Part 4. Badger culls work or make things worse !?

Historically a number of badger culls have taken place :- Hartland, Thornbury, Steeple Leaze, and the ISG (Independent Scientific Group's) "Krebs" or RBCT Randomised Badger Culling Trial .. in Ireland , assorted pilot culls (O'Connor 1998), East Offaly, Four Areas Trial .. which have been interpreted as proof that badger culls "work" ie. reduce cattle TB. Two others Sussex, Staffs were inconclusive in that regard, as were the Woodchester study area ones : - Beech tree cluster, North Woodchester/Atcombe; as well as perturbation studies in North Nibley and RBCT reactive cull Triplet E , Wilts (see earlier discussion ). Unfortunately, having ruled out cattle as the hidden problem, and also inadvertently by extension the impact of the effectiveness of cattle controls then "logically " culling a few TB badgers must have caused the decreases which were in fact simply due to cattle controls !

Hartland, Devon

1979-84 . As shown in Figure 6, TB was nearly eradicated by cattle controls by the early 1970s, going to 2-yearly testing allowed a flare up in cattle, which was resolved by intensive synchronised testing ( Richards 1972; Krebs 1997).. gassing badgers assumed to have been responsible second time around, but actually irrelevant.

Thornbury, Avon

Clifton-Hadley 1995, Gallagher 2000, Zuckerman 1980 p. 73, ISG p.114-8 . 1975-8, 104 sq.km. 158 herds, 12,000 cattle ; as part of the national decrease Figure 1, cattle controls got TB down to 1 breakdown 1970; a flare up to 16 herds 1973 resolved by cattle controls, so "clear" 1980-91.. in fact 1-3 unconfirmed breakdowns/a, until TB re-introduced after BSE 1993 peak with increased restocking. Sampling of badgers to justify gassing found 31 TB + out of 223 ..interestingly 18.6 % before gassing, only 5.4 % afterwards, ie. less cattle spillover after cattle TB resolved. There was three times the cattle TB in intervention /cull area, as in comparison area, so cattle controls took longer to work ( Zuckerman 1980) .There may have been 4.9 adult badgers /km , with 14.5 % TB prevalence .. so c. 510 badgers gassed, c. 74 with TB .. one sett had to be re-gassed 19 times ..but completely irrelevant in the cattle TB reduction (Cheeseman 1981). Recolonisation took over 10 years, badgers DO immigrate over "Hard boundary" motorways !(Kruuk 1985).

Steeple leaze, Dorset

Little 1982, Wilesmith 1982. 1973-5, 12 sq.km. a cluster of 6 problem herds ; solved by up to 5 tests /a , 244 reactors, (83 VL), so just 1 in 1976-7. A pilot cull found 15 TB+ badgers out of 39 culled , but none out of 41 culled after gassing and cattle problem already resolved, so no spillover from cattle. Gassing entirely irrelevant ; re-colonisation in MAFF reports .TB found in a few sika deer, foxes and rats as spillover from active cattle TB cluster .. but officially "all due to badgers" Figure 6. Only 136 TB badgers in 2126 culls & RTAs 1972-96 for whole of Dorset.. so hardly a self-sustaining reservoir (MAFF report).

Offaly, Ireland

Eves 1999, Mairtin 1998 a, b; Kelly 2008; O'Corry-Crowe 1993, ISG p. 114-8, 1989-94. The "hole =cull area" of 550 sq.km., with donut ring "control" area round it .. resulted in 91 % drop in the cull area, vs. 53 % in the control ring ; but that was simply because the latter was three times the size of the cull core :- so, 1458 reactors from 55,000 cattle vs. 5646 from 150,000 cattle . NB there was also a flare up in control area cattle, so reduction less . Hence 7000 reactors removed reducing cattle TB, versus cull of a mere 141 TB badgers . The Kelly 2008 study extended the data to 16 years.

Four areas trial, Ireland

Griffin 2003, 2005, Olea-Popelka 2006, ISG p.114-8. 1997-2002. Reductions of cattle TB of 51-76 % in the "control" ie. cull areas contrasted with comparable reference areas. However, there were rises due to stopping pre-movement testing in 1996, and the variable drops in cattle TB reflected differences in herd size, dairying/beef, previous TB history .. and there was twice the background TB in reference vs cull areas : 393/193 confirmed breakdowns. Badger data given separate from cattle paper, and unfortunately culls included buffer zone badgers, clearly not relevant to cull /reference cattle ; so extrapolating actual numbers .. just 286 TB badgers from 960 sq.km., Cork 115, Donegal 37, Kilkenny 59, Monaghan 75 .. the cull of a mere 37 TB badgers from 215 sq.km. in Donegal especially irrelevant, so the much lauded alleged "drop" of 96 % "due to the cull" WAS in fact, due to a flare- up in contiguous herds in the reference area (Olea-Popelka 2006).

