Loss of science quality in New Zealand is having dire consequences
Dr Jo Pollard (BSc (Hons, PhD))
Back in the 1990s, in at least one institution, rigorous government science was alive and well.
At Invermay Agricultural Research Centre, “Lab” meetings were being held where the scientist (or trembling student)
presented his or her proposed experiment: the background, hypotheses to be tested and methods. From these proposals were
hammered out, with the critical input of all to be involved (particularly sharp, insightful comments were usual from
some of the technical staff). Biometric approval was required to ensure the results from the research would be
meaningful.
In those days any hint of a prejudicial bias seriously undermined a scientist’s credibility. His or her worth was measured by the number of publications in high
quality, peer-reviewed journals.
Twenty-five years on, it seems a scientist’s selling ability matters most. Gaining funding and successfully delivering
results that generate more funding is vital to career development. And since the NZ government controls the money
(grants to universities, NGOs and its own departments) the government gets and selects what it wants.
Sadly, one of the things the NZ government wants and has been getting since the 1960s is widespread aerial poisoning
with the broad spectrum poison 1080. The government has argued it needs to kill introduced mammals claimed to spread
bovine tuberculosis (Tb) and threaten native wildlife, and widespread poisoning is the best way.
Respected scientists are expected to sit tight while the media, government departments and high ranking officials misuse
data to suit pre-ordained agendas. An example is the repeated claim by the Parliamentary Commissioner for the
Environment, Dr Jan Wright, that 1080 poison is “moderately humane”. The basis for the claim was a report commissioned
by the National Animal Welfare Advisory Council (Beausoleil et al., 2010). The report actually said that the word "humaneness" should be replaced with “animal welfare impact” because
truly humane control methods are rare, and that 1080 had a “severe to extreme impact lasting for hours”, meaning that
its effects on welfare were rated as “intermediate” relative to other poisons. Cyanide (which causes rapid loss of
consciousness and death) was at one end of the scale and anticoagulants such as brodifacoum (which has a severe to
extreme impact for days to weeks) at the other end.
In another example, respected ornithologist the late Professor Ian Jamieson & his student had to endure incorrect statements and gross extrapolation in the national media regarding the student’s
study of two groups of robins near Dunedin. Their research findings were held up as an example of how harmless 1080 was
and that it increased population growth (which it hadn’t).
New Zealand’s degraded science culture has spawned the rise of some disease-busting and biodiversity-conserving
specialists whose claims go unchallenged, have a very strong following and who are supporting ever-increasing poisoning
campaigns.
The most obvious victim currently is the kea, NZ’s own mountain parrot; a species known to be killed by 1080 poisoning
operations since the 1960s and now on the very edge of extinction. Estimates of total kea numbers as low as 1000 have
been made since 1986. A current “wild stab” guestimate provided recently was 3000-7000 birds, and numbers have declined
noticeably in recent years (Paul van Klink, speaking at Kea Konvention, Arthurs Pass, 2017).
Department of Conservation (DoC) monitoring has demonstrated that in every 1080 aerial poisoning operation, an average
of 12% of kea die. This incredible loss to such a rare species is not justified by science, but it is endorsed by DoC’s scientist in charge of kea management, Mr Josh Kemp.
This bird on the brink of extinction deserves top quality science but Kemp’s plan for further, increased poisoning to
save them falls far short. According to Kemp, increased 1080 poisoning is needed to kill stoats, which he says are a
major cause of the kea’s demise. This contrasts sharply with Kemp’s and other scientists’ previous findings (from the
1960s, 1990s and last decade) that despite living in the company of stoats, kea were “relatively immune” to predation.
Any predation impact was thought likely to have occurred long ago (stoats were introduced to NZ in the late 1800s).
It seems that Kemp is required to provide evidence that stoats (and even the largely vegetarian possum) are major predators of kea, and that 1080 is needed to kill them. Otherwise the government’s further, ever-increasing
poisoning plans would clearly be irresponsible. In March 2014 Mr Kemp stated:
”This year (2014-15) the DOC is gearing up to broadcast 1080 to poison pests over about 500,000 hectares of beech forest
inhabited by kea under the Battle for our Birds campaign. These operations are approved on the basis that benefit
derived to kea from stoat and possum control will outweigh the non target risk...” (Kemp 2014, unpublished).
