Robbing screens can clearly act as a diverting obstacle to bees (foreign or natal) attempting to enter a hive, causing them to “give up and go away.” But is that enough to completely deter investigating honey thieves?

Since discovering that robbing screens can (to our great surprise) prevent a substantial proportion of returning young foragers from figuring out how to re-enter their own hive, I was now curious as to how effective they were at deterring potential “robbers.” So I shifted our research to try to answer three questions:

Question #1: How effective are robbing screens at thwarting the perseverance of investigating scouts (potential honey thieves) — attracted to the scent emanating through a robbing screen — until they “give up and go away”?

Question #2: If a scout were indeed successful at entering through a robbing screen and then returned home to recruit additional foragers to seek that scent, could the screen then deter enough of the recruits to prevent mass robbing from occurring?

And then there’s the practical application:

Question #3: Which design of robbing screen is most effective?

We addressed these questions in separate experiments and investigations.

Experiment #1

Scientific design: So how does one go about testing robbing screens? It would be difficult (and expensive) to do so with hives full of bees, due to the biological variables of a colony’s attractiveness to robbers, and its strength and defensiveness. So I decided to test the effectiveness of robbing screens with us controlling the olfactory attractiveness, and without any colony (or its guard bees) present.

We began by day-to-day training of the foragers in the yard to the scent of anise, by open-feeding ~1/2 cup at a time of anise-scented 50% sugar syrup in small trays at my home apiary (with over 50 hives of bees scattered at various locations on the property). The result was that soon, anywhere on the property that we exposed even a whiff of anise scent, investigating scouts or scent-foragers would quickly appear — especially just after the syrup in the trays had all been consumed (we’d refill the trays from time to time throughout the day).

To test whether robbing screens could deter those “investigators” from gaining entrance into an empty hive body that was emitting the attractive scent (henceforth referred to as “hives”), I first tried building various designs of screened traps fitted inside the hive to count any entered bees — but the bees stubbornly refused to enter them. So we instead placed a framed Plexiglas viewing window over each test hive so that we could simply count the entered bees visually. Inside each hive we set a cotton pad with a drop of anise extract, next to a tiny computer fan to gently blow the fumes out of the entrance (Figures 1-3).

Fig. 1 To attract foragers trained elsewhere to the scent, we used a drop of anise extract and a small fan to simulate the aroma of nectar coming out of a hive containing bees. We did not include any syrup in the scent lure, to avoid the possibility of a bee getting a sugar reward and then exiting and recruiting others.

Fig. 2 We tested (L to R) the stainless steel RobberGard, the plastic BeeSmart, and a “traditional” self-fabricated wood and 1/8” hardware cloth device, each compared to a Control hive with a ¾” high open entrance.

Fig. 3 We ran paired tests of each robbing screen design against an open Control entrance.

We again used a crossover design, swapping robber screen types and positions back and forth between the pair of hives in each run to equalize any location, box, or scent variables. We temporarily screened the entrance of the Control hive with window screen between runs to confirm that bees were equally attracted to the entrances of the two boxes, before starting each test run.

We then counted the number of bees entering for several minutes. Since there was no syrup reward, we took our final counts before any disappointed bees began exiting. Our results are shown in Figure 4 and Table 1.

Fig. 4 The robbing screens worked! Compared to the unscreened Controls (gray), all three types of robbing screens greatly decreased the total numbers of investigating bees that entered.

Table 1 Since the intensity of the investigative pressure by bees seeking the scent varied from run to run (due to wind, time of day, amount of open feeding taking place, etc.), the durations of the individual runs varied considerably. In order to better compare the efficacy of each robbing screen type, I calculated the rates of bees entering per minute for easier visual comparison (Figure 5).

Fig. 5 Compared to their paired unscreened Controls (gray), all three types of robbing screens greatly decreased the rates of investigating bees entering per minute (you may also wish to compare the median values in Table 1).

