Dominic Woodfield is the Managing Director of Bioscan, a long established and well-respected consultancy specialising in applied ecology.
He is a life-long birder, a specialist in botany, habitat restoration and creation and in protected fauna including bats, herpetofauna and other species. He is also a highly experienced practitioner in Environmental Impact Assessment and Habitats Regulations Assessment. Most of his work is for the development sector, but he has also undertaken commissions for Natural England, the RSPB, Wildlife Trusts and campaign groups. He once mounted an independent legal challenge in defence of an important site for butterflies in Bicester, Oxfordshire, which resulted in planning permission for a five-hundred unit housing development being overturned. He lives in Oxford with his partner and family.
Note: this guest blog was first written in July 2021, shortly after Natural England had been forced to finally release the long-suppressed BTO report mentioned in the title. The blog was not published at the time to avoid any interference with a contemporaneous High Court case. The judgment in that case is shortly to be re-examined in the Court of Appeal, but as the key facts of the case referred to here have been broadly established, there is no longer the same sensitivity surrounding publication. The issues this blog raises are as pertinent now as they were just under a year ago.
How a suppressed BTO report became published as a ‘joint’ BTO/NE study.
In February 2021 a paper was published in the journal Bird Study, entitled “A comparison of breeding bird populations inside and outside of European Badger (Meles meles) control areas”. The paper had four authors: two ornithologists from the British Trust for Ornithology (BTO) (E. Kettel and G. Siriwardena) and two employees of Natural England (NE), both of whom happen to have been closely involved with the licensing of badger culling (I. Lakin and M. Heydon). The paper reported on a contract let by NE to the BTO to analyse geographically delimited subsets of data from the records collected annually by volunteer birders partaking in the national Breeding Bird Survey (BBS). The purpose of the analysis was to attempt to detect any perceivable effects on bird populations from badger culling.
One of the risks to ecosystems subjected to badger culling is a trophic cascade phenomenon known as ‘predator release’ or ‘carnivore release effect’. This is where removal of an apex predator species creates a vacuum which other predator species may expand in population to exploit.
A particular concern in the case of badger culling is that populations of predator or omnivore species such as foxes and hedgehogs, that are more efficient at predating the nests, young or adults of ground nesting birds, may expand in cull areas as badgers are removed. Given that culling aims to reduce badger numbers by 70-95% in cull areas, these local population changes and expansions have the potential to be substantial. These additional predators may then exert additional lethal and non-lethal pressures on vulnerable bird species in cull areas, an impact potentially of conservation importance if affecting precarious populations of rare species (e.g. stone curlew) or declining species of conservation concern (e.g. lapwing). While badgers also opportunistically take birds and eggs, foxes in particular are widely accepted to exert proportionately greater pressure – especially for certain groups of species such as ground nesting waders. There is a reason why the Wildfowl and Wetlands Trust and other conservation organisations such as the RSPB spend large sums of money annually on excluding foxes from reserves by fencing and/or on lethal fox control.
Concerns about the potential negative consequences from this ‘predator release’ side-effect of badger culling were raised by Defra in 2007, drawing on data collected during the Randomised Badger Culling Trial (RBCT) (Defra 2007). They were echoed again in a desk-study conducted by the Food and Environment Research Agency (FERA) in 2011 (FERA 2011). Culling started in 2013, but over the next five years to 2018 (when tens of thousands of badgers were culled over large parts of western England), and notwithstanding the findings and recommendations of Defra 2007 and FERA 2011, NE and Defra did precisely nothing to study or monitor such effects. In fact, it took the sharp scrutiny of legal challenges brought by Tom Langton in 2018 and the accompanying witness evidence of myself and others, to force NE into improving the robustness of its impact assessment processes, firstly in respect of European Sites and latterly in respect of SSSIs. In an effort to escape adverse judgment, Natural England promised a High Court judge in 2018 that they would look at the issue of risks from predator release more closely.
The BTO were contracted to analyse relevant BBS datasets and they reported that same year. Three years later, Kettel et al. 2021 purports to present the results of that 2018 study. The truth, however, is that Kettel et al. 2021 is the outcome of a suspiciously protracted process of ‘peer review’ during which the original 2018 BTO report to Natural England was withheld from public view, despite multiple requests for it to be released. Over the course of this period, many decisions on whether to grant culling licences were made by Natural England and a national policy decision on the continuation of the badger cull was made by the Secretary of State. The original BTO 2018 report was consistently cited as part of the evidence base informing such decisions. Yet the public were denied access to it.
