Electronic supplementary material: Localized reactive badger culling increases risk of bovine tuberculosis in nearby cattle herds. Flavie Vial and Christl A. Donnelly* MRC Centre for Outbreak Analysis and Modelling, Department of Infectious Disease Epidemiology, Imperial College London, Norfolk Place, London, UK. * c.donnelly@imperial.ac.uk 1. Reactive culling The RBCT reactive treatment involved a series of localized badger culls carried out in response to confirmed bTB herd breakdowns (Table 1). When bTB was confirmed in a cattle herd within a reactive trial area, field staff mapped the land used by the affected herd (“reactor land”). Survey data were then used to estimate the likely home ranges of badgers using this land, and to identify their setts (sometimes on neighbouring properties). The average reactive culling operation captured badgers within an area of 5.3km2 (Table 2). A detailed account of the RBCT reactive treatment can be found in a 2005 report to the Independent Scientific Group on Cattle TB [1]. Reactive culling was suspended in November 2003, after data on the incidence of herd breakdowns up to August 2003 [2] revealed that reactive badger culling was associated with an estimated increase of 27% in the incidence of confirmed cattle herd breakdowns (95% CI: 2.4% decrease to 65% increase after adjustment for overdispersion). Table 1 : Approximate dates of reactive culling, by triplet and culling year (defined to run from 1 May – 31 January). Triplet 1999 2000 2002 2003 A Jul-Nov 2000 Jan 2003 May 2003 B May-Dec 1999 Aug-Sept 2000 Sept 2002- Jan 2003 May-July 2003 C May-Aug 2000 Jul 2002- Jan 2003 D May 2003 Aug-Sept 2003 E Jun 2002-Jan 2003 Jul-Oct 2003 F Jul 2002- Jan 2003 Jun-Sept 2003 G Aug 2002- Jan 2003 Sept-Oct 2003 H Jan 2003 Sept-Oct 2003 I May-Sep 2003 Triplet J was eligible for reactive culling in 2003 but no culls had been performed when reactive culling was suspended in November 2003. No culling took place in 2001 during the foot-and-mouth disease outbreak. Table 2: Total numbers of badgers culled in reactive areas, by triplet and culling year. Triplet 1999 2000 2002 2003 Total A B 73 C 34 47 36 117 34 84 110 301 179 115 101 395 122 122 D E 62 126 188 F 145 291 436 G 172 84 256 H 17 143 160 94 94 I J Total 0 73 247 642 1,107 2,069 Triplet J was eligible for reactive culling in 2003 but no culls had been performed when reactive culling was suspended in November 2003. No culling took place in 2001 during the foot-and-mouth disease outbreak. Reactive trial areas Although the boundaries of the RBCT trial areas were at least 3 km apart, the reactive trial area and the proactive trial area in some triplets were geographically close. Some of the herds within the reactive areas were within <5 km of other herds in proactive areas and vice versa. The treatment boundaries of proactive areas extended slightly beyond the trial area boundaries as culling performed on the land parcels belonging to the herd near/on the trial boundaries would have affected badger groups with territories which extended beyond the boundaries of the trial areas. Since proactive badger culling started before reactive badger culling, some of our case herds within reactive areas have had badgers culled within 3km of their lands (those badgers were proactively culled inside the proactive trial areas) during the time period between the start of the first proactive cull and the first reactive cull. Although those badgers were proactively culled prior to the onset of reactive culled, they are still in the proximity of farms within the reactive areas and may therefore conceivably have influenced bTB risks for those reactive herds. This is the reason why the number of badgers proactively culled prior to the onset of reactive culling within 3 and 5kms of the reactive herds was considered in our analyses. Selection of case and control herds. The Defra animal health information system (VETNET) provided data on cattle bTB tests and herd breakdowns. Each case was matched to a control selected randomly from herds within the same triplet that had a clear herd test within a year of the breakdown disclosure date and that had no associated land within 5km of the reactor land. As a result of this study design, and because trial areas were roughly 10km across, the control herds thus selected tended to be located closer to the trial area boundaries than case herds. Median distance to the trial area boundary was 1094m (min = 0, max = 4470) for case herds, with control herds being somewhat closer to the edge of the trial area (median = 871, min = 0, max = 4541) (Figure 1). Figure 1: Stacked histogram of the distribution of herd distance to the trial area boundary (n = 1207) for case and control herds. 