarpy2016).
Many of the studies that explored the impacts of pollination stress on honey bees occurred in production systems where honey bees are transported hundreds to thousands of kilometres for pollination services (e.g., Cavigli et al. Reference Cavigli, Daughenbaugh, Martin, Lerch, Banner and Garcia2016; Alger et al. Reference Alger, Burnham, Lamas, Brody and Richardson2018). In the Canadian Maritime provinces of New Brunswick, Nova Scotia, and Prince Edward Island, honey bees are primarily used for pollinating lowbush blueberry, Vaccinium angustifolium (Ericaceae), although many beekeepers also sell bee products (e.g., wax and honey). Hives in this region are typically moved relatively short distances (less than 100 km) from the apiary and remain in blueberry fields for only 2–3 weeks before returning to noncrop apiary settings (McCallum and Cutler, unpublished data). The negative consequences for overall colony health may be less severe in the case of lowbush blueberry pollination compared to other pollinator-dependent crops in North America due to the shorter distance travelled and the shorter time period spent in blueberry fields.
At least two studies specifically examined the impacts of blueberry (Vaccinium spp.) pollination on honey bee health. Grant et al. (Reference Grant, DeVetter and Melathopoulos2021) found that sending bees to blueberry pollination increased the occurrence of European foulbrood disease by 41% and 53% during two field seasons. A second study, based in Québec, found that sending bees to pollinate lowbush blueberry significantly reduced brood production, which was potentially related to increased N. ceranae (Dufour et al. Reference Dufour, Fournier and Giovenazzo2020). Such effects may be due to the low nutritional quality of Vaccinium pollen (14.9% crude protein; Somerville et al. Reference Somerville, Nicol, Somerville and Nicol2006), poor weather conditions sometimes experienced during blueberry bloom (Tuell and Isaacs Reference Tuell and Isaacs2010), environment–pathogen interactions (Dufour et al. Reference Dufour, Fournier and Giovenazzo2020), or exposure to pesticide residues (Drummond et al. Reference Drummond, Lund and Eitzer2021).
Understanding how pollination practices affect Nosema prevalence could be useful in helping beekeepers make informed management decisions. In this study, we investigated whether sending hives to lowbush blueberry pollination would increase Nosema spore loads in honey bees. We predicted increased Nosema spore loads in spring and autumn, due to the natural life cycle of Nosema, where spore loads tend to be higher in the colder months when bees cannot leave the hive for cleansing flights and where older, infected bees die during the summer months and do not transmit spores to newly developed bees (Bailey Reference Bailey1955). We also predicted that Nosema spore loads would increase following movement of hives into blueberry fields for pollination. Based on recent studies documenting proliferation of N. ceranae across Canada (Williams et al. Reference Williams, Shafer, Rogers, Shutler and Stewart2008; Emsen et al. Reference Emsen, Guzman-Novoa, Hamiduzzaman, Eccles, Lacey, Ruiz-Perez and Nasr2016), we anticipated N. ceranae would be more prevalent than N. apis in our samples.
Material and methods
Data collection
The study was conducted from April to September 2020. Eleven beekeepers participated in the study, managing 12 beekeeping operations throughout the Canadian provinces of New Brunswick, Nova Scotia, and Prince Edward Island. Beekeepers were selected owing to their adherence to best management practices (e.g., for feeding, overwintering, and disease and pest management) and research experience. All beekeepers had previously partnered with the research team on applied research projects, and all hives included in the study met the regional pollination standard for colony strength (i.e., eight frames of bees, four frames of brood with 100% coverage, two frames of honey, and one laying queen; Atlantic Tech Transfer Team for Apiculture 2020). Due to restrictions posed by the COVID-19 pandemic, we were not able to standardise hives for similar levels of Nosema infection at the onset of the study.
