Previous Page Table of Contents Next Page


Workshop on the Assessment and Management of Deepwater Fisheries - Synopsis

This workshop discussed the assessment and management of deepwater stocks, issues related to trawl, acoustic and egg surveys; the use of catch and effort data; and tag and recapture data. The presentations referred to orange roughy, redfish, smooth oreo dory, Antarctic toothfish and a range of species from the northeast Atlantic. It was noted that many deepwater species have low productivity, aggregate, and often are found associated with underwater features. The reasons for their aggregating behaviour were discussed and it was agreed that more work on fish behaviour is needed. Several noted that the low productivity of deepwater fishes may not be universal as some such species have moderate levels of productivity.

The problems associated with surveys and use of remotely operated vehicles were reviewed as were concerns with understanding stock structure, fish distribution and movement. It was accepted there is no one best way to estimate abundance of deepwater fish and use of a range of methods offers the best way forward. In looking at biology, age and growth it was agreed that biological characteristics vary widely between species, with high longevity, slow growth rates, high age at maturity and low fecundity. For such stocks, sustainable yield levels will be relatively low and recovery from depleted states, slow.

Stock structure is generally poorly known for most deepwater species and their depth of capture means that direct methods to monitor distribution and movements can rarely be used. Biological parameters are often variable and poorly known. Several deepwater species do not spawn each year so that reliance upon gonad data to determine age at maturity may mislead. Transition zones in otoliths may mark onset of spawning and spawning frequency may determine the proportion of the stock available to capture or survey each year. Ecological processes and interactions and links between deepwater demersal fish and energy sources must be better understood as stock-recruitment relationships and recruitment remain poorly known for most deepwater species.

The need for accurate catch and appropriate spatial-scale data, valid indices of abundance and estimating absolute abundance was noted. Although fish age information was considered essential for estimating population productivity, until the development of a validated methodology it was not evident that orange roughy is characterised by unusually low growth rates, low natural mortality, high age at maturity and high longevity. The role of biological reference points, the ecosystem approach to fisheries management and data-poor situations were examined. It was agreed that the most important step towards an ecosystem approach to fisheries management is to get single species fishing mortality to appropriate levels where necessary. This needs better integration of assessment and management of marine resources with appropriate management frameworks that ensure single stock management.

Other presentations covered management arrangements for high-seas orange roughy, deep-sea and deepwater fisheries in New Zealand; development of high-seas fisheries in the western Indian Ocean and frameworks for management advice including setting of management reference points. The Workshop identified eight actions as priorities for improving assessment and management of deep-sea resources:

  1. formulate management objectives more explicitly
  2. incorporate biological reference points and performance measures into management procedures
  3. implement management systems that promote data collection (e.g. adaptive management)
  4. conduct cooperative research with the industry and other stakeholders
  5. implement management procedures that do not have high information needs (e.g. design decision “triggers” for opening and closing fisheries)
  6. integrate assessment and management using, e.g. Management Strategies Evaluation techniques
  7. ensure use of collective experience world-wide (using, e.g. “meta-analyses”) and
  8. adhere more closely to the precautionary approach.

Biological parameters of deepwater fishes

J.D.M. Gordon
Scottish Association for Marine Science
Dunstaffnage Marine Laboratory
Oban, Argyll PA37 1QA, United Kingdom

This topic may be addressed from the perspective that “deepwater fishes are long lived and slow growing, have a high age and size at first maturity and a low fecundity”. This sentence, or a variant of it, is frequently used in the context of deepwater fisheries. An appropriate question to follow is then, “is this statement valid for all deepwater fishes or simply those that are exploited?”.

In the Rockall Trough of the northeast Atlantic Ocean there are more than 100different species of demersal fishes that inhabit depths between about 400 and 2000m. Few of these species are marketed, but among those sold are the roundnose grenadier (Coryphaenoides rupestris), blue ling (Molva dypterygia), black scabbardfish (Aphanopus carbo), orange roughy (Hoplostethus atlanticus) and two deepwater squalid sharks (Centrosymnus coelolepis and Centrophorus squamosus). It follows that in the commercial fishery many species are discarded and it is highly probable that none of these discarded fishes will survive. Research trawls that have small mesh usually yield in excess of 40 species at any given depth and many of these would pass through the meshes of a commercial trawl. Deepwater species often have large scales and small amounts of mucus on their bodies. It is a widely held view that many fish will not survive after being damaged while escaping through a trawl. With the growing awareness of the ecosystem effects of fishing it is important to have an understanding of the biology of all the fish species. This presentation reviews the existing knowledge on the biological parameters of several families of deepwater species.

Family Scyliorhinidae

The blackmouth dogfish (Galeus melastomus) occurs on the upper slope and is exploited in southern Europe. When spawning it produces up to 30 large eggs per female but despite some research effort determining their age remains uncertain. Galeus murinus and several species of the genus Apristurus spp. are common on the lower slope but little is known of their biology.

Family Squalidae

There are about 12 species of squalid shark. All produce relatively small numbers of live young. Female Centroscymnus coelolepis mature at about 124 cm and ring counts on the spines suggest they attain (unvalidated) maximum ages of about 70 years. The birdbeak dogfish (Deania calceus) may attain an age of 35 years. These sharks are seldom found deeper than 2 000 m.