2. Local bTB risk factors A number of local herd-level risk factors have been identified inside all three treatment groups of the RBCT by previous analyses [3]. Some of these risk factors had also been described for herds outside the RBCT area. Herd size Herd size has repeatedly been identified as one of the major bTB herd-level risk factor [4-6]. Large herds tend to pasture on larger areas, with higher probabilities of contiguous herds thereby facilitating cattle to cattle spread of M. bovis [7]. A comparative case-control study in England between 1995 and 1999 revealed that herd size was a significant predictor of both transient and persistent bTB breakdowns and associated herd size with management-related risk factors such as turnover rates, stocking density, farm enterprise and feeding [8]. In our reactive herds case-control study, median herd size was estimated at 73 head of cattle (min = 0, max = 1073), with case herds being usually larger than control herds (Figure 2 & Table 5). Figure 2: Stacked histogram of the distribution of herd size (n = 1207) for case and control herds. 400 Count 300 type case 200 control 100 0 0 200 400 600 Herd size 800 1000 Historic incidence Historic incidence of bTB (Tables 3 & 4) was found to be a robust predictor of the rate of future outbreaks in both Irish [9] and British [6] herds. Herd breakdowns tend to be recurrent [10] possibly as a result of the failure to clear the source of the disease, especially from larger herds, by test and slaughter [11]. Subsequent breakdowns could therefore arise from undetected (tuberculinnegative) infected animals. This factor is probably exacerbated for dairy herds for which turnover is less important than for stores or beef enterprises. Other permanent factors (such as the presence of badgers and/or contiguous herds) may make these areas particularly prone to bTB reemergence. Table 3: Number of confirmed bTB breakdowns in the previous year for case and control herds. Number of confirmed Case Control bTB breakdown in previous herds herds 1 year 0 557 559 1 47 45 Table 4: Number of confirmed bTB breakdowns in the previous two years for case and control herds. Number of confirmed Case Control bTB breakdown in previous herds herds 2 years 0 433 491 1 153 106 2 18 7 Farm area Farm area has been found to be associated with an increased bTB risk beyond any effect of herd size as show in England [12]. Larger farms, regardless of the number of land parcels, may include more active badger setts or more contiguous herds, both risk factors identified in other studies [7, 12-14]. Median farm area (combined area of all the land parcels belonging to a farm) was estimated at 0.61 km2 (min = 0.01, max = 5.18), with case herds associated with larger farms (Figure 3 & Table 5). Figure 3: Stacked histogram of the distribution of farm area (n = 1187) for case and control herds. 500 400 300 Count type case control 200 100 0 0 1 2 3 Farm area (km^2) 4 5 6 Dairy herds have been found to be more at risk of a confirmed bTB outbreak (Table 5). Animals in dairy herds tend to have a longer life expectancy, and thus a longer exposure to bTB and increased risk of breakdown [15], than beef cattle that are slaughtered at a young age. Unlike beef farms that use a variety of breeds and crossbred animals, dairy farms in the UK predominantly use one breed of cattle (Ivan Morrison pers. comm.). A breed-related difference in susceptibility may ensue [16] although it is difficult to disentangle its potential effects from higher production stress under more intensive management conditions for dairy cattle for example [17]. Interestingly, dairy herds within the RBCT tended to be much larger than other enterprise types [13] . Table 5: The average herd size, % of dairy herds, farm area and number of confirmed bTB breakdowns in the previous year and the previous two years. Leading Case During Control Case Post1 Control Case Post2 Control Case Control Herd size 135.27 82.05 161.61 87.30 138.16 79.24 135.34 76.28 % dairy herds 43% 17% 34% 18% 33% 17% 33% 19% Farm area (km2) 0.79 0.53 0.93 0.53 1.04 0.59 0.99 0.56 Number of confirmed 0.035 0.078 0.08 0.063 0.12 0.09 0.10 0.08 0.23 0.21 0.26 0.17 0.42 0.23 0.40 0.23 bTB breakdowns (previous year) Number of confirmed bTB breakdowns (previous two years) 3. Summary of variables Data were analysed for four time periods, with both the breakdown disclosure date of the case and the clear herd test of the control required to be within the time period under analysis: a) from the completion of the initial proactive cull (Table 6) until the first reactive culling operation (Table 1) in each triplet (“leading”); b) from the first reactive culling operation in each triplet until the suspension of reactive culling in November 2003 (“during”); c) from the suspension of reactive culling in November 2003 to 1 year after the end of the last proactive cull (Table 6) (“post1”); and d) from the suspension of reactive culling in November 2003 until the compilation of the database on 21 January 2007 (“post2”). Table 6: Dates of initial and follow-up culls in proactive areas, by triplet. Triplet Initial cull 2nd cull 3rd cull 4th cull 