Abstract
1. Introduction
2. The rabbit as a
pest in New Zealand
3 Assessment of
current control measures
4 The effectiveness and suitability of RCD virus as a biocontrol agent in New
Zealand
5 The intended programme in which RCD is proposed to be used
6 The
effects, positive and negative, of RCD
7 The likely success and costs of measures which can be employed to ameliorate
negative impacts
8 Other relevant matters
9. Acknowledgements
10. References
Major Issues of
Ecology and Wildlife Management
to be considered for
RCD Import Decision Making
A report to the
Chief Veterinary Officer,
MAF, Wellington, New Zealand
By
H. Moller, K. Brown & N. Alterio
Ecosystems Consultants
P.O. Box 6161
Dunedin
Ecosystems Consultants Report No. 7
27 February 1997
Rabbits impose serious economic, social, and ecosystem level environmental effects in semi-arid and sandy soil areas of New Zealand. Some indigenous species of conservation importance may be threatened by ongoing high rabbit abundance in such areas, but others may be buffered from the effects of introduced small mammalian predators that preferentially feed on rabbits.
Existing technologies to kill rabbits are adequate and effective but too expensive to make them cost-effective for farmers. No more cost effective alternatives to RCD or existing technologies were identified by the RCD applicant group, peer reviewers or public submissions. RCD has the potential to impose sustained and long-term suppression of rabbit numbers in both low rabbit and rabbit-prone areas, but its effects may also be small and/or temporary. RCD is unlikely to make the rabbit problem worse in the areas currently experiencing low rabbit numbers.
Any strong and long-term reduction in rabbit abundance is likely to assist bovine tuberculosis suppression by decreasing the number of ferrets and the benefits of this will be shared widely throughout New Zealand society.
Removal of prolonged grazing pressure from rabbits will trigger profound habitat changes in some places (especially where domestic stock do not graze) which may assist some wildlife species and threaten others. Increases in woody weeds are likely with attendant increased costs. Prolonged and strong reduction in rabbits is likely to trigger increased abundance of stoats and rodents. Sudden reductions in rabbits by RCD outbreaks will definitely trigger short-term increased feeding on native species by predators, and prolonged reduction in rabbits may bring long-term equilibrium changes in diet that result in more native fauna being consumed. Any risks to indigenous biota could therefore last for as long as RCD suppresses rabbits, just as any benefits accruing to species conservation would also then be long-term. These same risks and benefits may be spread throughout New Zealand and are not restricted to the rabbit-prone land. Neither opponents nor proponents of RCD can rightfully claim that it will bring net conservation harm or gain for species conservation values. Diversion of funding from existing DoC work programmes to RCD monitoring, research or mitigation adds risk that the release of RCD will bring a net conservation loss.
The risk of not releasing RCD to solve current and ongoing conservation problems of species declines and ecosystem degradation in rabbit prone areas has scarcely been acknowledged by opponents to RCD.
In the absence of a certainty of conservation outcomes there is a need to plan and fund strong monitoring and mitigation measures to minimise unwanted environmental effects. These costs have been greatly underestimated by the IIA. The need for urgency in releasing RCD has been exaggerated on ecological grounds, but is proven for economic and social impacts. A moderate delay of at least 2 years to mount research and contingency plans, and to develop the necessary tools would minimise environmental risks. However even decades of research could never resolve several of the underlying ecological uncertainties about the long-term outcomes of RCD release. Formulation of a detailed and integrated wildlife and landuse management plan by representatives of all major stakeholders can help minimise risks and make it more likely that any benefits of RCD are captured for long term sustainability of economic, land and environmental values.
This report summarises the major issues of ecology and wildlife management arising from a decision on whether or not to release Rabbit Calicivirus Disease (RCD) to control rabbits Oryctolagus cuniculus). Wildlife management involves a multi-disciplinary scientific consideration of strategic planning options for the application of cost effective and environmentally sound pest control or conservation management interventions. It involves much more than simply ecology.
The potential biological control of a widespread and often abundant animal such as the rabbit has many and varied ecological and wildlife management consequences. Therefore, this report is only a summary of the main RCD issues and risks. It primarily considers the topics identified by MAF-RA as "potentially major issues raised by submitters" under the headings prescribed in "Matters to be Considered for Decision Making".
This report evaluates: (i) the public submissions to MAF-RA, (ii) the Taylor Baines report, (iii) MAF-RA's RCD Peer Review group reports, and (iv) the RCD Applicant group's IIA.
A few additional issues are raised by us under "Other Matters" on p16 of this report because we judge them to be especially relevant to the decision making process. In the interests of brevity, citations of the submissions, reports or scientific literature are not exhaustive.
This report does not address the question of whether or not the RCD virus should be introduced. This is the responsibility of the Chief Veterinary Officer.
2. The rabbit as a pest in New Zealand
Measure of the severity and distribution of the impacts on pastoral agriculture.
The case for serious direct economic impact of rabbits on pastoral production on moderate to extreme rabbit-prone land (semi-arid regions and sandy soils) is overwhelming and well quantified. However, it is certain that the majority of New Zealand sheep and cattle farmers do not have an economic problem with rabbits. On this basis, some submissions have used the latter to argue that RCD is an inappropriate tool to use. This reasoning confuses separate issues. Provided the risks and costs are not prohibitive for all stakeholders then it matters little that most of the considerable direct economic benefits flow to a minority of farmers and their associated rural communities.
There are also indirect effects of rabbits on overall economic return to the affected farmers and the wider New Zealand economy and export trade4. For example, several benefits of cheap biological control of rabbits on rabbit prone areas are not measured by simple calculations of the degree of competition for herbage between rabbits and stock. Several of the worst affected farmers are unable to invest in fertilisers and alternative land use regimes until the immediate burden of rabbit control costs has been removed. The true measure of the impact of rabbits must therefore include a mass of indirect effects and lost potential for economically and environmentally sustainable use of the land.
Though not proven with direct experiments, a growing body of circumstantial evidence suggests that bovine tuberculosis (Tb) is spread to cattle and deer by ferrets, and potentially also by the stoat. It is certain that the number of ferrets (and cats) is higher in areas with higher rabbit abundance, and that reduction in rabbits will cause a reduction in ferrets. This in turn is likely to reduce the risk of reinfection of cattle and deer following culling of infected stock. In some areas, reduction of ferrets is likely to make the difference between eradication of Tb and the accreditation of herds as Tb-free. Reduction of ferrets may also be necessary to remove the disease from the overall endemic region. Accordingly, costs of ongoing Tb vector control, Tb testing and culling will be reduced along with the overall risk to our dairy and cattle export trade from non-tariff trade barriers.
