Clicky

Most recent
 Bovine TB in GB. Latest SAM data, maps and older VETNET data.
Added or updated 2 months ago
PDF icon is-culling-badgers-effective
Added or updated 5 months ago
 An example of how badger culling in the UK is being cast in a negative light
Added or updated 11 months ago
PDF icon number-of-badgers-removed-up-to-end-of-2022
Added or updated 12 months ago
 Bovine TB in the UK, England, Ireland, Wales and New Zealand
Added or updated one year ago
 Cattle movements and TB restricted herds
Added or updated one year ago
 Does badger culling make economic sense?
Added or updated one year ago
 Data needed to achieve meaningful results in the 2013 badger culls
Added or updated one year ago
 Badger cull thoroughness
Added or updated one year ago
 TB in Great Britain and Scotland
Added or updated one year ago
 A critique of the Randomised Badger Culling Trial (RBCT)
Added or updated one year ago
 Launch of farmer-led TB governance in New Zealand
Added or updated one year ago
 Bovine TB in Ireland
Added or updated one year ago
 Historical record of bovine TB in Ireland and Great Britain
Added or updated one year ago
 Did culling affect TB infection in badgers?
Added or updated one year ago
 Is specificity of the skin test over-estimated?
Added or updated one year ago
 Number of cattle tested and slaughtered
Added or updated one year ago
PDF icon european-badger-responses-to-low-intensity-selective-culling-Using-mark-recapture-and-relatedness-data-to-assess-social-perturbation
Added or updated 2 years ago
PDF icon jahresstrecke-dachse-2019-2020
Added or updated 3 years ago
PDF icon 3-steps-to-implement-a-bovine-TB-contingency-plan-farmers-weekly-11-feb-2021
Added or updated 3 years ago

Pilot badger culls in Somerset and Gloucestershire and their value in planning future policy

Summary

This page looks at how the ability to reliably estimate a badger population in a cull zone has significantly improved in years subsequent to the Randomised Badger Culling Trial (RBCT 1998 to 2005). It looks at the reduction in the badger population in the pilot zones. It finally explains why accurately measuring badger number reduction is important for planning future TB control policy.

Proportion of badgers culled during the first 6 weeks of culling in 2013

Subsequent to the Randomised Badger Culling Trial (RBCT) a very simple and powerful method has become available which directly measures the proportion of animals removed from a population. This method is known as cull-sample-matching (CSM).

CSM was used in the first year of the pilot badger culls. DNA profiles from trapped hairs were used to identify badgers in an initial survey. Badgers were then culled and from samples taken from these culled badgers the number of profiles which matched unique profiles in the survey were found. This number divided by the number of unique profiles in the survey then directly gave the proportion of badgers removed. Dividing the number of badgers removed by the proportion found gave the initial size of the population.

In my view, this is an enormous step forward and a watershed moment in our ability to estimate the proportion of badgers removed and hence badger population size. One of CSM's big strengths is that it removes much of the need for human judgement which is a major concern with some other methods currently available particularly when judgement becomes very difficult due to changing season, weather and ground conditions.

The graphs below show the results of CSM used in the pilot culls in the Somerset and Gloucestershire zones during the first 6 weeks of culling. CSM found that in the first year of culling approximately 40% of badgers were removed in the Somerset zone and approximately 35% in the Gloucestershire zone. The ranges in each graph represent 95% confidence intervals.

Percentages of badgers culled in Somerset and Gloucestershire
The results shown in the above graph were extracted from Table 3.1 on Page 7 of Reference 1. The proportions spotted in the graph were calculated by dividing the number of ear tips matching hair profiles by the number of unique hair profiles in the survey. As noted at the foot of Table 3.1 in Reference 1, this quotient neither accounts for false negatives nor population churn. The 95% confidence intervals which are shown in the above graph were taken from the estimated cull efficacy stated in Table 3.1.

The initial number of badgers in the Somerset cull zone

According to Table 6.6 on page 30 of Reference 1, the area of the Somerset cull zone is 256 km2.

According to Table 3.1 on Page 7 of Reference 1,

Number of ear tips matching hair profiles = 87
Number of unique hair profiles = 216
Number of badgers reported culled = 865

In the number of badgers reported culled, 5 were found to be duplicates and 2 were found to be road kill.1 This left 858.

Hence CSM gave the number of badgers which existed before the cull in 2013 in the Somerset zone to be 858*216/87 = 2130.

This gives the number of badgers per square kilometre to be 2130/256 = 8.3.

The initial number of badgers in the Gloucestershire cull zone

According to Table 6.6 on page 30 of Reference 1, the area of the Gloucestershire cull zone is 311 km2.

