Badger vaccination in England: Progress, operational effectiveness and participant motivations

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2020 Crown Copyright. People and Nature published by John Wiley & Sons Ltd on behalf of British Ecological Society 1Animal & Plant Health Agency, National Wildlife Management Centre, Stonehouse, UK 2Centre for Science Studies, Sociology, Lancaster University, Lancaster, UK 3Environment & Sustainability Institute, University of Exeter, Penryn, UK

4. Estimates of badger trapping efficiency for non-government groups did not differ from those achieved by highly experienced government operatives. Our estimate of vaccine coverage (i.e. the average proportion of the target badger population vaccinated during an operation) was 57% (range 48%-63%).

Interviews and participant observation revealed a range of motivations among
individuals involved in badger vaccination including disease control, demonstration of an alternative to badger culling and personal or professional development.
Barriers to wider adoption of badger vaccination expressed by interviewees related primarily to a perceived lack of confidence among farmers and landowners in the effectiveness of badger vaccination for bTB control, but also to the limited availability of funding.
6. Our study suggests that badger vaccination led by non-governmental groups is practically feasible, and may achieve levels of coverage consistent with disease control benefits. Wider uptake of badger vaccination across England might potentially be achieved by addressing the knowledge gap of the effect of badger vaccination on cattle TB, working closely with farmers and vets to better communicate

| INTRODUC TI ON
Many diseases of humans and livestock are also shared by wildlife and, in some instances, management interventions in wild animal populations are required in the interests of public health, agriculture or conservation (Delahay, Smith, Smith, & Hutchings, 2009). In some cases, there may be substantial social barriers to implementing these measures. Hence, a cross-disciplinary approach, combining ecological and social science perspectives, can be valuable in understanding the practical challenges of managing disease at the interface between wildlife, livestock and human populations (De Vos et al., 2016). This may be particularly beneficial where the aim is for management to be implemented voluntarily, as understanding motivations for involvement and barriers to change, including stakeholder confidence in the efficacy of a particular approach, become pivotal. Indeed, from a pragmatic standpoint, it could be argued that in this context stakeholder enthusiasm for an approach is as important as its efficacy.
Bovine tuberculosis (bTB), caused by Mycobacterium bovis, is a chronic disease of cattle, which remains a critical issue in livestock farming in many parts of the world, including the United Kingdom (Palmer, Thacker, Thacker, Waters, Gortázar, & Corner, 2012). Badgers were first identified as a potential source of the disease in UK cattle in 1971 (Muirhead, Gallagher, Gallagher, & Bum, 1974) and there has since been much debate over how best to manage the risk to livestock.
Successive governments have commissioned reviews of the scientific evidence base and research into the effectiveness of different strategies to control disease risks from wildlife (Bourne et al., 2007;Godfray, Donnelly, Donnelly, Hewinson, Winter, & Wood, 2018;Krebs et al., 1997) which essentially comprise of badger culling (Jenkins, Woodroffe, Woodroffe, & Donnelly, 2010;McDonald, Delahay, Delahay, Carter, Smith, & Cheeseman, 2008), biosecurity measures (to reduce interspecies interactions) (Judge, McDonald, McDonald, Walker, & Delahay, 2011) and badger vaccination (Carter et al., 2012;Chambers et al., 2010). Vaccination of cattle is a potential option for the future but is not currently available (Chambers et al., 2014;Vordermeier et al., 2002). Despite considerable research investment, the relative efficacy of the different approaches is not clear and controlling the potential bTB risk posed by badgers remains a practically challenging and controversial issue.
Vaccination can contribute to disease control by reducing the numbers of either susceptible and/or infectious individuals in a population and thereby reducing the number of new infections.
Intuitively, if badgers are an important source of infection to cattle, then reduction in disease incidence in badgers should result in fewer new infections in livestock. However, the actual impact of badger vaccination on cattle TB incidence remains a significant knowledge gap. The UK Government has invested approximately £27 million in research and development of badger vaccines against bTB since the mid-1990s . This resulted in a licensed injectable vaccine (BadgerBCG © ) which has been shown to reduce the severity and progression of disease (and hence excretion of bacilli) in both captive (Chambers et al., 2011;Lesellier et al., 2011) and wild badgers (Chambers et al., 2011 was enacted to allow suitably trained 'lay vaccinators' (i.e. nonveterinarians) to vaccinate badgers for the purposes of disease control (Brown, Cooney, Cooney, & Rogers, 2013). Substantial progress has also been made in relation to the development of an oral vaccine for delivery in a palatable bait, akin to the highly successful approach taken to vaccinate wildlife against rabies in Europe and North America (Cross, Buddle, Buddle, & Aldwell, 2007). However, although there is evidence that orally administered BCG confers a level of protection against infection (Aznar et al., 2018;Chambers et al., 2017) and substantial progress has been made with respect to candidate bait and delivery systems (Carter et al., 2018;Gowtage et al., 2017;Palphramand et al., 2017;Robertson et al., 2016), a licensed oral vaccine is not yet available for use. Consequently, capture and injection is currently the only means of vaccinating badgers in the United Kingdom, and requires a licence issued by governments or their agencies. BEVS aimed to help create 'a protected badger population' between the High Risk Area where a relatively high proportion of cattle herds are infected with bTB, and the Low Risk Area which has a low incidence of bTB in cattle (Defra, 2014). The scheme provided part funding to successful projects, including free training and ongoing support and advice from government staff, loans of equipment such as traps, and vaccine supplies (Defra, 2014). To receive support, vaccination the evidence base (in order to increase confidence in badger vaccination as a viable disease management approach), and by increased financial support for new initiatives and the scaling up of existing projects.