Sussex

Wilesmith 1986 & Pritchard 1986 . 1981-4, 3 contiguous herds had 19 reactors . But the cull of 8 clans produced just 10 TB badgers out of 47 culled, since TB virtually eradicated in the rest of the county ; so the study found very few TB badgers elsewhere and no spillover to "other wildlife" sampled either. And 1972-96 just 68 TB+ /1319 culls & RTAs for whole county (MAFF reports); and a unique Irish spoligotype dating back to 1970s cattle imports (Krebs 1997 maps, Wilesmith 1983).

Staffordshire

Cheeseman 1985, Hewson 1987. 1980-4. Ten problem herds in this last non-southwest area to go "clear" under Area Eradication : Ipstones, Bradnop, and Cheddleton parishes (MAFF reports). The cluster of farms FGH had 36 reactors 1980-4, so in the study area , the 5 initial clans produced just 8 TB badgers out of 45 culled. A later breakdown farm B, extended the cull area resulting in a total of 16 TB badgers out of 145 culled .. with few RTAs in the 1972-96 total .... only 26 out of 604 culls and RTAs for for whole county so hardly a persistent self-maintaining reservoir (MAFF report , and see Figure 7 & Table with RBCT data, Triplet G, Staffs/Derbyshire produced 113 TB badgers out of 1252 culled ... after cattle TB reintroduced via Mad Cow replacements (Section 3).

The RBCT , Randomised Badger Culling Trial; 1998-2005 culls, plus 6 years after effects ..... see figure 7 ,with 2 tables reactive/proactive badgers culled.... ISG 2007, Final report; Donnelly 2007 IJID (And Suppl.), 2011; Jenkins 2010, & 2008 IJID; Morrison 2000; Viale 2011; Woodroffe 2005 a /2009 & 2005 b/, 2006 PNAS; Carter 2007; Macdonald 2006; Consultations DEFRA 2005/2010; see also EFRA Reports 2006, 2008; Given the avalanche/tsunami of highly complex statistics from the RBCT, few published Criticisms :- DEFRA : 2004 Godfray /2007 King .. More 2007, Hancox papers.

It is important to reappraise this study since it is the basis for the proposed 2 pilot culls arising from the second cull Consultation (DEFRA 2010).... the 150 sq.km. cull would it is claimed achieve a 12-16 % reduction in cattle TB , AFTER 9 years ... which actually is a mere 46 out of 292 expected breakdowns (Annex B, para. 20) .. the ISG 2007 figure balancing beneficial/detrimental effects was 22 breakdowns prevented out of 187 expected after 7 1/2 years, and the Welsh Intensive Action Area back then was 44 out of 374 breakdowns in c. 300 sq.km (Jenkins 2010) . BUT, it is very difficult to see how anyone can regard this as "A Meaningfull " contribution to the problem, given that there are some 7000 breakdowns a year in a hotspot area now encompassing half of GB !?? The CSL/FERA badger vaccine model admitted that this would do half as well as culls, and a 2 km. wide vaccination ring is hilariously absurd as an obstacle to outwards cattle movements . The Krebs report 1997, recommended this study, with culls encompassing (at that time), most of the alleged badger TB hotspots having most new and repeat breakdowns .The 10 triplets would contrast reactive and widespread proactive culls with survey only = no cull areas.. hence 30 trial areas altogether; with importantly a 2 km wide "badger- movement-Outside" area as a "control" of inside cull area effects ... and proof as to whether either cull strategy would help within some 5 years .. derailed by foot and mouth, 2001, final results nearer 10 years on, cost £50 million. Each of the 30 trial areas being 100sq.km, the outside rings 97 sq.km.

