Abstract
Varroa destructorAnderson and Trueman 2000 (Acari: Varroidae) is an ectoparasitic mite of the honey bee, Apis mellifera L. (Hymenoptera: Apidae). Honey bee colonies require extensive management to prevent mortality caused by varroa mites and the viruses they vector. New miticides (Thymovar and HopGuard) to manage varroa mites were evaluated during the spring and fall treatment windows of the Canadian prairies to determine their effectiveness as part of an integrated management strategy. Thymovar and HopGuard were evaluated alongside the currently used industry standards: Apivar and formic acid. Results demonstrated that Apivar and formic acid remain effective V. destructor management options under spring and fall conditions. Applications of Thymovar during spring were associated with a reduction in brood area, and therefore should be limited to the fall season. The miticide HopGuard was not effective in managing V. destructor, and alteration of the current delivery system is necessary. This study demonstrates the potential for new effective treatment options to supplement currently used V. destructor integrated pest management systems.
Honey bees, Apis mellifera L., are valued throughout the world as managed pollinators and are intrinsic to global food production (Klein et al. 2007). However, without rigorous management, parasitism by the varroa mite, Varroa destructorAnderson and Trueman 2000 (Acari: Varroidae), causes colony mortality and significant economic losses for beekeepers involved in pollination and honey production (Currie et al. 2010). Integrated Pest Management (IPM) is a strategy aimed at discontinuing prophylactic treatments of miticides through the judicious use of pest monitoring methods, economic thresholds, and management tactics so that miticide applications targeting varroa mites are only performed when necessary (Delaplane et al. 2005). Canadian varroa mite populations have twice developed resistance to synthetic miticides, first to the pyrethroid fluvalinate and second to the organophosphate coumaphos (Currie et al. 2010). Currently, the formamidine amitraz is highly effective as a synthetic miticide used in varroa mite IPM in Canada (Nasr et al. 2010). However, in the event that Canadian varroa mite populations develop resistance to amitraz, new management tactics for varroa mite IPM systems are needed.
In western Canada, there are two temporal windows during which colonies can be treated for varroa mites: April–June after overwintering but before the honey production season, and September–November before the onset of winter. Beekeepers commonly use a synthetic miticide in one of these treatment windows and a miticide such as an essential oil or organic acid in the opposite window, as needed (Nasr et al. 2010). The use of a nonsynthetic miticide in the opposite treatment window facilitates the use of a synthetic miticide only once a year, thereby minimizing the development of resistance to synthetic miticides and reducing residues in the colony environment. As honey is a product destined for human consumption, it is important to keep it free of miticide residues; consequently, varroa mite treatments must be removed from the colony before honey production begins. Therefore, treatment options that can be rotated with synthetic miticides while providing a reasonable level of varroa mite management are valued by western Canadian beekeepers.
Treatments incorporating the essential oil thymol are known to cause significant varroa mite mortality (reviewed by Imdorf et al. 1999). Thymovar contains thymol concentrated in cellulose wafers for prolonged activity. Hop (Humulus lupulus L.) extracts have recently stimulated interest as a potential management option for varroa mites (DeGrandi-Hoffman et al. 2012). HopGuard is a miticide containing hop extracts formulated in cardboard strips.
The main objective of this research was to assess the suitability of two new miticides, Thymovar and HopGuard, for use in IPM systems of western Canada. To assess this, it was imperative to determine 1) possible side effects on honey bee brood, 2) efficacy against varroa mites in both the fall and spring treatment windows, and 3) how the new miticides compare with the currently used industry management standards, Apivar (amitraz) and formic acid.
Materials and Methods
Apiary trials were conducted in Edmonton, AB, Canada, and replicated during the fall treatment window of 2011 and the spring treatment window of 2012. Colonies were housed in double-storey Langstroth hives situated on screened bottom boards with a removable tray allowing the varroa mites to fall through to sticky traps. All colonies were supplemented with sugar syrup in division board feeders as necessary, and were wrapped with insulated winter wraps during winter periods.
Colony Assessment.
To assess the strength of the colonies pretreatment, 2 (2011) or 3 (2012) wk posttreatment, and 6 wk posttreatment, all combs in the colony were removed, and both sides were visually assessed to estimate the area of capped brood, bees, and honey to the nearest 30 cm2 (after Skinner et al. 2001). Samples of ≈300 bees were removed from brood combs at each assessment and preserved in 70% ethanol to determine varroa mite infestations of adult bees using the alcohol wash method (De Jong et al. 1982). In the laboratory, samples were agitated on a 175 rpm orbital shaker for 15 min and then rinsed repeatedly with running water so that varroa mites fell through the strainer into a basin and were counted. The number of bees in each sample was then determined by weighing the sample and extrapolating from the weight of three subsets of 10 average bees. Varroa mite infestation was calculated by dividing the number of varroa mites found in the sample by the number of bees in the sample. Varroa mite mortality was also evaluated 2–3 d pretreatment, and periodically throughout the duration of each trial through the use of sticky traps. Varroa mites dying within the colony fell through the screened bottom board onto the sticky trap, which was collected to be counted every 3–7 d and replaced with a new sticky trap. Varroa mite drop per day was calculated by dividing the number of varroa mites found on the sticky trap by the number of days the sticky trap was in place.