Family Rajidae

Although there are numerous species of rays whose ranges extend to abyssal depths, most occur only in low abundances. Only the relatively small Raja fyllae occurs in appreciable numbers. All the rays produce small numbers of large egg cases. The age of the deepwater rays is unknown.

Family Chimaeridae

There are five species of Chimaerids in the Rockall Trough but only two, Chimaera monstrosa and Hydrolagus mirabilis, are abundant at the depth where the commercial fishery is prosecuted. Sometimes the larger Chimaera monstrosa are landed. As with the scyliorhinids and rays, they produce a few large egg cases. Nothing is known about their age and growth.

Family Alepocephalidae

This is a diverse family but in the Rockall Trough only Alepocephalus bairdii is abundant and at some depths it comprises a high proportion of the discards. Farther south another species, Alepocephalus rostratus, can be dominant in the catches from the mid slope area. Alepocephalids produce large eggs and, in the case of A. bairdii, in batches of 2 500 to 8 500. Unvalidated age estimates for this species suggest that it lives for about 35 years.

Family Argentinidae

The argentine, or greater silver smelt (Argentina silus), is a semi-pelagic species that is commercially exploited in the Rockall Trough. It grows to a length of about 60cm and has a maximum age of about 25 years. Female argentines mature at about 35 cm (6–9yr). The eggs are large and the fecundity is relatively low (16000–36000).

Family Synaphobranchidae

The cut-throat eel (Synaphobranchus kaupi) has a wide depth range, from about 500to 2000 m and exhibits a typical “bigger-deeper”size-distribution characteristic. Although seldom caught in commercial trawls it is the most abundant species at mid-slope depths in the catch of small survey trawls. Its fecundity is between 60000 and 165000 and this specices has a pelagic leptocephalus stage. They metamorphose at a length of about 9 to 10 cm. Their age is unknown.

Family Notacanthidae

The two dominant spiny eels, (Notacanthus bonapartei and Polyacanthonotus rissoanus), have a relatively low fecundity (30000 and 9 000 respectively) and also have a leptocephalus stage. The age is unknown.

Family Macrouridae

This is a diverse family that occupies all depths on the slopes of the Rockall Trough. The roundnose grenadier is one of the main exploited species. Its age has been validated up to six years and ages of up to 75 years have been reported. In the fishery most of the fish landed are less than 30years old. The age at maturity is about 10 years and the fecundity is relatively low at about 25000eggs. The other species of the upper to mid-slope tend to be smaller in size but a maximum, unvalidated age of 20 to 30 years appears to be the norm.

Family Gadidae

The blue ling attains a length of about 145 cm and a length at first maturity at about 90cm. It probably has a maximum age of about 30 years. In common with most gadoid fishes it is highly fecund producing millions of eggs.

Family Moridae

The only species of interest to the fishery is Mora moro. It attains an age of about 34 years and produces two to three million eggs. The smaller Lepidion eques, a significant discard in the fishery, reaches a maximum size of 40 cm and females mature at about 24 cm. Their fecundity ranges up to 100000 eggs.

Family Trichiuridae

The black scabbardfish (Aphanopus carbo) has been fished in deepwater since the 17thcentury yet many features of its life history remain unknown. Mature fish of length greater than one metre are found around Madeira and off mainland Portugal but the eggs, larvae and early life history stages have not been found. In the Rockall Trough the fish caught in the trawl fishery are all sub-adults and catches of juveniles are rare. Unvalidated age estimates suggest that this is a fast growing species with ages from about 10 to 25 depending on the ageing method that is used.

Family Scorpaenidae

The bluemouth (Helicolenus dactylopterus) is an important commercial species in southern Europe. Small bluemouth can be an important component of the discards from the bottom trawl fishery in the Rockall Trough. The maximum size is about 46cm and ages of up to 46 years have been reported. Females reach maturity at about 13 years and produce live young.

Family Trachichthyidae

The orange roughy (Hoplostethus atlanticus) can reach a maximum size of about 60cm but in the Rockall Trough they generally attain lengths of about 40 cm. Ages of up to 125 years have been estimated and females mature at about 53 cm. The fecundity ranges from about 28000 to 380000. The silver roughy (Hoplostethus mediterraneus) attains a maximum size of 42 cm and is a shallower living, more southerly distributed, species. Ages of 10 to 11 years have been reported and the fecundity ranges from 4000 to 100000.

In 2001 one of the terms of reference of the ICES Working Group on the biology and assessment of deep-sea resources (WGDEEP) was to rank the commercial species in order of vulnerability to fishing. On a scale of 1–5 (1 being the most vulnerable) the following species were ranked as follows; squalid sharks (1.5), orange roughy (1.6), roundnose grenadier (2.4) black scabbardfish (4.0) and blue ling (4.0). It was acknowledged, in view of the many uncertainties, that these were fairly crude estimates but that the main pattern as indicated was robust.

