New Zealand Threat Classification System: reply to Breen and Middleton

This article is related to:

Problems with implementation of the New Zealand Threat Classification System, as illustrated by Baker et al. (2010)

We thank the Editor of the New Zealand Journal of Marine and Freshwater Research (NZJMFR) for the opportunity to reply to the letter ‘Problems with implementation of the New Zealand Threat Classification System, as illustrated by Baker et al. (2010)’ by Breen and Middleton.

The letter criticises the procedural framework for the New Zealand Threat Classification System (NZTCS). None of the expert panel members (i.e. the authors of Baker et al. 2010) were authors of this framework. Our role was to apply the listing framework and criteria (Townsend et al. 2008) to the 56 species or subspecies of marine mammals in New Zealand waters, based on the best available information. The list facilitator's role is to ensure that standards are maintained in the quality of the data gathered for listings and that the threat system is applied consistently and without bias by each expert panel. Rod Hitchmough, as a scientist for the Department of Conservation (DOC), has carried out the list facilitator role for all three published cycles (Hitchmough 2002, 2013; Hitchmough et al. 2007) and continues to do so during the current 2012–2014 listing cycle.

The NZTCS, described in detail by Townsend et al. (2008), is designed to allow comparable assessments across all animals, plants and fungi in terrestrial, freshwater and marine environments, on a rolling 3-yearly cycle. During the last NZTCS review cycle, the risk of extiction was assessed for more than 12,000 taxa (summarised by Hitchmough 2013). The criteria and guidelines for such a large-scale exercise must be simple and general. Arguably they could be tailored for improved application to any single taxonomic group, but that would erode their comparability across groups. Reports of the outcomes for such a large number of taxa can only include limited summaries of the background information presented for each individual taxon and record the judgements of the panel in regards to the listing criteria. The frequency of the review of listings (usually every 3 years) and the opportunity for feedback both via correspondence and via a formal submission process advertised on the DOC website provides ample opportunity for errors to be addressed and quickly corrected. The Townsend et al. (2008) manual provides full definitions of the listing criteria and the qualifiers available for each taxon. This information is summarised in the published lists as well. Appendix 1 of Townsend et al. (2008) presents more detailed definitions of the terminology used.

The letter questions the notification process for finalised lists on the basis of the use of the words ‘propose that the endemic [New Zealand] NZ sea lion be uplisted’ (original emphasis) by Baker et al. (2010, p. 107). This ignores our very specific statement ‘This new list updates and supersedes all previous NZ marine mammal threat classification listings for NZ (Molloy & Davis 1992, 1994; Hitchmough 2002; Hitchmough et al. 2007) and remains valid from the date of publication until the next list is published’ Baker et al. (2010, pp. 102–103).

The letter advocates that panels should use ‘appropriate statistical techniques when extrapolating population trends, should not allow subsets of the available data to be selected arbitrarily, and should recognise that few populations will show a truly linear trend in abundance over three generations’. We agree, and this is already the NZTCS practice. All available trend information is considered for each taxon. We are very aware of the problem of distinguishing long-term trends from short-term fluctuations. For the NZ sea lion, worrying information about reduced pup production was already available to the expert panels which met in 2001 and 2005, but the sea lion was listed as Range Restricted (Molloy et al. 2002 criteria) not as declining, because it was acknowledged that these might have been short-term fluctuations. However, by the time the expert panel met in 2009, the decline had continued for the estimated length of a sea lion generation, as defined in the NZTCS (see below). We do not consider it unreasonable to infer then that the trend was ongoing.

Breen and Middleton's criticisms of the specific recommendations of Baker et al. (2010) focus largely on the revised listing of New Zealand sea lion to Nationally Critical from the previous listing as Range Restricted. This species is endemic to New Zealand and has faced a number of known threats, including recent disease epizootics and fisheries-related mortality (Robertson & Chilvers 2011). We consider that most of Breen and Middleton's criticisms are unfounded but, given space limitations, address only the following as most relevant to the listing decision:

1. ‘To review this classification, one needs to: see the data showing an ongoing trend; see a methodology that predicts rate of decline; and know the generation time. None of this was described.’

