Can Robber Screens Reduce Mite Immigration? Part 3
The Effect of Mite Immigration Upon the Efficacy of OAE. 1
Trial B –– A Comparative Trial of Effect of Robber Screens Upon Mite Buildup. 4
Can Robber Screens Reduce Mite Immigration?
Part 3
First Published in ABJ September 2024
Randy Oliver
ScientificBeekeeping.com
Concurrent with the field trial of robber screens that I wrote about last month, we ran a completely different sort of test –– to see whether having robber screens in place would increase the efficacy of extended-release oxalic acid (OAE) treatment in late summer.
The Effect of Mite Immigration Upon the Efficacy of OAE
I’ve previously found that extended-release oxalic (due to its relatively slow action) appears to perform best when colonies are located in isolated yards with no other hives in the neighborhood (Figure 1) —the assumption being that the increase in efficacy was due to lack of immigration of mites from other hives. It occurred to me that I could use this observation to test whether robber screens would improve the efficacy of an OAE treatment.
Fig. 1 In the summer of 2019 I was asked by a grad student to place hives next to high-elevation meadows in the Sierra Nevada (where there were no resident honey bees; our brown hives are barely visible in the background), to see whether adding honey bees to the landscape would have an effect upon the seed set of a native flower (the results were inconclusive).
For the project, we placed hives next to three meadows. Since there was a good nectar and pollen flow occurring, we expected that the mite infestation rates of the colonies would explode without treatment. So we took mite wash counts, and treated 45 hives with elevated mite levels with shop towels soaked in 1:1 OA:gly (this was before we discovered the advantage of using thicker matrices). Nevertheless, the treatments were astoundingly efficacious [[1]].
We were again asked to restock the meadows in 2020, so used the opportunity to test two doses of OAE on Swedish sponges –– either a half sponge (25 g OA) or a full one (50 g OA, 1:1 glycerin). Despite the colonies engaging in the copious rearing of drones over a long duration, the degree of mite control by the treatments again amazed us (Figures 2 & 3).
Fig. 2 This trial ran from 16 June through 18 August. Without treatment, the mite counts would have been expected to quadruple (if they had done so, the red columns (ending counts) would have been four times as high as the blue ones).
Fig. 3 We waited even longer to take final wash counts in these two yards (16 June through 1 September). Note the very high starting counts (from which colonies often can’t recover), going down to zeroes. Despite that, by the end of 77 days of treatment, the mites had nearly disappeared!
We got similar results (not shown) from a pumpkin pollination apiary that we stocked in an isolated valley in Nevada (ours were the only colonies present) –– instead of the usual major increase in mite infestation that we’d experienced in previous years (with our colonies going downhill), the hives came back strong and full of honey, with mite wash counts of zero.
Practical application: The results from the above isolated apiaries strongly suggest that OAE treatment can be extremely efficacious when colonies are not exposed to the immigration of mites from surrounding colonies (managed or unmanaged) in the neighborhood.
Trial B –– A Comparative Trial of Effect of Robber Screens Upon Mite Buildup
Experimental Design
The above finding regarding the apparent effect of mite immigration upon OAE efficacy gave me an idea for another way to test whether robber screens would reduce mite immigration –– by seeing whether they would improve the efficacy of an OAE treatment. We had a number of available test yards near home, surrounded by neighborhoods containing plenty of hobby and feral colonies, from which mite immigration would be expected.
For the trial, I didn’t want the OAE treatment to be so strong as to eliminate any observable impact from immigrated mites, but also didn’t want the mites to build up excessively in the Control hives not receiving robber screens. So we treated all the hives with only a half dose of OAE. We knew from previous trials [[2]] that a 25-gram dose of OAE on a half sponge would not be expected to be fully efficacious in our local outyards, especially with the expected immigration of mites from the neighborhoods surrounding our test yards. My hope was that the low OAE dose would allow the effect of immigrated mites to show in our unscreened Controls, while still holding back mite buildup enough that we could save them once the trial was over.
This experimental design would be an indirect method of testing the effect of robber screens, since we wouldn’t be directly measuring mite immigration. But since it would save us from performing tedious stickyboard counts, we could include a lot of colonies, and replicate the trial in a number of yards.
Practical application: This experimental design would actually be more relevant to beekeepers than Trial A, since instead of merely providing incoming mite numbers from stickyboard counts, it would demonstrate whether installing robber screens would improve the efficacy of a treatment with a miticide.