This review includes a range of the non-target species and reveals a similar pattern of wide variability in life history parameters with a tendency for high age, slow growth, high age at maturity and low fecundity.

The fishery management process for deepwater fisheries in New Zealand: now and in the future

Kate Graham, Charmaine M. Gallagher and R. Tinkler
Ministry of Fisheries, ASB Bank House
101–103 The Terrace
P.O. Box 1020
Wellington, New Zealand


Commercial fish stocks in New Zealand are managed under the successful and evolving quota management system (QMS). Individual transferable quotas (ITQs) ensure ownership as well as responsibility for the sustained use of a fishery. Annual catch entitlements (ACE) provide the assurance that ownership rights can be developed under an annual total allowable commercial catch (TACC) determination for a stock. The Minister sets an annual TACC representing the total amount of fish that commercial fishers can remove from a fish stock with reference to the highest yield that can be achieved over time while maintaining maximum productivity (MSY). A process is in place to impose a penalty for exceeding the ACE and a deemed value is paid when the catch exceeds the catch limit. Statutory and voluntary compliance is continually monitored through audits of fish returns and licensed fish receivers inspections. Scientific research and compliance services costs are recovered from the industry in proportion to the investigation of specific fisheries.

The 1996 Fisheries Act mandates the Ministry of Fisheries to provide for the use of fisheries resources while ensuring sustainability. The Minister is required to maintain stocks at, or above, a level that can produce the MSY. The Ministry of Fisheries provides the Minister with scientific advice on which to base these decisions. Scientific stock assessments identify a safe level of fishing that is not expected to reduce the biomass of the underlying population below 10 percent of the initial biomass 90 percent of the time. A consultative process ensures that stakeholders concerns and interests are addressed. It is this consultative process that facilitates fishery management decisions. A more complete and detailed description of the New Zealand QMS can be found in Sissenwine and Mace (1992), Annala (1996) and Batstone and Sharp (1999).


Deepwater fisheries

Major deepwater fisheries in New Zealand include: orange roughy (Hoplostethus atlanticus), smooth oreo (Pseudocyttus maculatus), black oreo (Allocyttus niger), cardinal fish (Epigonus telescopus), hoki (Macruronus novazelandiae), hake (Merluccius australis), ling (Genypterus blacodes), bluenose (Hyperoglyphe Antarctica), warehou (Seriolella punctata), southern blue whiting (Micromesistius australis), squid (Notodarus sloanii), and scampi (Metanephrops challengeri). These species dominate New Zealand seafood production with annual harvests near 525 000 t. This figure is 70 percent of the total annual catch of seafood products in New Zealand. The 2002 Export Value of deepwater species was $NZ 502 million FOB, representing 70 percent of the export value of live, chilled and frozen fish (SeaFIC 2002).

Commercial fishery descriptions

The deepwater fishing fleet is mix of over 150 domestic vessels, 50–60 foreign vessels (chartered by New Zealand companies), and up to 10 foreign licensed vessels. The deepwater fleet consists of (a) New Zealand owned ice “fresher boats”reliant on land-based fish processing and (b),factory trawlers, which can be domestic or foreign-flagged vessels, and spend up to 6 months at sea. Vessel sizes range from 10 to 100 m in length. Owners hold a portfolio of species quota to enable better use of their capital assets, processing and marketing infrastructure and to allow them to align their fishing strategies to market opportunities.


Sustainability information

Fishery managers use multiple sources of information to enable sound management decisions. Deepwater scientific stock assessments are used as performance indicators for sustainability of commercial fish populations. Deepwater stock assessments rely on acoustic surveys, trawl surveys and species-specific sampling for catch at age and length-frequency information. Accurate stock assessments are needed as technological advances in fishing fleets increase catch efficiency. Fishery managers are dependent on the scientific stock assessment process to ensure that stocks are at a level that can provide maximum sustainable yield. It is critical that all stakeholders in a fishery increase their understanding and trust in the stock assessment procedures, as the stock assessment is the defining scientific process for estimating the biomasses that determine the annual TACCs.

Fishery management welcomes the incorporation of multispecies models to address multispecies fisheries or species complexes. Ecosystem management has a tendency to be based on policy and environmental issues with little scientific application from a fishery basis. In most areas, hoki is the target species of a mixed trawl fishery with as much as 12 percent of the catch composed of bycatch species such as hake, ling and warehou. In non-selective trawl fisheries, addressing the incidental catch of marketable and non-marketable species can challenge achieving sustainability and utilisation objectives. Concerns over the sustainability of a bycatch species can become a constraint to the best use of the target species.

Environmental effects from fishing information

Orange roughy and deepwater oreo fisheries have developed on plateaus, seamounts and underwater topographic features at depths of 700–1 400 m. They have become the largest and most valuable fisheries in New Zealand and Australia landing over 500 000 t (Koslow 1997; Koslow et al., 2000). Smooth and black oreo aggregate at slope depths of 850–1 150 m on rough ground and seamounts. Their longevity and productivity are similar to orange roughy. Annual sustainable yield is estimated at less than 2percent of virgin biomass. Since the early 1980s the catch has fluctuated between 15–25 000 t. Indications from scientific assessments are inconclusive with respect to whether landings at these levels are sustainable. The International Cooperation for the Exploitation of the Sea (ICES) advise that orange roughy stocks cannot sustain high rates of exploitation and their fisheries should be strictly limited and the populations closely monitored.

