Squid Fishery Management Company Limited v Minister of Fisheries
[2004] NZCA 132
•7 April 2004
IN THE COURT OF APPEAL OF NEW ZEALAND
CA39/04
BETWEENSQUID FISHERY MANAGEMENT COMPANY LIMITED
AppellantANDMINISTER OF FISHERIES
First RespondentANDCHIEF EXECUTIVE OF MINISTRY OF FISHERIES
Second Respondent
Hearing:5 April 2004
Coram:Hammond J
William Young J
O'Regan JAppearances: B A Scott and G T Carter for Appellant
U R Jagose and R E Schmidt for Respondents
Judgment:7 April 2004
Reasons:13 July 2004
REASONS (DELIVERED BY WILLIAM YOUNG J) FOR THE JUDGMENT OF THE COURT
Table of Contents Paragraph Number Introduction [1] The statutory scheme [4] Background [11] The grounds upon which the Operational Plan was challenged [67] The history of the litigation [69] Overview [73] The reasons why we set-aside the 2003-2004 Operational Plan THE LEGAL FRAMEWORK [74] IRRELEVANCIES [81] GENERAL OBSERVATIONS [85] DEFECTS IN THE DECISION-MAKING PROCESS [88] THE LEGAL BASIS OF OUR DECISION [103] THE ARGUMENTS OF THE APPELLANT [105] THE ARGUMENTS OF THE RESPONDENTS [107] THE JUDGMENT OF FRANCE J [116] Introduction
[1] On 5 April 2004 we heard an appeal from the judgment of France J dismissing the appellant’s application for judicial review of the Minister’s determination (given effect to by what is known as the 2003-2004 Operational Plan) of the maximum allowable limit on fishing-related mortality (MALFiRM) for sea lions in the SQU6T fishery. Given the nature of the case it was necessary for us to give our decision in short order and we did this on 6 April.
[2] The result was that we allowed the appeal and set-aside the 2003‑2004 Operational Plan against the undertaking of the appellant to adhere to a MALFiRM of 124 for the then current fishing season. There were orders for costs in favour of the appellant.
[3] We said that we would give our reasons in due course, and we do so now.
The statutory scheme
[4] It is appropriate to start with the relevant legislative scheme.
[5] The Minister’s power to impose MALFiRMs arises under s15(2) of the Fisheries Act 1996, which provides:
In the absence of a population management plan, the Minister may, after consultation with the Minister of Conservation, take such measures as he or she considers are necessary to avoid, remedy, or mitigate the effect of fishing-related mortality on any protected species, and such measures may include setting a limit on fishing-related mortality.
[6] This case concerns fishing related mortality of the New Zealand (or Hooker’s) sea lion associated with squid fishing activity. Section 15(2) applies because sea lions are a protected species (being marine mammals as defined in s2(1) of the Marine Mammals Protection Act 1978) and there is no population management plan under the Marine Mammals Protection Act.
[7] For the purposes of s15(2) of the Act, “fishing related mortality” refers only to the death of sea lions in the course of fishing activity. It does not extend to impacts on the sea lion population associated with, for instance, competition for squid. Further, what is important is the impact of fishing on the sea lion population as a whole and the section does not provide for measures aimed simply at eliminating or reducing individual deaths.
[8] Section 15(5) provides:
The Minister may, by notice in the Gazette, prohibit all or any fishing or fishing methods in an area either—
…
(b)under subsection (2) of this section, for the purpose of ensuring that any limit on fishing-related mortality is not exceeded.
[9] We should also set out ss5, 8, 9 and 10 of the Act:
5 Application of international obligations and Treaty of Waitangi (Fisheries Claims) Settlement Act 1992
This Act shall be interpreted, and all persons exercising or performing functions, duties, or powers conferred or imposed by or under it shall act, in a manner consistent with—
(a) New Zealand’s international obligations relating to fishing; and
(b) The provisions of the Treaty of Waitangi (Fisheries Claims) Settlement Act 1992.
8 Purpose
(1) The purpose of this Act is to provide for the utilisation of fisheries resources while ensuring sustainability.
(2) In this Act—
“Ensuring sustainability” means—
(a) Maintaining the potential of fisheries resources to meet the reasonably foreseeable needs of future generations; and
(b) Avoiding, remedying, or mitigating any adverse effects of fishing on the aquatic environment:
“Utilisation” means conserving, using, enhancing, and developing fisheries resources to enable people to provide for their social, economic, and cultural wellbeing.
9 Environmental principles
All persons exercising or performing functions, duties, or powers under this Act, in relation to the utilisation of fisheries resources or ensuring sustainability, shall take into account the following environmental principles:
(a) Associated or dependent species should be maintained above a level that ensures their long-term viability:
(b) Biological diversity of the aquatic environment should be maintained:
(c) Habitat of particular significance for fisheries management should be protected.
10 Information principles
All persons exercising or performing functions, duties, or powers under this Act, in relation to the utilisation of fisheries resources or ensuring sustainability, shall take into account the following information principles:
(a) Decisions should be based on the best available information:
(b) Decision makers should consider any uncertainty in the information available in any case:
(c) Decision makers should be cautious when information is uncertain, unreliable, or inadequate:
(d) The absence of, or any uncertainty in, any information should not be used as a reason for postponing or failing to take any measure to achieve the purpose of this Act.
[10] Section 5(b) makes relevant the United Nations Convention on the Law of the Sea (“UNCLOS”) and in particular arts 61, 62 and 65. Article 61 requires coastal states to “ensure” that the maintenance of living resources in its exclusive economic zone is “not endangered by over-exploitation”. The same article addresses the maintenance or restoration of populations of species above levels at which their reproduction may become seriously threatened. Article 62 provides that coastal states:
shall promote the objective of optimum utilization of the living resources in the exclusive economic zone without prejudice to Article 61.
Article 65 provides generally for the rights of, inter alia, coastal states to prohibit, limit or regulate the exploitation of marine mammals.
Background
[11] This case concerns the SQU6T fishery. This encompasses the Auckland Islands and Campbell Island. The fleet operating in this fishery targets squid on the southern and eastern edges of the Snares shelf and on the Auckland Islands shelf. The squid are harvested by trawling.
[12] The season can run from January to June. In recent years, the fleet has voluntarily delayed the start of the season until 1 February, and the season usually finishes in April.
