Can Robber Screens Reduce Mite Immigration? Part 4
SIDE NOTE: A Practical Observation. 9
Can Robbing Screens Reduce Mite Immigration?
Part 4
First Published in ABJ October 2024
Randy Oliver
ScientificBeekeeping.com
I’m typing these words freshly returned from presenting in New Zealand, where “reinvasion” (aka immigration) by varroa from abandoned hives (due to the downturn in the manuka honey market) is a big concern. Beekeepers there have asked me whether placing robbing screens (aka robber guards) on their hives will reduce the number of mites coming in.
Introduction
When asked the question above, I couldn’t find any hard data to support an answer, so I set up the two field trials that I’ve covered in my previous articles to find out [[1]]. But due to the surprisingly low mite immigration counts in those trials, I couldn’t tease out an answer. So, being a glutton for punishment, I decided to give robbing screens a third chance to prove their benefit against mite immigration.
We again set up a crossover trial, in which we would swap the robbing screen treatment back and forth on the hives, but this time we brought in collapsing colonies to ensure that an informative amount of mite immigration would take place.
Materials and Methods
In mid-June last year we introduced 28 5-frame nucs (dribbled with oxalic acid during a brood break) with new queens into single deeps, and two days later applied one strip each of Apistan (fluvalinate) and Apivar (amitraz). Since the mites in our operation haven’t been exposed to either of these treatments, I expected the intense combination to completely eliminate all mites.
By 5 July the colonies had grown to an average of 9 combs of bees. But we still got a mite wash count of 4 in one sampled hive. So we balanced colonies for strength, added a second brood chamber of drawn comb, and treated them with a single strip of Formic Pro. Ten days later we applied a second Apivar strip along with an extended-release OA/gly sponge for good measure (Figure 1).
Fig. 1 We applied miticides with four different modes of action to eliminate the mites (formic acid not shown). We left the treatments in for the full course of the trial, to kill any incoming mites (which we then counted on stickyboards)
This strong treatment regimen retarded colony growth somewhat, so we supplementally fed the colonies to help them to build up into double-deep, moderately strong, mite-free hives. We installed stickyboards under screened bottoms, and baited for ants so that they wouldn’t remove any fallen mites.
From 16-19 July we took baseline stickyboard counts, followed by semi-weekly counts from 25 July on (Figures 2 -6).
Fig. 2 The mite-free colonies on 21 July, all on screened bottoms with stickyboards. The weather was hot, and the colonies often bearded during the day
Fig. 3 Since by this time we were tired of taking stickyboard counts, I built a carry tote to hold all the necessary equipment and data clipboard (including reading glasses and a hand lens), which made things easier. We found the hand tally counters useful to keep us from losing count!
Fig. 4 Every hive got a numbered stickyboard. I make them out of fiberglass-reinforced plastic, with divider lines drawn with a Marks-a-Lot felt pen. These boards are reusable and last for years.
Fig. 5 After counting the mites, we use a sheetrock knife to scrape them clean in two strokes.
Fig. 6 And then we roll them with a light coat of petroleum jelly and mineral oil blend for the next count.
By the end of July, the baseline mite drop counts were low – averaging less than two mites dropped per hive per week, successfully confirming that (1) the colonies did not contain reproducing mite populations, and (2) unfortunately indicating that very little mite immigration was taking place. So beginning on 30 July, we started moving in untreated high-mite “mite factories” adjacent to the test hives to serve as a source of mites (Figure 7).
Fig. 7 To increase mite immigration, we extended the bear fence, and throughout the month of August moved in a dozen high-mite hives on the far side of the 28 test hives (in the shade in this view; photo taken as we laid out the robbing screens on top of the 14 highest-immigration hives). We fed the mite factories pollen sub and syrup to help them produce mites. Several of them collapsed during the trial and some got robbed out.
Moving in the “mite factories” increased mite immigration, but by late August it was clear that half the hives consistently kept mite immigration to a minimum (Figure 8).
Fig. 8 By early September only half the hives exhibited substantial mite immigration counts. On the left are the 14 low-immigration hives, with counts too low to be of use in the test, so we ran them as a Control group. To the right are the high-immigration hives to which we applied the alternating robbing screen “treatment.”
SIDE NOTE: A Practical Observation
As you can see in Figure 2, we had laid out the 28 hives in a rough square enclosed by a bear fence, 22 around the perimeter with their entrances facing outwards, along with 3 pairs of hives in the center, entrances facing the mite-donor hives (which were added later within an extension of the bear fence). I’ve drawn a rough schematic in Figure 9.
Fig. 9 This is an approximate layout of the hives, with those hives having cumulative mite drop counts above 90 shaded in red. Note how most of the high-immigration hives were nearer to the mite donors and outside corners of the apiary, rather than the interior. This suggests that the majority of mite immigration was due to bees drifting from the mite donors, rather than from being carried back by foragers robbing the collapsing mite donors.
Back to M&M
So we ran only the 14 higher-immigration hives as Test hives for the crossover design test of the robbing screens, leaving the low-immigration hives in place as an untreated Control group to perhaps track the level of background immigration into undisturbed hives. We flipped a coin to randomly assign whether each of the 14 Test hives received a robbing screen (Figure 10).
Fig. 10 We installed guards (screens) on all 14 Test hives, but raised the end on half of them to allow bees to freely return through their accustomed entrance during their “screen off” periods (left). During “screens on” (right), only the small top entrance was open (not full width). We reversed the “on” or “off” treatment roughly every 10 days, to determine the screens’ effect upon mite immigration.