Reduction of ferrets and cats (following any prolonged reduction in rabbits by RCD) will probably trigger increased abundance of stoats. Stoats are known to carry Tb but their role in spreading Tb is poorly understood compared with ferrets. Therefore, the possibility exists that RCD may merely change the source of the Tb problem from ferrets to stoats in rabbit prone areas.
Most submissions have missed the important point that current direct control of ferrets for Tb mitigation is likely to trigger increased rabbit abundance in the areas currently with few or no problems from lost production to rabbits. RCD could provide farmers with an opportunity to target ferrets for Tb mitigation without fear of creating a rabbit problem.
Any reduction of Tb risk and costs from RCD will benefit farmers and the wider New Zealand economy and society well beyond the severely rabbit infested areas.
Measure of the severity and distribution of the impacts on environmental values.
There is a lack of well-quantified studies of impacts on "ecosystem values" such as soil nutrients, soil structure, erosion, siltation of waterways and representativeness of natural community structures. Nevertheless we believe such effects are very likely to be occurring because of a mass of circumstantial evidence that such problems of degraded ecological systems are severe in rabbit prone areas. It is logical that they are exacerbated by overgrazing and there is abundant evidence that rabbits reduce the standing crop of plants to extremely low levels in many semi-arid regions. Prolonged suppression of rabbit abundance is likely to ameliorate such effects in the long run provided that farm stocking rates are contained to prevent overgrazing in the absence of rabbits and fire is used sparingly. Several submission accept the impact of rabbits on these broad scale environmental values.
However, the IIA made an unwarranted generalisation that rabbits are a threat to species conservation values and general concerns for wildlife have been voiced. A few direct and a mass of indirect ecological effects of rabbits benefit some native species, but not others. DoC undertakes little direct expenditure on rabbit control compared to other species conservation threats, and that control is often for their "good neighbour" role rather than for conservation gain.
Measure of the severity and distribution of the impacts on other economic values.
The case for rabbit impacts on "intensive agriculture and horticulture" outside the rabbit prone areas was scantly covered and poorly quantified in the IIA but has been corroborated by several submissions.
A potential for increased losses to "intensive agriculture and horticulture" from RCD by increased hare abundance is not mentioned in the IIA, and is difficult to quantify.
3 Assessment of current control measures
Effectiveness and cost of current control measures in achieving stated pastoral/environmental goals.
Current control technologies for rabbits and Tb mitigation are generally effective provided high levels of effort and expenditure are maintained. The issues to be considered are mainly ones of cost. We know how to kill rabbits without having RCD to help - the problem is simply that we cannot afford to do it within the current economic constraints faced by a number of farmers in rabbit-prone areas. However, RCD has the potential to give widespread cheap control after initial high start up costs.
The IIA and many public submissions in favour of introducing RCD made much of a supposed problem from 1080 poison-shyness amongst rabbits, but the documented scientific evidence for this concern is largely absent. Reliance on large and frequent applications of a single toxin is likely to result in a build-up of long-term resistance and for this reason we concur that it is desirable to find as many alternative safe rabbit control tools as possible. However, there is little tangible scientific evidence that we have already reached a state where the current technologies are no longer effective, so there is no need for undue haste to establish an effective and safe biocontrol agent.
No new economically viable alternative methods of rabbit control have been identified by the submissions. Several submissions called for a disease vectored sterilisation agent but development of these is decades away, will be enormously expensive, and will entail several of the same risks as involved with RCD.
Incidence of costs
Costs of rabbit control by current methods are extreme, well documented in the IIA, and affirmed by several of the submissions received.
Declines of several native species are ongoing and predation is a major cause. Costs of trapping for predator control to achieve conservation goals are extremely high (Ratz et al. unpubl.) and the benefits localised. Poisoning methods are far more cost-effective but need urgent research to improve efficacy and check environmental safety risks (Alterio 1996; Alterio et al. in press). Indirect control of predators through reduction of rabbits by RCD has the potential to achieve much more widespread and effective predator control than is currently possible by DoC, NGOs, Landcare groups or private individuals.
Severity and distribution of any impact on native flora and fauna.
Current rabbit poisoning methods undoubtedly kill some non-target native and introduced species, but quantified data do not exist to quantify its frequency or ecological importance.
A current raging public debate on the potential impacts of 1080 on invertebrates is yet to be resolved with reliable scientific data.
4 The effectiveness and suitability of RCD virus as a biocontrol agent in New Zealand:
The general features (irreversibility, unpredictability, potential to spread everywhere), and potential disadvantages of biological controls were given scant regard in the IIA, whereas the disadvantages of conventional control methods were stressed repeatedly. Similarly the advantages of chemical controls (reversibility, control of timing and spatial extent of application, flexibility in different agencies response) were given scant regard in the IIA. Accordingly there were several concerns amongst the peer review panel and public that the overall risk sensitivity in the application was biased.
Several submissions argued strongly against introduction of RCD on general grounds that past introductions of mammals and some birds have proved to be ecologically deleterious. While the lessons from the past are a useful reminder of what could potentially go wrong, they hardly represent a fair basis for considering whether or not RCD should be introduced. There is now a science of ecology, a much stronger awareness of impacts of introduced species, and the practice of biological control as a whole has become enormously more rigorous in its consideration of unwanted effects. In addition, many past introductions were accidental. Release of multi-host species is no longer acceptable in biocontrol. Available literature asserts that RCD is either entirely host specific or very nearly so. This is entirely different from the conservation risks posed by the introduced mammals with broad diets which some submissions have emphasised in their broad scale condemnation of any future introductions20. Their analogies reflect a "bioxenophobia" comparable with the "chemophobia" used by fervent proponents of biocontrols.
Clearly, potential introductions must be considered in detail on a case by case basis rather than from a broad scale position of either support or opposition to biocontrols. Ironically, further introductions of host specific biocontrol agents may be the only way we can effectively and cheaply restore mainland ecological communities by reducing the ecological impacts of earlier ill-advised introductions (Moller 1989).
Assessment of the likely success of RCD in achieving the stated pastoral / environmental goal, with and without conventional control.
The goals of the proposed RCD introduction were poorly stated in the IIA and there was no companion plan from MAF, Regional Councils or DoC to explain how the follow-up management issues would be managed or funded. Accordingly, it is difficult to define and assess likely success.