[13] The appellant does not itself fish or own quota but it has been accepted that it is entitled to represent those who do (who are its shareholders).
[14] Sea lions feed on squid. Probably for this reason they are susceptible to being caught up in trawl nets.
[15] The sea lion population has breeding sites (or rookeries) on the Auckland Islands and the Campbell Island. The rookeries on the Auckland Islands are thought to account for around 90% of the total sea lion population of around 13,000 animals. There are several much smaller sea lion colonies on the South Island and Stewart Island.
[16] There is a 12 nautical mile exclusion zone around the Auckland Islands within which fishing is not permitted. This provides a substantial measure of protection for sea lions. But the foraging range of sea lions is such that the exclusion zone does not provide complete protection.
[17] Sea lions have been protected from exploitation since 1894. They are currently categorised as a threatened species under s2(3) of the Marine Mammals Act.
[18] The current Auckland Islands based sea lion population is likely to be at or near its carrying capacity (although this is not certain). The threatened status of sea lions is not related to the size of the current population, but, rather, the limited number of breeding sites. The worry is that an epidemic or perhaps an oil spill at one or more of the current breeding sites could have catastrophic effects.
[19] These concerns are real enough. In January 1998, there was what was described as a “mass mortality event” in the Auckland Islands with pup mortalities estimated at 60% and adult mortality at 20%. In 2002, approximately one third of the pups at one of the rookeries died as a result of a bacterial epidemic.
[20] From a conservation view point the ideal would be for sea lions to establish other breeding sites but it is not clear what, if anything, can be done to bring this about. One theory is that if sea lion numbers on the Auckland Islands and Campbell Island are maintained at or around carrying capacity, associated population pressure will force sea lions to colonise other breeding sites.
[21] In very broad terms, the underlying strategy of the Ministry has been to retain existing breeding locations, establish at least two other breeding sites and allow population size to increase with the aim of moving the species towards “non‑threatened status” over a 20 year period. This strategy seems to us to be consistent with both the Marine Mammals Protection Act and the Fisheries Act and is not challenged by the appellant, at least in the present proceedings.
[22] Population estimates are based largely on pup counts at the different breeding sites. It is not practicable to count adult sea lions given their foraging activities.
[23] Since 1992, the Ministry, in consultation with the Department of Conservation (“DoC”), has imposed operational restrictions on the SQU6T fishery intended to limit sea lion deaths. This has involved the fixing of an annual MALFiRM. A MALFiRM is a scientifically calculated estimate of the total catch of sea lions that is considered biologically acceptable within a particular fishing year. As indicated in para [1] above, the acronym stands for “maximum allowable limit on fishing-related mortality”.
[24] Until 2003/04, the Minister calculated MALFiRMs using a formula that estimated the potential biological removals (“PBR”) that could safely occur. The formula used was a generic one, applicable to any marine mammal population where certain conditions exist. This generic formula had been developed by Dr Paul Wade and reflected the requirements of American legislation. For the purposes of fixing MALFiRMs for the SQU6T fishery, there were limited changes to the formula but it remains very cautious in its assumptions. For instance, one of the assumptions is that the relevant marine mammal population is severely depleted. In the documents before us this formula was usually referred to as either the “PBR rule” or the “Wade rule”.
[25] The MALFiRM has been recalculated each year to take into account new population data (mainly annual pup count survey information).
[26] When the MALFiRM is exceeded, the fishery is closed. So there has to be a method of determining or estimating sea lion mortality associated with squid fishing.
[27] One method of determining whether the number of sea lion mortalities in a particular fishing year has exceeded the MALFiRM is the use of a “strike rate”, that is an assessment of the average number of sea lions that will be caught in a single deployment of the trawl net or tow. Other methods involve the use of observers on all or a proportion (preferably at least 20%) of the vessels fishing. Because of inherent difficulties in collating information, there can be some overshooting of the MALFiRM before closure of the fishery can be implemented.
[28] Where a strike rate applies, the effect of this methodology is that for each fishing year, dividing the MALFiRM by the strike rate will give the maximum number of tows able to be conducted before the MALFiRM is exceeded and the fishery closed.
[29] The use of sea lion exclusion devices (SLEDs) is a complicating factor. These are intended to operate so as to direct live animals out of trawl nets. They can also, however, direct out of trawl nets the bodies of dead animals and animals which are so severely injured that their long term survival is doubtful. A cover net is placed over the escape hatch which is intended to ensure that only vital sea lions can escape. Extensive research has been carried out with a view to determining the extent to which the use of SLEDs reduces deaths amongst sea lions who are caught in trawl nets. Until the 2004 season, no discount had been allowed for the use of SLEDs. This was on the basis that there was insufficient evidence to show that sea lions which are ejected are able to survive the process. Given that SLEDs eject dead animals, their unrestricted use means that observers are not be able to identify the number of sea lions killed by trawling operations.
[30] Another complicating factor is the extent to which the Campbell Island sea lion population is relevant. We will revert to this point later.
[31] For the 2002-2003 season a MALFiRM of 70 was struck. During this season, the fleet was required to operate with the cover nets on SLEDs closed. This meant that sea lions ejected through the SLEDs would nonetheless drown. The purpose of this requirement was to allow post-mortem examination of the dead animals with a view to ascertaining what percentage of sea lions would have been viable upon ejection through the SLEDs. This requirement was opposed by the industry and compliance with it was not universal. The season was closed on 29 March 2003 by which point the estimated sea lion mortality had reached 79 animals.
[32] The result was litigation in the High Court which resulted in Ronald Young J setting-aside the closure on an interim basis (which, given timing issues, was effectively a substantive determination of the case), see Squid Fishery Management Co Ltd v Minister of Fisheries (CP 20/03, High Court, Wellington Registry, 11 April 2003). The High Court decision turned on the methods of calculating estimated mortality and, for this reason and also because of its nominally interim nature, is of little direct relevance to the issues which we must address. It should be noted, however that the Judge took the view that sea lions which drowned in the closed cover nets should not be counted against the MALFiRM, a conclusion which was not in dispute before us and upon which we express no opinion.