Utilisation information

Monitoring and evaluation of effort in a deepwater fishery is necessary to monitor fleet behavior in response to changes in stocks or changes in management measures including the TACC. Aligning stock abundance and variability with fishing effort provides fishery mangers with some of the information regarding behavior with respect to existing and future fisheries. Effort data is important to fishery analysts who study the trends in activity and fleet behavior. In addition, changes in catch per unit effort (CPUE) over a season and between seasons show how the fleet adjusts to abundance and fishery management decisions. Additional monitoring of spatial effort and fleet distribution is necessary as CPUE analysis may not identify changes in fleet behaviour as local aggregations of orange roughy are reduced. Changes in CPUE may better reflect a change in fishing strategy, geographical distribution of effort or in fish accessibility than changes in stock abundance. CPUE series are not informative about the size of the effort reduction necessary to allow stock rebuilding to commence.

A consultative process is used to allow stakeholders to express their operational concerns with respect to specific fisheries. Fishery managers address the costs and benefits of proposed sustainability actions with respect to an individual’s right to improve their well being. The consultative process is composed of three parts. First, participation in the stock assessment working group where the scientific information is presented and discussed. Second, a plenary session that incorporates the stock assessment information and any additional sustainability issues. The third opportunity for participation is through response to publicly promoted initial position papers presented by the Ministry of Fisheries. Submissions are summarized and presented to the Minister of Fisheries in a final advice paper. In this process, the Minister can determine the issues and concerns of industry, environmental groups and individuals.

Fishery managers complete additional monitoring and specific evaluation of economic and fishery data that can provide critical information that may not be specifically presented in the working groups or plenary sessions. Stakeholders often provide information or identify inconsistencies or specific trends in fleet behaviour. Fishery managers analyse this and other relevant information in the formation of a position for the Ministry of Fisheries to present to the Minister. In the case where specific skills are required, fishery managers initiate additional research or request specialist review of policies to address contentious issues. Two specific fisheries are presented here to demonstrate specific sustainability and utilisation fishery management interactions in deepwater fisheries: the high-value orange roughy fishery and the high-volume hoki fishery.


The orange roughy fishery is one of New Zealand’s most important fisheries in terms of export value. In the year ending December 2002, orange roughy exports were worth $NZ14 534 a tonne, with total earnings of $NZ124.4 million, representing 17.2percent of finfish export earnings1. The largest overseas market is the USA followed by Australia. The major exported product form is frozen fillets, although some product is exported in chilled-whole or processed form. New Zealand’s orange roughy is a relatively low volume fishery currently constrained within a TACC of 15 221 t.

1 New Zealand Seafood Export Statistics (volume of live, chilled, frozen fin fish).


Orange roughy inhabit depths between 700 m and at least 1 500 m within the New Zealand Exclusive Economic Zone (EEZ). Orange roughy are thought to be extremely slow growing and long-lived, possibly living to 120–130 years. They reach a maximum size of about 50 cm and vary considerably in length from one stock to another.


New Zealand’s orange roughy fishery developed in the 1970s and although catches have declined since then New Zealand remains one of the world’s largest suppliers. Figure 1 illustrates the separate stocks of orange roughy in New Zealand waters and their catches from the late 1970s and early 1980s through 2002. Catches are taken mainly from the Chatham Rise, the southern part of the South Island’s west coast and some parts of the Challenger Plateau.

Characteristics and fishery

The orange roughy bottom trawl fishery is prosecuted by highly-skilled operators and has industry-led management regimes. Figure 2 shows the concentration ratios2 for selected fisheries. They reveal that the six largest quota holders for ORH 3A hold 95 percent of the quota for this QMA. As for the hoki fishery, the orange roughy fishery is dominated by a small number of large quota holders, as indicated by the Herfindahl index3 (Figure 3). The index for ORH 3B is the highest for any fishery in New Zealand.

2 A Concentration Ratio describes the proportion of total output in an industry that is produced by a given number of the largest firms in the industry. The most commonly used concentration ratios are the relative outputs for the four largest firms and the eight largest firms. Although ‘total output’ is not necessarily synonymous with quota holding, one might expect the correlation between the two to be high.

3 A Herfindahl index provides a measure of concentration of the production in an industry and is calculated as the sum of the squares of market shares for each firm. Although ‘market share’ is not necessarily synonymous with quota holding, one might expect the correlation to be high. The Herfindahl index is an alternative method of summarising the degree to which an industry is oligopolistic and the relative concentration of market power held by the largest firms in the industry. It is generally believed to give a better indication of the relative market control of the largest firms than can be found with the four-firm and eight-firm concentration ratios.