Conclusion

All three robbing screen designs were highly (but not completely) efficient at deterring investigating bees from entering an empty scent-lured hive body.

Practical application: A few investigators still did figure out how to get in. The question then is, if a scout were to be successful at entering through a robbing screen and then return home with a sugary reward and recruit additional “robbers,” would enough of those recruited foragers also be able to figure out how to get in, and as a result, exponentially progress to mass robbing? I ask this question since I’ve observed that once mass robbing has begun, installing a robbing screen will not stop it, since the returning robbers orient to the exiting bees and quickly figure out how to get in.

So let’s talk about bee foraging and recruitment behaviors.

Rallying the Troops

Once a scout or lucky forager finds a worthwhile source of nectar, pollen, or other food source, it returns to its own hive, and if it “feels that its find is of importance,” will perform waggle dances to share that information with its nestmates, and potentially recruit other foragers. The more enthusiastic the dancer is about the return on investment of that food source (effort vs. reward), the more vigorously it will dance.

The Dance Language

Karl von Frisch discovered how honey bees convey information via their dance language. He concluded that they used two distinct dances, with the “waggle dance” indicating the direction and distance to a desirable food source, whereas the “round dance” ostensibly indicated the presence of an attractive food source somewhere nearby. Since then, researchers have concluded that the “different” dances actually represent the two ends of a continuum, with the intensity of the distance signal and accuracy of the directional signal of the waggle dance each decreasing as the distance to the food source decreases. For those interested, I highly recommend an article on the dance language by Grüter & Farina [[1]].

There are four main informational components delivered by a forager performing a waggle dance:

The direction of the food source (relative to the sun)
The distance to the source
The “profitability” and desirability of the food source
The odor of the food source

Let’s take a look at these four components, relative to how mass robbing gets going in an apiary. Keep in mind that in a bee yard, the colonies are unnaturally close together, relative to how bees evolved that dance language (the opportunity to rob presumably seldom occurred in a “natural” situation).

Direction

Kirchner [[2]] and Gardner [[3]], independently using sugar syrup bait stations and filming waggle dances, collected data that show how dancers adjust the width of the directional component of the waggle dance (alternating the figure-8 directional run to the left and right), rather than to a specific spot (Figure 6).

Fig. 6 I charted data from Kirchner’s and Gardner’s studies to illustrate the amount of deviation (the MVL) that foragers use in the directional component of the dance. The closer the source, the wider the indicated range; the farther the source, the more precise. For close sources, there is minimal directional information, leading the recruits to just fly out and “sniff around.”

Practical application: In any bee yard, recruited potential robbers would be searching the area for the scent, rather than being directed to a specific hive.

Distance

The waggle component of the dance provides acoustical information about the distance to the food source [[4]]. Honey bees gauge the distance flown not in meters, but by means of the “retinal image flow” experienced while flying. Not only that, but they can account for detours, and the differences in distance between the outgoing and return flights [[5]].

Practical application: Not surprisingly, we noticed that there would be more “investigators” nearer to the feeding stations than farther away. Their appearance shifted with the wind, since they were flying upwind toward the scent, rather than showing up at a specific location.

“Profitability” and Desirability

The dancer, by the vigor of her dance, communicates her evaluation of the profitability of the food source (based upon energy expenditure vs. caloric gain), and upon the degree of positive feedback from the mid-age receiver bees and potential recruits receiving a sample of the nectar via trophallaxis [[6]].

There’s a reason why there generally isn’t much robbing when there are easy pickin’s during a nectar flow (although I’ve seen nucs being robbed dry during a flow). In general, foragers favor nectar with a 30-50% sugar content (such as when they suddenly shift from pear nectar at 10% sugar to the 30% sugar offered by apple blossoms).

Below 30% sugar concentration, bees are able to simply suck nectar (or syrup) up through their proboscis — but that nectar carries fewer calories. Above 50% sugar content the bee must work harder, since it must switch to energetic “viscous dipping” of its tongue [[7]^{, [8]}] (It’s not clear to me whether a forager needs to also add saliva to imbibe honey.)