5th cull A Jan 2000 May 2002 Nov 2003 May 2004 Oct 2005 B Dec 1998 Oct 1999 D Dec 2002 May 2003 Au 2000– Nov–Dec Jan 2001 2002 Aug–Nov Oct 2003 2002 Sep 2004 Jun 2003 C Nov–Dec 1999 Jan 2001 E May 2000 Jan 2001 Jun 2003 Jul 2004 F Jul 2000 May 2002 May2002– 2005 Jun Jan 2003 Dec 2003 Sep 2004 Jun 2005 G Oct–Nov 2000 Dec 2000 Jul 2002 Jul 2003 Jun 2004 Jun 2005 Jun–Jul 2002 Sep–Oct 2003 Jul–Aug 2003 Sep 2003 May 2004 Jul–Aug 2005 Oct–Nov 2004 Oct–Nov 2004 Jul 2005 H I J Sept–Oct 2002 Oct 2002 Jun 2004 6th cull 7th cull Jul–Aug 2004 Sep 2005 Oct 2005 Sep 2005 May 2005 Data on whether any reactive culling had taken place and the number of badgers culled nearby (defined as within 1, 3, or 5km) cases and controls were extracted from the RBCT geodatabase (ArcGIS version 9, ESRI). Because reactive badger culling operations were prompted by nearby confirmed bTB breakdowns, the number of nearby confirmed breakdowns within 1, 3 or 5km of each case and control was also extracted. The number of nearby tested cattle herds not under bTB-related movement restrictions (again within 1, 3 or 5km) were recorded as a measure of the herd population at risk. Each of these variables was calculated for one year prior to the date the breakdown was detected in the case herd and the herd test date of the control [see main text], and separately for the previous two years (Table 7). Variables were log transformed before being screened using conditional logistic regression and after adjusting for the local bTB risk factors listed above. Table 7: Average number of culled badgers, confirmed bTB breakdowns and tested herds not under bTB restriction for cases and controls by time period (all variables measured in the previous two years). Distance Leading During Case Control Case Post1 Control Case Post2 Control Case Control 3.40 Nearby RBCT 1 0 0.04 11.56 5.89 4.39 2.57 5.61 culled badgers 1 3 0.28 0.49 41.67 21.67 17 9.68 22.01 12.78 5 1.44 1.56 71.82 43.74 31.64 17.66 40.98 23.13 Nearby confirmed 1 5.96 4.03 6.73 7.90 8.15 bTB breakdowns 3 17.48 14.08 19.19 16.2 23.06 18.27 23.93 18.48 5 33.22 28.66 35.57 32.33 42.62 36.62 44.03 37.16 Nearby tested 1 13.04 10.23 11.79 9.51 10.64 8.14 10.40 8.10 herds not under 3 47.34 39.03 39.67 35.49 36.49 30.79 36.52 30.78 bTB restrictions 2 5 96.02 81.23 80.37 75.32 74.16 65.06 74.07 64.48 1 4.94 5.55 5.72 During the “Leading” period from the completion of the initial proactive cull until the first reactive culling operation small numbers of badgers were proactively culled within 3 or 5km of a case or control within a reactive area. Proactively culled badgers may contribute similarly to numbers in the later time periods. 2 Some of the herds which have been tested twice in the 24 months period under study were also counted twice. Estimating odds ratio (OR) corresponding to covariate. Throughout the text, we present the risk (OR and their 95% confidence interval) of a herd experiencing a confirmed bTB breakdown following a doubling in the value of one of the covariates tested. Imagine the estimate of the log odds ratio = 0.2 and the standard error (SE) of this estimate = 0.07. In this case the OR associated with a doubling of the covariate is given by: OR = exp(0.2 * ln(2)) = 1.149 And the 95% confidence interval is obtained by: Lower confidence limit = exp (0.2 * ln(2) - 1.96*SE*ln(2) ) = exp (0.2 * ln(2) - 1.96 * 0.07 * ln(2)) = 1.044 Upper confidence limit = exp (0.2 * ln(2) + 1.96*SE*ln(2) ) = exp (0.2 * ln(2) + 1.96 * 0.07 * ln(2)) = 1.263 4. Further results from analyses based on variables measured in the year previous to the bTB breakdown (main text). Univariable associations with the numbers of nearby culled badgers and confirmed bTB breakdowns in the previous year. While reactive culling was taking place, the univariable models showed strong associations between cases and increased numbers of badgers culled in the previous year (Table 8). The number of (proactively) culled badgers within 3 and 5km of a herd significantly decreased its bTB risk prior to the onset of reactive culling. Cases were consistently associated with increased numbers of nearby confirmed bTB breakdowns in all four time periods and all three distances (Table 9); with a doubling in the number of nearby confirmed bTB breakdowns always resulting in estimated increases in bTB risk from 34% to 231%. 