Mark commented on Kettel et al. 2021 here. He remarked that it didn’t seem a particularly robust or conclusive study and didn’t really answer any of the questions or concerns about the potential collateral ecological effects on birds from removing large numbers of an apex predator from the ecosystem. He also observed that it was not the 2018 contract report sent to NE by the BTO and on which licensing and policy decisions around badger culling and advice to Defra and Ministers from Natural England up to 2021 was based. Tom Langton (see here) and myself had repeatedly asked NE and the BTO to see a copy of the original 2018 report and in 2020 Mark joined the clamour (see here, here, here and here). We were told that despite it being cited with no mention of ‘draft’ and despite it being used to make policy and licensing decisions, it was a draft scientific paper that was being prepared for publication and that it was undergoing a process of peer review, and on that basis would not be released.
BTO’s former Chief Executive Andy Clements effectively said “out of our hands – NE’s the client and they decide” (notwithstanding that he had already joined the NE board by that time). Mark commented that this secret use of volunteer data was not a very edifying position for BTO to have got itself into. However, I can report that the original BTO 2018 report has now, three years after it was issued to NE, been dragged into the light.
Once again, it has taken legal pressure to force action. The release is a consequence of a ‘discovery application’ made directly to the High Court in connection with a current legal challenge by Tom Langton to the Secretary of State’s (SOS) decision to adopt a new badger culling policy. ‘Next steps for the strategy for achieving bovine tuberculosis free status for England’ (‘Next Steps’) sets out the Government’s policy on bTB control going forward from 2020 . BTO 2018 represents relevant evidence in that challenge, as it is the only study conducted by NE or the SOS/Defra into the collateral ecological effects of badger culling since 2012 and the only study available at the time ‘Next Steps’ was adopted. The SOS claims that its mere existence was evidence that the Minister considered wider effects on biodiversity in making his decision in 2020 and in accordance with his statutory duty under section 40 of the Natural Environment and Rural Communities Act 2006. NE and the SOS maintained that putting BTO 2018 in front of the court was not necessary as there was no material difference between it and Kettel et al. 2021 – despite the latter post-dating the impugned decision by around a year. However, in July 2021 two days before the hearing, they rescinded.
It had taken three years and countless requests, rebuffs, exchanges of correspondence and argument to get this report into the open. So, was it worth all the anguish and argument to get this report? Was the Secretary of State and Natural England right to say that BTO 2018 is not materially different from Kettel et al. 2021 and that it delivers the same conclusions? Well, on first pass it might be difficult to see much between them, other than noting that the original two BTO scientists who authored the 2018 report to Natural England have been joined by two new authors from Natural England (both of whom happen to have provided evidence to the High Court and Court of Appeal in support of Natural England’s approach to assessing the impacts of badger culling when issuing culling licences). However, once one drills down into the detail, one unearths a change of approach between the two papers that has arguably very significant implications for the scope, tone and meter of the discussions and conclusions and for any decision making based upon them.
To be clear, neither version of the study is an exemplar of robust science and neither suggests that decisions on assessing and mitigating the collateral risks of badger culling are being made on the basis of any solid evidence, let alone empirical, real-world evidence. Frankly, Kettel et al. 2021 is a rather embarrassing thing for a journal of the calibre of Bird Study to have within its pages. But beyond that is something much more interesting: the three years of ‘peer review’ have transformed the original BTO report to NE of 2018 into something that paints a somewhat more positive picture and, conveniently, better fits NE’s chosen narrative on this issue. Mainly this has been achieved by changing the methodology of the analysis.
In contrast to BTO 2018, the 2021 paper adopts an approach to assessing BBS datasets that clearly deviates from the scientific standard (Harris et al. 2018). This may be an attempt to broaden the scope of the study and overcome the crudeness of the base data and lack of data points, but its effect is to further reduce statistical power from a study already extremely underpowered. If one is being cynical, it is possible to read this shift in methodological approach as an attempt to massage the analysis to conceal some of the more unwelcome trends hinted at in the 2018 version. In any event, the result is the same: a different subset of species data gets taken forward for trend analysis in 2021 than in 2018, in the course of which negative and near-significant trends indicated for certain ground nesting waders in cull zones in 2018, disappear. Specifically, in Kettel et al. 2021, data from 15 BBS squares is used as a minimum threshold for analysis, rather than the normal accepted standard for BBS of 30 (which was applied in 2018 and follows Harris et al. 2018). In consequence, the crucial ‘ground nesting birds’ subset in 2021 excludes species such as oystercatcher (which showed a near significant negative trend in the 2018 study) and grey partridge (which did not). These instead become replaced instead by smaller passerine species, many of which nest close to but actually often above ground (e.g. yellowhammer, willow warbler) and some of which exhibited positive trends. Readers of this blog are not likely to need it pointing out that larger ground nesting waders are probably going to be of more relevance in a study of this nature as they are at inherently higher risk from predation or sub-lethal predator release effects than smaller songbirds and passerines. This is acknowledged by NE.
