This project seeks to:
1. Investigate the efficacy of robbing screens as a mechanism for preventing spread of Varroa between colonies.
2. Evaluate the economic feasibility of adopting robbing screens as a Varroa management tactic.
The purpose of this project is to investigate the efficacy and economic feasibility of a novel application of existing beekeeping equipment in reducing the spread of parasitic mites between honey bee colonies. This is an add-on to a funded project investigating the modes and magnitude of mite-driven movement of bees between colonies. Honey bees provide over $14 billion of pollination services in the US. Northeastern farmers producing crops such as apples and blueberries often rely on imported honey bee colonies to bolster pollination and improve yields. Recent high rates of colony losses have strained the supply of colonies needed to meet pollination demands. Beekeepers need effective, economical management options that bolster colony health and decrease mortality.
The leading cause of colony mortality in the US is the parasitic mite Varroa destructor. Organic and synthetic chemical miticides can control Varroa, but must be applied with precise timing and under specific conditions. This is critical in the Northeast because beekeepers have a short window to control Varroa loads before the colonies overwinter, and added pressure from Varroa significantly decreases winter colony survival . Furthermore, long term colony monitoring studies suggest Varroa populations can exceed pretreatment levels in as little as two weeks (Figure 1). This rate of population growth exceeds expected Varroa reproduction[6-8], indicating that mites are entering colonies from external sources, likely on bees from nearby colonies[9-11]. This phenomenon, referred to as horizontal transmission of Varroa, can diminish effects of management performed by beekeepers. Additional treatments, monitoring, and reduced colony health and survivorship are all potential consequences of horizontal transmission. We believe this phenomenon could be interrupted with a novel application of existing beekeeping equipment, the robbing screen (Figure 2).
Robbing screens are placed over the original colony entrance, but allow smells and sounds to pass through, tricking unfamiliar bees into thinking the only colony entrance is blocked by the screen. A notched opening in the top of the screen acts as a new colony entrance that is much less apparent to unfamiliar bees. Only the bees originating from the colony with the robbing screen learn the new entrance. The reduced opening also makes colony defense against robbers easier as their point of entry is restricted. We hypothesize these screens could prevent inter-colony transmission of Varroa by blocking foreign bees from carrying mites into colonies. This would increase efficacy of Varroa control methods implemented by the beekeeper and help ensure their fall treatments remain successful and Varroa loads remain low throughout winter. An economic analysis will aim to demonstrate the cost savings and increase in profits that robbing screen use can provide. This simple intervention has potential to decrease colony mortality in the Northeast, easing the burden on beekeepers and ensuring availability of local colonies for northeastern crop pollination.
This study was performed at the University of Maryland Clarksville Agricultural Research Farm. Nine apiaries of 4 colonies each were established on the farm. Four apiaries were placed ~0.5 miles roughly north, south, east, and west of center. Four additional apiaries were placed in the same direction ~1 mile from center (Figure 1). These 32 receiver colonies were established by making new splits from preexisting colonies at the UMD Beltsville Research Farm. All receivers were made of 4 frames of brood, 2 frames of honey, and with new Italian queens purchased from the same source placed into a single deep. Once the receiver apiaries were moved into place, they received a formic acid treatment (Mite Away Quick Strips) to render them as mite free as possible. Alcohol samples were taken right after the completion of treatment to test starting mite loads.
After the 32 receiver colonies were ready, the central apiary could be established. Two “donor” colonies with Varroa loads of >10 mites/100 bees were placed in a central apiary with two healthy colonies (<3 mites/100 bees). Donor colonies were chosen from pre-existing colonies located at the Upper Marlboro UMD Research Farm. Two colonies with mite loads > 10 mites/100 bees were identified by alcohol wash. These colonies were further stressed by removing honey supers and compacting the entire colony to 2 brood boxes. This was done to maximize the ratio of Varroa mites to adult bees to ensure the mite infestation remained intense. Healthy colonies were chosen from the same location as donor colonies and were also compacted to 2 brood boxes, but these colonies had mite load < 3 mites/100 bees. These 4 colonies were moved into the central location at the experimental station.