Practical application: This is why it is recommended to feed 50% sugar syrup, since at that concentration, they can transport it, share it, and process it with the least effort for caloric gain. Plus it supplies the nurse bees with the critical water required to produce jelly.

In any case, foragers investigate whatever’s available, and respond to the feedback from the mid-age receiver bees that unload them to “tell them what the colony needs.” When nectar is scarce, foragers will work to pillage high-viscosity honey (or even dry sugar).

ODOR

It appears to me that the most important part of the waggle dance is its conveyance of the odor associated with the reward. Forget our human perspective — bees exist in a world defined by odors. A successful forager shares the “odor cue” associated with the reward (on its body or via passing out tastes of the nectar) with interested recruits. As explained by Kirchner [[9]]:

Dances lead the recruits into the right area, where odor cues then … become more and more important for them finally to find the flowering plants.

Practical application: This is why it’s so easy to train bees to a scent. And also why robbing occurs at the end of a nectar flow, when hives are exhausting the scent of ripening nectar. Or when a beekeeper leaves unprotected honey out during a dearth (the recruited “robbers” will only seek honey having the same odor). Or why one must be careful not to start robbing by adding a scented “feeding stimulant” to sugar syrup.

An amusing anecdote Regarding Training Bees to A Scent

One August day, during our nectar dearth, I stopped by one of our outyards. On top of one hive was a hand-scrawled angry note from a neighbor saying that our bees had attacked him at his house. Incredulous, I drove over with a case of honey and a smile.

He came out apologetically, and explained what had occurred. He was an admitted Dr. Pepper junkie (the full sugar version). After drinking one, he’d toss the can into an open recycling can in his garage. One day he inadvertently left the garage door open, and some of my hungry foragers discovered the sugary remainders in the cans — soon filling his garage with a cloud of bees. As he willingly admitted, they showed no sign of aggression, and he hadn’t actually been “attacked.”

But here comes the funny part. After he had posted the note to our apiary, he went back to his house. Once the bees had all gone home for the evening, he closed the garage door.

The next morning he walked outside, and was delighted to not notice any bees (not realizing that there were hundreds in the air, “sniffing” for the scent of their disappeared sugar source). So to celebrate, he popped the lid on a can of Dr. Pepper. Immediately, those bees smelled Dr. Pepper, and (orienting to the odor) descended on the top of the can. As you might imagine, he was a more than a bit surprised and startled, and dropped the can — spreading the sugary lure in front of him, which the cloud of bees then swooped down upon. It was apparently an impressive scene!

Again, no stinging occurred. I congratulated him heartily for being such a good animal trainer and we shared a laugh. I gave him the case of honey, and suggested that he get a lid for his recycling can.

The Purpose of a Robbing Screen

Scouts (or foragers recruited to a scent) scour the neighborhood for that scent, flying upwind to see whether it leads to a sugary reward. If they land on a source of the scent and find that there’s no sugar reward, they fly off to continue their search. (We see this when our scented-syrup open feeding trays go dry and the bees immediately abandon them.)

Practical application: The main function of a robbing screen is to redirect and divert bees that are investigating the scent emanating from a hive, away from the hive entrance. If they fail to quickly reach a reward, they “just give up” in frustration and fly away to look elsewhere. The robbing screen thus challenges the investigating bees’ persistence and perseverance — some designs being more effective than others.

Since we had found that a few investigators do eventually figure out how to get through a robbing screen (not surprising, since most of the natal bees of a hive also figure out how to do so), I decided to see what would happen if we placed a scented sugar syrup reward inside and allowed foraging/robbery to progress.

Experiment #2

We set up four bee-free “hives” of similarly-weathered bottom boards and hive bodies. For each run we placed the boxes in rough circles, their entrances facing outwards, with no noticeable objects near them. Three test hives got different types of robbing screens, and one hive served as the unprotected Control. Into each hive we placed two pans of anise-scented 1:1 sugar syrup midway back, but no fans (Figure 7).