Orange roughy Quota Management Areas, Fisheries, and catch in tonnage from the late 1970sand early 1980s through 2002


Deepwater and inshore fishery 4-, 6- and 8-firm concentration ratios


Deepwater and inshore fishery Herfindahl indices


Bycatch in this fishery is categorized into two distinct groupings, commercial bycatch (fish that can be sold commercially) and other non-commercial incidental species. Commercial bycatch of cardinal fish, spiky oreo and other oreo species accounts for approximately 30 percent of the total catch, other bycatch of approximately 44non-commercial species accounts for around 0.5 percent by volume of the total catch (Anderson, Gilbert and Clark 2001). There are no known interactions between the orange roughy fishery and protected species.

Current orange roughy management regime

The New Zealand Government and industry take the conservation of orange roughy and other fish stocks extremely seriously. They are managed and harvested under New Zealand’s QMS, which provides predetermined catch limits for eight management areas. The QMS allows rapid responses to new information, provides economically rational rules regarding the methods of fishing and also rewards a long-term fishing perspective. Current catches have been set at, or below, sustainable levels, which are estimated to be six percent or less of the standing stock sizes. In cases where sustainability measures have been triggered catch levels and TACCs have been reduced and in some cases closed, often at the request of industry. Industry leadership and participation in the management of these fisheries contributes to the international success achieved by the QMS. Catch limits have been respected because of the onerous penalties that face rule breakers, including confiscation of catch, vessels, equipment and quota, coupled with severe fines. More importantly New Zealand fishers recognize and directly gain from the long-term benefits of managing fisheries at sustainable levels.

The long-term management strategy set under the Fisheries Act 1996 is to maintain each fishery stock at, or above, the level that can produce the MSY. Regular stock assessments monitor the status of each fishery (Francis, 1992), and there is strong participation from the industry and other stakeholders in the fisheries research and stock assessment process. For orange roughy, the MSY is interpreted as either the Current Annual Yield (CAY) or Maximum Constant Yield (MCY).

Sustainability measures used in the orange roughy fishery are a combination of measures imposed by the Fisheries Act (1996) and Fisheries (Commercial Fishing) Regulations (2001) and industry management initiatives endorsed by the Minister of Fisheries. Industry initiatives include the management of sub areas within QMAs, restricted areas, fishing gear restrictions and compulsory bycatch reporting.

The future of orange roughy management

The ongoing challenge for fisheries management arises from the great longevity of orange roughy and their relatively low productivity. Orange roughy are susceptible to overfishing and there are concerns over long-term damage to their stocks. The comprehensive stock assessment programme that has been underway for a number of years is to continue. Stock assessments and other research are funded by government levies and contracted to external science providers as well as directly purchased by industry. Examples of proposed future research projects that are initiated by the Ministry include:

The orange roughy quota holders collective, the Orange Roughy Management Company (ORMC) and the Tasman Pacific Fishing Company have initiated limited management responsibilities for orange roughy fisheries. Future opportunities for full-scale management through the development of fish plans may be developed.


The hoki fishery is New Zealand’s most important fishery in terms of weight and had a total allowable commercial catch (TACC) of 200 000 tonnes in the 2002–2003 season and an estimated value of $NZ1.1billion4.

4 Based on an interim value of $NZ580/tonne and 195 713 tonnes landed in the 2001–2002 fishing year.

Hoki is also New Zealand’s single biggest fish export earner; in the year to April 2003, 13200t of Hoki products worth $NZ47.1million FOB were exported representing 25percent of all fish exports by weight. The largest overseas markets are the USA, Australia, Germany, Japan, and the People’s Republic of China. Major exported product types include frozen fillets, fillet blocks, loins and portions and minced blocks.


Hoki is a member of the Merlucciidae family, which grows to an average length of 60–100 cm and an average weight of 1.5kg. Hoki are widely distributed throughout New Zealand waters from 34° S to 54° S from shallow water to around 900m. Juveniles are generally found in shallower waters and adults in waters deeper than 400m. Hoki normally live to around 12 years of age but may live to 25years old.


New Zealand’s hoki fishery was developed in the early 1970s by Japanese and Soviet vessels, and to a lesser extent by South Korean vessels in the late 1970s. Prior to 1978 domestic vessels caught less than 1000 tonnes a year. Until the 1987–88 fishing year domestic vessels were still catching less than 10 000 tonnes a year with the majority of the hoki quota being caught by foreign vessels chartered to New Zealand firms.

Characteristics of the hoki fishery

The hoki fishery, in common with other deepwater fisheries in New Zealand, is dominated by larger quota holders relative to inshore fisheries, although less so than in the case of oreo and orange roughy fisheries. Figure 2 shows that the 4-, 6-, and 8-firm concentration ratios are much higher than for the inshore fisheries; Figure 3 indicates that the Herfindahl indices for the deepwater fisheries are an order of magnitude higher than for inshore fisheries, reflecting the need for larger vessels to harvest these deepwater stocks.

The hoki season runs year-round but the major portion of the catch is taken from late June to late August–September when hoki aggregate to spawn around Hokitika Canyon off the South Island’s West Coast and in Cook Strait. Minor spawning grounds also occur off the Pegasus Bay on the South Island’s east coast and also off Puysegur Bank in the south-western-most part of the South Island. Non-spawning hoki are predominantly targeted in Cook Strait and on the Chatham Rise. In each case harvesting is by midwater or bottom trawling with most of the commercial catch taken between 200 and 800 m. Bycatch of several species is considered problematic and includes ling, silver warehou and in particular hake. The HAK7 (Challenger FMA) quota has been overfished in nine of the last 12years.