Krebs warned very clearly from the outset , that "it is not known if, how, or to what extent badgers contribute to cattle TB" .. so rather disappointing ( 3 members of isg, were part of Krebs), that the RBCT started with 3 key assumptions :- 1. that badgers do give cows TB (although they are not quite sure how, see Plausibility below), 2. and do constitute a major reservoir of TB, 3. so culls would have an effect..both a drop and an increase, since indeed, it was widely believed that the failure of decades of culls to reduce cattle TB or prevent the explosion since mid-1980s was due to badgers "perturbed" by culling , to spread TB more widely to cattle. Gassing allegedly spread TB in Wilts (Overend 1980), the N.Woodchester 1980s study showed loss of territoriality, and wider home ranges , and marked TB badgers were culled outside study area too (Cheeseman 1988) . This "Perturbation Problem" idea has taken hold as the new "religion", and entirely underpins the proposed cull rationale.... culls might work if some 70 % of badgers culled, the edge effect minimised by a big enough area, hard boundaries to stop migration, synchronised culls, over at least 4 years . Alas, as S. Holmes pointed out " it is a capital mistake to theorise in advance of the data, one tends insensibly to wish facts to suit the theory rather than theory to suit the facts ".... Irish culls, did find a breakdown in territoriality and wider home ranges eg. Offaly/kilmurry (o'corry-crowe 1993, sleeman 1993, 2005, more 2007), but found no perturbation effect ! it is very hard to see why the badger cull, with immigration into the vacuum area would cause a rise in outside area cattle ? ; and NB. there was a drop in the critical first 500 m of this outside ring anyway isg p. 99.

Rather importantly, the ISGs own data actually showed that :-

Reactive culls had nil effect :- 1 year after culls ended hastily in 2003, because they seemed to be making things worse , as compared to no cull/survey only areas , culling a mere 311 TB badgers from 900 sq.km. (Figure 7, Table) had no effect :- there were 356 /358 confirmed breakdowns accumulated; 175/172 unconfirmed breakdowns; and 56/59 repeat breakdowns (ISG p.249 Lefevre's update) ..AND PROACTIVE culls HAD NIL EFFECT on Unconfirmed breakdowns ISG p .96, 101. The ISG's own data p. 103, interpreted the study cull period alone over 10 triplets, for confirmed breakdowns, gave a reduction of just 116 breakdowns, and a perturbation rise outside of 102 , so a net difference of just 14 fewer breakdowns cull + outside area (1000+970 sq.km.) ...or JUST 1.4 breakdowns / 100 SQ.KM.... which is within the range of a chance random effect ! Conclusions :- Assuming the perturbation effect ..Confirmed breakdowns during the study (ISG 2007) :- reactive >27 % (-2.4 to 65, non-significant); proactive inside < 23 % (1 to 23)/outside >24.5 % (-0.6 to56, non-significant ) ......by August 2011 :-proactive : inside < 25.7 ( - 19 to 32)/ outside > 7.6 % (-14.2 to 35.1 non-significant).

What is "Perturbation" ? .

The RBCT study did :--
a. reaffirm the fact that culls disrupt territories, so larger home ranges INside cull areas;
b. show doubled prevalence in badger TB especially towards the outer part of proactive cull areas; BUT,
c. the assumption that this was due to greater badger-to-badger spread, then causing greater badger-to-cattle contact and a rise in cattle TB is dubious , compared to admitting there was greater cattle-to-badger spillover ! (Woodroffe 2006). In the proactive outside areas home ranges were undisturbed, so perturbation absent there, cannot have caused the rise.. although doubtless badger prevalence by spillover occurred there too ? The FMD cattle TB rise (Section 3), happened in both reactive cull & no cull areas of Triplet E, Wilts ( SE 3108; Macdonald 2006).

Perturbation :- Biological plausibility . Disentangling the truth amidst the somewhat selective interpretation of the tsunami/avalanche of "incomprehensible" statistics is daunting ....SO, where does the Alleged "perturbation " increase come from ? There are a number of areas where the "Biological Plausibility" of the alleged cull effects are very dubious :-

Too few TB badgers.

See Figure 7, and TABLES (ISG 50, 74-5, 205-9). reactive culls only 311 TB badgers or c. 1 / 3 sq.km. ; 76 culls reacting to 169 herd breakdowns (only half end total); 6 of the 9 areas after FMD, 2001, so two thirds culled in last 5 months, and the FMD delay meant a lag of up to 2 years since the initiating incident so not the same badgers supposedly responsible. Strikingly, Figure 7/Table .. in 7 of the 9 cull areas, 31 or fewer TB badgers / 100 sq.km. can hardly have made the slightest difference to cattle TB ! The Report SE 3108, on Triplet E, Wilts. found far too few TB badgers, too few badger movements between clans, and they were not the same DNA/Spoligotype as the imported cattle which caused the rise !