5. Analyses based on variables measured in the two years previous to the bTB breakdown. Nearby reactive culling and TB risk. The presence of any reactive culling activity within 1km of a herd, in the previous two years, significantly increased its bTB risk (OR: 1.81, 95% CI: 1.31-2,48, p < 0.001), even after adjusting for the number of nearby confirmed bTB breakdowns. The detrimental effect of reactive culling remained as the distance to the herd increased (3km: OR: 2.45, 95% CI: 1.59-3.78, p < 0.001; 5km: OR: 3.29, 95% CI: 1.69-6.39, p < 0.001). Associations with the numbers of nearby culled badgers and confirmed bTB breakdowns in the previous two years. The univariable models showed a strong association between case herds and a greater number of badgers culled in the previous two years both during and after, but not before, reactive culling (Table 10). Case herds were also associated with increased numbers of nearby confirmed bTB breakdowns in all four time periods (Table 11). Interestingly, when both variables were included in a multivariable model, the number of nearby badgers culled significantly decreased a herd’s bTB risk prior to the onset of reactive culling, while the number of nearby confirmed bTB breakdowns remained a significant risk factor for herds (Table 12). During reactive culling, the number of nearby badgers culled remained a significant bTB risk factor for herds within 3 and 5km (Table 12), although the number of nearby confirmed bTB breakdowns retained a non-significant trend towards an increased bTB risk. After the end of reactive culling, the number of nearby confirmed bTB breakdowns remained the only significant bTB risk factor (Table 12), although the number of nearby badger culled retained a non-significant trend towards an increased bTB risk. Table 8: Odds ratios (OR) with 95% confidence intervals for the univariable models of the association of cases with increased numbers of culled badgers in the previous year. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby RBCT ---- p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 ---- p = 0.10 p = 0.38 ---- p = 0.11 p = 0.43 OR: 0.45 OR: 0.74 OR: 1.31 OR: 1.33 OR: 1.38 OR: 0.74 OR: 0.92 OR: 0.77 OR: 0.93 (0.29-0.70) (0.54-1.02) (1.18-1.47) (1.20-1.47) (1.23-1.55) (0.51-1.06) (0.77-1.10) (0.56-1.06) (0.79-1.11) culled badgers Herd type p < 0.001 p < 0.001 p < 0.001 p = 0.02 p = 0.02 p = 0.02 p = 0.005 p = 0.003 p = 0.002 p = 0.01 p = 0.005 p = 0.004 [Dairy] OR: 7.05 OR: 8.72 OR: 8.19 OR: 2.07 OR: 2.03 OR: 2.06 OR: 2.19 OR: 2.25 OR: 2.27 OR: 2.18 OR: 2.24 OR: 2.27 (2.86-20.89) (3.33-28.11) (3.12-26.00) (1.16-3.81) (1.12-3.80) (1.13-3.87) (1.29-3.84) (1.33-3.92) (1.35-3.93) (1.15-3.93 (1.30-4.01) (1.32-4.03) p = 0.08 p = 0.11 p = 0.09 p = 0.57 p = 0.64 p = 0.86 p = 0.69 p = 0.63 p = 0.66 p = 0.69 p = 0.62 p = 0.66 OR: 0.88 OR: 0.89 OR: 0.88 OR:0.96 OR: 0.97 OR: 0.99 OR: 0.98 OR: 0.98 OR: 0.98 OR: 0.98 OR: 0.98 OR: 0.98 (0.76-1.02) (0.76-1.03) (0.76-1.02) (0.85-1.09) (0.86-1.10) (0.88-1.12) (0.90-1.07) (0.90-1.07) (0.90-1.07) (0.89-1.08) (0.89-1.07) (0.89-1.07) p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 OR: 15.77 OR: 13.85 OR: 14.75 OR: 21.06 OR: 18.64 OR: 20.56 OR: 11.55 OR: 11.23 OR: 11.96 OR: 10.53 OR: 10.34 OR: 11.03 (3.57-69.71) (3.00-64.06) (3.19-68.25) (6.98-63.55) (6.18-56.26) (6.76-62.54) (4.97-26.88) (4.83-26.12) (5.14-27.84) (4.30-25.77) (4.24-25.23) (4.50-27.03) bTB historic p = 0.02 p = 0.047 p = 0.055 p = 0.81 p = 0.71 p = 0.40 p = 0.79 p = 0.78 p = 0.78 p = 0.79 p = 0.77 p = 0.76 incidence OR: 0.37 OR: 0.41 OR: 0.42 OR: 0.93 OR: 0.88 OR: 0.75 OR: 1.07 OR: 1.07 OR: 1.08 OR: 1.08 OR: 1.08 OR: 1.09 (0.16-0.87) (0.17-0.99) (0.17-1.02) (0.51-1.70) (0.46-1.69) (0.39-1.45) (0.64-1.78) (0.65-1.78) (0.65-1.79) (0.63-1.83) (0.64-1.84) (0.64-1.85) Herd size Farm area Models are adjusted for herd type, herd size, farm area and historic incidence (in grey). Significant associations are denoted in bold. --- denotes that odds ratios were not estimable. Table 9: Odds ratios (OR) with 95% confidence intervals for the univariable models of the association of cases with increased numbers of confirmed bTB breakdowns in the previous year. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby confirmed p = 0.02 p < 0.001 p = 0.005 p < 0.001 p = 0.003 p = 0.02 p = 0.005 p = 0.002 p = 0.04 p = 0.003 p = 0.001 p = 0.03 bTB breakdowns OR: 1.34 OR: 2.08 OR: 2.31 OR: 1.50 OR: 1.72 OR: 1.73 OR: 1.44 OR: 1.80 OR: 1.63 OR: 1.53 OR: 2.01 OR: 1.75 (1.04-1.72) (1.39-3.12) (1.29-4.13) (1.20-1.88) (1.20-2.47) (1.08-2.77) (1.12-1.85) (1.24-2.62) (1.03-2.58) (1.15-2.02) (1.32-3.05) (1.06-2.88) Herd type p < 0.001 p < 0.001 p < 0.001 p = 0.14 p = 0.11 p = 0.09 p = 0.01 p = 0.05 p = 0.01 p = 0.01 p = 0.08 p = 0.02 [Dairy] OR: 6.12 OR: 6.79 OR: 5.80 OR: 1.54 OR: 1.58 OR: 1.64 OR: 2.04 OR: 1.75 OR: 2.03 OR: 2.05 OR: 1.72 OR: 1.98 (2.58-16.98) (2.81-19.18) (2.46-15.87) (0.87-2.76) (0.91-2.82) (0.93-2.93) (1.20-3.58) (1.00-3.14) (1.19-3.57) (1.17-3.72) (0.94-3.20) (1.12-3.60) p = 0.13 p = 0.10 p = 0.17 p = 0.87 p = 0.99 p = 0.95 p = 0.60 p = 0.78 p = 0.86 p = 0.73 p = 0.08 p = 0.94 OR: 0.90 OR: 0.89 OR: 0.91 OR: 1.01 OR: 1.00 OR: 1.00 OR: 0.98 OR: 10.99 OR: 0.99 OR: 0.98 OR: 1.45 OR: 1.00 (0.78-1.03) (0.77-1.02) (0.79-1.04) (0.90-1.13) (0.89-1.12) (0.90-1.12) (0.89-1.07) (0.90-1.08) (0.91-1.08) (0.89-1.108) (0.95-2.22) (0.90-1.10) p = 0.003 p = 0.002 p = 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 OR: 7.16 OR: 8.07 OR: 8.10 OR: 21.73 OR: 21.03 OR: 20.21 OR: 9.21 OR: 11.77 OR: 12.16 OR: 8.84 OR: 10.92 OR: 11.25 (1.96-26.13) (2.19-29.74) (2.29-28.66) (7.33-64.37) (7.23-61.20) (7.11-57.45) (3.80-22.34) (4.72-29.37) (5.03-29.42) (3.41-22.90) (4.08-29.24) (4.38-28.95) bTB historic p = 0.02 p = 0.01 p = 0.02 p = 0.55 p = 0.48 p = 0.64 p = 0.85 p = 0.62 p = 0.78 p = 0.72 p = 0.42 p = 0.65 incidence OR: 0.36 OR: 0.32 OR: 0.38 OR: 0.83 OR: 0.81 OR: 0.87 OR: 1.05 OR: 1.15 OR: 1.08 OR: 1.10 OR: 1.27 OR: 1.14 (0.16-0.82) (0.14-0.75) (0.17-0.88) (0.45-1.53) (0.45-1.47) (0.48-1.57) (0.63-1.76) (0.66-1.99) (0.64-1.83) (0.65-1.89) (0.72-2.27) (0.565-1.98) Number of herd tested p = 0.50 p = 0.59 p = 0.17 p = 0.18 p = 0.13 p = 0.09 p = 0.15 p = 0.10 p = 0.11 p = 0.23 p = 0.108 p = 0.12 not under bTB restriction OR: 1.11 OR: 0.88 OR: 0.68 OR: 0.85 OR: 0.78 OR: 0.74 OR: 1.23 OR: 1.48 OR: 1.51 OR: 1.21 OR: 1.56 OR: 1.55 (0.82-1.50) (0.56-1.39) (0.39-1.17) (0.67-1.08) (0.57-1.07) (0.52-1.05) (0.92-1.65) (0.93-2.33) (0.91-2.51) (0.89-1.66) (0.95-2.57) (0.89-2.71) Herd size Farm area Models are adjusted for herd type, herd size, farm area, historic incidence and the number of nearby herds (in grey). Significant associations are denoted in bold. Table 10: Odds ratios (OR) with 95% confidence intervals for the univariable models of the association of cases with increased numbers of culled badgers in the previous two years. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby RBCT ---- p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p = 0.07 p = 0.01 p = 0.007 p = 0.07 p = 0.01 p = 0.005 OR: 0.46 OR: 0.65 OR: 1.21 OR: 1.32 OR: 1.51 OR: 1.12 OR: 1.15 OR: 1.14 OR: 1.12 OR: 1.15 OR: 1.16 (0.30-10.71) (0.54-0.80) (1.10-1.33) (1.19-1.46) (1.30-1.75) (0.99-1.27) (1.03-1.28) (1.04-1.26) (0.99-1.27) (1.03-1.27) (1.04-1.28) culled badgers Herd type p < 0.001 p < 0.001 p < 0.001 p = 0.03 p = 0.02 p = 0.02 p = 0.005 p = 0.006 p = 0.004 p = 0.005 p = 0.007 p = 0.005 [Dairy] OR: 6.70 OR: 8.27 OR: 6.84 OR: 1.91 OR: 2.13 OR: 2.05 OR: 2.17 OR: 2.14 OR: 2.21 OR: 2.22 OR: 2.19 OR: 2.26 (2.78-19.19) (3.21-26.02) (2.78-19.67) (1.08-3.45) (1.17-3.98) (1.12-3.88) (1.28-3.77) (1.26-3.74) (1.30-3.88) (1.28-3.97) (1.26-3.94) (1.29-4.08) p = 0.17 p = 0.19 p = 0.15 p = 0.74 p = 0.65 p = 0.98 p = 0.58 p = 0.49 p = 0.50 p = 0.59 p = 0.48 p = 0.46 OR: 0.91 OR: 0.91 OR: 0.91 OR:0.98 OR: 0.97 OR: 1.00 OR: 0.98 OR: 0.97 OR: 0.97 OR: 0.97 OR: 0.97 OR: 0.96 (0.80-1.04) (0.79-1.05) (0.79-1.04) (0.87-1.10) (0.86-1.10) (0.88-1.13) (0.90-1.06) (0.89-1.06) (0.89-1.06) (0.89-1.07) (0.88-1.72) (0.88-1.06) p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 OR: 13.85 OR: 12.68 OR: 12.23 OR: 16.92 OR: 18.98 OR: 23.88 OR: 10.84 OR: 11.01 OR: 11.04 OR: 10.31 OR: 10.50 OR: 10.52 (3.37-56.92) (2.95-54.47) (3.00-49.82) (5.88-48.69) (6.34-56.80) (7.52-75.82) (4.59-25.57) (1.28-2.20) (4.63-26.33) (4.15-25.63) (4.17-26.45) (4.18-26.47) bTB historic p = 0.25 p = 0.64 p = 0.61 p = 0.50 p = 0.48 p = 0.83 p = 0.13 p = 0.08 p = 0.06 p = 0.36 p = 0.25 p = 0.22 incidence OR: 0.78 OR: 0.90 OR: 0.790 OR: 1.13 OR: 1.14 OR: 1.04 OR: 1.28 OR: 1.33 OR: 1.36 OR: 1.17 OR: 1.22 OR: 1.24 (0.51-1.19) (0.58-1.39) (0.58-1.38) (0.79-1.60) (0.79-1.66) (0.71-1.53) (0.93-1.77) (0.97-1.84) (0.98-1.87) (0.83-1.64) (0.87-1.72) (0.88-1.75) Herd size Farm area Refer to footnotes from Table 8 Table 11: Odds ratios (OR) with 95% confidence intervals for the univariable models of the association of cases with increased numbers of confirmed bTB breakdowns in the previous two years. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby confirmed p = 0.004 p = 0.007 p = 0.11 p < 0.001 p < 0.001 p = 0.006 p < 0.001 p = 0.001 p = 0.01 p = 0.001 p = 0.001 p = 0.008 bTB breakdowns OR: 1.48 OR: 1.72 OR: 1.60 OR: 1.82 OR: 2.16 OR: 2.05 OR: 1.72 OR: 1.96 OR: 2.02 OR: 1.73 OR: 2.09 OR: 2.25 (1.13-1.93) (1.16-2.954) (0.91-2.82) (1.35-2.44) (1.39-3.35) (1.23-3.42) (1.26-2.36) (1.30-2.97) (1.15-3.55) (1.25-2.40) (1.34-3.26 (1.23-4.10) Herd type p < 0.001 p < 0.001 p < 0.001 p = 0.10 p = 0.15 p = 0.12 p = 0.03 p = 0.06 p = 0.02 p = 0.03 p = 0.06 p = 0.03 [Dairy] OR: 6.18 OR: 6.12 OR: 5.48 OR: 1.62 OR: 1.53 OR: 1.59 OR: 1.86 OR: 1.72 OR: 1.89 OR: 1.90 OR: 1.74 OR: 1.89 Herd size (2.62-16.83) p = 0.19 (2.61-16.64) p = 0.20 (2.38-14.52) p = 0.28 (0.91-2.94) p = 0.99 (0.86-2.75) p = 0.99 (0.90-3.86) p = 0.88 (1.07-3.31) p = 0.44 (0.99-3.05) p = 0.63 (1.10-3.33) p = 0.70 (1.06-3.50) p = 0.57 (0.97-3.19) p = 0.70 (1.07-3.43) p = 0.79 OR: 0.91 OR: 0.92 OR: 0.93 OR: 1.00 OR: 1.00 OR: 1.01 OR: 0.96 OR: 0.98 OR: 0.98 OR: 0.97 OR: 0.98 OR: 0.99 (0.80-1.05) (0.80-1.05) (0.82-1.06) (0.89-1.13) (0.89-1.12) (0.90-1.13) (0.88-1.06) (0.89-1.07) (0.90-1.08) (0.88-1.07) (0.89-1.08) (0.89-1.09) p = 0.005 p = 0.002 p = 0.002 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p = 0.001 p < 0.001 p < 0.001 OR: 6.10 OR: 6.77 OR: 7.05 OR: 17.44 OR: 1.82 OR: 16.56 OR: 5.72 OR: 8.85 OR: 9.76 OR: 5.28 OR: 7.61 OR: 8.92 (1.71-21.80) (1.96-23.34) (2.09-23.75) (5.76-52.82) (1.35-2.44) (5.89-46.56) (2.26-14.50) (3.56-21.99) (4.01-23.76) (1.96-14.26) (2.87-20.16) (3.45-23.05) bTB