A survey of UK beekeeper’s Varroa treatment habits

The global spread of the parasitic mite Varroa destructor instigated a substantial decline in both managed and feral honeybee (Apis mellifera) colonies mainly across the Northern hemisphere. In response, many beekeepers began to treat their colonies with chemical acaricides to control mite populations in managed colonies. However, some countries or beekeepers allowed their bees to develop mite-resistance by adopting a “treatment-free” approach, rather than using selective breeding programs. Yet, the distribution and proportion of beekeepers either treating or not within the United Kingdom (UK) is unknown, as it is in most Northern hemisphere countries. Therefore, the aim of this study was to conduct a beekeeper survey to determine the current treatment strategies within the UK. We gathered 2,872 beekeeper responses from an estimated 30,000 UK beekeepers belonging to 242 bee-associations in the winter of 2020/21. The survey indicated that the majority (72–79%) of UK beekeepers are still treating their bees for Varroa, typically twice-yearly using chemical-based methods. Six percent or 1,800 UK beekeepers were treatment-free for six years or more. This is reflected by our finding that 78 associations out of 242 consist of responders who entirely treated, while only four associations had more than 75% of their members that were non-treating. Overall treatment status was not affected by association currently. Using the baseline data from this survey it will be possible in the future to observer if a shift towards treatment-free beekeeping occurs or not.

Line 201-203, that assumes that the survey reached and was completed by a representative wedge of the beekeeping population. Many treatment-free beekeepers tend to be younger and more active online, so this may have skewed the responses. These caveats should be mentioned. THE COMMENT ABOVE ONLINE ACTIVE HAS NOW BEEN ADDED TO THE RELEVENT PART OF THE DISCUSSION L 239-241. Line 218-219: This is not surprising at all. Most treatment free beekeepers in the United States expect annual losses of 50%. Thus, in order to rebuild the following year, they must keep more colonies in reserve. I am disappointed this survey didn't ask for selfreporting of annual loss rates or honey production per colony. The high level of treatment-free suggests that beekeepers and the much higher colony numbers suggests that the resistance is probably not enough for low annual losses. Is the 4% what beekeepers reported was specifically due to varroa? This would then be an underrepresentation of actual winter losses due to varroa. Numerous surveys have shown that when varroa levels are high, winter losses increase, even though beekeepers blame other factors for the loss. Many beekeepers with colony winter losses don't know what killed their colony, but beekeepers with more experience can often diagnose the contracted brood nest area and varroa frass. WE HAVE GIVEN MORE CLARIFICATION ON THE PRINCIPAL CAUSES OF COLONY LOSSES IN THE UK Line 254-255: Why is the low winter loss attributed to treatment-free beekeeping. There could be other explanations, such as very active beekeeping groups that teach how to monitor and treat effectively. The evidence presented does not warrant this statement, as there is no segregation of losses by those treating and those not treating presented. WE AGREE THAT NOT ENOUGH EVIDENCE IS PRESENT AT THE MOMENT SO THE SENTENCE HAS BEEN DELETED Line 256-260: These statements require the reader to go to the original articles and assess the numbers of colonies surveyed. Important information such as colonies assessed in the survey should be included with the paraphrased results, especially as these are not peer-reviewed studies but association results and so the statistics have not been evaluated. The way they are presented in this paper, the reader assumes they are peer-reviewed papers, unless the references are checked in detail. Reviewer #1: The submitted survey of UK beekeeper´s varroa treatment habits presents interesting new data and fills a gap in our knowledge on recent beekeeping practice. The study is well presented with a clear description of the data sampling and evaluation procedure. The discussion picks up relevant links to related surveys and the establishment of mite resistance in bee populations depending on the treatment strategies. The length of the different chapters is appropriate and well balanced. Some confusion may arise about the number of respondents for the data in table 3. The table reports on a total of 2,868 responses from 2,897 responses in general (L116). If 25 responses from the Isle of Man have been excluded, some of the 16 participants with zero colonies (L146f) seem to be included here. However, this wouldn´t make any sense for my understanding. The CORRECTED) In L120 Table S1 is mentioned but it should be Table 5 as far as I understand.
(deleted) In L 264 the second "by" has to be erased. Deleted Reviewer #2: The paper has the potential to be interesting but is currently missing the access to the data and the supplemental information, so difficult to assess. Also, the conclusions are very broad, and the data presented don't support the conclusions the authors make, as the survey did not include any assessment of colony losses. Without distinguishing what the actual loss rates of the treatment free beekeepers have, it is not possible to state that the bees are resistant to varroa or have a high tendency for survival. The higher colony numbers actual suggests the opposite. TREATMENT-FREE POPULATIONS VS THE NATIONAL AVERAGE. ANOTHER  SURVEY COLLECTS LOSS DATA BUT NOT AIMED AT SEPARATING OUT THE  TREATED AND NOT TREATED. WE HAVE ALSO INCLUDED THE MAIN REASONS  IN ENGLAND & WALES FOR COLONY LOSSES AND VARROA ACCOUNTS FOR  ONLY 4% ALTOUGH WE AGRREE THIS MAY BEEN UNDERESTIMATED BY THE  BEEKEEPERS. Also there are a number of odd sentence structures and grammatical errors. I have made note of the mores specific comments in a separate document. ALL SHOULD BE FIXED NOW Reviewer #3: The study aims to estimate the number of beekeepers in the UK who are not treating their colonies for varroa, and to estimate whether the existence of untreated and resistant colonies in the UK could be assumed. Information on beekeeping practices is currently limited in the scientific literature, and this question and the data the study provides are relevant and valuable. The article is generally clear and well written. The methodological choices are globally sound and with a high number of respondents to the survey, which is important to note. Still some information should be added in the methodological section, and the limits induced by some methodological choices should be identified and discussed as they can have important impacts on the interpretation of the results. Also, the findings should be discussed regarding the scientific literature on the subject. For these reasons I recommend major revisions.