There are many serious flaws in the claim that 1080 is needed to kill stoats. The stoat has to eat a poisoned animal to
die (the poison is delivered on cereal baits) and kill rates have been highly variable, e.g. “The aerial 1080 at Mt Arthur did not reduce the stoat FTI...Stoat FTI subsequently increased” (Kemp et al. 2014, unpublished). Stoats are also known to devastate native birds after poisoning in a response know as “prey
switching”, after a steady diet of rats has suddenly been poisoned off.
Another major flaw is that aerial 1080 poisoning has been repeatedly observed to cause increases in firstly mice, then
rats, in the poisoned areas. As Dr Andrea Byrom of Landcare Research bravely pointed out in 2013, such increases create
excellent conditions for the proliferation of....stoats! This is what was observed at Okarito in 2011, where a 1080
operation was followed by large increases in mice, then rats, then a stoat plague in late 2012. Tellingly, in a draft
2016 paper Kemp held this poisoning operation up as the “best example” of aerial 1080 improving productivity of kea and preventing mortality from stoats! He had already pointed out, in 2015,
that his Okarito study lacked replication, random assignment of the treatments, and observers blind to treatments.
Kemp has had to contort his data wildly, and make some extreme assumptions, to support the required claims. A graph he
shows of alleged stoat numbers is not that at all, rather it uses an undefined measure “based largely on seed rain”.
Never mind that seed rain does not always show a strong correlation with stoat numbers in the scientific literature.
Another piece of “evidence” used is a single kea carcass that swabbed positive for stoat saliva – stoats are scavengers
of dead animals, and will also prey on sick birds, so the saliva is not good evidence that stoats are a major predator
of kea. Neither is the finding of a dead kea “cached” underground (also used as “evidence”) since rats cache dead
animals too.
Kemp’s claims that only kea in human-inhabited areas die from eating 1080 pellets are simply untrue- in Kahurangi in
2014, a site chosen by Kemp for monitoring specifically because of its remoteness, 2/22 (9%) of monitored kea were killed by the poison.
Worryingly, Kemp’s kea predation studies rely on frequently visited nests, and there are plans for more monitoring. The
scientific literature and bird codes of practice warn against any interference with nests as this is likely to attract
predators and cause abandonment, and that any data from such nests must be treated as biased. Kemp’s nests that are
monitored with still cameras experience multiple visits from stoats and other species.
The Strategic Plan for Kea Conservation (2015, 2017), a collaborative project between DoC and the Kea Conservation
Trust, fails to achieve basic scientific standards. Research is cited without providing references, and bold,
unreferenced claims are made (e.g. “the recent population decline exacerbated by introduced predators”). Important
definitions (e.g. “predator”) are lacking. The Background omits important issues e.g. starvation, previously identified
as the major cause of death. There are no references to published research by Kemp despite his scientific involvement in
kea management since the 1990s. There is no consideration of welfare impacts of prolonged deaths from poisoning or
interfering with nests.
Positive, timely outcomes from the Strategic Plan seem unlikely. For instance the 2015 plan to “adequately” estimate the
number of kea remaining is still some way from being realised, with various options for methods still being discussed in
2017. There were plans to use an “experienced scientist” to develop a working population model, in 2015. In 2017 there
were still plans to “have a working population model developed and refined by an appropriate scientist.”
Kemp’s followers (kea lovers and DoC) are immediately and highly defensive of any criticism, to the point of leaping to
their feet and shouting down any hint that his work and endorsement of more poisoning could be at all flawed (author’s
personal experience at Kea Konvention, Arthur’s Pass 2017). The science-less nature of planning documents and
publications, the misuse of scientific mana among followers, and followers’ extreme reactions to any challenge are
typical, in fact diagnostic, of pseudoscience.
Readers of this concluding paragraph will be in a minority, since few now have time to read about and consider
scientific integrity. Hopefully the imminent demise of the kea will engage those few still thinking wisely to call
urgently for science in conservation management.
For references please see http://1080science.co.nz/scientific-reviews-of-1080/
Dr Jo Pollard (BSc (Hons), PhD) is an independent scientist, with particular interests in animal welfare, NZ’s ecology,
and scientific integrity.