historic p = 0.07 p = 0.08 p = 0.14 p = 0.39 p = 0.47 p = 0.35 p = 0.09 p = 0.06 p = 0.09 p = 0.28 p = 0.19 p = 0.25 incidence OR: 0.66 OR: 0.67 OR: 0.73 OR: 1.17 OR: 1.14 OR: 1.18 OR: 1.34 OR: 1.38 OR: 1.33 OR: 1.22 OR: 1.27 OR: 1.22 Farm area (0.42-1.03) (0.43-1.05) (0.48-1.11) (0.82-1.67) (0.80-1.62) (0.83-1.67) (0.96-1.86) (0.98-1.92) (0.96-1.84) (0.85-1.73) (0.89-1.82) (0.87-1.73) Number of herd tested p = 0.08 p = 0.19 p = 0.28 p = 0.34 p = 0.28 p = 0.12 p = 0.17 p = 0.43 p = 0.29 p = 0.14 p = 0.27 p = 0.32 not under bTB restriction OR: 1.38 OR: 1.47 OR: 1.50 OR: 0.85 OR: 0.78 OR: 0.66 OR: 1.28 OR: 1.24 OR: 1.37 OR: 1.32 OR: 1.38 OR: 1.39 (0.96-1.97) (0.83-2.61) (0.72-3.13) (0.62-1.18) (0.51-1.22) (0.39-1.11) (0.90-1.82) (0.73-2.11) (0.76-2.44) (0.91-1.92) (0.78-2.45) (0.73-2.62) Refer to footnotes from Table 9 Table 12: Odds ratios (OR) with 95% confidence intervals for the multivariable models of the associations of cases with increased numbers of RBCT culled badgers and increased numbers of confirmed bTB breakdowns in the previous two years. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby RBCT ---- p < 0.001 p < 0.001 p = 0.04 p < 0.001 p < 0.001 p = 0.54 p = 0.27 p = 0.07 p = 0.54 p = 0.35 p = 0.06 OR: 0.44 OR: 0.56 OR: 1.12 OR: 1.31 OR: 1.54 OR: 1.04 OR: 1.07 OR: 1.71 OR: 1.04 OR: 1.06 OR: 1.11 (0.29-0.66) (0.44-0.71) (1.00-1.25) (1.16-1.48) (1.31-1.81) (0.91-1.20) (0.95-1.21) (0.99-1.23) (0.91-1.20) (0.94-1.20) (0.99-1.25) culled badgers Nearby confirmed p < 0.001 p = 0.002 p = 0.01 p = 0.01 p = 0.52 p = 0.92 p = 0.003 p = 0.01 p = 0.07 p = 0.004 p = 0.01 p = 0.048 bTB breakdowns OR: 1.86 OR: 2.03 OR: 2.64 OR: 1.53 OR: 1.19 OR: 0.97 OR: 1.66 OR: 1.79 OR: 1.71 OR: 1.66 OR: 1.91 OR: 1.88 (1.36-2.55) (1.29-3.22) (1.26-5.55) (1.10-2.13) (0.71-1.98) (0.52-1.79) (1.19-2.32) (1.15-2.79) (0.95-3.09) (1.17-2.35) (1.18-3.09) (1.00-5.52) Herd type p < 0.001 p < 0.001 p < 0.001 p = 0.06 p = 0.02 p = 0.01 p = 0.03 p = 0.05 p = 0.02 p = 0.03 p = 0.06 p = 0.03 [Dairy] OR: 8.21 OR: 9.50 OR: 6.91 OR: 1.77 OR: 2.19 OR: 2.36 OR: 1.87 OR: 1.75 OR: 2.36 OR: 1.91 OR: 1.76 OR: 1.94 Herd size (3.05-26.78) p = 0.05 (3.38-33.44) p = 0.07 (2.63-21.40) p = 0.09 (0.98-3.25) p = 0.89 (1.17-4.23) p = 0.69 (1.23-4.73) p = 0.99 (1.07-3.33) p = 0.43 (1.00-3.11) p = 0.56 (1.23-4.73) p = 0.59 (1.07-3.51) p = 0.55 (0.98-3.25) p = 0.63 (1.09-3.57) p = 0.61 OR: 0.85 OR: 0.86 OR: 0.87 OR: 0.99 OR: 0.97 OR: 1.00 OR: 0.96 OR: 0.97 OR: 0.97 OR: 0.97 OR: 0.97 OR: 0.97 (0.72-1.00) (0.73-1.01) (0.75-1.02) (0.88-1.12) (0.86-1.10) (0.88-1.13) (0.88-1.06) (0.89-1.07) (0.89-1.07) (0.88-1.07) (0.88-1.08) (0.88-1.08) p < 0.001 p = 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p <0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 OR: 14.87 OR: 1.76 OR: 12.32 OR: 17.23 OR: 20.12 OR: 27.66 OR: 5.81 OR: 8.96 OR: 9.85 OR: 5.38 OR: 7.70 OR: 8.91 (3.10-71.42) (0.89-3.48) (2.83-53.58) (5.64-52.66) (6.65-68.85) (8.33-91.86) (2.29-14.77) (3.58-22.40) (4.00-24.21) (1.98-14.57) (2.88-20.59) (3.40-23.37) bTB historic p = 0.03 p = 0.18 p = 0.22 p = 0.51 p = 0.525 p = 0.89 p = 0.10 p = 0.07 p = 0.07 p = 0.31 p = 0.18 p = 0.22 incidence OR: 0.58 OR: 0.71 OR: 0.74 OR: 1.13 OR: 1.13 OR: 1.03 OR: 1.32 OR: 1.37 OR: 1.35 OR: 1.20 OR: 1.27 OR: 1.25 (0.35-0.95) (0.44-1.17) (0.46-1.20) (0.79-1.62) (0.78-1.64) (0.70-1.51) (0.95-1.84) (0.98-1.91) (0.97-1.87) (0.84-1.71) (0.89-1.82) (0.88-1.77) Number of herd tested p = 0.13 p = 0.10 p = 0.11 p = 0.39 p = 0.20 p = 0.17 p = 0.15 p = 0.38 p = 0.25 p = 0.12 p = 0.23 p = 0.25 not under bTB restriction OR: 1.36 OR: 1.76 OR: 2.17 OR: 0.87 OR: 0.74 OR: 0.67 OR: 1.30 OR: 1.27 OR: 1.41 OR: 1.34 OR: 1.42 OR: 1.46 (0.92-2.01) (0.89-3.48) (0.85-5.56) (0.62-1.21) (0.47-1.18) (0.38-1.18) (0.91-1.85) (0.74-2.17) (0.78-2.54) (0.92-1.96) (0.80-2.54) (0.76-2.79) Farm area Models are adjusted for herd type, herd size, farm area, historic incidence and the number of nearby herds (in grey). Significant associations are denoted in bold. --- denotes that odds ratios were not estimable. 6. Alternative analyses Permutation tests Herds geographically close to one another are more likely to have similar covariates. Furthermore, if bTB is present in an area, then the areas around it are more likely to also be affected. We investigated any impact of unmodelled spatial correlation on our results using permutation tests. For each univariable and multivariable model, paired herds were randomly relabelled as “case herd” or “control herd” and new estimates were derived. We fitted each model and calculated the proportion of permutations for which we observed, for each variable, an effect at least as significant as the one in the model fitted to the observed (non-randomised data) to obtain permutation-based p-values. We found that the standard p-values (as reported in the main text) were very similar to those obtained from the permutation tests (Table 13) and thus