Apiaries were distributed across two different treatments. In the first treatment, or “blueberry hives,” all hives within the apiary (n = 8) were rented out for lowbush blueberry pollination (June) before being returned to the home apiary for the remainder of the study (until September). In the second treatment, or “home hives,” all hives within the apiary (n = 4) remained at the beekeepers’ home apiaries for the duration of the study. In the case of “home hives,” bees could freely forage on floral resources within the surrounding landscape that would have included wild plants (e.g., goldenrod (Asteraceae), brambles, rhodora (Ericaceae)), agricultural crops such as forage feed (e.g., clover (Fabaceae), corn (Poaceae), and soybeans (Fabaceae)), and residential gardens. These home apiaries were at least 10 km away from commercial blueberry fields – outside of the foraging range of honey bees. None of the studied hives were treated with Fumagilin-B®, a registered chemical treatment for nosemosis, for at least 12 months before the study began. Otherwise, the studied hives were subject to standard best management practices, including optimal overwintering protection and preparation, spring feeding (i.e., sugar syrup and pollen patties), varroa mite management, and young, healthy, and vigorous queens. The economic disadvantage of keeping honey bee colonies “home” from pollination meant we could only secure four apiaries for the home hive treatment.
Planned monthly data collection visits at each site were not possible due to COVID-19 pandemic travel restrictions. Therefore, participating beekeepers collected monthly samples using sampling kits that we provided with detailed sampling and shipping instructions (Supplementary material, Instructions provided to participating beekeepers). Beekeepers were asked to collect one-quarter cup (approximately 60–65 mL) of bees from the inner cover (or outer frames, if an inadequate number of bees was found on the inner cover) of each labelled hive in their yard monthly. These samples were stored at –18 °C until shipping and at the same temperature upon receipt at our laboratory until processing could occur.
Twelve hives were labelled within each apiary, and data were collected repeatedly from each hive monthly from April to September. We considered each hive as an independent experimental unit. Over the six-month data collection period, we anticipated 48 monthly samples from home hives and 96 monthly samples from blueberry hives, resulting in a total of 288 samples collected from the home hives and 576 samples collected from the blueberry hives. Hives in the blueberry treatment were sent to lowbush blueberry pollination on different dates across the Maritimes, due to differences in bloom period across the region (4–30 June 2020) and remained in the field for 2–3 weeks.
Quantifying Nosema spores
We used standardised methods previously described in McCallum et al. (Reference McCallum, Olmstead, Shaw and Glasgow2020) to quantify Nosema spores in bees from each hive. To release contents from guts of bees, 30 bees were crushed within a sealable plastic bag. Using an eyedropper, 5 μL aliquots of the resulting liquid were deposited into each well of a standard haemocytometer (Reichert Bright-Line, Improved Neubauer, 0.1 mm depth; Hausser Scientific, Horsham, Pennsylvania, United States of America), and the eyedropper was thoroughly cleaned between samples. The eyedropper was used to collect 5 μL of clean water and was observed for any possible spore carryover every few samples as an extra precaution to avoid spore contamination between samples.
The spores were counted under 400× magnification. In each well of the haemocytometer, total number of spores in each corner and the centre square was counted. The mean spore load was calculated following Cantwell (Reference Cantwell1970).
Nosema identification using polymerase chain reaction
After quantifying Nosema spore loads, the remaining bees from the original sample were then prepared and processed for a second test. We removed 50 bees (sample size recommended by the National Bee Diagnostic Centre, Beaverlodge, Alberta, Canada) from the samples with the highest spore loads to ensure ample spore abundance for optimal Nosema species representation (Nosema apis versus Nosema ceranae). These samples were shipped frozen on wet ice to the National Bee Diagnostic Centre.
Nosema species identification performed by the National Bee Diagnostic Centre followed protocols developed by Hamiduzzaman et al. (Reference Hamiduzzaman, Guzman-Novoa and Goodwin2010) and Gisder and Genersch (Reference Gisder and Genersch2013). Nosema species were identified using polymerase chain reaction with a Multiplex Supermix (Bio-Rad, Hercules, California, United States of America). Briefly, amplification assays were executed by engaging 60 ng of genomic DNA and 0.4 ng of each primer in a Veriti thermal cycler (Thermo Fisher Scientific, Waltham, Massachusetts, United States of America). The protein-coding gene RPS5 was used as a reference housekeeping gene. The polymerase chain reaction conditions were as follows: 95 °C for 5 minutes, followed by 35 cycles of 1 minute at 94 °C, 1 minute at 58 °C, 1 minute at 72 °C, and 7 minutes at 72 °C. Amplification products were separated by 1% agarose gel stained with SYBR Safe (Thermo Fisher Scientific) and were finally observed under ultraviolet and blue light illumination.