Results
Scientists use the word “stochastic” to describe a phenomenon whose outcome is determined by one or more random variables. I can safely say that mite immigration into hives is wildly stochastic, dependent upon a number of variables. To enter a hive, a mite needs to hitchhike a ride in on a bee, but that bee can be a returning forager who was robbing another mite-infested colony, or a drifted bee or surreptitious robber from another hive. Flight of any of those bees is dependent upon weather (Figure 11).
Fig. 11 The amount of mite drop increased when we moved in “mite donor” hives, and then correlated largely with flight weather.
And then there is the attractiveness (presumably due to odor) of a hive to drifted bees –– all my observations indicate that some hives are magnets for drifted bees. And some colonies may practice more vigorous guarding –– preventing the entry of drifted or curious bees from other hives. In this trial, half the hives (the Control group) exhibited very little mite immigration. The occasional large spikes in mite immigration may have been due to a colony’s foragers robbing out a hive collapsing from varroa (a process that may take only hours). To see how much hive-by-hive variation there was in our semi-weekly mite drop counts, take a look at Figure 12.
Fig. 12 Semi-weekly mite drop counts for all 28 hives. Although there were trends (notably on 7 and 28 September and 8 October), note the occasional large spikes in the counts of individual hives (perhaps due to robbing?). Note that at the last count, most hives were dropping only 1 mite per day, confirming that they still had no appreciable endemic mite infestation.
So did having a robbing screen on or off make a difference? I’ve created a chart to help us see (Figure 13).
Fig. 13 In this crossover trial, we alternated whether the 7 hives in each test group had their robbing screens on or off. I’ve averaged the semi-weekly mite drop counts for each test group, dividing the unscreened low-immigration Controls into two groups (by hive number) to show the relative consistency of the Control hive counts. To interpret the results, look for inflections in the slopes of the red and green plots following the alternation of screen treatment (occurring at the four dashed columns). Robbing screens were on when the plotlines are dashed.
All hives were unscreened up until 11-Sep, with Group A happening to exhibit consistently higher mite drops (a result of our random assignment), which continued throughout the trial. We installed screens on both Test groups on that date, and although mite counts dropped, they also did so in the unscreened Controls. Starting with the 20-Sep swap, only one of the test groups had screens on.
At the 20-Sep switch, there appeared to be a possible benefit to having screens, since Group A’s counts went up more than the other groups when its screens were removed.
But then at the 28-Sep switch, the mite drop of Group A increased when we screened it! After the last switch on 11-Oct, it’s not clear whether there was a benefit to Group B from being screened.
Discussion
How frustrating! To my considerable surprise, we have been unable to tease out any obvious positive effect of robbing screens upon mite immigration –– in three separate field trials in different yards in different years. Although there are testimonials by beekeepers that they thought they observed a benefit, I’ve yet to see hard data to support that conclusion.
When in New Zealand, I heard over and over again that they were having a huge problem from “reinvasion” of mites from unmanaged hives –– but when I asked, that was only an assumption not based upon any hard data. My own hard data from California, and that of other collaborating beekeepers, suggests that it would take an awful lot of collapsing colonies in the neighborhood to produce enough mite “reinvasion” to make a large difference in the rate of mite buildup in a colony [[2]].
As an example, in this trial there were 12 donor hives that nearly or completely collapsed during the trial, some of which got robbed out. Yet the cumulative amount of mite immigration into the 28 adjacent experimental hives averaged only 93 mites (Figure 14).
Fig. 14 Note the wide range in cumulative mite drop counts (all or most of which was presumed to be from immigrated mites)(I excluded one mostly zero-count hive that had issues).
The average total mite drop per hive was only 93 (from which any endemic resident mites would need to be subtracted to quantify immigration) –– hardly enough to make a difference in the infestation rate of a “normal” colony with mites. As an example, I’ve run a couple of simulations (Figure 15).
Fig. 15 In the upper simulation I entered a figure for mite immigration of zero (figures in blue in the lowest row), and applied two 95% efficacy treatments. In the lower simulation I set the amount of immigration to a total of 1000 mites. If you look at either total mites in the colony (red plot line) or the infestation rate (blue alcohol wash), even having a thousand mites coming in didn’t make a heckuva lot of difference!
Now I’m not saying that mite immigration or “reinvasion” can’t be a problem – I lived through the collapse of the feral population in California, and the Aussie beekeepers are starting to experience it. And I’ve heard any number of beekeepers say that when their neighbors finally started controlling mites in their hives, that mite control became easier for them too. And it appears that extended-release oxalic acid is more efficacious in apiaries where there is no mite immigration taking place. So in our own operation, we blast any extreme high-mite “outlier” hives with formic acid and requeen them, to eliminate the “mite diffusion” coming from them.
The above said, we need to be careful about blaming our problems solely upon “mite immigration.” The numbers simply don’t work!
Next
I’m not done yet with robbing screens (aka robber guards). We’re currently running experiments to determine exactly how and why they work, and whether there are ways to improve their design to reduce the entry of foreign bees. Teaser: some of our findings are surprising. Stay tuned!
Citations and Notes
[1] Published in the last two issues of ABJ.
[2] https://scientificbeekeeping.com/a-study-on-bee-drift-and-mite-immigration-part-6/ ABJ July 2023.