[33] The whole exercise is complex and uncertain. The following table, which we have taken from a paper produced by the Ministry, indicates the many variabilities:
Fishing year
SQU6T TACC (t)
SQU6T catch (t)
Sea lion MALFiRM
Estimated mortalities
Closure date
1987-88
32,333
7,021
-
33
-
1988-89
35,933
33,462
-
141
-
1989-90
42,118
19,859
-
117
-
1990-91
30,190
10,658
-
21
-
1991-92
30,190
10,861
32
82
-
1992-93
30,369
1,551
63
17
-
1993-94
30,369
34,534
63
32
-
1994-95
30,369
30,683
69
109
-
1995-96
30,369
14,041
73
101
4 May
1996-97
30,369
19,843
79
123
28 Mar
1997-98
32,369
7,344
63
62
27 Mar
1998-99
32,369
950
64
14
-
1999-00
32,369
6,241
65
71
8 Mar
2000-01
32,369
3,254
75
67
- a
2001-02
32,369
11,502
79
84
13 Apr
2002-03
32,369
6,847
70
79
-b
a The fishery was not officially closed in 2000/01. Industry voluntarily withdrew the majority of vessels on 7 March 2001. Some observed vessels with closed cover nets remained in SQU6T for a short period in an effort to obtain SLED performance data.
b Under the Operational Plan the SQU6T fishery was closed on 29 March 2003 when the MALFiRM count reached 79 sea lions. A High Court Ruling in April 2003 allowed for continued fishing in SQU6T and established a separate procedure for estimating sea lion mortalities resulting in the 39 mortalities indicated. This estimate (39) is derived from a different procedure from that set out in 2002-03 Operational Plan. Fishers had voluntarily withdrawn from SQU6T as at the end of June.
[34] Work to develop a model specific to the New Zealand sea lion population has been underway since 1996. This work culminated in 2003 with the completion of what is now known as the Breen Kim model and which followed on from three earlier models, two of which had been published internationally with international peer review. All parties accepted that the Breen Kim model was the best available information upon which the Minister should base his decisions.
[35] For the purposes of the application of the Wade rule up until the 2002‑2003 season, the Campbell Island population was taken into account. The Breen Kim model uses only the Auckland Islands sea lion data. This is because the extent and quality of information about the Campbell Island sea lion population is too limited for use in the Breen Kim model. The exclusion of the Campbell Island sea population means, according to Dr Paul Breen, one of the co-authors of the Breen Kim model, that results produced by the new model tend to be conservative. On the other hand, there is a paucity of evidence as to the extent to which sea lions from Campbell Island are killed by the relevant squid fishing activities. If it is the case that Campbell Island sea lions are not at risk, then leaving them out of the exercise is appropriate.
[36] Work preliminary to the 2003-2004 Operational Plan was carried out under the auspices of the Aquatic Environment Working Group (“AEWG”). The AEWG is convened and chaired by the Ministry and comprises Ministry, Department of Conservation, industry and environmental representatives.
[37] For the purposes of the development of the 2003-2004 Operational Plan, the AEWG agreed to test a number of alternate by-catch rules using the Breen Kim model as against overall conservation management objectives which were provided by DoC. These objectives were to ensure that the sea lion population remains above 90% of its carrying capacity (K) or else remains above 90% of the level it would obtain in the absence of fishery by-catch, 90% of the time in 20 and 100 year runs.
[38] Those objectives were reflected in three criteria that the AEWG developed to test the alternate by-catch rules:
1.Crit100. A threshold of 450,000 of the set of 500,000 runs was required to satisfy the criterion mentioned in para [37] above - the number of mature animals in the population remains above 90% of K or above 90% of the numbers that would have occurred with no fishing.
2.Crit20. A threshold of 90,000 of the set of 100,000 runs was required to satisfy the same criterion.
3.Nmat/K. Whether the mean number of mature animals in the population are 90% of K, averaged over the second 50 years of all 5,000 model runs.
[39] Potential rules were proposed by different stakeholder groups and from there a set of rules to be modelled was selected by the AEWG. Those alternate rules were:
1.Rule 0 – no fishing;
2.Rule 1 – unconstrained fishing;
3.Rule 3 – three in-substance variants of the Wade rule:
(a)Rule 310 (which approximates the then current MALFiRM rule but applied only in relation to the Auckland Islands sea lion population);
(b)Rule 305 (half the current MALFiRM rule); and
(c)Rule 320 (double the current MALFiRM rule);
4.Rule 4 – the adaptive rule (by which the rule would vary depending on the level of sea lion population, so that at low population levels, half the current MALFiRM would be allowed, at intermediate levels the current MALFiRM would be allowed, and at high population levels twice the current MALFiRM would be allowed).
[40] Rule 1 assumed unconstrained fishing involving a level of effort assessed by reference to actual experience in the relevant fishery and this was treated as being 2,871 tows per season. This is a somewhat artificial concept of unconstrained fishing. Rule 310 is an analogue of the Wade rule. Rules 305 and 320 are variants on rule 310. Rule 320, however, can be seen as the Wade rule altered so as not to assume the necessity for recovery of a severely depleted population but rather maintenance of a population at or near carrying capacity.
[41] Dr Breen was instructed to assess the performance of these rules against the agreed criteria using the Breen Kim model. In a report of 30 June 2003 he noted that all rules, save for unconstrained fishing, met all of the criteria and, in fact, unconstrained fishing only narrowly failed to meet one of the criteria (the crit100, with a score of 89.5% against the required 90%). The average annual by-catches were calculated as being 99 under rule 1, 31 under rule 305, 53 under rule 310, 76 under rule 320 and 75 under rule 4.
[42] The Ministry issued its initial position paper (IPP) on 15 August 2003.
[43] The IPP recommended that the Breen Kim model be applied in the future with “little or no modification”, but subject to further refinement of input data. It also noted that the model only applied to the sea lion population at the Auckland Islands as no allowance had been made for the Campbell Island sea lions. The IPP then set out the rules tested using the model, the criteria against which the rules were evaluated and the results of the modelling.
[44] The IPP noted that “unconstrained” fishing only “marginally” failed to meet the crit100 criterion and that it met the primary management objective. The paper went on:
This indicates that even when the fishery is unconstrained (at current fishing effort levels), the population of sea lions remains at or increases through time to attain over 90% K, with high certainty.