K E Y W O R D S
badger, bovine tuberculosis, vaccination, voluntary and community sector, wildlife disease management projects are required to cover a minimum area of 15 km 2 of largely contiguous, accessible land.
In 2015, the only company manufacturing BadgerBCG experi- Previous research has suggested that landowners and farmers have, in general, little confidence that it is possible to capture a sufficient proportion of the badger population to make vaccination worthwhile (Naylor et al., 2017;Warren, Lobley, Lobley, & Winter, 2013). Here we describe the deployment of badger vaccination in England from the point of licensing in 2010 until 2017 by government and non-government groups including: voluntary groups, commercial operators and farmers. We address some of the key knowledge gaps on practical aspects of vaccine deployment (Naylor et al., 2017), including estimates of trapping efficiency and vaccine coverage. Through dialogue with key stakeholders we also explore the motivations of groups and individuals participating in vaccination projects, and the potential barriers to wider implementation.

| Quantitative data sources and analysis
In England, all badger vaccination projects are required to submit records of their activities as a condition of the Natural England licence under which they operate. These records are held in an SQL database hosted by the Animal and Plant Health Agency. The current study draws on 7,282 records (the total number of records available) from 2010 to 2017. A record is defined here as a given night of badger trapping in a specific area undertaken by a vaccination group.
Reported data include numbers of traps deployed, days of trapping, badgers captured and doses of BCG vaccine administered. Statistical analyses were carried out using GLMM in R version 3.3.2 (R Core Development Team, 2016).

| Injectable vaccine deployment
Badger vaccination involved trapping animals in baited cage traps followed by intra-muscular injection of the vaccine (the full process is described in Box 1). We mapped the spatial distribution of badger vaccine deployment projects in England (2010-2017), aggregated by year and vaccination group type (government, commercial operators, landowners, farming sector and VCS). It is assumed that the number of doses delivered in any given year equates to the number of badgers vaccinated, as each animal was temporarily marked when captured to avoid revaccination within the same year (see Box 1).