The same day the central apiary was moved into place, the bees in those colonies were painted. The donor colony bees were painted red, and the healthy colony bees were painted blue. Painting was achieved by shaking bees of of all frames in the colony into a plastic tub covered with a lid. Once the majority of the adult bees in the colony were in the tub, we scooped ~500 bees at a time into a modified CO2 Varroa tester (https://www.my-honey.nl/winkel/co2-varroa-tester/). The screen was removed from the tester so the mites would not become separated from the bees. Bees were then anesthetized with CO2 and spread onto a flat surface where they could be marked with Sharpie Paint Pens of the appropriate color. We repeated this process 8 times per colony, so that each colony had ~2,000 painted bees at the end of the process.
Once all 36 colonies were in place, sticky boards were placed underneath them to monitor the change in mite load throughout the experiment. Since receiver colonies were new splits, we opted for sticky board monitoring over repeated alcohol washes as we were concerned about removing too many adult bees from the relatively small colonies. Sticky boards were changed approximately every 3 days throughout the duration of the experiment, with some exceptions due to weather. At this time, robbing screens were placed on 2 of the 4 colonies in each receiver apiary. All colonies also had entrance reducers engaged with the larger opening.
Finally, camera sensors were installed on all 36 colonies to detect the presence of painted bees. Cameras consisted of a simple computer (Raspberry Pi 3B+) and camera module (Pi Camera Version 2) (Figure 6a). Cameras were placed over either the opening in the entrance reducer (on colonies without robbing screens) or above the opening in the top of the robbing screen (on colonies with robbing screens) (Figure 2). All colonies had a piece of white cardboard glued to the bottom board underneath the camera to act as a white back drop for photos. Once all cameras were installed, this marked the beginning of the experiment on September 28, 2019. Cameras were turned on daily and sticky boards replaced every three days until the weather became too cold for bees to migrate on November 10, 2019. A final alcohol wash sample for Varroa was taken from all colonies on that date.
Throughout the experiment, the central apiary was checked for whether colonies had collapsed. By the end of the experiment, one of the donor colonies had only a handful of bees remaining and was functionally dead. Adult bees were taken from donor and healthy colonies for viral analysis. Samples were placed immediately on dry ice and then stored at -80 degrees Celsius. Viral samples were taken near the start of the experiment (10/3/19), at the middle (10/15/19), and towards the end (11/5/19). Samples will be analyzed for viral prevalence and copy number using PCR techniques at a later date.
Results presented in this report are preliminary, and statistical analyses are forthcoming.
Despite the extremely intense mite load donor colonies were experiencing, they did not rapidly crash as we had assumed they might. All colonies were technically alive at the end of the experiment after 45 days. However, both donor colonies had decreased significantly in population size (from an average of 10 frames of bees to an average of 2 frames of bees), indicating they could have still donated a significant number of bees to receiver colonies. Photos taken by cameras are still being analyzed. It appears there were some complications, mainly with cameras placed above robbing screens. Some of these cameras were detecting blue and red colors very frequently, when no painted bees were present. This is due to colored objects appearing in the background of photos. Often the white cardboard used as a backdrop can look blue, especially in fading afternoon light. This means some cameras have an exorbitant amount of photos (~75,000). The script is being refined to filter out extraneous photos so that relevant ones, with painted bees present, can be detected.
Even still, from the cameras we have been able to process, we have found photos of painted bees on cameras outside the central apiary. A total of 33 blue bees and 11 red bees have been found in 4 apiaries. So far, all painted bees have been found in apiaries 0.5 miles from the center, and none at apiaries 1 mile from the center. Photos will continue to be analyzed until we feel we have a full picture of the results.
We have started to analyze differences in mite load between colonies which had robbing screens and colonies which did not. Mite loads in both groups increased from the start of the experiment to the end, and colonies with robbing screens have numerically lower mite. Whether these differences are significant will be determined once we perform statistical analyses.
Education & Outreach Activities and Participation Summary
The on-farm demonstration listed was the annual UMD College of Agriculture Farm Tour at the Clarksville Research Farm. At one of the stops on the farm tour, I showed an example of an experimental apiary, how the cameras were set up, and described the methods, need, and expected results of this study. This meeting was attended by a few farmers, but mostly agriculture extension and administrative personnel.
I have given many talks about this study to beekeeping groups. These are attended entirely by beekeepers of varying operation sizes from hobbyist to commercial. The following are the talks I have given since this grant began on September 1st.
- Southern Pennsylvania Beekeepers Group, Hershey PA, n = 75
- Inland Empire Beekeeping Association, Spokane WA, n = 150
- Florida State Beekeepers Association, Orlando FL, n = 20
- Apimondia International Beekeeping Convention, Montreal CAN, n = 50