Statistical analysis
We used a completely randomised design, with the factor of interest being pollination with two levels – blueberry and home – and with hives being the experimental units. The number of replications for blueberry was n 1 = 96 and for home was n 2 = 48. Because the response values (Nosema spores in millions) were measured repeatedly (monthly), repeated measures analysis was completed to determine the effect of pollination on Nosema spores and how the effect changed during the six months. Akaike information criterion (Littell et al. Reference Littell, Henry and Ammerman1998) was used to determine the most appropriate co-variance structure to be “unstructured.” The repeated measures analysis was completed using the “mixed procedure” of SAS (SAS Institute, Inc. 2014). The validity of model assumptions (normal distribution and constant variance of the error terms) was verified by examining the residuals as described in Montgomery (Reference Montgomery2020), which showed the normal distribution assumption was violated in the original data. However, a fourth root transformation that was applied to Nosema spore values met the assumption. The P-value for the interaction between pollination and month effect was 0.006; therefore, multiple means comparison was conducted using the “lsmeans” statement of Proc Mixed at the 5% level of significance to generate letter groupings. The means reported in Fig. 1 are back-transformed to the original scale.
Mean Nosema spp. spore loads in honey bee hives across three Maritime provinces, Canada, during spring and summer of 2020. Error bars represent standard deviation. Means sharing the same letter are not significantly different at the 5% level of significance. “Blueberry” hives were brought to lowbush blueberry fields during blueberry bloom, whereas “home” hives remained at an apiary. The area highlighted in yellow (blueberry bloom) represents the time when lowbush blueberry fields were flowering.
We also described trends in Nosema species identification using percentages of detections (e.g., percentage of N. ceranae detected in total samples submitted) as a function of all samples submitted for detection by polymerase chain reaction.
Results
Nosema spore loads in blueberry pollination versus home apiary treatments
We found a significant interaction (F 5,139 = 3.40, P = 0.006; Supplementary material, Table S1) between pollination treatment and month. Nosema spore counts in home hives decreased sharply from April to June, whereas spore counts in blueberry hives increased from April to May and then decreased from May to June (Fig. 1). During the first sampling period (April), Nosema spore loads in prepollination blueberry hives (1.70 ± 0.12 million spores, mean ± standard deviation) were 44% lower than those in home hives (3.12 ± 0.09 million spores, mean ± standard deviation). Spore loads in prepollination blueberry hives more than doubled in May samples but did not significantly differ from spore loads from home hives during that period (Fig. 1). In June, blueberry hive Nosema spore loads (1.04 ± 0.15 million spores, mean ± standard deviation) were 170% greater than spore loads in home hives (0.39 ± 0.22 million spores, mean ± standard deviation; P < 0.05; Supplementary material, Table S2). Spore loads through the remainder of the year remained low and consistent (range 0–0.112 million spores) among all hives (Fig. 1).
Nosema species identification
We had anticipated submitting 72 samples to the National Bee Diagnostic Centre for Nosema species identification, but only 54 samples across the six collection periods were submitted by beekeepers (36/54 = blueberry hives, 18/54 = home hives). Of the samples analysed, 76% (41/54) tested positive for N. ceranae (Fig. 2). Only 2% of samples (1/54) tested positive for N. apis. One sample (1/54) was positive for both N. apis and N. ceranae. Twenty per cent of samples (11/54) were negative for Nosema.
Nosema spp. detection via polymerase chain reaction from honey bees from hives used in blueberry pollination “Blueberry” and hives that remained at a home apiary, “Home,” in the Canadian Maritime provinces during spring and summer 2020.
Discussion
Consistent with our predictions and with other published studies (e.g., Traver et al. Reference Traver, Williams and Fell2012; Dufour at al. Reference Dufour, Fournier and Giovenazzo2020; McCallum et al. Reference McCallum, Olmstead, Shaw and Glasgow2020), Nosema spore loads were highest in the spring (April and May) and decreased over the summer months. Bees leave their hive more often as seasonal temperatures rise, resulting in more cleansing flights and less faecal matter in the hive (Winston Reference Winston1987; Retschnig et al. Reference Retschnig, Williams, Schneeberger and Neumann2017). With less infected faecal matter around the colony, uninfected bees have reduced exposure and probability of Nosema infection. Spore loads in April–May frequently exceeded the economic threshold of 1 million spores per bee, reaching levels above 25 million spores per bee, a level nearly 25 times higher than untreated hives in a regional study conducted in 2018–2019 (McCallum et al. Reference McCallum, Olmstead, Shaw and Glasgow2020). Differences between spore loads in these studies could be due to a variety of biotic (e.g., genetic differences in brood stock) or abiotic (e.g., weather differences) factors.