| Trapping efficiency
To investigate trapping efficiency during vaccination operations we used data on the numbers of traps set and the proportion of traps that captured badgers on each trapping night at a particular location (e.g. a badger sett, with multiple such locations in each vaccination area). A sufficient number of records was only available for projects undertaken from 2010 to 2015 inclusive (n = 6,352). Too few records were available for operations in 2016 and 2017 because of an interruption in vaccine supply. To investigate predictors of trapping efficiency, we constructed GLMMs using the r package lme4 (v1.0-5; Bates, 2010) with a binomial error structure. The response variable (trapping efficiency) was the number of traps that were occupied (by at least one badger) divided by the total number of traps set. Operational year (i.e. year of the project, indicating F I G U R E 1 Bovine TB risk map for Great Britain (Accessed from TBHub website: https://tbhub.co.uk/guida nce/testi ng-and-compe nsati on/testi ng-areas / 8 January 2018) how long a group had been carrying out vaccination), vaccination group type (government or non-government) and calendar month were included as fixed effects. Inclusion of operational month in the model served to assess whether some months of the open season for badger trapping (May-November inclusive in England) were more favourable than others. Interaction terms were not included in the model as the distribution of data was such that certain combinations of predictor variables were not represented. Vaccination group ID was included as a random effect to account for additional variation between different non-government vaccination groups.
Initial analyses indicated that the models were over-dispersed, so an additional record-level random effect was added to control for this (Harrison, 2014). The significance of fixed effects was evaluated by step-wise model simplification using chi-squared test statistics and a threshold for p of 0.05.

| Vaccine coverage
To estimate coverage in vaccinated areas, it was first necessary to estimate the size of the population subject to each trapping operation. To this end we used the Lincoln-Petersen (LP) index (Lincoln, 1930;Petersen, 1896) with the Chapman adjustment (Chapman, 1951), a simple capture-mark-recapture approach that can be used to estimate population size based on just two visits: a single capture and marking (first night of trapping) and an opportunity for recapture (second night of trapping). The assumption underlying this approach is that the ratio of marked individuals in the original sample to the total population size is the same as the ratio of marked individuals to the total sample size in the second sample. We recognize that a proportion of the population is likely to evade capture and that by this method it is therefore only possible to estimate the size of the 'trappable' population. Vaccination records were aggregated by group and year such that the estimated trappable population size for a given group and year = (total vaccinated and marked on 1st night of trapping × total trapped on 2nd night of trapping)/(total marked badgers trapped on 2nd night of trapping).
We used records from the first and second nights of vaccination operations only (occasionally trapping was extended to a third or fourth night when no badgers were captured on the first two nights). As the LP Index is known to be biased by small sample sizes, we also excluded data from groups where less than seven marked badgers were caught on all second nights of trapping (Robson & Regier, 1964) following the recommendation in Robson and Regier (1964). In some cases, animals captured on the first night of trapping at a given location were already marked indicating that they had been previously trapped and vaccinated as a consequence of a separate trapping operation in a neighbouring area. Hence, to avoid double counting, any marked animals trapped on the first night of a given operation were excluded from the analysis.
We used the 'fishR' function within fsa package to estimate the trappable population size based on the LP index for all remaining vaccination group by year combinations (n = 33), with 95% confidence intervals (Ogle, Wheeler, & Dinno, 2020). In an attempt to account for the 'untrappable' component of the badger populations vaccinated (and thereby get closer to the true population size), we adjusted these estimates upwards by 13%, following the recommendations of Smith and Cheeseman (2007). We then estimated vaccination coverage with 95% confidence intervals.

| Vaccination group motivation and barriers to wider vaccine deployment
Here we draw on findings from participant observation of badger vaccination and interviews undertaken in 2016 and 2017 as part of a wider social research project on bTB 'disease control' practices.
Participant observation was undertaken with a VCS vaccination group in the Edge Area of England in summer and autumn 2017. This involved two visits (each of 2-3 days) with nine volunteers and two badger vaccination group representatives, to document conversations and practices during badger trapping and vaccination. The data considered here are derived from field notes from these visits and nine semi-structured interviews with individuals from across England who were involved in badger vaccination (see Table 1). The aim of the interviews was to learn about opinions and experiences related to bTB and hence topics varied depending on the participant, but questions related to badger vaccination included: • What is your opinion on badger vaccination?
• What are your motivations for taking part (or not taking part) in vaccination?