Fig. 7 Once we saw that roughly equal numbers of bees were investigating the fronts of the hives, we unplugged the entrances, and sat nearby and recorded the number that entered each hive.

We ran a few runs at each of four different locations on the property. After counting the bees for each run, we’d close the entrances and shoo out the bees inside, then independently swap the boxes, guards, bottom boards, and syrup dishes around to randomize their locations relative to each other. We’d then wait (and adjust if necessary) until we observed equal interest by investigating bees in front of the boxes, prior to opening the entrances.

Then the action would start happening! Invariably (and hardly surprisingly) syrup-lapping bees would build up in the unscreened Control box first (Figure 8).

Fig. 8 Unlike as with in the unbaited scent trays in Experiment 1 (in which there was no reward), incoming investigators would hang out at the syrup to tank up. At the left is an example of early “robbing,” and at the right what it would look like when we called the count. We concluded each run once there were too many bees to count with reasonable accuracy.

Results

I’ll first show our findings graphically in Figure 9, then numerically in Table 2.

Fig. 9 There were always more robbers in the Control hives (gray columns). Note how quickly invasion got going in the first runs (earlier in the day, and more proximate to the open-feeding syrup stations). As we moved further away (and later in the day), invasion took longer (being bored, we checked less frequently, and recorded a count as 60 once the dishes were overflowing with bees).

You may notice that the later runs were for longer periods. This was because again and again we found that interest in scented syrup was most intense in the morning, and waned later in the afternoon. I have no idea why.

Table 2 Some robbing screens appeared to be more efficient than others.

The Question of Recruitment

This experiment was not designed to simulate the initiation of the robbing of a particular hive, but rather what would occur in an apiary should there be investigating foragers seeking a particular scent that they’d already associated with a potential reward (such as at the end of a nectar flow, or had a colony already been robbed).

That said, I was curious as to whether recruitment of additional bees into a hive would be accelerated via waggle dances or the presence of “robbers” already in the hive. As far as recruitment by waggle dances, take a look at the times involved in each run. In several runs, the dishes of the Control box would be filling with bees within a few minutes — long before any forager could have returned to its hive to recruit others.

This observation is supported by the findings of Dr. Adrian Wenner, who also experimented with scented sugar syrup and found that recruitment to nearby (within several hundred feet) scented syrup sources did not appear to be the result of directional orientation by waggle dances, but rather due to scent alone, especially when associated with a reward from previous experience [[10]]. That appeared to be the case in our runs, since we’d been open feeding to train the foraging force to the scent — so the invading bees could have come from any number of hives in the yard.

But that doesn’t mean that the presence of “happy” bees slurping up the syrup couldn’t be drawing others (although investigating bees couldn’t see them). We did not notice any sort of “recruitment pheromone” at the outside feeding sites, nor by the feeding bees within the hives. No Nasonov glands were exposed [[11]], and bees ignored the dishes or combs as soon as the syrup was all taken.

Despite regularly moving the test “hives” to new locations on the property, at some locations very quick “invasion” of the test hives occurred as soon as we uncovered the entrances. This appeared to be somewhat associated with proximity to the open feeding stations, wind, sun, or other features of the landscape.

The number of entered bees in a hive appeared to progress in a linear fashion, rather than exponentially. The wait for the twentieth bee to show up was often agonizingly long — we didn’t notice a pickup of the pace of invasion with time, strongly suggesting a lack of recruitment by either waggle dances, or some signal from the feeding bees. This is not to say that more waggle dance scent recruitment couldn’t have taken place had we run our tests for longer duration. But by that time the “random” entry of foragers from many hives would have already occurred.

Practical application: The robbing of a particular hive in an apiary is more likely due to the robbers orienting to the scent, rather than precise directions from waggle dances.