Nonetheless, rule 1 (which permitted unconstrained fishing) was still put to one side (given that it did not meet the crit100 criterion) and so too were rules 0 (no fishing) and 305 (half the current MALFiRM). Of rules 310 and 320 and 4, the IPP observed:
MFish feels that the remaining three rules are reasonable options for consideration in setting the MALFiRM, presenting a range of perspectives on acceptable sea lion mortalities, utilization of the squid fish stock, and the flexibility of the decision rule to variation in the pup count.
[45] The IPP assessed the likely impact of the various possible decision rules on utilisation. Rule 310 was estimated to produce 31.5% lost fishing effort and 53% of seasons closed as against 11.4% lost fishing effort and 23% of seasons closed for rule 320 and 12.2% lost fishing effort and 24% of seasons closed for rule 4.
[46] The recommendation was for rule 320 producing a MALFiRM of 124 sea lions. The IPP also proposed two options, either a predetermined strike rate (of 5.3%) or the use of dedicated MALFiRM vessels using nets without operational SLEDs to permit monitoring of a minimum of 20% of all tows. In both cases there was to be a 20% discount for the use of SLEDs. The effect of this discount would be to increase what would otherwise be the maximum number of tows.
[47] The IPP contained the following table:
Table 3 Rule performance against key indicators, from the base case projections from Breen and Kim, 2003.
Performance Indices Fishing Rules
0
1
310
320
4
Crit20 a
N/a
97,781
100,000
99,989
99,997
Crit100 b
N/a
447,570
499,052
487,109
489,846
Nmat/K c
98.2%
91.7%
95%
93.4%
93.5%
Lost fishing effort d
100%
0%
31.5%
11.4%
12.2%
Seasons closed
100%
0%
52%
23%
24%
Maximum bycatch (100 yr runs)
0
545
77
151
169
Mean annual bycatch
0
99
53
76
75
a pass level for this index is 95,000 out of 100,000 projection-run years
b pass level for this index is 450,000 out of 500,000 projection-run years
c pass level for this index is 90% of K
d based on average annual fishing effort (2,871 tows) conducted during the years 1988-2003[48] Submissions were then received on the IPP from stakeholders, including the appellant.
[49] The appellant’s position was that there was no scientific basis for a MALFiRM and that unconstrained fishing should be permitted. If a MALFiRM was to be struck, a 5.3% strike rate was supported but with a 40% allowance for the use of SLEDs. The New Zealand Seafood Industry Council made submissions which supported generally the submissions of the appellant but which also challenged the methods proposed in the IPP for assessing the extent of sea lion mortality.
[50] DoC’s argument (in line with that advanced by various environmental groups) was that rule 310 should be preferred. DoC also contended that no allowance should be made for the use of SLEDs.
[51] On 12 September 2003 the Ministry completed its final advice paper (FAP). This was made available to the Minister.
[52] The FAP noted that only rules 305, 310, 320 and 4 passed the performance criteria set by the AEWG. It rejected the industry submissions in favour of unconstrained fishing in these terms:
… For Rule 1, the model outcome shows that over 100 years, an average level of sea lion take of 99 animals per year significantly reduces the probability of the population growing to levels that are considered acceptable for management of the population, in order to allow it to attain non-threatened status in the near future. Further, this indicates that, irrespective of where the current population is, relative to K, there is a high probability that the difference between the level of the population that would have been attained in the absence of fishing, and that with fishing under Rule 1 is significantly more than the accepted 10%.
[53] On the whole the FAP adhered to the position set out in the IPP save that the preference for rule 320 was perhaps not as strongly expressed as in the IPP. Importantly for present purposes, the paper observed:
The three MALFiRM options presented in the IPP all satisfy the interim management criteria relating to sea lion conservation objectives. There is a much larger set of decision rules that would equally justify these criteria.
The FAP also noted that a “significantly higher” MALFiRM could be supported given the size of the sea lion population but did not relate that comment to the preference for rule 320.
[54] At this time, the Ministry also received and circulated to the stakeholders a review of the Breen Kim model by Dr Goodman, the Director of the Environmental Statistics Group at Montana State University. Although Dr Goodman’s review was essentially supportive, he expressed some caution about the statistical methods used in the modelling and the level of data used to generate the model. He suggested that the results of the Breen Kim model not be used to justify the changing of the by‑catch rules to rules 320 or 4. In his view, such a decision would not constitute a robust management procedure.
[55] The Minister discussed Dr Goodman’s concerns with officials on 17 September 2003 and subsequently received two reports from Dr Susan Waugh (the chief scientist employed by the Ministry). An aspect of Dr Goodman’s reservations (related to what is referred to in the papers as the “delay factor”) involved assumptions built into the Wade rule which were not particularly relevant to the fixing of a MALFiRM for sea lions. This was addressed in the reports from Dr Susan Waugh.
[56] It is apparent that Dr Goodman did not have the New Zealand legislative scheme clearly in mind and was not aware of the full details of the provenance of the Breen Kim model. It is fair to say, as the Minister has, that there do not appear to be serious errors in the Breen Kim model. On the other hand, there are grounds for caution given the uncertainties highlighted by Dr Goodman.
[57] The meeting between the Minister and his officials on 17 September 2003 was wide-ranging. The Minister by then had the IPP and an addendum (addressed to the SLEDs issue and post-mortem evidence), the FAP and the submissions made on the IPP as well as Dr Goodman’s report. According to the affidavit of Dr James Cornelius, a senior official who was present at the meeting, the Minister asked him whether there could be a series of rules arising from rules 310 and 320 which effectively tripled or quadrupled the MALFiRM and, after observing that only three options had been presented to him, asked how he should deal with all possible points between those options. Dr Cornelius said in his affidavit that his answer was:
… that other options that met the management objectives were evaluated (such as half the rate of rule 310, or no fishing) but not recommended to him because of the overly restrictive impacts on squid utilisation opportunities.
[58] If the interaction between the Minister and Dr Cornelius was exactly in those terms, then this was an unfortunate response. As we will indicate, the evidence from Dr Breen shows that a “rule 390” (ie providing for 9 times the by-catch permitted by rule 310) would also have met all management objectives. The response which Dr Cornelius gave the Minister could have been taken as suggesting that rule 320 produced an outcome which was near the point at which the management objectives were not met.