By July, the mean spore load was below the economic threshold and remained low until September, which is consistent with many other Canadian studies (e.g., Copley et al. Reference Copley, Chen, Giovenazzo, Houle and Jabaji2012; McCallum et al. Reference McCallum, Olmstead, Shaw and Glasgow2020). Contrary to our expectations, we did not detect an increase in Nosema spore loads during the September collections, which is consistent with results by Punko et al. (Reference Punko, Currie, Nasr and Hoover2021). This may be because September 2020 was warmer than historical temperature averages for the region. Data from three weather stations close to the apiaries (Kentville, Nova Scotia; Mactaquac Provincial Park, New Brunswick; and New Glasgow, Prince Edward Island) show that the mean temperatures during September 2020 were 1.0 °C higher than the 1981–2010 mean (Supplementary material, Table S3). In warmer weather, continued regular cleansing flights may delay the onset of enhanced Nosema spore numbers.
We expected Nosema spore loads to be greater in blueberry hives than in home hives following blueberry pollination, but this was not observed. We did see a sharp increase in Nosema spore counts in blueberry hives in May (coinciding with a decrease in spore counts from home hives). However, this was well before the hives went to blueberry fields for pollination, and therefore, the act of hives being moved to or residing in blueberry fields did not cause this increase in spore counts. We have no quantitative or qualitative information pointing to differences in home hive versus blueberry hive management before movement that could explain the spike in Nosema spore counts in May. Beyond specific treatments for Nosema, the use of pollen substitutes has also been shown to potentially affect Nosema incidence. Pollen substitutes (i.e., pollen patties) are protein-rich formulations used for stimulating brood production and colony growth, and use of these has been shown in some cases to correspond to increased Nosema incidence (DeGrandi-Hoffman et al. Reference DeGrandi-Hoffman, Chen, Rivera, Carroll, Chambers and Hidalgo2016; Jack et al. Reference Jack, Uppala, Lucas and Sagili2016). In our study, beekeepers were not asked to provide information regarding the details of supplemental feeding, but blueberry hives and home hives both would have received pollen patties of varying origin and composition. Further exploration of the interaction between early-season supplemental feeding and Nosema infections would be useful in understanding how Nosema infections are affected by management.
Numerous biological and management factors can affect disease pressures in honey bee colonies, and the results of studies examining the effects of hive transport for crop pollination have been variable. A study by Zhu et al. (Reference Zhu, Zhou and Huang2014) found that Nosema spore loads were 2.5-fold greater in hives that were transported 275 km for highbush blueberry pollination relative to hives that remained stationary. Migratory colonies transported between bee yards in Spain had greater Varroa and N. ceranae loads than stationary colonies did in the period following initial transportation but a lower viral load of deformed wing virus (Jara et al. Reference Jara, Ruiz, Martín-Hernández, Muñoz, Higes and Serrano2020). Although Simone-Finstrom et al. (Reference Simone-Finstrom, Li-Byarley, Huang, Strand, Rueppell and Tarpy2016) did not examine specific diseases, these authors found transporting hives for pollination had a significant negative impact on adult bee lifespan. The results of our study, compared to others, suggest transportation or placement of honey bee hives in agricultural settings for pollination can have variable effects on pathogen loads.