It is clear that economic and environmental benefits are localised and short-term following conventional controls in rabbit-prone areas. There is uncertainty about whether RCD will achieve better control than the existing management regimes, but there is at least a potential that RCD could provide much more effective, widespread and cheaper control than currently possible.
The risk of not capturing this potential benefit has been given scant regard by those submitters opposing RCD on several other grounds.
Possible increased rabbit abundance as a result of myxomatosis, in areas currently with very low numbers, has been presented by Gibb & Flux (1983) and is reiterated in some public submissions (ie. reduction of rabbits by RCD in current low rabbit populations could trigger declines of predator abundance, followed by rapid resurgence of rabbits out of control because predators cannot breed fast enough to reassert control). Wider oscillation of rabbit abundance around a higher average level is therefore possible. There is no scientific evidence to either rule out this possibility, nor to consider it likely. It is a speculative scenario because (i) it assumes rabbit abundance will indeed decline in low rabbit areas because of RCD, (ii) it assumes predator abundance will decline further once already low rabbit abundance is reduced even further , and (iii) the supposed inability of the lowered predator population to reassert control on rabbits following their initial knock-down has neither been demonstrated by experiment nor mathematically simulated to assess its likelihood.
Intuitively we judge that the above rather complicated scenario is unlikely to occur, and therefore that introduction of RCD is unlikely to make the rabbit situation worse either in rabbit-prone or low rabbit areas. Nevertheless it cannot be ruled out categorically. The low level of risk must be scaled against the large area potentially affected and the unwanted economic impact, should RCD trigger rabbit irruptions in the areas currently experiencing no economic problems from rabbits.
Assessment of the ability of RCD to maintain the stated pastoral / environmental goal over time.
The potential for rapid return to pre-RCD rabbit levels has been given scant treatment in the IIA but is emphasised by several submissions from the public. It deserves consideration in the overall risk assessment when potential benefits are weighed against potential costs.
Data are scarce but the available evidence from Spain is for a return to 40% of pre-RCD levels within 5-6 years (B. Cooke pers. comm. in Moller 1996a). Modelling by Barlow (1996) underscores that we know very little about how effective it will be in New Zealand. Plausible scenarios from collation of all known parameters overseas predicts anything from prolonged suppression of rabbits without intervention to an instantaneous but short-lived knockdown without persistence of the disease (Barlow 1996). The IIA unwarrantedly implied that RCD will spread and persist in NZ, with only occasional warnings that this is an unknown.
Estimates of the size of initial knock-downs in rabbit abundance by RCD are anecdotal and have a poor statistical basis. Peer reviewers and some submissions have pointed out that the IIA made unwarranted and repeated assertions that the efficacy of RCD will be higher in high density rabbit populations and that it will not persist in low density populations. Some field evidence from Spain is to the contrary and the mathematical basis for the prediction is doubtful.
Recent media reports that RCD does not impact as strongly as expected on rabbits in Australia should be disregarded entirely by the Decision Maker. The data as reported concern establishment success after a single or few initial releases, not a measure of the impact on rabbits once established. There are several examples of biocontrol agents and new organisms that were extremely difficult to establish in the receiving community (Moller et al. 1993; Moller 1996c). Indeed, it would be astonishing had RCD established and impacted on all sites immediately after a single release site.
The multiple factor contributions (rabbits, avoidance of overstocking, reduction in use of fire, increased use of fertiliser) to the environmental problems in the semi-arid high country were given extremely scant coverage in the IIA, and several submissions have raised their (in our view legitimate) concerns that the planning in the RCD proposal has only tackled part of the overall problem. The IIA does not mention the risk that many farmers will abandon other control methods if RCD is very effective in the long-term, nor the possibility that farmers will simply increase stocking rates to recreate the same environmentally unsustainable levels of grazing pressure as currently imposed by rabbits. If RCD provides partial or intermittent control of rabbits and less secondary control effort is applied, the economic and environmental benefits promised may not be realised.
It is logical (indeed sound wildlife management practice from a cost-effectiveness standpoint) that farmers should greatly reduce secondary control efforts if RCD offers prolonged protection against rabbits, but then the benefits of using multiple tools as stressed in the IIA may not eventuate. There already are signs that farmers are holding off applying conventional methods on the expectation that RCD is imminent and will be effective. We note the repeated and heartening assurances of several farmers in their submissions that they stand ready to apply follow-up controls even after RCD is released. Others assure that they will use the resulting "window of opportunity" to instigate more sustainable landuses in future. Several other members of the public have expressed little confidence that these safeguards will occur and provide the long-term sustainability and the environmental benefits promised. The absence of detail in the IIA about how these safeguards will be achieved has underscored these concerns and made it difficult for many members of the public to evaluate the risk of failure to capture the environmental benefits and opportunities for sustainable land use.
Humaneness
There is consensus that RCD is relatively more humane than Myxomatosis, but divergent opinion about its relative acceptability compared to poisoning and/or shooting. There are few absolute measures of humaneness that will be accepted by all stakeholders in the RCD debate, so there is little prospect of resolving the differences of opinion remaining. The absence of a large number of submissions challenging RCD as unacceptably inhumane, and the absence of literature detailing symptoms of unacceptable suffering suggest that there are no broad scale ethical concerns for the release of RCD on humane grounds.
Prolonged suppression of rabbits by RCD would rapidly decrease the number of rabbits currently experiencing suffering from existing control methods.
5 The intended programme in which RCD is proposed to be used
Details of how RCD virus will be deployed
Several peer reviewers and public submitters have provided valuable suggestions for how the strategy for release (Part II of the IIA) could be improved. The ecological and management basis for several choices must be better explained than in the IIA and critically evaluated by a multi-disciplinary team involving wildlife scientists and managers, representatives of all stakeholders (Farmers, MAF-RA, MAF Qual, Regional Councils, DoC, Conservation NGOs, Animal welfare groups). The planning needs to be enormously more detailed and long-term than provided to date.