Proactive 1204 TB badgers culled or c. 1 / sq.km. 51 culls, but nearly half had 15 or fewer badgers / 100 sq.km. ( see Table). SO, just 1515 TB badgers out of 11,000 culled from 1900 sq.km. total. And since the start population was actually unknown, the 70 % cull estimated would from 1204, leave c. 400 missed by the few culls/year SO Why did'nt these escapees cause a perturbation upswing as supposedly in reactive areas .. still badger activity in areas despite culls. From personal knowledge of the Glos. /Exmoor areas, it seems to me the Triplet A proactive cull of just 362 badgers was a tiny fraction of the population, 20 adults/sq.km. would be 2000 badgers; but the Triplet H cull of 593 from badger- poor -quality moorland probably was a truer figure. A member of the Wildlife cull team, Paul Caruana, was very critical of cull efficacy in EFRA report 2006. Dartmoor long regarded as a "due to badgers" hotspot (Figure 6) but like Exmoor fewer badgers, the real problem is shared common grazing, without testing the cattle population as one unit. Rather ironically, the 2 Proposed Pilot "shooting free running badgers culls" were called off at the 11th hour in Autumn 2012, largely owing to a doubling of the guestimated badger population numbers !

Too few infectious badgers. Just 166 multi-lesion excretors out of 9919, p. 77; Jenkins 2008.. hence ISGs suggestion NVL badgers must be infectious : Wrong Woodchester clinical sampling found infectious sputum, faeces, urine from very few VL badgers and Gallagher's studies too found excretors mainly in badgers with generalised florid TB ( 1998, 2000).

Transmission

The ISG, seem rather confused as to transmission amongst both badgers and cattle p. 121,151, 172-173. Their understanding of cattle TB is influenced by experimental infection of calves .. a nasal innoculation unsurprisingly renders them "sputum positive" nebulisers /infectious early on, then declining ( Neill, Cassidy, McCorry studies) .....but they dont recognise that infectious cases via droplet aerosols in exhaled lung breath comes late on from VL, visible lesion cattle , active spreader cases ( Morrison 2000, Woodroffe 2005a, p. 861). Also, high dose infection may make reactors within 3 weeks, but natural contact spread with reactors takes 7 mnths for c. 1/2 to react to IFN p.234-5.. 6-12 months in Svensson study (M'Faydean 1910), So claiming a " badger cull to reactor effect" within 2-3 months is unlikely (Woodroffe 2005 b); often cited Francis 1958 , reactor by 8-51 days....BUT in truth it probably takes 12-18 months for most cattle: to catch TB, then become reactors , and have their next annual test (Godfray, King in DEFRA reports ).. so there was not enough time for the cull to show up in initial and later reactive culls; nor in initial proactive cull SEE Graph based on Consultation Figure 7; point A in fact spans 1998-2002; 3 areas started after FMD. ISG's own data p. 109, show the 23 % (non-significant) reactive rise happened BEFORE the cull anyway ! cattle to badger transmission, rather belatedly, the isg realised that " tb cattle pose a risk to badgers " (Woodroffe 2006).... which was most clearly shown by ISGs data on initial cull numbers :-

Table for first proactive culls

Triplet Reactor cattle in previous year Initial cull TB badgers Initial cull TB clans TB badgers per clan (Woodroffe 2005 a & 2009)
A 57 8 7 1.1
B 70 13 7 1.0
C 62 4 4 1.0
D 187 102 44 2.2
E 34 29 17 1.3
F 14 13 7 1.1
G 23 29 13 1.7
H 36 12 8 1.0
I 154 82 29 2.0
J 215 65 32 1.6
Totals 852 357 168 1.7