More than three-quarters of our samples contained N. ceranae, with only two samples (4%) testing positive for N. apis. In our region (Maritime Canada), N. ceranae was first detected in samples collected in 2006 (Williams et al. Reference Williams, Shafer, Rogers, Shutler and Stewart2008), but it is possible that N. ceranae has been established for a longer time, having been misidentified as N. apis. Dominance of N. ceranae has been observed in numerous other studies (e.g., Klee et al. Reference Klee, Besana, Genersch, Gisder, Nanetti and Tam2007; Chen et al. Reference Chen, Evans, Smith and Pettis2008; Williams et al. Reference Williams, Shafer, Rogers, Shutler and Stewart2008; Emsen et al. Reference Emsen, Guzman-Novoa, Hamiduzzaman, Eccles, Lacey, Ruiz-Perez and Nasr2016; McCallum et al. Reference McCallum, Olmstead, Shaw and Glasgow2020) and has become a common honey bee pathogen around the world (Grupe and Quandt Reference Grupe and Quandt2020). Although the global spread of N. ceranae was caused by human actions, the displacement of N. apis by N. ceranae may also be due to the ability of N. ceranae, but not of N. apis, to infect wild hosts (Martín-Hernández et al. Reference Martín-Hernández, Botías, Barrios, Martínez-Salvador, Meana and Mayack2011), a faster reproduction rate of N. ceranae compared to N. apis (Williams et al. Reference Williams, Shutler, Burgher-MacLellan and Rogers2014), combined with the ability of N. ceranae to thrive across a wide range of environmental conditions (Martín-Hernández et al. Reference Martín-Hernández, Bartolomé, Chejanovsky, Le Conte, Dalmon and Dussaubat2018). Given that only one antimicrobial product is currently available to treat nosemosis (Fumagilin-B®; Can-Vet Animal Health Supplies Ltd., Guelph, Ontario, Canada; DIN 02231180), it is important to know the identity and quantity of the parasite for effective treatment.
Due to COVID-19 restrictions, we relied on cooperating beekeepers to collect and send samples, using standardised sampling methods and frequent communication with beekeepers. This was largely effective, but we did experience some challenges. Some hives were excluded from the study throughout the season if they swarmed, became queenless, became weak, or died, or if the hive label was lost, and sometimes beekeepers did not submit records as to why a hive was removed from the study. This means we would not have known if a hive failed due to Nosema pressure. Inconsistency of sample submissions by beekeepers also would have led to some degree of variability in the seasonal variation in Nosema spore loads, although reporting was consistent from May to July, which marked the point before and after bees were moved to and from lowbush blueberry fields (Supplementary material, Table S2). In addition, due to shipping delays associated with COVID-19, receipt of some samples was delayed and not every sample had enough bees, precluding Nosema species diagnostics by the National Bee Diagnostic Centre; however, the planned large sample size helped to mitigate this challenge. Logistical issues associated with the COVID-19 lockdowns meant that samples could not be sent on dry ice, and we cannot rule out whether potential degradation of sample quality during shipping affected our results and interpretation.
We found no evidence that movement within Canada’s Maritime provinces of honey bee colonies to lowbush blueberry fields for pollination affects Nosema spore loads. Future studies might consider how these relatively small-scale movements of hives from apiaries to lowbush blueberry fields, and associated management practices (i.e., supplemental feeding), might affect other end points of honey bee health (e.g., colony growth, brood and honey production, and disease and pest pressures). Because most research on movement and pollination stress focuses on higher-impact (e.g., longer-distance) hive migrations, future research efforts could benefit from examining smaller-scale movements of bees for use as commercial pollinators. Improving understanding of the consequences that sending bees for pollination has on bee health, hive productivity (e.g., honey production), and resilience (e.g., overwintering success) could be useful for beekeepers when deciding whether to participate in pollination.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.4039/tce.2022.30.
Acknowledgements
The authors thank the beekeepers who dilligently collected and mailed in the samples for this project; without their efforts, this project would not have been possible. The authors also thank two anonymous reviewers and the editorial team at The Canadian Entomologist for their constructive comments and thorough critiques of earlier versions of this manuscript. Funding for this research was provided by the Sobeys Undergraduate Research Award and by the Atlantic Tech Transfer Team for Apiculture, which is supported by the Canadian Agricultural Partnership, the provincial governments of New Brunswick, Nova Scotia, and Prince Edward Island, Bleuets NB Blueberries, New Brunswick Beekeepers Association, Nova Scotia Beekeepers Association, Wild Blueberry Producers’ Association of Nova Scotia, Prince Edward Island Wild Blueberry Growers Association, and the Prince Edward Island Beekeepers Association. The authors thank P. Wolf-Viega and her team at the National Bee Diagnostic Centre for Nosema species identification and J. Harrison and A. Byers for their work with the Atlantic Tech Transfer Team for Apiculture during this project.
Competing interests
The authors declare no competing interests.
Open access