any unmodelled spatial correlation within the data had not affected the significance of our reported results. Inclusion of the herd’s distance to the trial boundary We observed an apparently protective effect of the number of badgers proactively culled before the onset of reactive culling (main text Table 2). A possible explanation could be that, as a result of the selection process for control herds (see above), control herds were more likely to be located close to the trial area boundary and therefore more likely to have had badgers proactively culled in their vicinity. To allow for this potential effect, we re-fitted all models adjusting for herds’ distances to the trial area boundary. We found that the output from the models adjusted for the distance to the trial area boundary (Table 14) were very similar to those without this adjustment (main text). We also observed that the number of badgers proactively culled before the onset of reactive culling remained a significant protective factor, even after adjusting for the herds’ distances to the trial area boundary. Table 13: P-values of the permutation tests for the multivariable models of the associations of cases with increased numbers of RBCT culled badgers and increased numbers of confirmed bTB breakdowns in the previous year. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby RBCT ---- p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 ---- p = 0.047 p = 0.21 ---- p = 0.04 p = 0.22 p = 0.003 p = 0.001 p = 0.02 p = 0.10 p = 0.67 p = 0.75 p = 0.003 p < 0.001 p = 0.03 p < 0.001 p < 0.001 p = 0.02 Herd type [Dairy] p < 0.001 p < 0.001 p < 0.001 p = 0.03 p = 0.01 p = 0.004 p = 0.03 p = 0.11 p = 0.02 p = 0.04 p = 0.15 p = 0.03 Herd size p = 0.03 p = 0.04 p = 0.03 p = 0.64 p = 0.53 p = 0.62 p = 0.64 p = 0.80 p = 0.86 p = 0.79 p = 0.87 p = 0.92 Farm area p < 0.001 p = 0.003 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 bTB historic p = 0.003 p < 0.001 p = 0.01 p = 0.83 p = 0.74 p = 0.62 p = 0.95 p = 0.74 p = 0.84 p = 0.82 p = 0.52 p = 0.71 p = 0.44 p = 0.43 p = 0.27 p = 0.13 p = 0.03 p = 0.49 p = 0.31 p = 0.12 p = 0.12 p = 0.51 p = 0.07 p = 0.12 culled badgers Nearby confirmed bTB breakdowns incidence Number of herd tested not under bTB restriction Refer to footnotes from Table 12 Table 14: Odds ratios (OR) with 95% confidence intervals for the multivariable models of the associations of cases with increased numbers of RBCT culled badgers and increased numbers of confirmed bTB breakdowns in the previous year and accounting for herds’ distance to the trial area boundary. Leading During Post1 Post2 Distance (km) 1 3 5 1 3 5 1 3 5 1 3 5 Nearby RBCT ---- p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 ---- p = 0.06 p = 0.26 ---- p = 0.06 p = 0.32 OR: 0.42 OR: 0.53 OR: 1.26 OR: 1.35 OR: 1.47 OR: 0.66 OR: 0.90 OR: 0.68 OR: 0.91 (0.26-0.70) (0.38-0.72) (1.12-1.42) (1.20-1.53) (1.27-1.69) (0.43-1.01) (0.74-1.09) (0.46-1.01) (0.75-1.10) culled badgers Nearby confirmed p = 0.02 p = 0.003 p = 0.04 p = 0.13 p = 0.76 p = 0.60 p = 0.002 p = 0.001 p = 0.03 p = 0.001 p < 0.001 p = 0.02 bTB breakdowns OR: 1.46 OR: 2.18 OR: 2.09 OR: 1.22 OR: 1.20 OR: 0.86 OR: 1.51 OR: 1.93 OR: 1.70 OR: 1.63 OR: 2.15 OR: 1.84 (1.08-1.97) (1.30-3.66) (1.03-4.23) (0.44-1.59) (0.69-1.66) (0.49-1.50) (1.16-1.96) (1.30-2.87) (1.06-2.74) (1.21-2.19) (1.37-3.36) (1.09-3.08) Herd type p < 0.001 p < 0.001 p < 0.001 p = 0.02 p = 0.007 p = 0.003 p = 0.02 p = 0.09 p = 0.02 p = 0.03 p = 0.14 p = 0.03 [Dairy] OR: 8.98 OR: 10.85 OR: 9.23 OR: 2.19 OR: 2.45 OR: 2.89 OR: 1.91 OR: 1.65 OR: 1.97 OR: 1.90 OR: 1.61 OR: 1.92 (3.22-31.24) (3.65-41.75) (3.15-34.23) (1.18-4.20) (1.31-4.78) (1.48-5.95) (1.10-3.40) (0.93-3.01) (1.14-3.47) (1.06-3.52) (0.86-3.04) (1.08-3.49) p = 0.02 p = 0.04 p = 0.02 p = 0.40 p = 0.33 p = 0.32 p = 0.64 p = 0.80 p = 0.88 p = 0.80 p = 0.89 p = 0.98 OR: 0.82 OR: 0.83 OR: 0.82 OR: 0.94 OR: 0.93 OR: 0.93 OR: 0.98 OR: 0.99 OR: 0.99 OR: 0.99 OR: 0.99 OR: 1.00 (0.70-0.97) (0.70-0.99) (0.69-0.97) (0.83-1.08) (0.82-1.07) (0.82-1.07) (0.89-1.07) (0.90-1.08) (0.91-1.09) (0.89-1.02) (0.89-1.10) (0.90-1.10) p < 0.001 p = 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 OR: 19.35 OR: 14.31 OR: 17.22 OR: 27.66 OR: 26.18 OR: 34.66 OR: 8.87 OR: 10.19 OR: 11.23 OR: 8.72 OR: 9.37 OR: 10.49 (3.83-97.65) (2.80-73.13) (3.50-84.68) (8.63-88.68) (8.09-84.67) (10.05-119.57) (3.60-21.85) (4.05-25.67) (4.57-27.60) (3.28-23.18) (3.45-25.42) (3.98-27.67) bTB historic p = 0.007 p = 0.006 p = 0.02 p = 0.85 p = 0.99 p = 0.83 p = 0.93 p = 0.74 p = 0.80 p = 0.81 p = 0.51 p = 0.65 incidence OR: 0.28 OR: 0.25 OR: 0.29 OR: 1.06 OR: 1.00 OR: 0.92 OR: 1.02 OR: 1.10 OR: 1.071 OR: 1.07 OR: 1.21 OR: 1.14 (0.11-0.70) (0.09-0.67) (0.10-0.81) (0.56-2.02) (0.51-1.99) (0.45-1.90) (0.61-1.72) (0.63-1.90) (0.63-1.81) (0.62-1.84) (0.68-2.17) (0.66-1.97) Number of herd tested p = 0.63 p = 0.47 p = 0.33 p = 0.10 p = 0.04 p = 0.08 p = 0.33 p = 0.10 p = 0.13 p = 0.51 p = 0.08 p = 0.16 