Quota management areas

Hoki is managed as a single stock, denoted as the quota management area (QMA) HOK1 with a nominal fishery, HOK10, in the Kermadecs. An informal subdivision within HOK1 into an eastern and western stock enables fine-tuning of catch effort to reduce pressure on the two juvenile stocks as required.

Management of the hoki fishery

Management methods of the hoki fishery are well developed and it is administered by a combination of statutory fishing rules as established under the Fisheries Act 1996 (‘the Act’) and the Fisheries (Commercial Fishing) Regulations (2001), both administered by the Crown through the Ministry of Fisheries and a voluntary Code of Practice (CoP) established by the Hoki Fishery Management Company (HFMC). The Ministry of Fisheries perceive that generally, statutory and non-statutory rules work well in tandem and that agreement with industry is often sufficient to achieve the Government’s sustainability objectives, rather than regulating for changes in catch, effort, for example.

The HFMC is a private company comprised of 40 or so shareholders who hold around 99percent of the hoki quota. The objectives of the HFMC are to improve the management and economics of the hoki fishery by representing the interests of quota holders with the government responsible for fisheries research and stock assessment programmes, assisting with the balancing of catch against annual catch entitlements and implementing and monitoring fisheries management programmes.

The Hoki Fishery Management Company obtained certification in 2001 from the Marine Stewardship Council (MSC) for the New Zealand hoki fishery. This process involves independent, third party scientific certification of sustainability measures employed in a fishery. Achieving certification is a means of assuring consumers of the environmental sustainability of the fishery and offers potential for improved market access and value-adding opportunities. New Zealand hoki was awarded this certification subject to undertaking certain corrective actions identified in the independent assessment process. The corrective actions planned by the HFMC address environmental risks and ecological impacts associated with the fishery, including seabird and marine mammal interactions and marine habitat issues cited in the Ministry’s sustainability review.

The sustainability measures used in the hoki fishing industry are a combination of measures imposed by the Fisheries Act 1996, e.g. the TAC/TACC setting (s13), requirement to hold a fishing permit (s89); statutory regulation of the Fisheries (Commercial Fishing) Regulations (2001), which, inter alia, prohibit the use of nets with a mesh size less than 100mm mesh size (s 71), vessel length restrictions; requirements to furnish returns, Fisheries (Reporting) Regulations 2001 and voluntary measures imposed by the HFMC code of practice (CoP). Voluntary measures include the requirement to redirect fishing effort to another area at least 3 nautical miles away for at least three days should more than 10 percent of the hoki catch be less than 60 cm total length; towing at depths more than 450 m and avoiding areas where small hoki are known to aggregate.

The future of hoki management

The comprehensive stock assessment programme that has been underway for several years is to continue. Stock assessments and other research are funded by government levies and contracted out by MFish to external science providers and are also directly purchased by the industry. Examples of proposed future research projects initiated by MFish include the following.

The HFMC has stated5 that they intend to develop a fishery plan to better manage the hoki fishery. Under s11A of the Fisheries Act there is provision for each fishery to develop an alternative management regime to the necessarily broad plan administered by MFish.

5 See their website <>.

According to the HFMC, “The new management plan …will build on the current model and should manage the fishery in a long-term sustainable way as well as monitoring the commercial viability of the fishery. The new plan will undoubtedly provide more detail on procedures for such things as mitigating bycatch, trawling methods, ecological risk assessments and processing stock assessments in a timelier manner. However, developing the plan will involve scrutiny of changes to process, outcomes and information flows, through a process of consultation and with the agreement of relevant government bodies.”This ambitious goal is consistent with government policy, namely to devolve responsibilities to the industry who, with the right frameworks in place, are better able to manage the fisheries resources at a micro-level than government can hope to achieve.


The Strategic Plan of the Ministry of Fisheries describes the goal of fishery management as to maximize the value New Zealanders’obtain through the sustainable use of fisheries resources and to protect the aquatic environment. Pearce (1991) suggested that the Ministry transfer the responsibility for managing the resource users’operations to those who hold the resource rights and charge government with the responsibility for protecting the broader public interest in resource conservation and environmental protection.

The future governmental role will be to act in the public’s interest by setting the ground rules to ensure sustainability of marine species based on scientific and environmental standards. The Government can provide an enabling framework to stimulate innovative opportunities for fishery utilisation as well as environmental protection. The Ministry of Fisheries can identify the public interest affected by fishing and protect the public interest by defining enforceable ground rules upon which those who have the rights to fish can organise themselves and exercise their rights. Public interests include stock conservation, protection of seabed habitat, seabirds, marine mammals, biodiversity and biosecurity. The Ministry must design specifications that set out the constraints on how resources are to be used. Performance standards will be developed to facilitate decisions about harvesting, stock management and allocation to those who hold the property rights.