The Question of Visual Recognition or Training

One factor that we couldn’t control was that an experienced “robber” might learn to associate the visual image of a particular type of robbing screen with a reward, and thus orient to that image (as well as know how to enter it!). This was especially a question with regard to the white BeeSmart devices, which stood out against our dark test boxes.

In order to see whether there was actually a learning curve associated with that device, I graphed out the rate of bee entry over the course of all our runs, in order to see whether there was an increased progression of entry over the course of our runs that day — since it is likely that some of the same bees would have kept showing up at our scent-baited hives as we ran run after run in the same yard (Figure 10).

Fig. 10 It doesn’t appear that the bees got any better at invading through the BeeSmart device with “practice” over time. So visual orientation didn’t appear to be the case.

Conclusions

We’ve clearly found that a robbing screen can indeed make it difficult for potential robbers to enter a hive (but unfortunately at the cost of losing some bees returning from their first flights).

But we’ve also found that a robbing screen cannot alone prevent robbing from getting underway. Honey bees have a talent for “sniffing out” any odor associated with a sugary reward, and the perseverance to figure out how to get at it (as any beekeeper with a honey house or stack of supers quickly learns).

We also found that some designs appeared to work better than others.

Coming Up

In order to test these devices without the confusion of foragers returning to their own hive, I’d so far excluded the contribution of the critical players in a colony’s defense against robbing — the vitally important guard bees that normally patrol the entrance.

Practical application: A colony is dependent upon its guards to defend itself against robbers — a robbing screen is merely a device that decreases the number of investigative potential robbers that the guards have to deal with.

A good robbing screen should also be designed to work in concurrence with a colony’s guard bees. But our observations suggest that the screen designs that we’ve tested may actually work against the guards!

Next month, I’ll focus on the guard bees, and make suggestions as to how we might improve the design of a robbing screen — based upon physics, guard behavior, and close field observations, rather than assumptions and “common sense.”

Citations and Notes

[1] Grüter, C & W Farina (2009) The honeybee waggle dance: can we follow the steps? Trends in Ecology & Evolution 24: 242–247. https://www.socialinsect-research.com/resources/Gr%C3%BCter%26Farina2009b.pdf

[2] Kirchner, W, M Lindauer & A Michelsen (1988) Honeybee dance communication: acoustical indication of direction in round dances. Naturwissenschaften 75: 629-630.

[3] Gardner, K, T Seeley & N Calderone(2008) Do honeybees have two discrete dances to advertise food sources? Animal Behaviour 75(4): 1291-1300.

[4] Kirchner, W, op cit.

[5] De Marco, R & R Menzel (2005) Encoding spatial information in the waggle dance. Journal of Experimental Biology 208(20): 3885-3894.

[6] Farina, W & A Wainselboim (2005) Trophallaxis within the dancing context: a behavioral and thermographic analysis in honeybees (Apis mellifera). Apidologie 36(1): 43-47.

[7] Wei, J, et al (2020) Sucking or lapping: facultative feeding mechanisms in honeybees (Apis mellifera)Biol. Lett.1620200449 http://doi.org/10.1098/rsbl.2020.0449

[8] Wei, J, et al (2023) Honey bees switch mechanisms to drink deep nectar efficiently. Proceedings of the National Academy of Sciences 120(30): e2305436120. https://www.pnas.org/doi/pdf/10.1073/pnas.2305436120

[9] Kirchner, W & A Grasser (1998) The significance of odor cues and dance language information for the food search behavior of honeybees (Hymenoptera: Apidae) Journal of Insect Behavior 11: 169–178.

[10] Ibid. See Chapter 10, and especially Fig. 10.2 and Table 10.1.

[11] Wenner concluded that Nasanov pheromone does not serve as an attractant to sugar syrup. Wenner, A & P Wells (1990) Anatomy of a Controversy, Columbia University Press, p.318.

Category: Varroa IPM Strategies, Varroa Management