Specific issues raised for consideration include:
the basis for the choice of up to 12 release sites in the pilot phase
the basis for, and the exact criteria to be met for a decision to stop or proceed to general release after the pilot phase have not been declared
the basis for choice of a number of pre inoculated rabbits for release is not declared
a two week pilot phase is a impractically short period to allow full evaluation of the pilot phase. A year may be more appropriate so that the behaviour of RCD in a wide range of seasons within New Zealand conditions can be gauged. Also, the behaviour of RCD in the first few weeks will give no indication whatever of its true impact on the population.
much might be learned if the transition to main release is delayed sufficiently to observe spread of relatively few main foci of initial release
constraints and opportunities for release in different seasons are problematical and cannot be known before release in New Zealand
inclusion of a well replicated series of pilot release sites in low density rabbit populations would be extremely valuable to assess the scale and risks of conservation impacts throughout much of New Zealand, thereby facilitating planning and research on how to protect conservation values. The release strategy seems to be based upon the economic benefit objectives and to ignore the conservation goals and risks.
a pivotal decision to release in the North Island should be delayed as long as possible to minimise risks and use the preliminary monitoring of RCD in the South Island to guide prediction of what will happen in the North Island.
inclusion of only 6 monitoring sites is scant replication
several issues of standardising of monitoring methods and co-ordination of research and management have not been considered or are treated scantly, including:
(i) methods of assessing changes in rabbit abundance
(ii) required sample sizes and statistical power
(iii) replication requirements for scientific interpretation
(iv) stratification of sites along ecological gradients
(v) measures of virulence and attenuation
(vi) impacts on non-target species, vegetation and ecosystem processes.
The financial underpinning of the programme
The financial underpinning of the whole RCD endeavour was not made adequately transparent in the IIA, and several submissions have expressed major concerns that the funding for research, monitoring, management and mitigation of unwanted effects may be inadequate. The public cannot weigh the risks and likelihood of capturing the benefits without the financial underpinning clearly identified.
Similarly, they cannot weigh whether there will be net conservation benefit or harm if it is not certain that DoC will be given adequate additional funds beyond their current allocation to deal with the new tasks associated with RCD. If they are forced to divert funds from their existing conservation work programme it is more likely that a net conservation harm might occur. Provision of extra funding for predator control has been rightly urged by several submissions.
The distribution of costs and benefits cannot be apportioned by the stakeholders until the detail about financing is known. Several submissions are concerned that several of the costs of RCD will be picked up by the wider community, and Vote Conservation, whereas the bulk of benefits flow to private farmers. The effect of RCD would then be a de facto taxpayer subsidy for a small private sector group.
The financial needs for adequately monitoring the programme and mitigating any associated effects will be much larger than those detailed in the crude cost benefit analysis appended to the IIA. Financial needs will be especially high in the first five years to allow monitoring methods to be developed, tested, and if necessary perfected; to obtain pre-release base-line measures; and to gauge the effects of the virus as it spreads and its ecological trajectory settles more towards an equilibrium situation. It is clear that subsequent monitoring of the ecological effects will have to be very long-term (decades) if is to capture the real impacts of RCD on natural systems. Similarly, expenditure on mitigation of any unwanted side effects may be needed indefinitely or for at least as long as RCD suppresses rabbit abundance.
Some agencies may consider research an expensive luxury that could be avoided. We consider it essential for the long-term lessons to be learned for future biocontrol attempts of this magnitude. Research also allows us to determine the expenditure needed for mitigation of pastoral/environmental effects; the level of follow-up rabbit control required; and to understand the ecology of the RCD tool itself.
Australia has allocated AUS$1.9 million per annum simply for monitoring RCD and its impacts. So far New Zealand has allocated approximately NZ$190 000 per annum through its FfRST research programme. The latter is a totally inadequate sum to measure the potential outcomes. We cannot simply import the lessons from the greater Australia investments because ecological conditions and impacts are likely to be very different between the countries.
Integration of RCD with other control measures
No detailed plans have been tabled for public consideration on how this will be achieved. Several submitters have called for institutionalised responses and the generation and distribution of educational material to encourage farmers to take advantage of any "window of opportunity" presented by RCD.
How the short and long term effectiveness of the programme will be monitored
This is given scant and inadequate coverage in the IIA. In part it reflects a lack of crisply stated goals and objectives for the entire proposal. There is a clear need to convene a multi-disciplinary group with representatives from all stakeholder groups to critically evaluate and cost what is required.
6 The effects, positive and negative, of RCD
Non-target species
Judging from the huge number of submissions received, infection of humans and other animals is an area of major concern for the public. We have insufficient relevant expertise to guide the Decision Maker concerning the chance of such direct effects of the virus on other species.
However, we advise as ecologists that the indirect effects on non-target species are potentially many, strong, unpredictable and could be both beneficial and deleterious to several species of conservation importance. All these aspects were inadequately dealt with by the IIA which seriously misrepresented the evidence by claiming that net environmental gain must occur from successful suppression of rabbits by RCD. There was a clear and independently formulated consensus from all the ecological/environmental peer reviewers before the IIA was released that this was a problem. Several submissions have subsequently raised the same concern. We cover these issues in more detail below.
Target species
Population impacts of RCD on rabbits are likely to be initially large but it is impossible to predict the longer term outcome with certainty because:
persistence and spread in New Zealand conditions cannot be known for certain until it arrives and has had time to become well established;
there is a poor statistical basis for measures of initial knock-down overseas;
that initial knockdown may be little guide to subsequent population density of rabbits because of the young rabbits gaining immunity and transmission of the disease may be density dependent;
genetic resistance may build in the longer term in all sections of the rabbit population;
the virulence of RCD may change in response to selective pressures (probably it will decline), but it is unknown by how much or how soon;
reduction of rabbits is likely to have a profound effect on the habitats in which they live, which in turn is likely to alter rabbit mortality rates (from agencies other than RCD), reproductive rates and movement patterns.
Accordingly, the effects of RCD on rabbit abundance may be extremely light or severe; they may be temporary or long term; and the effects may vary from place to place. This uncertainty does not arise from academic conservatism or dithering. It reflects ecological reality Ä the impacts cannot be predicted precisely no matter how much we wish to be able to advise more precisely on the outcome.
The Decision Maker therefore has to weigh the potential risks, costs and benefits within a very broad range of potential outcomes. The IIA and several submissions have emphasised extremes of these uncertainties in support of their stances either for release (here perceptions of benefits have been maximised but risks and cost minimised) or against release (here unknowns and perceptions of costs and risks have been maximised but benefits minimised, and the cost of doing nothing has largely been ignored). In view of the very real ecological uncertainty about outcomes, both on target and non-target species, there has been ample opportunity for each camp to justify their pessimism or optimism and the result is an extremely polarised stand-off.
Indigenous flora and fauna and natural communities.