According to Table 3.1 on Page 7 of Reference 1,

Number of ear tips matching hair profiles = 80
Number of unique hair profiles = 233
Number of badgers reported culled = 708.

Hence CSM gave the number of badgers which existed before the cull in 2013 in the Gloucestershire zone to be 708*233/80 = 2062.

This gives the number of badgers per square kilometre to be 2062/311 = 6.6.

As noted before, the above numbers neither account for false negatives nor population churn.

Proportion of badgers culled during the complete cull period in 2013

Culling in the Somerset zone was extended by a further 3 weeks during which time a further 90 badgers were removed, taking the total across the whole cull period to 940. See Page 9 of Reference 8.

Culling in the Gloucestershire zone was extended by a further 5 weeks and 3 days during which time an additional 213 badgers were removed bringing the total to 921. See Page 9 of Reference 8.

Hence using the initial populations estimated above, the proportion of badgers which had been removed over the complete cull periods were as follows.

Proportion in Somerset = 940/2130 = 44%.
Proportion in Gloucestershire = 921/2062 = 45%.

It should be noted that the proportion of the additional removed badgers which were road kill in the two zones is an unknown. It follows that the extent to which these proportions over-estimate the proportions of badgers which were actually culled is an unknown.

Causes of cull-sample-matching errors

Reference 10 concluded that hair-trapping by means of unbaited barbed-wire traps, placed at sett entrances and well-used runs, offers a method of censusing badgers that is relatively accurate and precise, comparatively non-invasive, potentially applicable in a variety of habitats and at different population densities, and not prohibitively expensive. However CSM, which is a method which uses hair-trapped samples, leads to estimates of badger proportions culled which are lower than other methods used to-date. This section lists reasons which may explain why the CSM results may be in error. It then looks at how hair samples were collected so that the feasibility of these reasons may be assessed.

Reasons for low estimates include the following.
  1. Hairs from badgers which were already dead were collected in the survey.
  2. Hairs from badgers outside the cull zone were collected in the survey.
  3. Hair trapping effort was concentrated where culling was not.
Section 2.1 on Page 5 and Section 7.7 on Page 27 of Reference 1 supplies some background as to how hairs were collected in the survey as follows.

We aimed to sample as many badgers as possible in 50 1km x 1km cells selected at random from a grid covering approximately 300km2 in each of the two pilot areas. This aimed to provide a representative sample, covering 17% of cells and 12% of the population, assuming an even distribution and 70% trapping/sampling efficiency.
...
Hair traps consisted of a loop of barbed wire fixed to wooden stakes, on fence-lines or natural objects such as logs and suspended approximately 20cm above the ground. They were placed at active setts and areas of obvious badger activity such as runs crossing land boundaries and at latrines. Each hair trap was given a unique alphanumeric identifier and its location taken with a GPS unit. Samples from hair traps were collected on each day for up to 18 days. During the first three days hair traps were moved or additional hair traps were placed to maximise sample collection. Only samples from the last 15 days were used to estimate population sizes. Each sample consisted of hairs from a single tuft. Where a hair trap had tufts on multiple barbs all tufts were taken and stored in separate plastic ziplock bags.

A fourth reason why CSM may lead to an erroneous low estimate of the percentage of badgers culls is failure to identify when matching samples match. This is commonly referred to as a false negative.

Capture Mark Recapture

As with cull-sample-matching, capture-mark-recapture (CMR) uses DNA analyses on trapped hairs. According to Page 7 of Reference 12, it works as follows.

The pre-cull badger population was estimated based on the number of setts in the area and the average number of badgers per sett based on the frequencies with which badgers were repeatedly hair-trapped. Then, the number of badgers culled was compared against this estimate of pre-cull population size, in order to provide an estimate of the proportion of badgers removed.

CMR is considerably more involved and complex than CSM and is considered to be less reliable than CSM.12

The number of badgers culled

The following thumbnail shows the number of badgers culled per square kilometre in the pilot zones and compares these numbers each year with the numbers culled in the Randomised Badger Culling Trial (RBCT). Click the thumbnail to clearly see the graphs.



Data in the above graph were sourced as follows.

In the RBCT the numbers of badgers culled in each of the 10 proactive triplets were sourced as shown in the following table.

CullReference
1stPage 10 of Reference 2
2nd-4thTable 2.4 on Page 49 of Reference 4

These numbers of badgers culled were summed across all 10 proactive triplets and divided by 1000. One thousand was assumed (a) to be the total size of the proactive triplets in square kilometres when summed, and (b) to be consistent with the areas stated in Table 6.6 on page 30 of Reference 1 for the pilot zones.