Treatment Application.
A complete randomized block experimental design was used when assigning treatments to account for variation in colony strength; colonies were blocked by total brood area as determined by the pretreatment colony assessment. Each of the five treatment groups consisted of eight colonies for a total of 40 colonies at the beginning of each trial. During the fall 2011 trial, all treatments were placed on 5 September; remaining treatments were removed on 17 and 19 October. During the spring 2012 trial, treatments were placed on 7 May and removed 18–20 June. The five treatments used for both trials were—Apivar (Véto-pharma, Villebon-sur-Yvette, France), Thymovar (Pronatex Inc, Richmond, QC, Canada), Hopguard (BetaTec Hop Products, Washington, DC), 65% formic acid, and an untreated control. Apivar was applied once according to label directions, with one strip for every four frames of bees in each brood chamber; the strips were left in place for 6 wk. In the fall 2011 trial, one application of two wafers of Thymovar were applied 4 cm from the periphery of the brood area according to the label in the top brood chamber and left in for 6 wk. In the spring 2012 trial, the first application of Thymovar wafers was replaced with a second application of new wafers after 3 wk. Hopguard was applied once in the fall 2011 trial using the application rate of one strip for every four frames of bees in each brood chamber. The Hopguard strips were reapplied weekly for 3 wk during the spring 2012 trial. For the formic acid treatment, two 40 ml Dri-Loc pads (Sealed Air Corp., Elmwood Park, NJ) saturated with 65% formic acid were applied weekly for 3 wk during both fall 2011 and spring 2012 trials. The formic acid Dri-Loc pads were placed on the near the top bar frame rests in the upper brood chamber opposite to the entrance of the hive.
Finishing Treatment.
To calculate the efficacy of each miticide, the number of varroa mites remaining in the colonies was quantified using a “finishing treatment.” For the fall 2011 trial, a finishing treatment of two applications of oxalic acid was applied once the colonies were broodless, on 14 November and 28 November. The oxalic acid was applied as fumes using the Mitexx machine developed by Nasr et al. (2008). The applicator tip was inserted into the hive entrance, and lengths of burlap were placed in the hive entrance to prevent fume loss. Two grams of oxalic acid applied as oxalic acid tablets (0.5 g each) were placed into the machine where they sublimated (Nasr et al. 2008). The oxalic applicator was removed after ≈60–90 s when fumes started to emit from the top entrance of the wrapped colonies. Sticky traps were inserted to determine varroa mite mortality during the finishing treatment of oxalic acid.
After the spring 2012 trial, the finishing treatment consisted of one strip of Apivar for every four frames of bees, left in the colonies for 6 wk as per label instructions. The mortality of varroa mites during the finishing treatment was assessed by counting the varroa mites falling to the sticky traps during the finishing treatment of Apivar.
Statistical Methods.
Results
Impact of Treatment on Brood.
There were no significant differences among treatment groups in the amount of brood present in the colonies before treatment in the fall 2011 trial (Fig. 1) or the spring 2012 trial (Fig. 2). The total area of brood 6 wk posttreatment during the spring 2012 trial was significantly higher in the formic acid treatment group than the Apivar group (F=3.47, df=4, 29, P=0.0195).
Average (±SE) total brood area throughout the fall 2011 trial. Different letters indicate significant differences between treatments within each sampling date according to one-way ANOVA followed by Tukey means separation (P<0.05).
Average (±SE) total brood area and brood area in the top chamber throughout the spring 2012 trial. Different letters indicate significant differences between treatments within each sampling date according to one-way ANOVA followed by Tukey means separation (P<0.05).
Three weeks posttreatment in the spring 2012 trial, the area of brood in the top chamber was significantly differently among treatments (F=19.67, df=4, 30, P < 0.0001), with ≈86 ± 7% less brood in the Thymovar group on average than the control (Fig. 2). This trend was also evident in the total brood area, which varied significantly among treatment group means (F=5.99, df=4, 30, P=0.0011), with the average total brood area in the Thymovar colonies 52 ± 14% less than the control (Fig. 2).