BOX 1 Badger vaccination: An overview of the process
• What has worked well in the badger vaccination project?
• What have been the main challenges you have faced?
• How have you addressed these challenges? Can anything else be done?
All field notes (FN 01-05) and interview transcripts (Int 01-09) were entered into NVivo (version 11; QSR International Ltd). The data were organized into codes and subcategories; for example, the 'badger vaccination' code was categorized by motivations, barriers and type of activity. Subsequently, these themes were analysed into subthemes. The names of all research participants have been changed to preserve their anonymity.

| Reflections on combined use of quantitative and qualitative data sources
The bringing together of two such contrasting sets of empirical data in this paper presented a number of challenges to the authorship team. For the co-authors who primarily work with quantitative ecological data, a lack of familiarity with methodological approaches, presentational and publishing norms when dealing with qualitative data represented a key learning opportunity. More broadly, the differing aims of the research approaches were highlighted; whereby quantitative approaches more traditionally employed by ecologists involve seeking general patterns, trends and relationships between variables whereas social science research seeks to define and understand complexity rather than reduce it (Creswel, 2009). However, the contrasting datasets used in this context were directed towards a common end; to support decision-making and to facilitate and inform the future practice of badger vaccination.

| E THI C AL S TATEMENT
Ethical approval for the collection of the qualitative data presented in this paper was obtained from Lancaster University, and each participant read and signed a consent form. The names of all research participants have been changed to preserve anonymity. All badgers were trapped under licences issued by Natural England.

| Trapping efficiency
On average, across all vaccination group types, 40% of traps deployed on a given trapping night resulted in the capture of a badger.

| Motivations of vaccination group representatives and vaccination project participants
The motivations of research participants for their initial and ongoing involvement in badger vaccination were summarized into themes  Table 3) and error bars represent standard deviation around means F I G U R E 5 Estimated proportion of the trappable population vaccinated during badger vaccination projects in England (2010)(2011)(2012)(2013)(2014)(2015). The most extreme 95% confidence intervals of estimates from a given year are indicated (Table 4). Some of these themes are closely aligned to government objectives, for example to contribute to the management of bTB, while others appeared unrelated to disease management, for example to build friendships (Table 4).
In interviews, three badger vaccination group representatives (Int 04, Int 05 and Int 06) in the Edge Area spoke of undertaking badger vaccination to try to prevent the spread of bTB: Well 'cos what we're really doing is trying to have like a prevention, so I can see the benefit for vaccination there, in that we don't want it to establish in [Edge Area county] (Int 04) When asked why they wanted to prevent the establishment of bTB in the local area, a badger vaccination group representative who worked for the Wildlife Trusts said it was to protect the rich biodiversity associated with dairy farming: that area that we are doing the vaccination in, you know you get different wildlife, you've got the nice bushy hedgerows, you've got the pasture fields, so it does make a difference. So you know we want to support the dairy farmers there […] 'cos if their industry collapses, it's likely they're then going to turn to arable which is typically worse for ecosystems (Int 06) Her motivation for co-ordinating badger vaccination was to prevent the spread of bTB to help the dairy industry remain in the local area, thereby reducing the likelihood of the land being used for monoculture arable production and protecting the current ecosystems. Her motivation to prevent the spread of bTB was linked to the Wildlife Trusts' wider aim of 'improving life for wildlife and people together' (The Wildlife Trusts, 2018).
Another coordinator of a badger vaccination project said that badger vaccination was being undertaken to build relations with local farmers in the hope that they may work together in trying to manage bTB: its kinda strategic. We work together on vaccination, build bridges and hey presto, we've got relations where we can work together to manage TB. We might not share the same views, but we share the same aims (Int 06) The motivation of 'disease control' was shared by two farming representatives involved in badger vaccination (Int 02, Int 03). They were supportive of badger vaccination as a stop-gap until they secured a badger culling licence for the local area. During interview, a farming representative said: We are not against vaccination, it just needs to be done the right place at the right time. I worked with