AS MENTIONED ABOVE WE HAVE A PARAGRAPH COMPARING LOSSES IN
Major comments Some methodological limits should be identified and discussed, especially i) the possible influence of the sampling through the beekeeping associations (are these  associations involved in prescribing varroa treatments?)  THERE IS NO PRESCRIPTION OF TREATMENTS VIA ASSOCIATIONS SINCE  TREATMENT ADVICE COMES CENTRALLY FROM THE NATIONAL BEE UNIT &  THE MAP INDICATES WE HAD A VERY WIDE SPREAD OF RESPONSE and ii) the lack of information that the survey provides about the colony losses and about the beginning year of the beekeeping activity. Without this last information, the share of beekeepers who have "never treated" cannot be interpreted as it is done currently regarding the possible existence of surviving colonies (those beekeepers could have started recently, and their colonies may not have faced this absence of treatment for varroa for a long time). It should not be suggested that these colonies had never known any treatment (it could be the case, but the data of the survey do not  allow to conclude on this question).  THERE ARE A GROWING GROUP OF BEEEKEEPERS IN THE UK THAT NEVER  TREAT SUCH AS OXFORDSHIRE NATURAL BEEKEEPING ASSOCIATION AND  THEIR LOSSES FOR TWO WINTERS WERE SIMILAR TO THE NATIONAL  AVERAGE. THIS INFORMATION IS PROVIDED IN THE MS AS WELL AS A NEW  SUPPLEMENTAL FIGURE SHOWING THE NATIONAL AVERAGE COLONY  LOSSES FROM 2007 TO 2021. The existence of colonies surviving without treatment for varroa and the possible losses that the absence of treatment may induce should also be discussed regarding the scientific literature on these questions.