At 6 wks posttreatment in the spring 2012 trial, the brood area in the top chamber was also significantly different among treatments (F=7.59, df=4, 30, P=0.0002). The brood area in the top chamber for the Thymovar colonies averaged 77 ± 13% less than the control colonies (Fig. 2). The total brood area also varied significantly among treatment means 6 wk posttreatment (F=2.71, df=4, 30, P=0.0487).
Impact of Treatment on Varroa Mite Infestation.
Average varroa mite infestation before treatment in the fall 2011 trial was 3.77 ± 0.62% and was significantly higher in the colonies subjected to the Thymovar treatment than the other groups (F=3.63, df=4, 33, P=0.0161, Tukey). Repeated measures analysis showed a significant effect of treatment over time on varroa mite infestations (F=15.55, df=4, 28, P < 0.0001), with significantly lower infestations observed over the course of the fall treatment period in the Apivar, formic acid, and Thymovar groups than in the control or HopGuard colonies (Fig. 3).
Average (±SE) percent varroa mite infestations in response to treatments throughout the fall 2011 trial.
Before treatment in the spring 2012 trial, the average varroa mite infestation was 4.49 ± 0.57% and did not vary significantly by treatment (F=0.92, df=4, 35, P=0.4640). Repeated measures analysis indicated that treatment affected the varroa mite infestation (F=18.4, df=4, 30, P < 0.0001) during the spring 2011 trial when all treatment groups were significantly less infested over time than the control (Fig. 4).
Average (±SE) percent varroa mite infestations in response to treatments throughout the spring 2012 trial.
Impact of Treatment on Varroa Mite Drop.
Varroa mite drop was ≈21 per day before treatment in September 2011, increased substantially to ≈200–300 per day after the first application in all treated groups, and then fell to below 100 mites per day after the first week in the HopGuard and control groups, while peaking with reapplications in the formic acid group and fluctuating in colonies subjected to Thymovar and Apivar. By 6 wk posttreatment, mite drop in the control colonies averaged ≈100 per day (Fig. 5).
Average varroa mite drop per day (±SE) in response to treatments throughout the fall 2011 trial. Arrow 1 indicates the time of application of all treatments; arrows 2 and 3 indicate timing of reapplication of formic acid. A temperature graph indicating the average daily temperature at the research site is included for reference.
In early May of 2012, mean varroa mite drop levels were ≈33 per day before treatment for all colonies, but following treatment application varroa mite drop increased rapidly to ≈150–200 per day for all colonies except the controls. Spikes in mite drop for 3 d were evident at each reapplication of HopGuard. From mid-May to mid-June, mean mite drop numbers gradually declined in all colonies except the controls (Fig. 6).
Average (±SE) varroa mite drop per day in response to treatments throughout the spring 2012 trial. Arrow 1 indicates the application of all treatments, arrow 2 indicates the time of reapplication of HopGuard and formic acid, and arrow 3 indicates reapplication of HopGuard, formic acid, and Thymovar. A temperature graph indicating the average daily temperature at the research site is included for reference.
Varroa mite drop did not vary significantly among treatment means before treatment during the fall 2011 trial (F=0.90, df=4, 33, P=0.4743) or the spring 2012 trial (F=0.66, df=4, 35, P=0.6236). Repeated measures analysis showed that treatment affected varroa mite drop over time during the fall 2011 trial (F=4.18, df=4, 28, P=0.0089; Fig. 5) and the spring 2011 trial (F=8.05, df=4, 30, P=0.0002; Fig. 6).
Efficacy of Treatments.
Treatment efficacies were as follows for the fall 2011 trial: Apivar (87.07 ± 2.69%), formic acid (78.48 ± 8.47%), Thymovar (88.91 ± 8.47%), HopGuard (42.96 ± 6.46%), and the control (28.69 ± 7.33%). Treatment efficacies were as follows for the spring 2012 trial: Apivar (74.93 ± 3.18%), formic acid (71.90 ± 6.52%), Thymovar (82.33 ± 3.32%), HopGuard (43.56 ± 3.18%), and the control (24.09 ± 3.89%).
Discussion
Both trials confirmed that HopGuard causes insufficient varroa mite mortality to be of use for IPM in western Canada. Thymovar was not found to be suitable for the spring application window, but functioned effectively in the fall application window. Apivar and formic acid remain integral components of varroa mite IPM in western Canada.