not under bTB restriction OR: 1.09 OR: 1.23 OR: 1.45 OR: 0.81 OR: 0.70 OR: 0.71 OR: 1.16 OR: 1.47 OR: 1.49 OR: 1.12 OR: 1.58 OR: 1.51 (0.77-1.54) (0.70-2.16) (0.69-3.05) (0.63-1.04) (0.50-0.98) (0.49-1.04) (0.86-1.57) (0.92-2.33) (0.89-2.49) (0.80-1.55) (0.95-2.63) (0.85-2.67) Distance to trial area p = 0.04 p = 0.30 p = 0.17 p = 0.02 p = 0.03 p = 0.006 p = 0.67 p = 0.82 p = 0.45 p = 0.44 p = 0.62 p = 0.25 boundary OR: 2.24 OR: 1.55 OR: 1.47 OR: 2.24 OR: 2.19 OR: 2.83 OR: 1.14 OR: 1.08 OR: 1.27 OR: 1.29 OR: 1.19 OR: 1.47 (1.05-4.78) (0.67-3.56) (0.79-3.92) (1.14-4.39) (1.08-4.43) (1.36-5.92) (0.62-2.11) (0.57-2.03) (0.68-2.35) (0.67-2.48) (0.60-2.34) (0.76-2.85) Herd size Farm area Models are adjusted for herd type, herd size, farm area, historic incidence, the number of nearby herds and the distance to the trial area boundary (in grey). Significant associations are denoted in bold. --- denotes that odds ratios were not estimable. 1 Le Fevre, A. M., Donnelly, C. A., Cox, D. R., Bourne, J., Clifton-Hadley, R. S., Gettinby, G., Johnston, W. T., McInerney, J. P., Morrison, W. I., Woodroffe, R. The impact of localised reactive badger culling versus no culling on TB incidence in British cattle: a randomised trial. 2005 [cited 2011]; Available from: http://collections.europarchive.org/tna/20081027092120/http:/defra.gov.uk/animalh/tb/isg/pdf /lefevre1005.pdf 2 Donnelly, C. A., Woodroffe, R., Cox, D. R., Bourne, J., Gettinby, G., Le Fevre, A. M., McInerney, J. P., Morrison, W. I. 2003 Impact of localized badger culling on tuberculosis incidence in British cattle. Nature. 426, 834-837. 3 Vial, F., Johnston, W. T., Donnelly, C. 2011 Local cattle and badger populations affect the risk of confirmed tuberculosis in British cattle herds. PLoS One. 6(3):e18058., 4 Kaneene, J. B., Bruning-Fann, C. S., Granger, L. M., Miller, R., Porter-Spalding, B. A. 2002 Environmental and farm management factors associated with tuberculosis on cattle farms in northeastern Michigan. Journal of the American Veterinary Medical Association. 221, 837-842. (doi:10.2460/javma.2002.221.837) 5 Porphyre, T., Stevenson, M. A., McKenzie, J. 2008 Risk factors for bovine tuberculosis in New Zealand cattle farms and their relationship with possum control strategies. Prev. Vet. Med. 86, 93106. 6 White, P. C. L., Benhin, J. K. A. 2004 Factors influencing the incidence and scale of bovine tuberculosis in cattle in southwest England. Prev. Vet. Med. 63, 1-7. 7 Griffin, J. M., Martin, S. W., Thorburn, M. A., Eves, J. A., Hammond, R. F. 1996 A case-control study on the association of selected risk factors with the occurrence of bovine tuberculosis in the Republic of Ireland. Prev. Vet. Med. 27, 217-229. 8 Reilly, L. A., Courtenay, O. 2007 Husbandry practices, badger sett density and habitat composition as risk factors for transient and persistent bovine tuberculosis on UK cattle farms. Prev. Vet. Med. 80, 129-142. 9 Olea-Popelka, F. J., White, P. W., Collins, J. D., O'Keeffe, J., Kelton, D. F., Martin, S. W. 2004 Breakdown severity during a bovine tuberculosis episode as a predictor of future herd breakdowns in Ireland. Prev. Vet. Med. 63, 163-172. 10 Krebs, J. R., Anderson, R., Clutton-Brock, T., Morrison, I., Young, D. 1997 Bovine tuberculosis in cattle and badgers. London: Ministry of Agriculture, Fisheries and Food. 11 Woodroffe, R., Bourne, F. J., Donnelly, C. A., Cox, D. R., Gettinby, G., McInerney, J. P., Morrison, W. I. 2003 Towards a sustainable policy to control TB in cattle. In: Conservation and Conflict: Mammals and Farming in Britain. London, UK: Linnean Society. 12 Johnston, W. T., Gettinby, G., Cox, D. R., Donnelly, C. A., Bourne, J., Clifton-Hadley, R., Le Fevre, A. M., McInerney, J. P., Mitchell, A., Morrison, W. I., et al. 2005 Herd-level risk factors associated with tuberculosis breakdowns among cattle herds in England before the 2001 foot-and-mouth disease epidemic. Biology Letters. 1, 53-56. 13 Vial, F., Johnston, W. T., Donnelly, C. In prep. Local cattle and badger populations affect the risk of confirmed tuberculosis in British cattle herds. 14 Denny, G. O., Wilesmith, J. W. 1999 Bovine tuberculosis in Northern Ireland: a case-control study of herd risk factors. Vet Rec. 144, 305-310. 15 Humblet, M. F., Boschiroli, M. L., Saegerman, C. 2009 Classification of worldwide bovine tuberculosis risk factors in cattle: a stratified approach. Vet. Res. 40, (50 10.1051/vetres/2009033) 16 Allen, A. R., Minozzi, G., Glass, E. J., Skuce, R. A., McDowell, S. W. J., Woolliams, J. A., Bishop, S. C. 2010 Bovine tuberculosis: the genetic basis of host susceptibility. Proceedings of the Royal Society B: Biological Sciences. 277, 2737-2745. 17 Ameni, G., Aseffa, A., Engers, H., Young, D., Gordon, S., Hewinson, G., Vordermeier, M. 2007 High Prevalence and Increased Severity of Pathology of Bovine Tuberculosis in Holsteins Compared to Zebu Breeds under Field Cattle Husbandry in Central Ethiopia. Clin. Vaccine Immunol. 14, 13561361.