The consequence of changing the methodology and deriving a different subset for trend analysis is that Kettel et al. 2021 comprises fourteen species of, on average, lower inherent risk from carnivore release than the fourteen species making up the ‘ground nesters’ subset in BTO 2018. It is for this reason that BTO 2018 paints a somewhat less ambiguous and somewhat more interesting or even concerning picture than the public-facing Kettel et al. 2021. Obviously both studies are unable to disaggregate the potential causal factors underlying the observed non-statistically significant trends (i.e. distinguishing badger culling from other covariables), but there is enough in BTO 2018 (and, crucially, more than in Kettel et al. 2021) to suggest that a continued highly precautionary approach to assessing the impact of predator release on ground nesting waders and waterfowl is merited. That of course was and remains common sense. It is also consistent with the findings of Defra 2007, FERA 2011 and indeed the Godfray Review of 2018. But it doesn’t sit very well with the position of NE as set out in internal guidance for processing badger cull licences and the opinion of the two Natural England authors, who are largely dismissive of this risk (despite the contradiction of their being authors of the guidance that requires that the same risk must be assessed in respect of European Sites and SSSI).
One very pointed question arising out of all this is where does this leave the assessments of possible impacts on ground nesting waders and waterfowl occurring outside designated sites, for example lapwing and curlew and the many others for which the Kettel studies simply had insufficient or no data (e.g. redshank, snipe)? NE acknowledge that predator release is a threat that needs assessing and, in some instances, mitigating in respect of these species, but only in respect of protected sites. BTO 2018 supports the case that concern should be exercised outside protected sites, even if Kettel et al. 2021 is rather more equivocal on the matter. In point of fact, there are no such assessments. NE consider that there is no threat to such species outside designated sites. Perhaps others might consider the differences between the original BTO 2018 report and the public-facing Kettel et al. 2021 revision to be trivial.
At the end of the day, neither version of the study is robust, neither identifies clear, statistically significant effects but neither identifies a clear absence of effects (the part of that Ying Yang that NE seem keen to overlook) and consequently neither delivers any reassurance that possibly only small magnitude (but potentially highly significant in conservation terms) impacts, worthy of ours and Ministers’ attention, aren’t occurring in badger cull areas as a consequence of increased predation pressure on populations of vulnerable and/or scarce ground nesting species such as redshank, snipe and so on.
The study NE elected to do in relation to its promise to look more closely at this issue is tantamount to taking a count of the number of red cars on the way to work and then seeking to claim that local traffic volumes are going down from those data. As Mark said back in February 2021, “I’m not sure I would pile any more money into doing more of this [sort of study] on the basis of the results so far.”. I agree. As soon as I deduced back in 2018 what NE were up to with this methodology, I could see it was little more than a feint at investigative science, an excuse not to do something more meaningful and a means to kick awkward and difficult to answer questions into the long grass. If Defra and NE seriously want to demonstrate that this acknowledged risk is no risk at all, they need to actually design a proper monitoring experiment involving targeted collection of relevant, real-world data, not borrow coarse-grained datasets from somewhere else, apply meaningless analyses to them as a surrogate, and then tinker with those analyses to try and elicit a better answer. It is possibly for reasons of cost that NE have gone for cheap and cheerful analyses of already available data. But it may also be a more worrying thing: prejudicial dismissal of concerns which Defra and FERA before them, and the Godfray Review in 2018, specifically suggested ought to be subject to further study. But probably the most interesting and potentially alarming thing about seeing these two papers, at last, side by side is the question it raises about why the differences are there. Why were some of the points of note arising out of the analysis in BTO 2018 allowed to disappear as a consequence of the change in methodology in Kettel et al. 2021?
- Why was the lumping of passerine species that nest on or near the ground together with ground nesting waders/waterfowl (thus masking collective trends in respect of the latter that might otherwise be a cause for further investigation or concern) permitted and/or not explained?
- Why were the problems of poor experimental design (reliance on a dataset with methodological problems e.g. bias towards songbirds and expressly not designed for this sort of extrapolation) not discussed?
- Why are potentially significant negative effects on oystercatcher identified in 2018 and not in 2021?
- Why is there no comment on how the analysis in 2021 skews consideration towards species at inherently lower risk of negative effects from predator release than those extracted for particular attention in 2018?
- Are the two ‘new’ authors from Natural England solely responsible for these revisions? Are their backgrounds and role in facilitating badger cull licensing within NE a matter of any relevance here?
As Mark said back in February 2021 – you decide.