Page 2 of Dr Henrik Moller's responses to specific points raised by Colin O'Donnell

My comments refer to numbered locations marked on Appendix 4 (attached - see rhs column), continuing at Section 3 before returning to the Summary & Recommendations (Section 1) at the end.

3.0 Mitigation of impacts on foraging versus breeding habitats.

4.         More research in TWC red beech forests would indeed be an added safeguard that the general patterns observed elsewhere do indeed operate there. However the study proposed (" a study of foraging preferences ") could never by itself answer the critical applied questions - whether or not foods limit the wildlife problems, and certainly not whether the proposed logging will significantly alter population levels. There are also serious theoretical problems with inferring ecological needs from preference studies (Gray & Craig 1991). Detailed study of bird and bat feeding and nesting preferences will be a valuable first step, but I caution that by themselves this information will not be sufficient to evaluate putative logging impacts.

3.0 Mitigation of impacts of logging versus predator control.

5.         This statement is true; there have been repeated demonstrations of lowered bird abundance in response to past logging methods. However, in the context of the new style of "natural forest management" proposed, the DoC Critique's statement misleads because (i) the types of habitat changes proposed by TWC's plans are fundamentally different from the type of changes imposed by logging in the past (this is partly acknowledges by the DoC Critique at the beginning of the last paragraph on the next page), (ii) even in such cases there have been no studies to partition the relative importance of habitat modifications and predation or competition in declines.

However, lack of evidence of impacts for the new style of tree extraction is not proof that they will not occur. A rigorous adaptive management approach is needed to evaluate the letter in the long term, and the short-term comparative study of bird and bat use of artificially created, naturally occurring and random forest locations ( as proposed by Ecosystems Consultants for this year) can give an immediate indication of this potential risk.

6.         The threatened status of kaka is well researched, and their low abundance and infrequent success in breeding in the South Island are major causes for concern. Historically and on a very large spatial scale, habitat removal undoubtedly caused declines in the distribution of kaka. However clearfelling or over-cutting is not proposed by TWC, so it is a non sequitur to link these historical patterns to predictions of what might happen now.

7.         A Landcare Research team led by Dr Peter Wilson has tested the hypothesis (first suggested by Beggs and Wilson 1991) that reduction of honeydew by introduced social wasps reduces kaka breeding rate and/or success through competition. A supplementary feeding experiment has now finished that discounted the competition hypothesis (Wilson et al. 1998). Instead predation of chicks and sitting adult females at nests by stoats is now considered to be the key threat to South Island kaka. This was identified as the probable key determinant of extinction probability in a Population Viability Analysis (PVA) done in 1993 (Seal et al. 1993). I contacted Dr Ron Moorhouse, DoC's science and research scientist studying South Island kaka (31 August 1998) to solicit his opinion of the key threat to the bird. He holds the view that predation of adult female kaka is indeed the key threat and has suppressed the population well below the level where nest hole availability has any influence on the abundance of kaka. In support of this view he cited the Landcare research team as finding virtually no re-use of the same holes for nesting in their 8 year research programme. I also contacted Dr Wilson (5 October 1998) to check that no recent contrary evidence, nor significant doubts about the interpretation of their experiment emerged in the anonymous peer review of their submitted paper. He confirmed that no such challenges emerged, and reiterated his belief that predation remains the urgent and most important threat to the subspecies. He also mentioned that an updated PVA analysis has recently been prepared that reconfirms the overwhelming importance of predation of adult females in depressing kaka numbers. Colin O'Donnell (in litt., 5 October 1998) concurs with this overall conclusion that predation is the key variable, but points out that he observed two nests being re-used in the Windbag area (South Westland) after 15 years.

O'Donnell & Rasch (1991) referred mainly to competition with possums for mistletoe. The supplementary feeding experiment above also refutes this hypothesis.

8.         This statement is based on a general "ecotone" idea promulgated overseas. The Eglinton predator research team could find no evidence that stoats concentrated their activity near to roads in 1992/93, but in 1991/92 traps along roads actually caught significantly fewer stoats (Dilks et al. 1996). The only direct test of this "habitat edge" hypothesis in New Zealand so far failed to find any change in probability of predation of yellow-eyed penguin chicks on edges c.f. middle of forest breeding areas (Ratz 1997). The habitat patch sizes may have been too small in that study to detect such and (sic) effect, or it may be that New Zealand biota are so vulnerable to predation that any slight concentration of predators on habitat edges is irrelevant (since nestlings in centres are also nearly always eliminated). Nevertheless TWC commissioned an application to FRST's Public Good Science Fund for research into this potential impact of roads. Unfortunately insufficient funds were available for FRST to support the research, so TWC are now seeking funds elsewhere to check out potential unwanted effects. Roads may also benefit biodiversity by allowing cost effective predator controls (Moller & Alterio 1998).
In any event, the new style of helicopter logging proposed by TWC reduces requirements for roads by about 15-fold (Kit Richards, pers. comm.).