[59] On 23 September 2003, the Minister made the key decisions which were to be implemented through the Operational Plan for the 2003/2004 fishing year. He adopted a MALFiRM of 62, a strike rate of 5.3% and a 20% discount for the use of SLEDs. He prepared notes recording the reasons for his approach. In these notes he observed:
The general tenor of the MFish advice is that we can afford to move towards more utilisation, in the utilisation-conservation spectrum. …
He concluded by saying:
One last point. Despite the quality of the advice, the decision is a very complex one to make. There is much contention. What is K? What does the pathology say? Are the criteria robust? Why choose the rules you chose? (ie they are constructs).
All of that can be argued away. What is left is ‘how good is the model?’ It seems that Breen is an improvement, and I understand the industry’s frustration that it has not been dusted off earlier. Now it is in play, and we will no doubt learn to use, trust and perhaps amend it over future years.
[60] Following a formal consultative process with DoC, the Operational Plan was signed-off by the Minister on 3 October 2004.
[61] The Minister’s decision was made public on 7 October 2003 accompanied by a press release. In the press release the Minister referred to having taken a “conservative approach” to the setting of the MALFiRM. He noted:
This year the official number of sea lion deaths allowed before the fishery is closed will be 62, the lowest number recommended in advice to the Minister. The least conservative option would have allowed 124 deaths. Last year the limit was 70.
[62] The appellant requested written reasons for the decision and these were provided in two stages, first in the form of his decision notes made on 23 September 2003 and secondly by way of a formal letter of 4 December 2003. We have already referred to the decision notes (see para [59] above). The Minister’s position as advised in the 4 December 2003 letter was that although the Breen Kim model suggested that a higher MALFiRM may not pose a threat to the sea lion population, caution was required given the Goodman review and that he had adopted a “precautionary approach”. He also noted that he had allowed for additional utilisation in the form of the 20% allowance for the use of SLEDs.
[63] Since the decision was made, the appellant has commissioned Dr Breen to identify the highest MALFiRM which would meet all of the agreed criteria. This “cusp rule” produces a MALFiRM of 571 sea lions, 923% of the MALFiRM settled on by the Minister. To put this in context, unconstrained fishing would result in an average by-catch of only 99 sea lions.
[64] There are two other factors we should mention.
[65] The first is that if the exercise had been calculated on the basis used up to the 2002‑2003 season (ie using the Wade rule but including the Campbell Island sea lion population) the MALFiRM would have been around 70.
[66] The other point relates to the overall effect on utilisation of the decisions made by the Minister. This was discussed before us primarily in relation to SLEDs. The 20% allowance for the use of SLEDs against the strike rate is reasonably conservative on the evidence available to the Minister. We are not suggesting that the allowance was wrong as we accept that the evidence on this point was limited. On the other hand we do not see this allowance as some sort of bonus for the industry - an informal quid pro quo for a low MALFiRM. The same approach is also applicable to the default strike rate struck by the Minister. At 5.3% this was distinctly more favourable to the industry than the strike rate for the previous year (9.4%). But again, there is no proper basis for regarding this as an off-set for a low MALFiRM.
The grounds upon which the Operational Plan was challenged
[67] The appellant’s statement of claim pleaded five causes of action, of which three remain live:
1. The Minister’s exercise of his discretion to impose a MALFiRM of 62 was:
(a)unlawful in that the Minister failed to act in accordance with ss10(a), 8 and 5(a) of the Fisheries Act;
(b)unlawful in that the Minister failed to take into account relevant considerations; and
(c) irrational.
2. The Minister acted on a mistake of fact in setting the MALFiRM at 62; and
3.The Minister’s reliance on the Wade rule to set the MALFiRM was unlawful.
[68] As put by the appellant in both the High Court and in this Court, the case largely turns on whether the Minister was properly advised and/or misled by information provided by the Ministry.
The history of the litigation
[69] The proceedings were commenced on 15 December 2003. The substantive hearing took place before France J on 11 and 12 February 2003 and she dismissed the claim in her judgment, which she delivered on 27 February 2003.
[70] The issues before us are narrower than those presented to France J. To the extent that it is necessary to do so we will refer to her judgment later in these reasons.
[71] On the appeal to this Court, the appellant sought an interim order deferring the closure of the season. The appellant’s application to the High Court was denied by France J but a further application to this Court was granted on 22 March this year.
[72] As is apparent from what we have said, we gave judgment, without reasons, on 6 April in favour of the appellant. An immediate judgment was required given the limited amount of time available before the season would, in any event, have come to an end.
Overview
[73] Given that we are now giving our reasons for a judgment which has already been delivered and the not entirely easy subject nature of the case, we think that it is appropriate at this point to explain why we reached the view that the 2003‑2004 Operational Plan should be set-aside and then relate this, where appropriate, to the arguments of the parties and conclusions reached by France J in the High Court.
The reasons why we set-aside the 2003-2004 Operational Plan
The legal framework
[74] In her judgment, France J discussed the relevant principles in a detailed and insightful way. Given that the relevant principles were not really in dispute for us, there is no need for us to deal with the legal framework in so elaborate a manner.
[75] The Minister, as is often the case under the Fisheries Act, was required to balance utilisation objectives and conservation values. In the context of a harvestable species, this requires utilisation to the extent that it is sustainable, see Westhaven Shell FishLimitedv Chief Executive, Ministry of Fisheries [2002] 2 NZLR 157 at para [46] and Kellian v Minister of Fisheries (CA150/02, 26 September 2002) at para [34].
[76] It is clear that the Minister’s adoption of a “precautionary approach” meant that he did not pick as the MALFiRM the best estimate of an allowable sea lion by‑catch which would maximise utilisation without infringing conservation values. This meant that his precautionary approach largely resolved uncertainties against utilisation and in favour of conservation.
[77] On a simplistic approach to the requirement to permit utilisation unless sustainability is threatened, it might be thought that the Minister was required to adopt a cusp rule. If this approach were right, then the MALFiRM picked by the Minister was plainly inappropriate. But we do not think such an approach is appropriate. The point of the exercise is not to arrive at a number of sea lions which can be harvested sustainably, and thinking associated with sustainability of a harvestable species is not appropriate. So the remarks made in the Westhaven and Kellian cases (which concerned harvestable species) must be read with caution in this context. Given the underlying uncertainties involved in exercises of this sort, any MALFiRM chosen is likely to carry some degree of risk (perhaps negligible) to the population in question. Optimum usage does not equate to maximum usage. We are not aware of a simple method by which risk on the one hand can be balanced against utilisation advantages on the other. A precautionary approach to the required balancing exercise is open to the Minister.