WE HAVE EXPANDED THE SECTION ON COLONY LOSSESS IN THE UK AS REQUESTED
Other comments L. 57: "due to almost universal treatment by beekeepers": this statement should be either nuanced or supported by a reference about American beekeepers' practices. E.g. see Thoms  L. 100-105: If I understand correctly, there was no question about the colony survival (cf. L100-105). It would have been a valuable information, as the emergence of resistant colonies is only possible if the non-treated colonies survive (in the case of backyard beekeepers replacing regularly their colonies that died from varroa, these non-treated colonies would not be more resistant than other). SEE EALIER COMMENTS ABOUT COLONY LOSSES L. 106-113. As the beekeeping associations who forward the survey to their adherents play a major role in the sampling, it would be useful to provide more information about these associations (as such beekeeper associations and their role can differ from one association or one country to the other). Especially, can these associations interfere with their adherents' treatment choices (e.g., by advising or not advising to treat for varroa) ?  THIS IS SOMETHING THAT IS NOT DONE THE ASSOICATIONS ONLY  FORWARED THE SURVEY LINK TO ALL ITS MEMBERS USING THERE  (CONFEDENTIAL DATABASE). IF THESE TYPE OF BEHAVIOR DID OCCUR, WE  WOULD HAVE PICKED IT UP IN THE ANAYLSIS. FURTHERMORE, THE  ASSOICATIONS THAT SEND OUT THE LINK HAVE NO WAY OF SEEING WHAT  THEIR MEMBERS WORTE. L. 123. One of the main analysis choices is the investigation for a possible spatial pattern. As many other factors could possibly influence beekeepers' choice about their treatment for varroa (beekeeping experience, age, etc), this hypothesis of a spatial pattern could be explained and justified. EACH ASSOICATION HAS A WIDE RANGE OF AGES, GENDERS, AND  EXPERIENCE TO REQUEST THIS TYPE OF DATA AND CONSDIER THEM IN THE  ANAYLSIS LIES WAY BEYOND THE SCOPE OF THIS STUDY. L. 130. Some biomechanical and natural methods that were reported by beekeepers (cf. Table 5) are not efficient against varroa. As the study focuses on the possible existence of untreated / resistant colonies and not on the choice of beekeepers to treat or not to treat (reasons for such choices, etc), these "low-efficacy" treatments could have been considered jointly with the treatment-free group, or a third variant for the response variable could have been considered. So, the choice to gather all the treatments together in a single response variable could be explained .  THE REVIEWER IS CORRECT IN THAT THE DIFFERENT TREATMENTS RANGE  WIDLY IN THEIR EFFICACY'S. TO ERR ON THE SIDE OF CAUTION WE HAD TO  MAKE IT VERY CLEAR TO UNDERSTAND WHAT 'TREATMENT-FREE' MEANT.  OTHERWISE, WE COULD HAVE RUN THE PROBLEM OF BEEKEEPERS  DECIDING WHAT THEY CONSIDER IS TREATMENT FREE. WE AGREE THE LOW-EFFICACY TREEATMENTS COULD BEE ALLOWING NATURAL RESISTANT TO  DEVELOP AND THIS ONE OPTION FOR BEEKEEPERS TO BECOME TREATMENT  FREE. IN THE FUTURE IT WOULD BE INTRESTING TO SEE IF THE USAGE OF  THESE LOW-EFFICACY METHODS INCREASED. L. 133. As one of the questions of the study is the existence of untreated and possibly resistant colonies, it would be useful to know how many colonies in total these 2,872 beekeepers manage and what percentage of the total number of colonies in the UK it represents. L. 160-162. Grouping the beekeepers who indicated that they have "never treated" with those indicating a specific number of years since the last treatment (here 10 years and more, or 6 years and more L. 201) is questionable as the date of installation of the beekeepers was not in the survey and is not known. Some of them could have started beekeeping recently, and the fact that they have "never treated" for varroa does not presume that their colonies had to face a significant period without treatment. This group of beekeepers with an unknown number of years without treatment should be considered separately and the confusion with the groups where a long treatment-free period is known should be avoided. L. 163-164. The fact than some association gather more than 75% of treatment-free beekeepers raises questions about the role of these associations, and about the possible exchanges related to varroa treatments that its members may have. Even if it was not the objective of the study to understand the determinants of the absence of treatment, it would be interesting to discuss the possible role of associations on this point given their central place in the survey sample. L. 178. "the majority are treated": the majority of colonies? YES, SO CHANGED L. 180-184. It would be useful to add the global percentage of beekeepers not using any treatment and to more clearly distinguish if the percentages given for the types of treatments are exclusive of each other or not. E.g., can the 3% of beekeepers using biomechanical methods be also in the 78% using chemical? Or are they only using biomechanical methods? The authors received no specific funding for his work.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
This statement is required for submission and will appear in the published article if the submission is accepted. Please make sure it is accurate.