The number of badgers culled in the pilot zones were sourced as shown in the following table.

CullAreaNumber of badgers culledReference
2013gloucestershire_1921Page 9 of Reference 8
2013somerset_2940Page 9 of Reference 8
2014gloucestershire_1274Page 1 of Reference 3
2014somerset_2341Page 1 of Reference 3
2015gloucestershire_1432Page 1 of Reference 15
2015somerset_2279Page 1 of Reference 15
2015dorset_3756Page 1 of Reference 15
2016gloucestershire_1252Page 2 of Reference 17
2016somerset_2217Page 2 of Reference 17
2016dorset_3502Page 2 of Reference 17
2016cornwall_4711Page 2 of Reference 17
2016cornwall_5851Page 2 of Reference 17
2016devon_62038Page 2 of Reference 17
2016devon_7833Page 2 of Reference 17
2016dorset_83000Page 2 of Reference 17
2016gloucestershire_91858Page 2 of Reference 17
2016herefordshire_10624Page 2 of Reference 17
2017gloucestershire_1172
2017somerset_291
2017dorset_3257Page 2 of Reference 18
2017cornwall_4213Page 2 of Reference 18
2017cornwall_5358Page 2 of Reference 18
2017devon_6727Page 2 of Reference 18
2017devon_7246Page 2 of Reference 18
2017dorset_81166Page 2 of Reference 18
2017gloucestershire_91012Page 2 of Reference 18
2017herefordshire_10394Page 2 of Reference 18
2017cheshire_11736Page 2 of Reference 18
2017devon_121874Page 2 of Reference 18
2017devon_131237Page 2 of Reference 18
2017devon_14708Page 2 of Reference 18
2017devon_15763Page 2 of Reference 18
2017dorset_163450Page 2 of Reference 18
2017somerset_171123Page 2 of Reference 18
2017somerset_18489Page 2 of Reference 18
2017wiltshire_192252Page 2 of Reference 18
2017wiltshire_201040Page 2 of Reference 18
2017wiltshire_211229Page 2 of Reference 18
The area of the Dorset cull zone was taken from Page 6 of Reference 16.

Reference 13 states the number of badgers culled in 2013 and Reference 3 states the numbers culled in 2014 to be as follows.

Year Area Initial 6 weeks Extension Total
Number cage trapped Number free shot % culled by free shooting Number cage trapped Number free shot % culled by free shooting
2013Gloucestershire165543771496731924
2013Somerset50036542494146955
2014Gloucestershire10816661---274
2014Somerset19414743---341

Evidently the percentage of badgers culled by free shooting in the initial 6 weeks was significantly higher in Gloucestershire than in Somerset for both 2013 and 2014. Protester activity in Gloucestershire was reported to be more than in Somerset3 and such activity may have had more of an impact on cage trapping than free shooting. However during the extesions in 2013, the percentage culled by trapping in Gloucestershire was significantly higher than by free shooting. One explanation for this is that protester activity was less in the extension than in the initial 6 weeks thus allowing cage trapping to work more effectively.

Why have contractors failed to cull 70% of the badgers in the pilot cull zones?

The government set a target to cull 70% of badgers as shown in the following extract taken from Section 5.34 on Page 27 of Reference 6.

Culling must remove a minimum number of badgers each year as specified below:
  • In the first year of culling, a minimum number of badgers must be removed through an intensive cull which must be carried out throughout the land to which there is access, over a period of not more than six consecutive weeks. This minimum number should be set at a level that in Natural England's judgement should reduce the estimated badger population of the application area by at least 70%.
  • A minimum number of badgers must also be removed in subsequent years of culling through an intensive cull which must be carried out throughout the land to which there is access. This minimum number should be set at a level that in Natural England's judgement should maintain the badger population at the reduced level achieved through culling in the first year.

This target is largely based on TB reductions which followed the RBCT trial and the estimated overall proportion of badgers which were removed during that trial to achieve that reduction.12

Limiting factors which influenced contractor ability to reach the target are outlined in Table 6.1 on Page 38 of Reference 5. Listed below are a couple of contributing factors which are not referred to in Table 6.1.

Techniques based on DNA analysis were used in the cull pilots to estimate the proportion of badgers removed. Such methods were in their infancy during the RBCT trial and were not used during that trial. As such it may not be safe to compare proportions determined during the RBCT with proportions determined in the pilots. As such a 70% target may be more difficult to reach in the pilots on account of the more advanced methods used to determine that proportion.