[78] We should make it clear that the appellant did not argue for the adoption of a cusp rule and this was plainly appropriate.
[79] Recognising as we do, that a value judgment was in the end called for and that a precautionary approach was available to the Minister, we are satisfied that the legislative framework required the Minister to form a view as to the extent to which (or perhaps the point at which) utilisation of the squid resource threatened the sustainability of the sea lion population. We see this as implicit in the way in which s15(2) is expressed; the Minister may only take measures which he or she “considers necessary” in terms of avoiding, remedying or mitigating adverse effects of fishing on a protected species.
[80] In this exercise, the information principles provided for in s10 of the Act were applicable. So the Minister was required to take into account the principle that decisions “should be based on the best available information” but with the appropriate allowances for uncertainty and caution where information is uncertain, unreliable or inadequate.
Irrelevancies
[81] It is quite apparent that there are some background factors which, while technically irrelevant, nonetheless complicated the process in issue in this case.
[82] At the heart of this case is the killing of marine mammals as an incident of the commercial exploitation of a fishery. It is clear that the opposition of some who have criticised the proposed adoption of rule 320 would have been equally strong even if it was crystal clear that fishing-related sea lion mortality posed no risk to the sea lion population.
[83] Running through the material we have read is a sub-theme that there may well be substantial support for a policy which protects sea lions as individuals. Associated with this are concerns sometimes mentioned in the papers as to the political acceptability of an increased MALFiRM.
[84] We see these considerations as irrelevant to the current exercise.
General observations
[85] We think it is perfectly clear that the Minister did understand that each of the decision rules which was tested was an input rather than an output of the Breen Kim model and that they were simply a series of points on a continuum. We also think it reasonably clear that the Minister understood that there was a difference in practical terms between rule 320 and rule 1 (unconstrained fishing).
[86] We also accept that there are uncertainties about the nature of the sea lion population, in particular, whether it is at or near carrying capacity, and that there are associated uncertainties as to the extent, if any, of any likely long term impacts associated with fishing activities.
[87] It is nonetheless perfectly clear that the preponderance of evidence is now far more firmly in favour of utilisation than it was previously. It is odd that when the evidence is moving firmly in the direction of greater utilisation, the MALFiRM has moved in the opposite direction and, indeed, according to the table referred to in para [33] the MALFiRM struck in the 2003-2004 Operational Plan was the lowest it had been since the 1991-92 year. We interpose here that on the evidence we have seen we are not sure that the 1991-92 MALFiRM was actually of operational significance.
Defects in the decision-making process
[88] With those observations we now turn to what we see as the defects in the decision-making process and its outcome.
[89] The Wade rule was recognised on all sides as producing outcomes which, in the present context, were more restrictive than the legislation required. This is because that formula is calibrated to United States and not New Zealand legislative requirements and is built on assumptions as to the underlying population, which, on the best evidence available, are not true of the sea lion population in issue.
[90] The Minister’s decision to strike the MALFiRM at 62 involved, at least in substance, utilisation of the Wade rule. This is because rule 310 was simply an analogue of the Wade rule applied to the Auckland Islands sea lion population.
[91] Other than its analogue relationship to the Wade rule and its compliance with the agreed performance criteria, there was no reason to choose rule 310 as the decision rule.
[92] In relation to all of this Mr Scott made the very strong submission that the Minister had said one thing (that he was accepting the Breen Kim model as a basis for decision making) but he had done another (by adopting the Wade rule).
[93] There is a semantic answer to Mr Scott’s submission in that the Breen Kim model was used to test rule 310 against the agreed performance parameters. But there is nonetheless substance to Mr Scott’s contention. Effectively by definition, rule 310 was necessarily going to meet the performance requirements given that it was an analogue of the Wade rule. The Minister knew that the Wade rule’s assumptions did not accord with the best available evidence as to the extent to which fishing threatened the viability of the sea lion population. We think that the corollary of the Minister’s adoption of rule 310 (which was merely an analogue of the Wade rule) as the decision rule was that he took a course which did not involve making an assessment (cautious or otherwise) of the point beyond which utilisation of the squid resource posed an unacceptable risk to the sea lion population. In this respect he departed from what we regard as the requirements of the legislative framework. If we are wrong on this point, and the Minister was endeavouring to make such an assessment, it necessarily follows that in using the Wade rule (which he was in substance) he was not acting on the best available evidence.
[94] A very important and disputed issue in the case is whether the material placed in front of the Minister indicated the extent of the head-room available between rule 320 and the point at which a MALFiRM would not meet the three agreed conservation criteria, at least according to the Breen-Kim model.
[95] Ms Jagose, who appeared for the Minister, argued in her written submissions, that the material before the Minister did show that a MALFiRM in the order of 500 sea lions or more would have met the agreed performance criteria. She advanced this submission on the following basis:
1. Table 3 of the IPP (see para [47] above) showed that the maximum by-catch under unconstrained fishing would be in excess of 545 animals with a mean of 99.
2. Since the Minister knew that unconstrained fishing narrowly missed satisfying the performance criteria, he would have been able to infer that a MALFiRM in the order of 500 sea lions would have satisfied the criteria.
3. The response from Dr Cornelius to the Minister’s question could not sensibly be regarded as negating the other information which the Minister had available to him.
[96] This submission (and particularly points 1 and 2 as just summarised) drew a strong and convincing response from Mr Scott for the appellant which was along these lines.
[97] The Minister did not say in his affidavit that he understood that a cusp rule would have produced a MALFiRM of around 500 animals. Indeed the drift of the material which originates from the Minister is not really consistent with this being his understanding. The question and answer exchange between the Minister and Dr Cornelius very much suggests that neither appreciated that a cusp rule would produce a MALFiRM which was nine times greater than that produced by rule 310. As well, Dr Waugh in her response to Dr Breen’s evidence on the result of the testing of the cusp rule noted that “it produces an interesting result” but commented that this material was not before the Minister; again suggesting that the likely result of the application of a cusp rule was not recognised at the time the key decisions were made.