Widespread effects are likely
In our opinion, the number of potential RCD effects on indigenous fauna and flora have generally been underestimated by the IIA, by DoC and by several submissions. Rabbits are a classic "keystone species" that have an inordinate variety of far-reaching ecological effects within the ecological community. If RCD imposes prolonged and effective suppression on rabbits, by definition RCD has become the new keystone species in the community. Profound and long-term changes are likely, especially in freely draining soils and/or semi-arid areas not currently grazed by stock. Indirect impacts on non-target species through predator/prey changes are possible in the majority of other areas in New Zealand where rabbits are currently at low levels. It is not an exaggeration to expect RCD release to potentially bring the most widespread and profound ecological changes to New Zealand's open habitats seen this century.
Many more species are potentially affected than the few listed by DoC and the IIA. Insects and lower vertebrates have been given particularly scant consideration. Restriction of the mitigation procedures merely to "Category A" species in high rabbit areas is a bureaucratic nicety that does not necessarily reflect absolute ecological need. Several species may change from categories A to B or C, or vice versa because of RCD. Several others are of unknown status (Categories I and X), or are endangered in New Zealand but secure in other countries (Category O). Some people may still place a high priority on maintaining those species in New Zealand.
Species and communities are vulnerable to change throughout the country. There has been a potentially mistaken assumption that effects of RCD on non-target indigenous wildlife will be greatest or evenly localised entirely within areas currently experiencing high rabbit abundance. Changes in predator numbers and their diets may be triggered by relative (proportionate) changes in rabbit availability, not the absolute level of change in rabbit numbers. For example, the change in use of native species in the diet of a cat may be just as great (possibly even greater) if the number of rabbits is reduced by 50% from 2 to 1 rabbit per hectare, as it would for a cat in a semi-arid region where rabbits declined by 50% from 30 to 15 per hectare. The field evidence from Spain was that RCD had an even bigger relative impact on low density rabbit populations (some local populations went extinct), so there is no guarantee at all that the ecological effects requiring mitigation, or the species gaining a benefit from RCD, will be restricted to the rabbit prone areas.
Potential direct effects of rabbit removal on native birds of prey have been omitted from consideration so far. Such effects are well documented overseas.
Substitution of sheep grazing to ameliorate unwanted effects of rabbit removal are mentioned in the IIA; this tool is theoretically workable but will take time and money to perfect. It has been used in Britain where some small nature reserves have sheep grazing leases to achieve conservation goals, with strict limits on the time of grazing and number of sheep present.
Impacts of prolonged rabbit decline on plant successions have been largely neglected by the IIA, mitigation plans and cost benefit frameworks. Information on the status of rare plants is scant, and hence the potential threat of RCD to plants by changing succession and plant competition is largely unknown. Finding out whether there is a significant threat here will take time and money. We concur with several of the public submissions that there is a likely and potentially significant risk of an increase in woody weeds that could also impact on native flora and fauna. This was a common feature in the UK following the introduction of myxomatosis (Sumption and Flowerdew 1985).
Diet switching
There is clear evidence of "diet switching" by predators and increased predation of ground nesting birds immediately following rabbit control (Pierce 1987, Norbury & Heyward 1996; Rebergen et al. subm.).
Diet changes will be long-term
The potential impacts of diet switching has been understated in the current debate by potentially wrongful assumptions that it will be temporary. Potentially far more important than the temporary intense diet switch just after rabbit reduction is the longer-term equilibrium diet change that is likely once rabbit abundance has been permanently suppressed by RCD. Smith (1994) has demonstrated that the use of non-rabbit foods by ferrets increases steadily in areas with naturally lower rabbit density compared to in high density areas. Norbury & McGlinchey (1996) showed that there were fewer rabbits per predator after the decline in predator numbers following rabbit poisoning. This predicts that the predators will each eat more native fauna even after there has been a reduction in the number of predators present following rabbit control. Similarly, undisturbed rabbit populations in the Mackenzie Basin had fewer rabbits per cat in sites of low overall rabbit abundance compared to sites with very high rabbit abundance (Clapperton, Fletcher & Moller, unpubl.). Collectively these inferences suggest that per predator consumption of native prey will increase if and for as long as RCD imposes reduction in rabbit abundance.
It is unknown whether the increased per predator intake of native prey will lead to an overall increase in predation pressure on native species. Diet changes may be counteracted entirely by the reduction in the number of predators present because of lowered rabbit abundance. Long-term increased predation pressure on species of conservation importance may result, but so might long-term reduction in predation pressure. Myxomatosis triggered increased predation pressure on some birds in the UK for at least 20 years until its effects on rabbits waned (Sumption & Flowerdew 1985).
The IIA has seriously misrepresented this potential threat by repeated implications that diet switches are temporary and that eventual reduction of predators following RCD must necessarily then bring about a long term reduction in predation of native fauna.
Diet switching due to RCD may have different effects than that of current poisoning methods
The IIA has further made an unwarranted assumption that diet switches already occur with current technologies so that RCD will merely repeat rather than exacerbate an existing threat. However, (i) RCD will potentially control rabbits in areas not currently experiencing rabbit control by conventional means, (ii) the timing of the fluctuations in rabbit abundance may be very different for RCD than for other methods, (iii) the spatial extent of the knockdowns by RCD are likely to be much bigger than the smaller patches poisoned or shot over, (iv) the speed and degree of knockdown of the rabbits may differ, (v) there will be no secondary poisoning of predators by RCD, and (v) RCD may trigger the long-term suppression of rabbits, and consequently the indefinitely elevated equilibrium levels of use of native fauna by predators as discussed above.
Accordingly, if RCD triggers recurring fluctuations in rabbit numbers then it may bring (a) more frequent diet switching, and (b) diet switching to new areas of conservation importance. If RCD triggers long-term and stable suppression of rabbit numbers the long term equilibrium diet change may bring prolonged increased or prolonged decreased predation pressure on native species. There are no reliable data to predict which is the most likely, and by how much net predation pressure might change. The effect is likely to vary in strength from place to place and potentially also between indigenous species.
The potential threat/benefit to native biota has also been seriously misrepresented by asserting that RCD is likely only to persist in high density rabbit areas. As stressed above, field evidence from Spain implies the reverse, and the mathematical model on which the IIA made its prediction is too preliminary to give adequate guidance on this matter.
The net direction of changes cannot be predicted
Accordingly, the Decision Maker must weigh a possibility that RCD may trigger long-term elevations of predation pressure on native biota over much of the rabbit's range in New Zealand against a possibility that it might also confer long-term benefit to the native biota over that wide area by triggering long-term decreases in predation pressures. This is an entirely different risk analysis from that portrayed in the IIA and therefore reflected in much of the public submissions that any predator prey switching must necessarily be short term, definitely followed by a net decrease in predation pressure, and that any such effects are likely to be localised to areas currently experiencing moderate or high rabbit abundance.