   As discussed above, Baker et al. (2010) was intended to outline the NZTCS process and to give an overview of the information that led to each specific threat classification. It was not a population assessment of the NZ sea lion. The article was peer reviewed through submission to the NZJMFR and the information considered by the expert panel was publicly available, as Breen and Middleton implicitly acknowledge by reference to the notes of the panel's deliberations obtained through an Official Information Request.

2. ‘The 2010 pup count, available after the classification but months before publication, showed a 21% increase from the previous year.’

   The time frame of the NZTCS guidelines and the meeting of the expert panel should not be confused with the much longer timeline of a peer-reviewed scientific publication. The panel met in May 2009 and the manuscript was submitted for publication in early December 2009. At that time, the latest available information was the pup counts from 2009.

3. ‘Baker et al. (2010) did not discuss generation time and did not mention that the Auckland Islands are only a part of the breeding population of this species: only 72% of pups were born there in 2009.’

   The generation time used by the expert panel is defined and accepted both by the NZTCS and by the International Union for Conservation of Nature (IUCN) in its widely cited Red Book listing (http://www.iucnredlist.org/details/17026/0). The IUCN states: ‘The mean age of reproduction of female New Zealand Sea Lions is 10.75 yrs’ and, consequently, assumes that 30 years represents three generations. We clearly stated that the NZ sea lion ‘has shown a 50% decline in pup production at the main breeding area, the Auckland Islands (Enderby, Dundas and Figure of Eight Islands), in the last 11 years’. This is referenced to Chilvers et al. (2007) and Chilvers (2009), which further outlines the proportion of breeding occurring at all known breeding sites.

4. ‘They presented no information to support the Range Restricted qualifier they applied.’ The definition of Range Restricted is: ‘Taxa confined to … geographic areas of less than 1000 km² (100,000 ha); this is assessed by taking into account the area of occupied habitat of all sub-populations (and summing the areas of habitat if there is more than one sub-population) …’ (Townsend et al. 2008, p. 29). Appendix 1 of Townsend et al. (2008, p. 32) further defines area of occupancy as, ‘the area occupied by the taxon, taking into account the fact that a taxon may not occupy all areas throughout its range because of unsuitable habitat. The smallest area essential at any stage in the life cycle of the taxon will be used (e.g. colonial nesting sites)’ (our emphasis). The breeding colonies of the NZ sea lion clearly occupy less than 1000 km².

5. ‘The expert panel chose to ignore alternative predictors of trends in New Zealand (NZ) sea lions, such as Breen and Kim (2006) and unpublished (but available) model results presented to Ministry of Fisheries’ working groups. Such reports had been specifically drawn to the attention of the expert panel.’

   The expert panel considered the information mentioned by Breen and Middleton and a summary letter, dated 5 May 2009, submitted by Middleton as Chief Scientist for the New Zealand Seafood Industry Council Ltd. The panel also considered that, given the dates of the publications (Breen et al. 2003; Breen & Kim 2006), these alternative predictors of trends did not include the recent evidence for the decline in pup counts. Instead, these earlier publications concentrated on modelling the potential causes of declines in NZ sea lions, particularly the role of fisheries-related mortality. It is not the mandate of the NZTCS process or the expert panel's responsibility to assess the specific causes of a decline (e.g. Robertson & Chilvers 2011) but, rather, to assess whether or not a decline is occurring and, if so, at what rate.

6. ‘Baker et al. (2010) stated that the sea lion population was fewer than 3000 mature animals. This bald statement omits some important calculations made by Chilvers (unpubl. data): she calculated females only, based on pup counts and the breeding percentage, and reduced the result by 20% to account for non-breeding. As written, the paper was misleading.’

   As outlined in Chilvers (unpubl. data), to which Breen and Middleton refer, this is the estimated number of breeding males and females, taking into account the polygamous mating system and the reproductive cycle of females. This is consistent with the definition of ‘mature individuals’ in Townsend et al. (2008, Appendix 1). The estimate of population size did not affect the placement of the sea lion into the Nationally Critical category–the listing was based on rate of population decline alone, using pup counts as the index for the trend.

Finally, the letter of Breen and Middleton was submitted 4 years after the convening of the expert panel and 3 years after Baker et al. (2010) was published. In accordance with the guidelines of the NZTCS, the listing is intended to be reviewed every 3 years to take into account new information and analyses. In fact, the expert panel was reconvened on 8 May of this year (2013) and a new listing should be available later this year.