| Barriers to wider implementation
The perceived barriers to badger vaccination, as identified by research participants were also summarized into themes (  In addition, practical issues related to badger vaccination were considered to be obstacles to the involvement of farmers. One vaccination group representative (H) said she was proactively changing the practice of badger vaccination to be less intrusive to the landowner to try to make it more attractive to local farmers: Next year, H says they will not do the activity survey before pre-baiting, but rather will do it when they We can't charge farmers for vaccination as many won't pay. Many are only agreeing to vaccination until they get a cull. Charging them for the service will be a sure fire way to reduce the number of sites where we can vaccinate! Saying that, a few large estates are ok paying for peanuts. It's just I don't know who will say yes and who will say no, and I don't want to make vaccination even more unattractive to them (Int 05) Demotivation of volunteers arising from the expansion of the bad-  The UK government commissioned an independent review of bovine TB policy which concluded in 2018  and have recently issued their response to the review findings (Defra, 2020).

| D ISCUSS I ON
Badger vaccination is heavily referenced throughout the response including a desire to gradually replace intensive culling of badgers with Government supported badger vaccination over the coming years. The findings of the present paper are particularly timely therefore in supporting these goals which are important contributors to the government's stated aim of achieving Officially TB Free status for England by 2038 (Defra, 2020).
Previous research suggests that landowners and farmers have, in general, little confidence that it is possible to capture a sufficient proportion of the badger population to make vaccination worthwhile (Naylor et al., 2017;Warren et al., 2013). Data from the present study however indicated that on average, levels of trapping efficiency did not consistently differ between operations led by highly experienced government staff and non-government-led operations.
This suggests that the training framework and the ongoing support system offered by government experts to lay vaccinator groups has been generally effective. Trapping efficiency for volunteer groups increased over time, probably owing to a number of factors, not least of which is likely to be the gradual improvement of staff field skills and increasing knowledge of the target population (number of badgers caught at each sett in previous years etc.). Annual variation in trapping efficiency may reflect badgers becoming habituated to the trapping process over time (Griffiths, 2011) or a gradual increase in the skill or experience of trappers. Our results also indicate seasonal variation in trapping efficiency which was highest in July and lowest in November. Lower availability of natural foods (Garnett, Delahay, Delahay, & Roper, 2002;Tolhurst, Delahay, Delahay, Walker, Ward, & Roper, 2009) and the presence of cubs (which tend to be more likely to be captured than adults, Tuyttens et al., 1999) in summer may explain such seasonal variations.
The results of a recent field trial in the Republic of Ireland suggested vaccination of over 30% of the target badger population with an oral vaccine would make eradication of bTB feasible, given maintenance of existing cattle controls (Aznar et al., 2018). Estimates of annual vaccination coverage from the present study exceeded 50% of the target population (average 57%, minimum annual estimate 48%).
This is consistent with an estimate for vaccine coverage of 55% (95% CI -44%-65%) achieved in a single year of a Welsh Government led program of vaccination by trapping and injection (Smith et al., 2017). trying to influence current government policy) or personal basis. All vaccination group representatives stated they were, at least in-part, motivated to prevent the spread of disease into their local area. This motivation strongly aligns with Defra's aim to fund vaccination projects in the Edge Area 'to support the creation of a protected badger population in uninfected areas' (Defra, 2014, p. 1). Many volunteers were also motivated to undertake badger vaccination for the purpose of disease control; however, this was often related to trying to prevent culling. Volunteers were also motivated by other factors unrelated to bTB, for example to gain work experience and build friendships. The expansion of badger vaccination across the country by VCS groups has likely led to more volunteers being involved for reasons unrelated to disease control. Many vaccination projects rely on volunteers, and so it follows that motivations unrelated to disease control may be important to the overall success of these initiatives.