Unfunded studies
Enter: The author(s) received no specific funding for this work.  Ethical clearance was given by the University of Salford, UK ethics committee. All data were analyzed anonymously. Yes -all data are fully available without restriction  Click here to access/download;Manuscript;Survey paper V5 PlosONE REVISED9AV.docx

Abstract 25
The global spread of the parasitic mite Varroa destructor instigated a substantial decline in 26 both managed and feral honeybee (Apis mellifera) colonies mainly across the Northern 27 hemisphere. In response, many beekeepers began to treat their colonies with chemical 28 acaricides to control mite populations in managed colonies. However, some countries or 29 beekeepers allowed their bees to develop mite-resistance by adopting a "treatment-free" 30 approach, rather than using selective breeding programs. Yet, the distribution and proportion 31 of beekeepers either treating or not within the United Kingdom (UK) is unknown, as it is in 32 most Northern hemisphere countries. Therefore, the aim of this study was to conduct a 33 beekeeper survey to determine the current treatment strategies within the UK. We gathered 34 2,872 beekeeper responses from an estimated 30,000 UK beekeepers belonging to 242 bee-35 associations in the winter of 2020/21. The survey indicated that the majority (72-79%) of UK 36 beekeepers are still treating their bees for Varroa, typically twice-yearly using chemical-37 based methods. Despite this, an estimated 3,000 UK beekeepers are treatment-free for over 38 six years or have never treated. This is reflected by our finding that only 78 associations out 39 of 242 consist of responders who entirely treated, whereas only four associations had more 40 than 75% of their members that were treatment-free. Although overall treatment status was 41 not affected by association currently, we predict that in time the shift towards intervention- Varroa and allowed mite-resistance to appear naturally. In both countries 1,000's of colonies 70 were lost initially, but losses declined after several years as resistance to the mite developed. 71 Whereas, in Brazil the evolution of mite-resistance by Africanised bees (Apis mellifera 72 scutellata) was not observed as it was probably masked by the losses from untreated non-73 varroacides or a combination of them [18]. Despite this, anecdotal evidence from beekeepers 88 suggests that numbers of non-treating beekeepers are more than expected [19]. 89

90
The aim of this study was therefore to conduct an online survey of UK beekeepers via their 91 associations to assess individual treatment habits, as it was the most efficient way to reach 92 beekeepers easily. The survey will provide crucial empirical data to support or refute 93 common perceptions around treatment and non-treatment beekeeping practices. 94 95 96

Materials and methods 97
The survey 98 The survey was constructed using Google Forms as it allowed an unlimited number of 99 responses whilst also allowing the incorporation of the University logo to add credibility to 100 the survey. The survey consisted of a brief description outlining the study, its aims, and our 101 definition of treatment, which was, "any form of external or additional control administered 102 to bees by beekeepers aimed at reducing Varroa numbers". Then followed six questions; 1) 103 association name, 2) number of colonies, 3) if they treat or not, 4) number of times a year 104 they treat, 5) number of years since last treatment, 6) type of treatment (see Fig S2 for  105 survey). Answers were either multiple-choice questions or open questions that all helped 106 assess the beekeeper's treatment habits. The survey was kept short in duration to increase the 107 response rates [20]. 108 The contact details for 325 beekeeping area associations across the UK (Fig. 1a)  area associations and those that participated (Fig 1a). To investigate if any spatial patterns in 129 treatment-free vs treated existed, the proportion of beekeepers falling into either category 130 were calculated from each association that had at least five responses to avoid illegible pie 131 diagrams on the figure (Fig.1b) 132

Statistical analysis 133
A GLM was conducted in R [21] using treatment status as the predictor, association as the 134 fixed effect and colony number as the random effect. A binomial family was used to fit the 135 model due to only two variants of the response variable (i.e., treating or treatment-free). 136 137

138
The 2,856 beekeepers who responded represents almost 10% of the estimated 30,000 139 members belonging to the four UK beekeeping associations and were wildly distributed 140 across the UK (Fig. 1a). An estimate of the number of colonies surveyed was calculated by 141 multiplying the median colony group size (Table 1)  The majority (67%) of beekeepers that responded managed between 1 to 5 colonies and only 155 3% had over 30 colonies (Table 1). Our data indicated that across the five colony size groups 156 the proportion of treatment-free beekeepers was greatest in the 1-5 group, although all other 157 groups were within 5% of that group ( Table 1). The GLM (Table 2) showed that treatment 158 status was not affected by association but did indicate a significant effect of colony group 159 size on treatment status. This due to a significant difference between the 1-5 and 6-15 groups 160