Thus, the 3 triplets recruited after the FMD jump shown in 2002, before intensive annual testing introduced ( to equalise cull effects).. D I J Had the most reactors with spillover to the most TB badgers, most clans, and most TB + /clan. The ISG, hence noted a doubling in badger prevalence, including reactive areas AND unperturbed "unculled" RTAs, p. 73-8, 84, 243; and the Welsh RTA study found that Powys previously with no TB badgers now ; as cattle TB arrived had up to 25 % infected... the badger TB hotspot similarly "migrated" from traditional Devon/ Cornwall to Glos. & Hereford/Worcs. Unexpectedly, higher prevalence in reactive cull badgers closely linked to breakdown, than in wider proactive culls (Tables in Figure 7). In addition to this greater spillover from cattle, there were more badgers culled later on in the outer part of the cull area 72 sq.km., than from the initial core area culls from 28 sq.km (SEE below, Serial culls/cattle controls in core/outer/outside rings)... although some were also immigrants from the Outside area .. both due to Immigrating INto a vacuum, and on DNA evidence, p. 69, 90. Badgers not Emigrating into outside area due to territoriality still intact, home ranges undisturbed , Woodroffe 2005 b :- a key flaw, in the whole perturbation argument , why would badgers immigrating into cull area cause an imaginary "perturbation " rise in cattle tb in the outside area !??

Attribution

It is surprising the ISG accepted the supposed " proof" culls work from the Thornbury, Offaly, Four Areas studies (dubious as above) ; a drop in latter of 51-76 % "due to the culls", p. 80-3, 114-8. Curiously, they hence interpret the clear spillover data in D I J in Table above, by inverted logic as TB badgers causing between 16 and 71 % of breakdowns... in fact fewest in F, the infamous West Penwith / Lands End , "problem area" , with decades of ineffective badger culls (wonder any Brocks left !) , which more significantly was the longest on intensive annual testing (Figure 6).

Statistics

Curious that data are given for Vetnet and ISGs own database, these are supposedly the same cattle breakdowns, but an unexplained difference of 10-20 % is the actual range of supposed cull effects. A log linear Poisson regression analysis , may not be ideal, since breakdowns are not random, but associated with previous single or cluster of TB herd breakdowns. And, whilst it may be rigorous statistical methodology, finding for eg. a "mean" minus 46.0 % "drop", ( 95 % Confidence Limits - 85.7 to 103.6 ie. range of 189.3 % across 10 triplets ) is of rather dubious value to the farmer with mud on his wellies !. It was "logical " to present cull trial data in terms of serial proactive culls 1 to up to 7, BUT 1st cull start ranged from 1998 to 2002, hence the point A jump in outside area was an average over 10 triplets, heavily skewed by the last 3 recruited after FMD ( D I J as above) .. the actually meaningful annual cattle control tests can be "super-imposed" on the Consultation Graph Fig.7.

Effects

Are the claimed Beneficial and Detrimental effects actually "due to " culled badgers/perturbed badgers " :- decrease inside the proactive cull area, and increase in reactive areas plus outside ring of proactive areas (Graph in Figure 7) ? It is worth remembering that cattle controls , ie. Annual testing reduces within herd spread and hence number of reactors , so hard to see how the Offaly badger cull supposedly did this (above); whilst movement restriction of breakdown herds stops them from exporting new cases to further herds. How is a proactive cull of 1204 TB badgers, from 1000 sq.km. ie. c. 1.2 /sq.km. realistically going to prevent the many new breakdowns supposedly "due to badgers" ? So re-examining :-

Beneficial / Reduction in cattle TB. The drop of 60 % by B in Figure 7 , Graph noted by Jenkins 2008, is actually due to the cattle controls. And cattle controls were more effective in the slightly smaller outside area versus inside area , both from A to C 70 vs 49 %; and from C to E 42 vs 4 % Aug. 2011 = 6 years since culls ended 2005; Very unlikely after effects so long after culling so few TB badgers. The ISG were puzzled , having "NO Clear explanation for the reappearance of beneficial effects " after the 2008 cattle TB jump to 40,000 reactors Figure 1 in Cattle Section 3). Cattle controls :- annual testing + movement restrictions Work ; hence equalising inside/outside levels by D in Graph , 2009. In addition, Jenkins 2008 , Suppl. noted that by January 2008, confirmed breakdowns in triplets had achieved 0 or just a single breakdown in cull / no cull ie. survey only areas :- Inside , proactive /survey only ..............outside , proactive / survey only
to 0.................4 / 2 ............................................... 6 / 4 .................... = 16
to 1.................3 / 1 ............................................... 4 / 3..................... = 11
detrimental /increased cattle tb due to perturbed badgers ( Reactive and Outside proactive areas ). Alas, the ISG , seem to have not fully appreciated the impact of the lack of testing during FMD jump in 2002 ; nor the 2008 tighter cattle controls jump to 40,000 reactors Figure 1 ( big increase in parishes brought back to annual testing, plus pre-movement testing), as outlined in Section 3 Historical overview.