Materials and Methods
We replicated the trial in six yards in the foothills (133 hives in total), in the expectation that the colonies in at least some of the yards would experience appreciable immigration of mites from outside. We treated every hive with a half sponge to retard mite buildup, but still allow an opportunity for robber screens to prove their benefit over the course of 51 days (from 9 August and 29 September –– a period during which mite populations typically explode in our area) (Figures 4 & 5).
Fig. 4 After first taking mite wash counts from every hive, we flipped a coin to determine which of each adjacent pair of hives would get an entrance screen (wide opening up), adjusting treatment assignment if necessary to even out the starting mite counts.
Fig. 5 Many of our outyards are squeezed into property corners in semi-rural wooded neighborhoods. In most of the test yards the hives were closely spaced, which would favor between-hive bee and mite drift. All the test yards were in areas with an abundance of hobby beekeepers and plenty of tree cavities for feral colonies.
We ran the trial for 51 days, after which we took final mite counts to compare to the starting counts.
Results
To our surprise, despite the half-dose OAE treatment, there was very little overall increase in mite infestation in any of the six yards (A – F); in fact the counts actually went down in many of the hives!
Practical application: This result confirmed our findings from our 2020 trial [[3]] –– that a half sponge would generally “hold” the starting mite count (of unscreened hives) at 50 days. In retrospect, we should perhaps have applied an even lower dosage (but, my sons aren’t crazy about me allowing mites to build up in hundreds of experimental hives).
Since the mite wash counts were very low and barely increased in the vast majority of the hives (going from an average of 2.3 to 2.4 mites per half cup of bees), I calculated the mean and median absolute changes in count (as opposed to proportional changes) per yard in Table 1.
As usual, the median values are more relevant than the means, because they aren’t influenced by a few wild outliers. Since the results were similar for all yards, I combined the data in Table 2.
The median values for mite wash counts in Table 2 did not differ between the Control and Screened hives. More visually, look at a histogram comparing the distributions of counts from the two test groups (Figure 6).
Fig. 6 Since the final mite counts averaged only around 2 mites per half cup of bees. I plotted the distribution of absolute change in mite count (rather than proportional change). For both test groups, mite counts mostly either did not change, or went up by 5 or fewer mites. Although the Screened columns are both slightly lower the those for the Controls, they were not significantly different (t-test p = 31%). The distribution of outliers was similar whether screens were installed or not, other than that two control colonies exhibited unusually high ending counts –– perhaps from being highly attractive to drifting bees, or from robbing out collapsing colonies in the neighborhood.
The histogram again suggests that having robbing screens on a hive didn’t make much difference. Not giving up, I worked hard to try to tease out evidence that the robber screens were of benefit, by graphing the data out in a number of different ways. For example, I wondered whether there was a relationship between a hive’s starting and ending mite counts (Figure 6).
Fig. 7 Columns going upward from the 0 baseline indicate that the mite count increased, those going downward that it decreased. There were more changes up or down in hives with higher starting counts (increasing from left to right), but there was no clear correlation between starting mite count and ending count (keep in mind that there can be no decrease in a hive that starts with a zero count), nor obvious benefit from having a robber screen on. As we’ve found in previous field trials, some colonies are mite drift magnets, and this could be the case for the four hives (two in each treatment group) whose mite counts increased appreciably.
Discussion
I was surprised by the lack of mite buildup across the board. Could it have been that the half-sponge OAE treatments performed better than expected? Or, as I elaborated on in a recent article [[4]], one factor might have been that we are finally showing some success from our selective breeding program for mite-resistant bees. This is making it more difficult for me to set up field trials of mite control methods, since I can no longer assume that the mite infestation in a colony will increase without treatment –– I now need to first confirm that any test colony is not innately resistant.
Next
Doing all this work without clear results was frustrating, and try as I might, we’ve so far been unable to produce evidence that robber screens reduce mite immigration. But I decided to stick with it, and the next year gave robber screens one more chance to prove that they could reduce mite immigration. Read about it next month!
Acknowledgements
My great appreciation to my helper Rose Pasetes.
Citations and Notes
[1]https://scientificbeekeeping.com/extended-release-oxalic-acid-progress-report-2019/ ABJ December 2019
[2] https://scientificbeekeeping.com/mite-control-while-honey-is-on-the-hive-part-2/ ABJ December 2020
[3] Ibid, Figure 12.
[4] https://scientificbeekeeping.com/selective-breeding-progress-report-2023/ ABJ September 2023