It is possible that a large number of contractors hired to shoot badgers in the cull zones were restricted in the hours they were able to shoot due to having to go to work the following morning. Badgers are still very active into the early hours of the morning so badgers which emerge late from their setts are not likely to be seen by contractors who retire early. The RBCT only culled badgers through cage trapping which could be set during the day. Hence cull rates in the RBCT would not have been affected by badgers emerging late. Although a query was sent to the National Farmers Union (NFU) which asked about contractor working arrangements (see Reference 7), no response has been forthcoming. As disclosed in Section 6 of Page 7 in Reference 8, the NFU carried out the cull (through hiring contractors) on behalf of farmers in both pilot zone areas.

How has contractor effort been distributed across the cull zones?

Reference 9 shows a Freedom of Information (FoI) request and the reply received from APHA in reply to a question regarding the distribution of effort across the cull zones. In reply APHA offered no helpful information and as such went against the spirit of responding to FoI requests. This unhelpful response together with the emphasis on lack of consistent contractor effort in the reasons given in Table 6.1 on Page 38 of Reference 5 leaves open to question the uniformity of effort across the cull zones. [Editor note: This comment was made on 5th Feb 2015] In fact in May 2015 Natural England confirmed that in fully accessible, 1km by 1km squares in the Gloucestershire cull zone in 2014 no badgers had been removed from a quarter of the squares.14

Will cull-sample-matching be used in the final year of culling to establish what proportion of badgers have been culled?

As outlined above, it is now possible to estimate initial and hence final populations of badgers in a cull zone more reliably with substantially less dependence on human judgement. In view of this, a revealing question to ask DEFRA would be:

Will cull-sample-matching (CSM) be used in the fourth year to allow badger number reductions to be examined in another year in each cull zone and to allow badger numbers in the first and last years to be compared?

Knowing how badger numbers drop will be important because comparison of the impact of the Thornbury badger eradication/recolonisation exercise and the partial cull performed in the RBCT revealed a substantial difference in the impact of the badger culling exercises.11 The substantial difference in the level of apparent benefit implied that cull thoroughness is key. If the extent to which badger populations were reduced is in doubt, interpretation of the impact of the culling in these pilots will be subject to uncertainties and speculation. As such these pilots will be of less value in planning future policy and may in fact lead to poor policy. Since the mid 1980's, policies in the UK have failed to control TB and, in fact, during this period, long term trends in TB levels have soared as shown in the thumbnail below. Click it to see the graphs at full size.



References

  1. Monitoring the efficacy of badger population reduction by controlled shooting during the first six weeks of the pilots. AHVLA. Report to Defra. 31st January 2014.
  2. Association between Levels of TB in Cattle Herds and Badgers. C.A Donnelly and J Hone. 2010.
  3. Summary of badger control monitoring during 2014. DEFRA. December 2014.
  4. Final Report of the Independent Scientific Group on Cattle TB
  5. Pilot Badger Culls in Somerset and Gloucestershire. Report by the Independent Expert Panel. Chair: Professor Ranald Munro. Presented to DEFRA Secretary of State Owen Paterson MP, March 2014.
  6. The Government's policy on Bovine TB and badger control in England. DEFRA. December 2011.
  7. Badger culls and contractors hired. An email sent to nfu.org.uk on 18th Jan 2015.
  8. Badger Culling: Controlled Shooting Pilots. SN/SC/6837. E Ares. Last updated: 11 January 2015.
  9. Bovine TB: Distribution of effort in badger cull pilots Year 2. Reply from DEFRA to a FoI. ATIC0507. 21 January 2015.
  10. Estimating social group size of Eurasian badgers Meles meles by genotyping remotely plucked single hairs. LJ Thomas et al. Source: Wildlife Biology, 13(2):195-207. Published By: Nordic Board for Wildlife Research. 2007.
  11. Relative impact of the Thornbury and RBCT badger culling exercises.(No longer available)
  12. Setting the minimum and maximum numbers for Year 2 of the badger culls. Advice to Natural England. Supplied by DEFRA. August 2014.
  13. Badger Cull Costs and Culled Badger Numbers. RFI 6456. DEFRA. 1st May 2014.
  14. Distribution of badger removals across the cull zones. FoI response from Natural England. RFI 2911. May 2015.
  15. Summary of badger control monitoring during 2014. DEFRA. December 2015.
  16. Setting the minimum and maximum numbers in Dorset for Year 1 of the badger cull. DEFRA. Advice to Natural England. August 2015.
  17. Summary of badger control monitoring during 2016. DEFRA. December 2016.
  18. Summary of badger control monitoring during 2017. DEFRA. December 2017.
Javascript is disabled