(( " footnote " the variable efficiency of cattle controls are the real reason behind supposed badger cull effects :- IT is important to understand that the numbers of new breakdowns in an area are critically determined mainly BY 3 KEY factors, number of baseline herds in the area, previous breakdowns in 1 or 3 year period , and efficacy of cattle controls/annual testing .... so given the wide variation in these bases, and start culls over 5 years , 1998- 2002; it was not in the least surprising the variation in the effectiveness of cattle controls in preventing new breakdowns . AND SO, whilst it reveals the highly complex difficulty in interpreting the cattle statistics, for the sake of completeness it is worth highlighting the following data ; ISG p.88, 94, 97, 100. THUS, comparing the proactive cull/no cull (survey) areas; both inside/ outside as regards baseline herds // 3 year previous breakdowns // study period breakdowns ...there were slightly more survey area herds inside & fewer outside .. so more/more background TB.. but more/less during study inversely ; still clustered inside proactive areas , but less so in reactive /outside proactives; BUT the few dozen breakdowns differences are well within expected controls efficacy :-

Baseline herds.......... // 3 year previous breakdowns..................// breakdowns during study (to 2005)
INSIDE AREA..proactive 1221 /survey only 1276...// confirmed & unconfirmed :- pro./surv. previous 345, 363 ...//during study 692, 833
..............................................................................// confirmed :-pro./surv. 257, 253... //472, 582
..............................................................................//unconfirmed :-pro./surv. 88, 110... // 220, 251
OUTSIDE AREA ..proactive 961 / survey only 923 ..//confirmed & unconfirmed :- pro./surv. 172, 205...// 514, 477
.............................................................................//confirmed :- pro./surv. 115, 154...// 362, 321
.............................................................................//unconfirmed :-pro./surv. 57, 51 ..// 152, 156;
A similar conclusion was attained by Viale 2011 who found from initial proactive cull to 1 year after culls ended 2005 : --1154 confirmed breakdowns had been accumulated , ie. ; 1306 baseline proactive herds gave 343 breakdowns / 1320 reactive herds gave 403 breakdowns / and 1380 survey only = no cull herds gave 408 breakdowns ... again unexpected given controls efficacy. and, there was considerable variation in background TB within the three treatment areas per triplet for the 3 Year previous ; For confirmed breakdowns figures for each triplet : survey only, reactive, proactive cull... ranged from just 9 to up to 54 ; and .. with the total no. per triplet, 43 to up to 130 .....Thus :--- A Glos. TOTAL 130 =37, 54, 39 /.. B Devon/Cornwall 120 = 35, 37,48 /.. C E. Cornwall 51 = 16, 21, 14 /.. D Heref. 55 = 27, 9, 19 /.. E Wilts.56 = 14, 18, 24 / ..F W.Cornwall 62 = 25, 17, 20 /.. G Staffs/Derby. 51 = 14, 18, 19 / ..H Devon/Somerset 43 = 17, 11, 15/.. I Glos. 74 = 20, 31, 23 /.. J Devon 55 = 15, 21, 19. These cattle control effects were also reflected in the pattern of breakdowns ; still clustered breakdowns inside proactive areas, but introduced cattle = less clustering in reactive & outside proactive areas .... the "badger/cattle" clustering effect with odds ratio probabilities in reactive areas ISG p. 109-12, were supposedly linked to culls at 1, 3, 5 km distance ; but badgers dont move 3 /5 km so were actually correlated with the breakdown which triggered the cull .. likewise later correlations (Viale 2011). )).

Considering these Supposed perturbation increases in more detail , by areas separately :-

Reactive Not enough time for cull effect to show as reactor cattle either for the initial rise, nor the end increase.. two thirds of badgers culled in last 5 months, so only A B C ran long enough to show any effect. Indeed, the 23 % rise was non-significant, and happened Before the cull anyway p. 109, so a bit daft to say it was due to the cull. Rise in J, not culled, and I only ran 1 week before reactive culls ended. Nov. 2003. Triplet E Wilts, had a FMD rise in both cull/no cull areas (Macdonald 2006), and the SE 3108, Report showed increase via imported cattle, which were not same spoligotype as local badgers .