Unfortunately, several of the submissions opposing RCD have tended to categorically assume that these changes will necessarily be bad for the non-target species. They have failed to emphasis the certainty that several species are currently threatened and declining, and thereby have not acknowledged the risk of turning down RCD.
The following medical analogy may help the public and the Decision Maker fully appreciate the importance of considering the risk of doing nothing. Imagine you knew you were ill or even dying. All the currently available therapies are only slowing your decline. If a new potential tool has been discovered but is untried, would you prefer your doctor to administer it even though there was no scientific surety that it will cure you, and even a possibility that it may make you worse?
People in general, and especially conservationists, are naturally risk averse and many would prefer to not take an active step (such as release of RCD) if there is some chance that it might go wrong. But doing nothing is also an active decision and choice with several attendant risks for wildlife. Failure to objectively weigh the risks of acting with the risks of not acting is a common feature of many wildlife management decisions involving the general public, and it is very evident in the submissions on the current RCD debate.
Integrity of indigenous ecosystems
The potential for RCD to restore ecosystem flow rates of energy and nutrients, soil structure, natural water flows and to minimise erosion are well substantiated by the IIA and submissions received. These benefits are all contingent on prolonged suppression of rabbit abundance (an unknown), lack of overgrazing by sheep, little use of fire and some fertiliser input and may be very slow to emerge. These latter contingencies emphasise the need for long-term and improved management of the semi-arid lands if the benefits provided in the IIA are to be captured.
Intrinsic value of ecosystems and ecosystem stability
See comments in earlier sections of this report.
7 The likely success and costs of measures which can be employed to ameliorate negative impacts
Control of mustelids (stoats, weasels and ferrets) and cats has been traditionally by trapping, but this method is expensive, time-consuming, short-term and only partially effective. Trapping predators, particularly stoats is not always effective and may not provide adequate protection for native wildlife in some circumstances (Murray 1992). These critical issues are ignored by the IIA.
Direct poisoning of these predators is estimated to be 3-7 times cheaper than trapping (Moller et al. 1996). Methods of direct and indirect poisoning are under development, but safe and proven poisoning methods are not currently ready. This is another critical point largely ignored by the IIA.
Further funding and research is urgently required to intensify and hasten the development and trialing of new predator poisoning technologies.
Intensive, widespread and long-term predator control may be required in numerous pastoral habitats to ameliorate any negative impacts on native wildlife resulting from release of RCD because:
(1) predator guilds (groups of predators living together in the same areas) may change: stoats and rodents may replace ferrets and cats and rabbits in pastoral habitats (Moller and Ratz, unpublished.);
(2) stoats and rodents are considered more serious threats to some native species than ferrets or cats (Brown in press; Alterio and Moller in press; submission 795 made by Department of Conservation);
(3) new predator-prey guilds may be less stable than the existing ones, especially if large fluctuations in rabbit numbers are triggered by RCD, as reported in Europe and the sporadic outbreaks of RCD observed in Australia are repeated in New Zealand;
(4) diet switching by ferrets and cats may be long-term since there may be fewer rabbits per predator (Moller and Ratz, unpublished);
(5) predators disperse and range widely when prey are scarce (Erlinge 1977; Debrot and Mermod 1983; King 1990); and
(6) predators re-invade cleared areas rapidly (Moller et al. 1992; Alterio 1996).
Changes in predator diet, movements, abundance and guilds will potentially add new and long-term threats to native wildlife communities. This critical point is largely ignored or underestimated by the IIA, and has serious implications for predator control efforts to ameliorate any unwanted impacts on native wildlife resulting from release of RCD. For example, costs of predator control and monitoring have been vastly under estimated in the IIA. Inadequate funding for predator control was raised by many of the public submissions (whether they were in favour or not of releasing RCD).
The importance of having adequately funded monitoring and mitigation plans in place is again well illustrated in our medical analogy on page 14. The dying patient may be much more willing to risk the potential remedy provided that full medical support was in place to mitigate any unwanted side-effects. Equally, the doctor may feel ethically more able to apply an uncertain treatment provided full support was available to minimise the symptoms if things went wrong.
To alleviate public concerns in this area, increased funding for predator control programmes and development of more effective predator control technologies are required. Existing predator control methods are unlikely to provide cost-effective widespread control in the long-term, so it is imperative that new cheap, safe and effective alternatives are developed and ready for use.
Minor omissions in the IIA concerning predator control include:
assessing the efficacy of trapping, ie. stoats are extremely difficult to trap in spring;
costs of controlling rodent populations following RCD;
risks of predator controls to non-target native animals ie. unwanted impacts on native raptors which are illegal to control;
potential unwanted impacts on native species by Little Owls.
The need for more consultation with Maori
Very little true consultation was achieved with Maori in the preparation of the IIA and therefore in the way their putative views were represented to all New Zealanders through the IIA. The submissions subsequently received from Maori groups are uniformly opposed to the release of RCD at present (refer to Taylor Baines 1996). However, relatively few submissions were received with an identifiable Maori representation, so the need for a more exhaustive consultation with Maori to assist the overall RCD decision is clearly signalled.
Is a Precise Risk Analysis feasible?
Some submitters and peer reviewers have called for a more formalised and precise risk assessment of the RCD issues. We concur that a more complete and thorough cost benefit analysis than provided by the IIA and its appendices would have been desirable. Nevertheless we believe that the potential ecological outcomes are extremely variable and many are unknowable before release. There is no immediate prospect of assigning precise probabilities to the various outcomes, partly because of a lack of detailed research of New Zealand's ecology, but partly because community ecology outcomes of this magnitude are too complex and inherently unknowable (Moller 1989, Yodzis 1988).
The need for urgency
There have been several claims for haste amongst protagonists for RCD release and there is concern amongst peer reviewers and public submissions that unreasonable haste brings ecological risks.
The wish for fast relief from rabbit pestilence and the associated expenses of conventional controls are understandable, especially in view of the currently low price for wool and high costs of farm mortgages. These are quite distinct reasons from the ecological and wildlife management concerns which have been claimed by the IIA as added imperatives for, in our view, undue haste.