Government objectives for fishery management are outlined in the 1996 Fisheries Act to provide for the utilisation of fisheries resources while ensuring sustainability. Three strategies are designed to achieve the government objectives: (a) protect the health of the aquatic environment, (b) enable people to get the best value from sustainable and efficient use of fisheries and (c), ensure the Crown delivers on its obligations to Maori with respect to fisheries.

Operational objectives

The Ministry of Fisheries is initiating an evaluation of its present fishery management system with the expectation of identifying and changing policies that do not meet the objectives for the governments’future vision for fishery management. Specifically, deepwater fisheries policies will be assessed with respect to how they achieve the objectives of sustainability, environmental effects and utilisation. This evaluation will encompass the services that fishery mangers rely on for specific stock management. Services such as research, scientific observation, compliance, enforcement, monitoring, education, cost recovery and deemed values are to be reviewed with respect to the roles government, industry and other stakeholders play.

A stock strategy will provide an outline of the Government’s ground rules and statutory requirements with respect to specific stocks under the 1996 Fisheries Act. It will provide the Ministry’s direction for the future of specific fisheries. It will demonstrate the extent to which the Government will participate in the prosecution of a fishery and outline where further opportunities exist for stakeholders to maximize the value of their wellbeing in fisheries. It is the intent of the Ministry of Fisheries to provide the operational framework that can facilitate the long-term development of such stakeholder fish plans.


Anderson, O.F., D.J. Gilbert & M.R. Clark 2001. Fish Discards and Non-Target Catch in the Trawl Fisheries for Orange Roughy and Hoki in New Zealand Waters for the Fishing Years 1990–91 to 1998–99. New Zealand Fisheries Assessment Report 2001/16.

Annala, John H. 1996. New Zealand’s ITQ system: Have the first eight years been a success or a failure? Reviews in Fish Biology and Fisheries, 6: 43–62.

Batstone, C. J. & B. M.H. Sharp 1999. New Zealand’s quota management system: The first ten years. Marine Policy, 23(2): 177–90.

Francis, R.I.C.C. 1992. Use of risk analysis to assess fishery management strategies: A case study using orange roughy (Hoplostethus atlanticus) on the Chatham Rise, New Zealand. Canadian Journal of Fisheries and Aquatic Science, 49: 922–930.

Koslow, J.A. 1997. Seamounts and the ecology of deep sea fisheries. American Scientist, Vol. 85: 168–176.

Koslow, J.A., G.W. Boehlert, J.D.M.Gordon, R.L. Haedrich, P. Lorance & N. Parin 2000. Continental slope and deep-sea fisheries: Implications for a fragile ecosystem. ICES Journal of Marine Science, 57:548–557.

SeaFIC 2002. New Zealand Seafood Exports Seafood Exports Calendar year to December 2002. By species by year by country. Report 7. New Zealand Seafood Industry Council.

Pearce, P. 1991. Developing property rights as instruments of natural resources policy: The case of the Fisheries. Environmental Economics (91) 30. Chapter 5.

Sissenwine, M.P. & P.M. Mace 1992. ITQ’s in New Zealand -the era of fixed quota in pertetuity. Fishery Bulletin, 90(1): 147–60.

Hyperdepletion in orange roughy fisheries

A. Hicks
University of Washington
School of Aquatic and Fishery
PO Box 355 020
Seattle, WA, USA

Catch per unit effort (CPUE) is commonly used as an index of abundance in stock assessments. However, it is possible that CPUE declines more slowly than abundance (hyperstability), or more rapidly than abundance (hyperdepletion). Harley, Myers and Dunn (2001) compiled CPUE and independent abundance data from the International Council for the Exploration of the Sea (ICES) and found evidence supporting hyperstability. However, hyperdepletion could not be completely ruled out and was most evident in hake fisheries.

In this study, the linearity between abundance (N) and CPUE (U) was tested using a log-transformed power function within a measurement model assuming independent errors, i.e.,

log(U) = log(q) + β log(N)


Estimates were found using maximum likelihood and confidence intervals for β were found with likelihood profiles.

Data from six orange roughy stocks were used to obtain CPUE and fishery independent abundance estimates: three from New Zealand, one from Australia and two from Namibia. Fishery independent abundances were either the results from stock assessments, which did not use CPUE as the dependent variable, research trawl surveys or research acoustic surveys. Within some stocks, there were alternative estimates of fishery independent abundance (i.e. different assumptions in a stock assessment were available), resulting in multiple analyses for that stock. Inconsistencies with the Namibian CPUE data precluded that stock from being analysed; thus only the New Zealand and Australian stocks were analyzed with the measurement error model.

The CPUE from three of the four analyzed stocks showed significant hyperdepletion when compared to the biomass outputs from the stock assessments. However, results from direct comparison of CPUE to survey data were not as significant, which may be attributed to fewer paired CPUE/abundance observations. The observation of hyperdepletion instead of hyperstability may be surprising, but it is consistent with orange roughy fisheries experience that in some cases stocks disappear as soon as fishing starts, but in places like Namibia, when fishing stops the fish come back. Further, hyperdepletion can occur in fisheries that are characterized by localized high density aggregations and a large low density abundance (Hilborn & Walters 1992), similar to hill fisheries typified by three of the four stocks analyzed. In contrast, the stock that showed a linear trend between CPUE and abundance consisted of a fishery targeting a single large spawning aggregation over flat ground. It appears that hyperdepletion is common in orange roughy fisheries and is related to the nature of the fishery. Understanding and accounting for hyperdepletion will have important consequences in the management of these fish.