In this study, individuals involved in badger vaccination identified barriers to its wider deployment related to low levels of confidence in its efficacy among landowners and farmers. Furthermore, a farming representative expressed scepticism about the efficacy of vaccination due to the absence of data from a large-scale field trial. Reporting on telephone interviews with 341 farmers (Enticott et al., 2012) state that 61% of farmers disagreed with the statement that 'Vaccinating badgers is better than culling badgers to control bTB'. Furthermore, in the present study we identified a perception that badger vaccination may reduce the likelihood of securing a cull licence for a given area, potentially creating another barrier to farmers and landowners supporting vaccination. An empirical demonstration of the effects of badger vaccination on the levels of bTB in cattle may help to inform the debate about bTB control in badgers  and, dependent on the results, could help to reduce scepticism about badger vaccination. Interviews with VCS groups in the present study also identified cost as a potential barrier to the expansion of badger vaccination, with several groups struggling to finance projects despite partial funding from Government.
Gloucestershire Wildlife Trust estimate that the average cost of vaccination in their VCS led project was £264/badger vaccinated (Gloucestershire Wildlife Trust, 2015).
Only limited badger vaccination has been carried out by farmers or landowners themselves, consistent with a general lack of confidence within the farming community (Enticott et al., 2014(Enticott et al., , 2012Naylor et al., 2017). This translates into a general unwillingness in the farming community to pay for badger vaccination (Enticott et al., 2014;O'Hagan, Matthews, Matthews, Laird, & McDowell, 2016), which was identified in the present study as a barrier to wider implementation of badger vaccination for VCS groups and is presumably also a key factor in the very limited interest from commercial operators.
The absence of any long-term investment in badger vaccination by farmers may be, at least in part, because it is viewed as a 'stop-gap' until a culling licence can be secured (Int 02 and Int 03). The present study identified potential expansion of badger culling into vaccination areas as demotivating for volunteers, limiting their ambitions to expand vaccination projects over larger areas.
The results of the present study demonstrate that it is possible to train significant numbers of lay badger vaccinators, and to expect that levels of vaccination coverage and efficiency achieved by non-government-led groups to be comparable to government operations. We conclude that non-government-led badger vaccination is therefore practically feasible, and could potentially contribute to bTB control in badgers. However, the current scale of badger vaccination projects is limited and we have identified multiple barriers to its expansion. Initiatives that might be expected to facilitate further uptake of badger vaccination for bTB control include increasing the availability of financial support, an empirical demonstration of the impact of badger vaccination on levels of disease in cattle and working closely with farmers, and vets, to disseminate the evidence to date and increase confidence in badger vaccination.

ACK N OWLED G EM ENTS
The Badger Vaccine Deployment Project was funded by the U.K.
Department for Environment, Food and Rural Affairs. Data on badger vaccination remain under the ownership of Natural England as the licencing body and we thank them for granting us access. The authors express their thanks to all who contributed views and data presented.
Social research findings in this paper are based on doctoral research on the practices of bovine TB control funded by an ESRC studentship (award no. 1539516) held by JP at Lancaster University.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

AUTH O R S ' CO NTR I B UTI O N S
C.H.B. and J.P. designed the paper with extensive input from R.J.D., A.R., F.A.P.S. and R.A.M.; J.P. was solely responsible for the collection and reporting of all qualitative data; C.H.B. carried out quantitative analysis with input from A.R. and F.A.P.S.; R.J.D., R.A.M. and G.W. provided comments on the manuscript. We are grateful for the comments of Dr Rodney Calvert at Natural England as well as those of two reviewers and the Associate Editor.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data from this study have been archived in the Dryad Digital