The IIA asserted, and statements from the RCD Applicant Group in the media have repeated that immediate deliberate release is now highly desirable because of the risks of bringing in myxomatosis, Foot and Mouth Disease, other rabbit diseases, or sub-optimal strains of RCD. However, there is no vector for myxomatosis in New Zealand, Foot and Mouth Disease does not occur in Australia, and other rabbit diseases are not common in Australia (MAF's RCD Disease Risk Assessment, 1996). The RCD strain free-running in Australia is precisely that intended for import into New Zealand anyway, so any RCD imported from free-living rabbits in Australia will be equivalent to that brought in from a laboratory source. If we are to get an illegal or accidental introduction it is now likely to be from Australia rather than Europe, and from all the arguments above this is a very much safer source than elsewhere. Accordingly, we think there is less cause for urgency now that the virus is free-running in Australia rather than more cause as argued by the IIA, at least as far as the disease considerations above are concerned.
If detailed contingency plans are made now, there is no reason why rapid intervention with planned releases cannot take over soon after any accidental or illegal deliberate release. In this way several of the putative advantages of deliberate release cannot be subsequently captured. We hope that MAF and the NZ government develop these plans now and prepare empowering legislation to allow them to rapidly move into pro-active management after any unauthorised or accidental release.
There were repeated assertions in the IIA that planned releases allow several advantages on ecological grounds for maximum effectiveness and to minimise any unwanted environmental effects. Little concrete detail was provided to evaluate these claims and the behaviour, spread and impact of the virus is so unpredictable in New Zealand conditions that we doubt that several of these advantages can be realised. Modelling shows no evidence that any season is any better than another for initial impact. The behaviour of any introduction at the establishment phase does not predict final outcomes, and getting it started may sometimes be slow. Its rate of spread to sites of especially conservation importance is unknown. The pattern of outbreaks is very patchy and irruptions erratic in Spain (Moller 1996a), so there is potentially no real control on its impact and any unwanted effects once release starts in New Zealand. The IIA claims that deliberate releases could reduce attenuation are bogus. Finally it is the long-term equilibrium outcomes that will determine the overall cost/benefit outcomes from RCD release and these will be little affected by the initial release strategies.
Accordingly we challenge the scientific and practical basis of any call for extreme urgency in RCD release, while acknowledging the social and economic reasons for badly affected farmers may be judged by the Decision Maker to be well proven and urgently in need of relief.
The value of delay
Several submissions have urged a delay in the release decision until we know more about the Australian experience with RCD.
Recent media reports of dramatic gains for wildlife from RCD effects are a nonsense and should not be given any validity by the Decision Maker. There simply has not been enough time for RCD effects to be fully felt yet. The initial changes following rabbit reduction are likely to be very different from subsequent ones. Diet switching to increased use of native fauna has already been found in Australia (Moller 1996a). It will take several years (even decades for some species) before the experience of Australian wildlife is known.
We see the main values in a moderate (2 to 5 year delay) as allowing:
adequate time for detailed planning of the wildlife management and scientific aspects of the programme, thereby maximising potential opportunities from RCD and minimising risks.
time for a better risk assessment for invertebrates, reptiles and plants, the groups considered inadequately in the IIA.
time for a more accurate costing of the consequences of a decision to release and a decision not to release.
development of detailed plans for how the potential benefits of RCD can be "locked in" to secure long-term sustainable management of the High Country.
rapid research on better and cheaper predator poisoning methods to assist mitigation and minimise environmental risks
the public to assess funding sources, institutionalised structures to guide management of RCD after at and after its release, and a detailed environmental mitigation plan alongside the RCD release strategy (or to combine them in an integrated plan)
all stakeholders time to become involved at a foundation level to work through the problems together rather than simply having a release strategy imposed on them
DoC to minimise disruptions to their existing work and financial planning
better consultation with Maori
rebuilding trust from the general programme by instituting a more considered and fully consultative process for evaluating the decision
time for science providers to establish methods and baseline measures to guide the programme and determine any necessary mitigation methods
time for a more considered and unbiased IIA to be prepared and put before the public
time to develop a more robust mathematical model of the RCD impact on rabbits for New Zealand conditions
checks on whether RCD impacts on and persists in low density rabbit populations in Australia. This lesson has an enormous impact on the spatial extent of any potential environmental risks and unwanted impacts on rabbit populations within most of the rabbit's distributional range in New Zealand. The risk sensitivity is fundamentally changed according to this outcome.
a thorough evaluation of the Rabbit & Land Management Programme and its success or otherwise in fostering more sustainable land uses
development of a multi-disciplinary team to identify ecological knowledge that can be gained in a specified number of years with a specified number of resources to aid risk assessment prior to introduction.
While several of the above suggestions will help decision making, minimise risks and improve our ability to mitigate unwanted effects, it is important to recognise that several ecological uncertainties are bound to remain even if there is a delay. Decades of research will not resolve some of the questions about long-term ecological responses to RCD. Meanwhile the environmental, economic and social costs of not acting continue. Accordingly we urge that any delay in the release decision is used constructively for directed efforts to reduce risk rather than as an open ended warrant to research issues that will not resolve uncertainties.
If a decision is delayed, we urge that the exact criteria for the subsequent decision are spelled out now so that the farmers and the Applicant Group are left in no doubt about the subset of issues to be resolved for the final decision within a fixed time frame. If this is not done there is a danger that nothing has been gained from the great effort put in by all concerned so far. Identifying the exact decision criteria for the final decision now will also allow better focus to any subsequent debate and research. It will bring public confidence that the critical minimum criteria have been identified and will be resolved. It will give a concrete planning horizon for the most affected stakeholders, the high country farmers.
We are grateful to Rachel Keedwell for assistance in editing drafts of this report.
Alterio, N.J. (1996). Secondary poisoning of feral stoats (Mustela erminea), ferrets (Mustela furo) and cats (Felis catus) by the anticoagulant poison, brodifacoum. New Zealand Journal of Zoology 23:331-338.
Alterio, N. and Moller, H. (in press). The diet of stoats (Mustela erminea), feral house cats (Felis catus) and ferrets (Mustela furo) in grassland surrounding Yellow-eyed penguin (Megadyptes antipodes) breeding areas. Journal of Zoology (London).
Animal Health Board (1995). National Tb strategy: proposed National Pest Management Strategy for Bovine Tuberculosis. Management Plan.
Barlow, N.D. (1996). Modelling RCD in rabbits. Unpublished report to MAF Policy.
Brown, K.P. (in press). Predation at nests of New Zealand endemic passerines; implications for bird community restoration. Pacific Conservation Biology
Debrot, S. and Mermod, C. (1983). The spatial and temporal distribution pattern of the stoat (Mustela erminea L.). Ä Oecologia 59: 69-72.