Due to an absence of any lasting activity in the fall 2011 trial, HopGuard was applied three times in the spring 2012 trial. In both trials it was evident that HopGuard activity was limited to 3 d after application; the bees removed much of the cardboard strip substrate during this time. This result is consistent with other studies involving oils formulated in cardboard strips which were shredded by the bees within 1 wk (Skinner et al. 2001). The findings in this study regarding HopGuard are significant, as it is the first study that involves the application of HopGuard to colonies with brood. DeGrandi-Hoffman et al. (2012) applied HopGuard to five-frame colonies with caged queens and found that the majority of varroa mite mortality occurred within 2 d of application, and they recommended reapplications in the presence of brood. The efficacies observed for HopGuard during both trials of this study were much lower than the efficacy of 93.5% reported for HopGuard applied to winter colonies with no brood (Rademacher and Harz 2011). Although large initial varroa mite drop (Figs. 5 and 6) was observed within 3 d of treatment with HopGuard, it was generally not sufficient to provide any longer-term varroa mite management statistically different from the control colonies even with repeated applications during the spring 2012 trial.
HopGuard does not cause sufficient varroa mite mortality to be effective in its present formulation. Although a cardboard strip formulation has potential in niche areas such as in the treatment of packages of bees (DeGrandi-Hoffman et al. 2012), or broodless colonies (Rademacher and Harz 2011), our study shows that it is limited in its application to large reproducing colonies. It is possible that repeated applications of HopGuard (likely at least six) could cause sufficient varroa mite mortality. However, such a demanding treatment schedule is unlikely to be favored by commercial beekeepers, and current label recommendations for HopGuard preclude more than three treatments per year. As our data show that HopGuard is capable of causing substantial initial varroa mite mortality, a delivery system that could deliver the hops beta acids over an extended period of time would be more effective as a colony miticide.
In both trials it was evident that Thymovar was effective for managing varroa mite populations. The varroa mite management with the Thymovar treatment was statistically similar to Apivar in terms of varroa mite infestation, varroa mite drop, and overall efficacy. However, treatment with Thymovar caused a drastic reduction in brood production that was most evident in the top brood chamber during the spring 2012 trial (Fig. 2). Floris et al. (2004) also found substantial brood reduction while using other thymol-based miticides and recommended that they not be used during times when colonies are population building. Removal of honey and brood directly below the Thymovar wafer was reported by Baggio et al. (2004). As the intensity of brood rearing in spring is directly correlated with honey production in summer (Szabo and Lefkovitch 1989), the brood reduction seen in this study (51% reduction in total brood 3 wk after treatment) would undoubtedly have a negative effect on summer colony performance and honey yields.
The reduction in brood rearing in response to the treatment of Thymovar in the spring 2012 trial is concerning; therefore, it is not recommended for the western Canadian spring treatment window. However, there was no significant effect of Thymovar on brood production in the fall 2011 trial. Therefore, it is likely that Thymovar can be used safely in the fall treatment window, as any effects on brood production would be minimal due to an overall decline in brood production and tendency for the honey bee cluster to move to the bottom brood chamber during this time.
It is likely that the finishing treatments used in both trials underestimated the efficacy of the treatments. To maximize the chances of killing all the remaining mites, oxalic acid could not be applied until 6 wk after the fall 2011 trial when the colonies were broodless, thus leaving a period in which varroa mite populations could recover. Similarly, according to label directions, the finishing treatment of Apivar was applied for a full 6 wk after the spring 2012 trial, thereby giving substantial opportunity for varroa mites in nearby infested colonies to reinvade the experimental colonies (Kraus and Page 1995, Gregorc and Planinc 2005). The varroa mite reinvasion in tandem with the reproductive potential of colonies during spring build-up undoubtedly allowed for elevated finishing treatment mortality during the 6 wk application of Apivar. While both finishing treatments used allowed for some degree of mite population growth, their timing maximized the mortality of most of the remaining varroa mites in the colony and thereby provided realistic efficacy data.
In summary, this study determined that two widely used miticides, formic acid and Apivar, remain effective management options for both treatment windows on the Canadian prairies. Thymovar is recommended for use during the fall season only, when bees are transitioning to the bottom brood chamber, and brood production is decreasing. Although HopGuard holds promise for the integrated management of varroa mite infestations in honey bee colonies, an alternative delivery system is necessary to increase its effectiveness.
This article is dedicated to the memory of Lloyd Dosdall who passed away during the publication of this research. We thank Samantha Muirhead, Anna Murray, Ali Panasiuk, Eric Papsdorf, Charlotte McCartan, and Keith Pudwill for their technical support. We also thank Rong-Cai Yang for assistance with statistical analyses. The Natural Sciences and Engineering Research Council of Canada awarded L. P. Vandervalk an Alexander Graham Bell Graduate Scholarship. This research was funded in part by grants from the Alberta Crop Industry Development Fund (ACIDF), the Beekeeping Commission of Alberta, Alberta Agriculture and Rural Development, BeeMaid Honey, Southern Alberta Beekeeper's Association, Poelman Apiaries, Canadian Bee Research Fund, Bayer CropScience, and Pioneer Hi-Bred.
References
