Department of Conservation statistician Ian Westbrooke comments on 1080 use and the impact on birds and responds to the Whiting O’Keefe paper.
Based on the fact that 1080 is an acute toxin, the primary focus for DOC on terms of native animals has been establishing if there is an immediate impact from 1080 operations, with the main focus on birds. The assessment is that if there are no immediate impacts it is very unlikely that there are longer term impacts.
To assess acute impacts, often an effective starting point is to compare counts of individuals of a species before and after an operation. If no change is observed as is typically the case, there is no requirement for comparison with non-treatment (control) sites. If a change is observed, it is not necessarily due to the operation (except where it is possible to recover a dead animal and autopsy for cause of death), but the results can provide the basis for estimating the maximum impact of the operation on the species in question. This approach has allowed the identification of a small group of species that have potential risk of impacts from 1080 operations
The evidence up to 2000 is well-summarised in Spurr (2000). I have combined and in some cases updated the summary tables of results, and derived some additional statistics. The studies all looked for immediate impacts by comparing numbers of individuals that could have been affected by aerial 1080 at a time point before and after an interval afterwards in which immediate impacts would have been observed. The table shows the numbers counted before and after 1080 impacts would be expected, along with descriptive information. The last three columns show the estimated rate of impact, together with confidence bounds, ordered by increasing size of impact, and for those with no impact by increasing upper confidence bound. An upper 95% confidence bound is given in all cases, which provides a 95% confidence interval where no impact was observed. For situations with an observed impact, there is also a 95% lower bound for the impact given, and together two the bounds provide a 90% confidence interval for impact in those cases.
For most species studied – short tail bats, kaka, brown kiwi, blue duck, robin, great spotted kiwi, fernbird, and NZ falcon – absolutely no immediate impact was observed. While some sample sizes are small, others are very large. The impact of sample size on reliability of the results is indicated by the size of the confidence bound. Mortality impacts greater than that confidence bound are incompatible with the data.
A few species show impacts in the table. The results for kokako involve a very small reduction in a large number of territories, some or all of which may be attributable to causes other than the 1080 operation. There are relatively small impacts evident for weka and for morepork. The authors of the submission totally misunderstand the evidence when they claim that short-tailed bats are “likely being killed in substantial numbers”, when zero impacts were found in one large study following an operation, and in two other studies bats were found not to be eating baits.
Robins and tomtits
Of greatest concern are the results for robins and tomtits, and that is precisely where substantial DOC research effort was concentrated from the late 1990s.
Research in Pureora, lead by Powlesland, showed that any substantial immediate impacts on robins was controlled or eliminated with properly screened carrot baits, and that robin population showed increases after a year even after a poorly screened carrot operation led to the substantial mortality in 1996. (The submission misses the key point in relation to population monitoring after this operation. The results show that population in the 1080 treatment area increased despite the substantial mortality observed amongst the banded birds.)
In contrast with robins, tomtits continued to experience deaths immediately after an aerial 1080 operation with properly-screened carrot baits in 1997, but not after a cereal operation in 1998. Therefore, tomtit became a focus for further research which established that appropriate cereal bait operations have little if any immediate impact on tomtit operations, but that carrot operations can have a negative impact on tomtits. These results have been taken into account in DOC’s management of aerial 1080 operations where tomtits may be at risk.
The tomtit results have been subject to seriously distorted reporting Claims were made that the 2001 Tongariro operation showed “a maximum negative effect of cereal baits on tomtit population of -35% at the 95% confidence level” . In fact, the paper showed that any impact was near zero and that “loss rates greater than 8.4% due to 1080 were incompatible with the data (95% one-sided confidence bound)” (Westbrooke, Etheridge & Powlesland, 2003.)
While indefinite amounts of conservation funds could be spent on researching 1080 impacts on native animals, there has been a variety of work done on modest budgets that has shown that risks are generally near zero, Where impacts have been observed, these have become targets for further research, and steps have been taken to reduce risk as close to zero as possible.
Click on table below for full-size view