Proactive . The increase in cattle TB , Supposedly due to perturbed badgers arises from tendentious selective interpretation of 4 cattle lines of evidence , plus 1 badgery one : -

1. Cattle effect.

The ISG focus on the " first culls" start jump, but it is an average over 10 areas starting 1998-2002, heavily skewed by DIJ recruited AFTER FMD jump (see above Point A in Graph in Figure 7 & Table). But, there was an initial transient rise both inside/outside, for both proactive and NO cull areas , Tables and accompanying text p. 88, 94, 97, 100 .. including a FMD blip inside (Letters ISG to Minister, 2005 Sep./Oct.) . Also, a rise beyond the 2 km wide outside ring, some areas on 2-4 year testing (IJID 2007 Suppl.). The apparent outside rise was non-significant after the start point A on graph (43 % up).. ie. cattle controls began to bite, so the alleged perturbation effect disappeared after 18 months (although badger culls/perturbation continued). The ISG also fail to explain why there was Actually a DROP in the first half km. outside the boundary of trial area which was based on badger territories, whilst actual herds overlap , so drop part of inside reduction, p. 99. The ISG also do not explain the 2008 jump in Graph, point C, rise to 37 % nearly as bad as start-up (43 %); absolutely nothing to do with badgers (see Figure 1, 40, 000 reactors in Historical overview in Section 3).

2.Serial culls/distance from boundary edge .

Important to realise each triplet study area comprised 3 concentric rings 2 km wide.. resulting in a "core " cull area 28 sq.km , an "outer" cull area of 72 sq.km// then the "outside" no cull "control" ring of 97 sq.km. NOT surprising that the initial cattle control/annual test "seemed " to have the biggest effect in core area, later effects more in the outer/outside areas twice/threefold size ...SO, for confirmed breakdowns (IJID 2007 suppl.), the proactive/survey areas had core area :- 107/ 132 with confirmed breakdowns ; outer area 291/346; outside area 307/ 263 ... and serial badger culls likewise found more in outer area = bigger population , so a pseudo-edge effect ! Exactly why the Offaly cull supposedly worked ! (see above).

3. Hard

Hard impermeable boundary p. 90. Much is made of a sea boundary, Lands End but obviously badgers dont live at sea ..the biggest "effect of culls" was actually because F had been the longest on annual testing , an intractable hotspot since the 1970s, Figure 6. Anglesey too bounded by sea, but although badgers migrated in over the bridge, new breakdowns introduced in a different area (see Cattle Section 3 Maps 1 & 3). With hindsight , lessons from Thornbury with 2 motorways & Bristol channel boundary , and immigration in Offaly; it was suggested major roads/rivers would stop migrating badgers. Hence choice of cull/control areas for Four areas & RBCT , but the motorway/river hard boundary is likewise dubious .. badgers do cross motorways , RTAs .. and even swim over rivers Bristol study & Suffolk release. It took over 10 years for the cleared Thornbury area to regain its former population , but recolonisation must have been across two M-ways (Kruuk 1985).So the hard boundary , effect c. 12 % of study area is greatly over-emphasised, and in any case badgers were Not emigrating into outer ring so couldnt have caused the cattle TB rise there !

4 . Synchronised culls

The biggest FMD jump nearest to 2002 culls as seen in partial culls in triplets B C E, p. 48.

5. Badgers

As noted above under Transmission , and Serial culls above; more badgers culled from outer ring in later culls, and more with TB since after the FMD jump spillover and resultant doubling in prevalence.. Wider home ranges in proactive and reactive areas (Woodroffe 2005 b), but more new scattered breakdowns , so spillover less clustered ; Some immigrants too based on DNA evidence;

So, these various elements SEEM to indicate an Edge effect, with badgers perturbed by the cull near boundary; and ISG chose to interpret higher TB in badgers there as due to greater contact =perturbation (instead of simply more spillover from cows) .. SO ASSUME more TB transmission and resultant rise in cattle ( instead of simply the efficacy of cattle controls ). BASED on these somewhat selective interpretations of the data, they hence interpret PERTURBATION as the reason why REACTIVE culls wont work, patchy, localised, asynchronous, short term ; and PROACTIVE culls might work if the edge effect overcome by a big enough area, with hard boundaries, synchronised culls of c. 70 % of badgers for at least 5 years .. so 9 years adding in the amazingly long lasting supposed post-cull effects of a mere 1515 TB badgers from 1900 sq.km. .. a tragedy that more culls "justified" based on such flimsy evidence.