Erlinge, (1977). Home range utilisation and movements of the stoat, Mustela erminea. Ä Wildlife Biology and Ethology, XIIIth Congress of Game Biologists, pp. 31-42.
Gibb, J.A. and Flux, J.E.C. (1983). Why New Zealand should not use Myxomatosis in rabbit control operations. Search 14: 41-43.
King, C. M. (1990). "Stoat." Pp 288-312 in The handbook of New Zealand Mammals. Ed. C.M. King. Oxford University Press, Auckland, New Zealand.
Moller, H. (1989). Towards constructive ecological engineering; the biological control of pests for the restoration of mainland habitats. New Zealand Ecological Society Occasional Publication No. 1, Dunedin.
Moller, H. (1996a). Evaluation of the Rabbit Calicivirus Disease Import Impact Assessment and Disease Risk Assessment. A report to the Chief Veterinary Officer, MAF, Wellington, New Zealand. Ecosystems Consultants Report No. 5.
Moller, H. (1996b). Evaluation of 1st Revision of the Rabbit Calicivirus Disease Import Impact Assessment. A report to the Chief Veterinary Officer, MAF, Wellington, New Zealand. Ecosystems Consultants Report No. 6.
Moller, H. (1996c). Lessons for invasion theory from social insects. Biological Conservation 78: 125-142.
Moller, H., McKinlay, B.M., Alterio, N. and Ratz, H. (1992). Control of mustelids and cats to protect Yellow-eyed penguins. In: B. M. F. D. Veitch, J. Innes and E. Murphy, (ed.), National Predator Management Workshop. Department of Conservation, Wellington,. Threatened Species Occasional Publication No. 1, pp. 54-55.
Moller, H., Townsend, C.R., Ragg, J.R. and Bannister, P. (1993). Invasions and environmental impacts of New Organisms; can they be predicted. University of Otago Environmental Policy and Management Report 1. Dunedin.
Moller, H., Showers, J. and Wright, M. (1996a). Sodium monofluroacetate (1080) poisoned jam baits laid for possums (Trichosurus vulpecula) also kill ferrets (Mustela furo). New Zealand Journal of Zoology 23: 135-141.
Moller, H., Norbury, G. and King, C.M. (1996b). Ecological and behavioural constraints to effective control of ferrets (Mustela furo). In: D. E. Wright, (ed.), Workshop on Ferrets as Vectors of Tuberculosis and threats to Conservation. Possum and Bovine Tuberculosis Control National Science Strategy Committee, Royal Society of New Zealand, Christchurch.
Molloy, J. and Davis, A. (1992). Setting priorities for the conservation of New Zealand's Plants and Animals. Department of Conservation, Wellington.
National Science Strategy Committee (1995). Annual Report from the Possum/Bovine Tuberculosis Control National Science Strategy Committee.
Norbury, G. and Heyward, R. (1996). The response of ferrets to rabbit control. In: Proceedings of a NSSC workshop on ferrets as vectors of tuberculosis and threats to conservation, The Royal Society of New Zealand, Wellington, 27-28 March 1996.
Norbury, G. and McGlinchey, A. (1996). The impact of rabbit control on predator sightings in the semi-arid high country of the South Island, New Zealand. Wildlife Research 23: 93-97.
Norbury, G. and Murphy, E. (1996). Understanding the implications of rabbit calicivirus disease (RCD) for predator-prey interactions in New Zealand: a review. Unpublished report, Landcare Research, Alexandra.
Pierce, R. (1987). Predators in the Mackenzie Basin: their diet, population dynamics, and impact on birds in relation to the abundance and availability of their main prey (rabbits). Unpublished report, Wildlife Service Wellington.
Ragg, J.R., Moller, H. and Waldrup, K.A. (1995a). The prevalence of bovine tuberculosis (Mycobacterium bovis) infections in feral populations of cats (Felis catus), ferrets (Mustela furo) and stoats (Mustela erminea) in Otago and Southland, New Zealand. New Zealand Veterinary Journal 43: 333-337.
Ragg, J.R., Waldrup, K.A. and Moller, H. (1995b). The distribution of gross lesions of tuberculosis caused by Mycobacterium bovis in feral ferrets from Otago, New Zealand. New Zealand Veterinary Journal 43: 338-341.
Ragg, J.R., Moller, H., Waldrup, K.A. and Mackintosh, C.M. (1995c). Ferrets (Mustela furo) and bovine tuberculosis (Mycobacterium bovis); the need for a multi-species approach. In: F. Griffin and G. M. de Lisle, (eds.), Tuberculosis in Wildlife and Domestic Animals. Otago Conference Series No. 3, University of Otago Press, Dunedin. pp. 290-293.
RCD Applicant Group (1996). Import Impact Assessment and Application to the Director General of Agriculture to approve the importation of Rabbit Calicivirus under the Animals Act 1967 and to issue and Import Health Standard under the Biosecurity Act 1993. Ministry of Agriculture, Wellington, New Zealand.
Rebergen, A.L; Keedwell, R. and Moller, H. (submitted). Breeding Success and Predation at Nests of Banded Dotterel (Charadrius Bicinctus) on Braided River Beds in the Central South Island, New Zealand. New Zealand Journal of Ecology.
Smith, G.P. (1994). Ferret (Mustela furo) diet in relation to prey abundance and Tb infection in pastoral habitats of Otago. University of Otago Wildlife Management Report No. 57. Dunedin.
Sumption, K.J. and Flowerdew, J.R. (1985). The ecological effects of the decline in rabbits (Oryctolagus cuniculus L.) due to myxomatosis. Mammal Review 15: 151-186.
Taylor and Baines Associates (1996). Analysis of Submission on the Importation Impact Assessment for the RCD Virus. A report prepared for the Chief Veterinary Officer, MAF Regulatory Authority, Wellington, New Zealand.
Walker, R., Reid, B.T. and Crews, K.B. (1993). Bovine Tuberculosis in predators in the Mackenzie Basin. Surveillance 20: 11-14.
Yodzis, P. (1988). The indeterminacy of ecological interactions as perceived through perturbation experiments. Ecology 69: 508-515.
Contact for Enquiries
Manager, Strategic Science Team
MAF Biosecurity New Zealand
PO Box 2526
Wellington
NEW ZEALAND
Phone: +64 4 894 0115
Fax: +64 4 894 0731
Contact this person