[98] Mr Scott also made the very telling point that the “maximum by-catch (100 year runs)” referred to in table 3 is not to be equated as just over the MALFiRM which would be struck with a cusp rule. We will put this point in his own words:
… The heading is actually misleading, because the performance indice is the mean of the maximum annual bycatch in each of the 5,000 simulations of 100 year runs. Within each 100 year run there will be 100 actual annual bycatches. The highest annual bycatch in each 100 year run was then aggregated with all the other highest bycatches from the 5,000 simulations. This produced a maximum catch range between 300 to over 1,000, the mean of which was 545 … . It is merely coincidental that the mean of the annual maximum catches was similar to the MALFiRM on the cusp. The mean could have been anywhere between 300 and 1,000, depending on where the numbers fell within that range. The cusp MALFiRM is not calculated by reference [to] the mean of the maximum bycatch. They are different things. One is a mean of the maximum actual catches. A MALFiRM is a maximum annual limit. …
[99] At the hearing of the appeal Ms Jagose recognised that the written submission advanced on this point was not sustainable.
[100] It would have been perfectly possible for the Ministry to have tested a series of possible decision rules using the Breen Kim model to pick the maximum MALFiRM which would have satisfied all the conservation criteria. Had this exercise been carried out, the Minister would have recognised that, according to the Breen Kim formula, a MALFiRM of approximately nine times that produced by rule 310 met the agreed conservation criteria. No one is arguing for the view that a MALFiRM of that order of magnitude should have been struck. The relevance, as we see it, of the testing of a cusp rule, is that it would have shown the Minister that the very comfortable degree of head-room above the rule 320 approach would, in all probability, have accommodated any legitimate doubts and concerns as to uncertainties associated with the operation of the Breen Kim formula.
[101] On the evidence, it appears that thinking along these lines prompted the Minster’s apparent inquiry addressed to Dr Cornelius as to the possible extent of the family of rules associated with rule 310. But this inquiry was answered (according to the affidavit of Dr Cornelius) in a way which pointed away from the reality that on the best available evidence, rule 320 was itself very conservative. Although we are reluctant to place too much weight on the terms of the press release (see para [61]), it is perhaps implicit in that document that the Minister saw rule 320 as not conservative (albeit that literally and in the comparative terms in which the press release was expressed, what was said was correct).
[102] In accordance with the general approach advanced by Mr Scott we conclude that it is more likely than not that the Minister did not appreciate the extent of the head-room between a MALFiRM struck in accordance with rule 320 and a MALFiRM which would result from the application of a cusp rule.
The legal basis of our decision
[103] To put all of this in legal terms, the grounds (which to some extent are overlapping) for our judgment are as follows:
1. The approach the Minister took did not address the extent to which, or the point at which, utilisation of the squid resource conflicted with conservation of the sea lion population. In this respect, his approach was not consistent with the requirements of the legislation which authorised only measures which he considered “necessary” for avoiding, remedying or mitigating fishing effects on the sea lion population.
2. In fixing a MALFiRM of 62 the Minister in substance acted on the basis of the Wade rule and in this respect he did not take into account the principle that decisions should be based on the best available evidence and was thus in breach of s10.
[104] As is apparent from what we have said, we see it as more likely than not (and thus proved for present purposes) that the Minister did not understand the degree of head-room available between rule 320 and the point at which the agreed conservation criteria were not met. On this basis there would appear to have been a fundamental flaw in his approach to the issue he was required to determine. This could be (and indeed this was the argument of Mr Scott) an independent ground for judicial review. But this line of argument is so intimately tied up with the conclusions we have just expressed that we do not think it necessary to explore the not entirely easy jurisprudence as to the circumstances in which a fundamental mistake of fact warrants judicial review.
The arguments of the appellant
[105] It is not really necessary for us to refer in any detail to the arguments advanced on behalf of the appellant by Mr Scott; this is essentially because we have found in favour of the appellant.
[106] This cursory reference to Mr Scott’s arguments is also associated with the fact that, to a very large extent, these reasons for our judgment involve the acceptance of his careful, thorough and balanced submissions.
The arguments of the respondents
[107] We have already addressed some of the key submissions advanced by Ms Jagose but it is appropriate to address her arguments in a more comprehensive way.
[108] She challenged thinking along the lines set out in para [66]. She said that the overall effect of the 2003-2004 Operational Plan when compared to that for the previous year permitted greater utilisation. She got to this point by saying that a MALFiRM of 62, a strike rate of 5.3% and a 20% discount for SLEDs was the equivalent, in terms of permitted fishing effort, of a MALFiRM of 74 if no discount was allowed for the use of SLEDs (as had been the case up until the current year).
[109] For reasons already given, we do not accept that this is an appropriate way of looking at the issue. We agree that the overall effect of the scheme is important. But a rational assessment of the individual components of the scheme was required. There is no basis for suggesting that the 20% discount for SLEDs was other than a realistic assessment of the admittedly limited evidence. It would be irrational to design a scheme on the basis of the false premise that all animals ejected via SLEDs from trawl nets die as a result of the experience. So we do not see that it is credible to suggest that in some way the allowance for the use of SLEDs can be thought to compensate for an inappropriately low MALFiRM.
[110] Ms Jagose argued that a cautious approach was justified given the uncertainties raised by Dr Goodman as to the Breen Kim model. The Minister was aware that the various decision rules proposed were simply points on a continuum, that the Breen Kim model indicated that movement toward the utilisation end of the sustainability continuum was possible and that there were other possible decision rules which would, on the Breen Kim model, meet the agreed criteria but provided for more utilisation than rule 320.
[111] As is apparent, we agree with those submissions. We are, however, left with view that the material made available to the Minister did not show the nature of the continuum which he was dealing with and where the proposed decision rules fell on that continuum. And, in any event, the Minister did not appreciate that a MALFiRM nine times that permitted by rule 310 would have satisfied all agreed conservation criteria.
[112] Counsel also argued that the Minister did not use the Wade rule. This we see as something of a semantic point. Rule 310 was an analogue of the Wade rule and that was its sole rationale. We think it is obvious that the striking of a MALFiRM of 62 implied the in-substance adoption of the Wade rule.