Treatment-free beekeeping 175
A total of 596 (21%) beekeepers stated they were not treating, whereas 2,260 (79%) 176 beekeepers were treating their colonies against Varroa. When asked about the duration 177 elapsed since their last treatment was applied, 72% had treated within the last year and 166 178 (6%) responders had not treated for 10 years or never treated (Table 3). The spatial 179 distribution (Fig. 1b)  However, many UK beekeepers are often hesitant to reveal that they are not treating as they 224 are in the minority and those not treating have been seen as 'bad' beekeepers by those 225 treating. This may explain the 67 beekeepers that did not declare an association due to their 226 unorthodox treatment method or lack of any mite-treatment method. In the UK there are 227 active and growing "treatment-free" communities. For instance, the Westerham beekeepers 228 are approaching their 5 th year of becoming treatment free and starting to bring in neighboring 229 bee clubs. Finally, the 'Natural Beekeeping Trust' has links to 35 like-minded groups and a 230 small number of their members fall into the 'never-treated group', but these beekeepers have 231 not always been welcomed by those adopting treatment regimens [24]. The study indicates 232 that despite hesitation there appears to be a growing number of beekeepers that already have 233 or are switching from using treatments to becoming treatment-free over the past decade, 234 although their numbers remain small relative to those treating. These treatment-free widespread distribution of treatment-free beekeeping found by this study (Fig. 1b). This is 243 important, since the two surveys used different sampling methods. The BBKA and this study 244 surveyed similar numbers of beekeepers, except the BBKA survey targets members randomly 245 selected from their membership each year, whereas this study approaches the beekeepers via 246 their associations using email so are self-selecting. Despite this the two outcomes are similar 247 in many aspects indicating that the survey has not disproportionally been returned by 248 treatment-free beekeepers that may be more active online. In fact, the opposite is the case 249 since the BBKA survey reported significantly more ( 2 = 29.3, p<.001) treatment-free 250 beekeepers than the current survey. 251 Nonetheless, the majority (72-79%) of UK beekeepers are still treating their colonies to 252 control Varroa numbers. We found beekeepers in the UK are predominantly using a single 253 chemical treatment, with the order of popularity been Oxalic acid, Thymol, Amitraz and 254 Formic acid. These treatments were used once or twice yearly. Only 20% of beekeepers in 255 this study are adopting a combination of methods approach (Table 5) although treatment  256 decisions appear to be based on seeing mites on bees or on the floor which is surprising given 257 the high amounts of endorsement that IPM receives [9,26,27]. The use of one method may 258 be more appealing to beekeepers because it is less time consuming than attempting a 259 combination of methods over an extended period such as in an IPM strategy. 260 The popular chemical Varroa treatments like Formic and Oxalic acids reported in this study 261 and other UK surveys [6,25] are also the preferred methods used in Europe [28] and the USA 262 [29]. These compounds have a high efficacy but without the stigma of synthetic compounds 263 or mite resistance, which could explain why many beekeepers are choosing to adopt these 264 methods [30]. However, the impact of these "natural" treatments should not be ignored. the UK and elsewhere have been able to stop treating as their honeybees have learnt to detect 301 mite infested cells and remove the pupa to prevent mite-reproduction, which leads to 302 decreased mite fertility and population size. However, the majority (72-79%) of UK bees are 303 still treating thus it will be many years before most beekeepers in the UK and elsewhere can 304 stop treating for Varroa. In the USA a recent survey found only around 63 (3%) of 2275 305 respondents stated no advantages to Varroa management and 92% of this group did not treat 306 for Varroa [37], a very different situation than found in currently in the UK. Although, most 307 beekeepers in the Northern hemisphere have long wished for a silver bullet for the Varroa 308 problem, however, it turns out that the bullet is their own bees, they just need to give their 309 bees time to develop mite resistant as has been done so successfully elsewhere in the world. The global spread of the parasitic mite Varroa destructor instigated a substantial decline in 26 both managed and feral honeybee (Apis mellifera) colonies mainly across the Northern 27 hemisphere. In response, many beekeepers began to treat their colonies with chemical 28 acaricides to control mite populations in managed colonies. However, some countries or 29 beekeepers allowed their bees to develop mite-resistance by adopting a "treatment-free" 30 approach, rather than using selective breeding programs. Yet, the distribution and proportion 31 of beekeepers either treating or not within the United Kingdom (UK) is unknown, as it is in 32 any Northern hemisphere country. Therefore, the aim of this study was to conduct a 33 beekeeper survey to determine the current treatment strategies within the UK. We gathered 34 2,872 beekeeper responses from an estimated 30,000 UK beekeepers belonging to 242 bee-35 associations in the winter of 2020/21. The survey indicated that the majority (72-79%) of UK 36 beekeepers are still treating their bees for Varroa, typically twice-yearly using chemical-37 based methods. Despite this, ese an estimated 3,000 UK beekeepers are treatment free for 38 over 6 years or have never treated. This is reflected by our finding that only 78 associations 39 out of 242 consist of responders who entirely treated. Furthermore, a marked significant 40 increase in treatment-free beekeepers occurred as the number of colonies they managed 41 increased. Although overall treatment status was not affected by association currently, we 42 predict that in . In time the shift towards intervention-free approaches could lead to increasing 43 numbers of Varroa-resistant feral and managed UK honeybee colonies.  varroacides or a combination of them [18]. Despite this, anecdotal evidence from beekeepers 91 suggests that numbers of non-treating beekeepers are more than expected [19]. 92

93
The aim of this study was therefore to conduct an online survey of UK beekeepers via their 94 associations to assess individual treatment habits, as it was the most efficient way to reach 95 7 group, which produces 21,200 colonies surveyed that is around 10% of the 220,000 colonies 146 estimated by the Center for Ecology and Hydrology to be in the UK. The majority (678%) of beekeepers that responded managed between 1 to 5 colonies and only 151 3% had over 30 colonies (Table 1). OInterestingly our data indicated that across the five 152 colony size groups the proportion of treatment-free beekeepers was greatest in the 1-5 group, 153 although all other groups were within 5% of the highest group similar ranging from 17% to 154 2392% (Table 1).of the 1-5 colony group consisted of beekeepers that treated their colonies. 155 Whereas the most treatment-free beekeepers fell into the 16-30 colony group. The GLM 156 (Table 2) showed that treatment status was not affected by association but did indicate a 157 significant effect of colony group size on treatment status. This due to a significant difference 158 between the 1-5 and 6-15 groups ( 2 [1, n=2628] = 9.81, p =.0017).This explained why there was 159 a significant association with colony number and treatment statusincrease in treatment-free 160 beekeepers as the numbers of colonies they managed increased, although treatment status was 161 not affected by association ( year and 166 (6%) responders had not treated for 10 years or never treated (Table 323). The 188 spatial distribution (Fig. 1B)  An estimated 4,093 treatments per year were administered by 2,238 beekeepers with the aim 204 of reducing Varroa numbers. The majority (70%) are treated once or twice a year (Table 434) 205 using a single chemical method (Table 545). The most popular chemical treatment method is 206 Ooxalic acid, followed by commercially produced Tthymol, and Amitraz. The current study 207 found 78% of beekeepers use chemical treatments (Oxalic acid, Thymol etc), 3% use 208 biomechanical methods (dDrone brood removal, sugar dusting etc) and less than 1% use 209 other methods (rRhubarb leaves, etcetc.) (Table 545). Overall, 80% use a single method of 210 treatment and only 20% use a combination of treatment methods (Table 545) UK beekeepers are often hesitant to reveal that they are not treating as they are in the 234 minority and those not treating have been seen as 'bad' beekeepers by those treating. This 235 may explain the 67 beekeepers that did not declare an association due to their unorthodox 236 treatment method or lack of any mite-treatment method. In the UK there are active and 237 growing "treatment-free" communities. For instance, the Westerham beekeepers are 238 approaching their 5 th year of becoming treatment free and starting to bring in neighboring bee 239 clubs. Finally, the 'Natural Beekeeping Trust' has links to 35 like-minded groups and a small 240 number of their members fall into the 'never-treated group', but these beekeepers have not