[113] Counsel argued that there is no obligation on the Minister to choose a MALFiRM based on a cusp rule, that is one which permits the most fishing while meeting the conservation criteria. As is apparent we agree with that submission and very much for the reasons which she gave (which are largely adopted earlier). We are left with the view that an assessment of the extent of the risk posed by fishing to that population was fundamental to the Minister’s task.
[114] Ms Jagose challenged the view that the cusp modelling exercise was relevant as it was not before the Minister and further argued that the Minister’s decision cannot fairly be impugned on the basis of the testing of a limited number of rules as this was pursuant to a process which the appellant had agreed to. We see some force in this line of argument, but it remains our view that the Minister could not squarely address the task laid down for him by the legislation without understanding where the proposed decision rules stood in terms of a potential threat to the sea lion population.
[115] Although we did not accept the key elements of the submissions made by Ms Jagose, it is right that we pay tribute to their thoroughness and the care with which they were presented to us.
The judgment of France J
[116] In a commendably short time after the hearing before her, France J released a lengthy and comprehensive judgment in which she analysed in detail and rejected the arguments advanced by the appellant.
[117] France J concluded, as we have, that in general terms at least, the Minister knew that the proposed decision rules were inputs into and not outputs from the Breen Kim model, that there was a continuum and that the model indicated that there was at least some head-room available above rule 320. She was concerned about the answer given by Dr Cornelius to the Minister but took the view that this exchange had to be put “in context”, which she described in this way:
[134] … First [the answers] reflect a process based on a certain number of rules being modelled. This was a process in which the relevant stakeholders had participated. Second, the Minister was aware of the overall context, and, in particular that greater utilisation was possible. Finally, the Minister’s use of the 20 percent discount illustrates that he understood his options were flexible.
[118] On the use of the Wade rule, France J said this:
[151] Finally, I deal with the claim that the Minister completely disregarded or rejected the Breen Kim model.
[152] The applicant acknowledges that management measures will usually (as here) produce a continuum of potential outcomes or options. Indeed, Mr Cade for the applicant says that it would not be here if the MALFiRM had been set at 124 as recommended by the Ministry. Mr Scott explains that this does not mean the Squid Company could not bring a case to challenge a MALFiRM of 124 but is acknowledging that the industry would have accepted that.
[153] Nor, the applicant says, can the Minister be criticised for exercising a degree of caution and the applicant is not arguing here for unconstrained fishing.
[154] The applicant says that it is not possible to have a MALFiRM of 62 because that could only be achieved by using the equivalent to the Wade model or the PBR rule. That model is no longer legally permissible as Breen Kim represents the best available information but its inappropriate assumptions have been carried over.
[155] Ultimately, for the applicant to succeed, it must be said that ss8 and 10 mean that the Minister has to adopt a higher MALFiRM than 62. More particularly, once it is accepted as it must be, that there is an available continuum, the applicant is really saying that the Minister should have set the figure at higher than 62, possibly at the first step at 70 (if the Campbell Islands figures were included) or presumably anything above 62 and up to 124. If the SLED discount is included, the MALFiRM is 74. Viewed in this light, there is merit in the respondents’ argument that the applicant is really just asking the Court to take a different view on the merits or weight.
[156] Further, s 10(a) only requires the Minister to “take into account” the principle that decisions are to be based on the best available information. (Authorities on a similar phrase are discussed in Greenpeace New Zealand Incorporated v Minister of Fisheries (High Court Wellington, CP492/93, 27 November 1995, Gallen J, at 25-26)).
[157] The applicant submits that s 10(a) reflects a trend internationally to require states to act on the best scientific information when making these sorts of decisions. The applicant also says that the objective of s 10(a) is to preclude decisions being based on “political whim” or expediency. Paul Starr, a scientist who gave evidence for the applicant, suggests that the approach to the model means “science” was only part of the Minister’s decision and that there are other, “largely unstated”, objectives influencing the Minister. That was not a matter developed in argument and the Minister does explain the other, legitimate, factors he considers.
[158] In considering the requirement to take into account the best available scientific evidence in Article 61(2) of UNCLOS, Burke says this means,
“the scientific information needs only be taken into account as part of the wider context which determines the proper measure.” (WT Burke, “The New International Law of Fisheries UNCLOS 1982 and Beyond”, at p56).
[159] Burke makes the point there are two elements to the requirement, namely, giving weight to the information; and the quality of the information. As to the latter, Burke says Article 61(2) “does not demand perfection” (at p 57).
[160] In the Fisheries Act, the principle is that the decision is to be “based” on the relevant information. The ordinary dictionary meaning of “base” suggests using as a foundation or ground (see Funk & Wagnalls New Standard Dictionary (1924)).
[161] It follows from the earlier discussion that I consider the Minister has taken into account this principle. The decision does rest on the platform set by the Breen Kim model but with some cautions. Dr Breen’s “main disagreement” with the Goodman review is with Goodman’s suggestion that more data, more modelling, and more development of by-catch rules are necessary before the results should be used in actual management. But it was open to the Minister to take cognisance of the fact this was the first time the model was being used and so to query the degree of certainty associated with it and the information it analysed.
[162] It is also relevant that the Breen Kim model is just that, a model. It necessarily is based on various assumptions albeit those assumptions have been the subject of scrutiny by the AEWG.
…
[213] It follows from my earlier conclusions that the Minister has not applied the PBR rule. The result of his use of the Breen Kim model has produced the status quo, although only if the SLED discount is excluded. But it cannot be said there has been reliance on the PBR formula as such. The Minister has chosen one of the rules modelled by Breen Kim albeit Rule 310 approximates the PBR rule. He has done this following his consideration of the Breen Kim model. He has chosen something within the available range. The Breen Kim analysis gave the Minister up to date information on whether the rule would meet the management objectives and the impact of the rule on fishing efforts. There has been no error under this head.
[119] For reasons already given, we differ from France J in relation to the significance which she places on the SLEDs discount. We also disagree with her conclusion that “the Minister has not applied the PBR rule”. More importantly we think that her reasoning did not squarely address what we regard as the requirement for the Minister to at least form a view as to the point at which utilisation posed an unacceptable risk to the sea lion population. As is apparent from what we have already indicated, we do not believe that he could properly use the Wade rule or its rule 310 analogue to do that.
Solicitors:
Chapman Tripp, Wellington for Appellant
Crown Law Office, Wellington for Respondents
2
0
0