Harley, S.J., R.A. Myers & A. Dunn 2001. Is catch-per-unit-effort proportional to abundance? Can J.Fish. & Aquat. Sci. 58:1760–1772.

Hilborn, R. & C.J.Walters 1992. Quantitative fisheries stock assessment: choice, dynamics, and uncertainty. Chapman and Hall, New York, 570 pp.

Issues in the management of high seas orange roughy fisheries in the New Zealand region

J. Annala1 and M. Clark2
1Ministry of Fisheries
P O Box 1020, Wellington, New Zealand
1National Institute of Water & Atmospheric Research
Private Bag 14–901, Wellington, New Zealand
<> (corresponding author)


Orange roughy are a common deepwater species in the southwest Pacific and fisheries for this species developed in New Zealand and Australian waters in the 1980s and 1990s (Clark 2001, Koslow et al. 1997). In addition to the major fisheries inside the 200 mile EEZs of these countries, a number of fisheries occur on the high seas in the region, on the Lord Howe Rise, Northwest Challenger Plateau, West Norfolk Ridge, South Tasman Rise and Louisville Ridge (Figure 1).

The New Zealand region, showing location of major fisheries for orange roughy outside New Zealand and Australian EEZs

(1000 m depth contour shown around New Zealand)


These fisheries have at times been substantial with total catches reaching over 15000t and regularly being around 4000–5000 t (Figure 2). The fisheries have been largely unregulated. In this paper we briefly describe these fisheries, trends in catch and effort, and discuss how the fisheries have fared in the absence of management and consider some issues for their future sustainability and management.


2.1 Lord Howe Rise

The Lord Howe Rise extends from the northwestern margin of the Challenger Plateau, off the west coast of New Zealand, out to Lord Howe Island in the western Tasman Sea. The ridge is mostly in international waters, although it does extend into both the Australian and New Zealand EEZs. A major fishery for orange roughy developed on the Lord Howe Rise in 1988. A number of countries fished the area in the late 1980s, but during the 1990s it was fished mainly by New Zealand and Australian vessels. Annual catches were initially as high as 4 000 t, but in recent years the fishery has become small with catches between 100 and 300 t.

Estimated catches of orange roughy from the high seas fisheries in the New Zealand region

Data are primarily from the New Zealand Ministry of Fisheries and Australian Fisheries Management Authority. 1988 on the x-axis, e.g. is the 1987–88 fishing year.


Unstandardised CPUE indices from the Lord Howe Rise orange roughy fishery1988–1989 to 2002–2003

1990 on x axis is the 1989–90 fishing year (from Clark 2004)


Catch-per-unit-effort (CPUE) has also decreased over time. Unstandardized and standardized CPUE analyses have been carried out on New Zealand data (Clark and O’Driscoll 2002, O’Driscoll 2003) and show a peak in the early 1990s, declining rapidly to low levels and remaining relatively consistent at less than 1 t/tow to the present (Figure 3). The number of vessels in the fishery has fluctuated between 4 and over 20 and effort has also varied with time of year. It has made CPUE hard to use confidently as an index of abundance, but this mesure still indicates a strong decrease in the stock size.

There has been no management of the fishery, although New Zealand and Australia have been discussing management options since the early 1990s. The low catch rates in the fishery have led to a sporadic distribution, and level, of effort, and many of the New Zealand vessels involved in the fishery in the last ten years have been small boats, which often hold little deepwater quota inside the New Zealand EEZ.

2.2 Northwestern Challenger Plateau

The northwestern corner of the Challenger Plateau has a large number of small seamount-like features arising from the continental slope, and these and flat ground in between have been the focus of this fishery since the early 1990s. As catch rates on the Lord Howe Rise started to decline, effort shifted to this region and it has seen more effort and larger catches than Lord Howe. Over 60 different vessels have fished the area since 1991 making over 11000 trawls. There have been some marked changes in the distribution of fishing over time. In the first two years, most trawling was carried out on flat ground, but tows then became progressively shorter as the vessels increased their targeting on the small seamounts. Over the last three years effort has shifted back to the flat area and average tow duration has increased from one hour/tow to over four hours/tow.

The maximum annual catch was 2900t in 1992–93 and although catch levels of 1000–2000t have occurred in recent years, this has been with almost double the number of tows undertaken in the early-mid 1990s. Catch-per-unit-effort has decreased over time. Unstandardized and standardized CPUE analyses have been carried out on New Zealand data (Clark and O’Driscoll 2002, O’Driscoll 2003), and show mixed results. Unstandardized CPUE over the entire year peaked in the mid-1990s, then declined to the present (Figure 4). Standardized CPUE showed an increasing trend over time (O’Driscoll 2003), but when core vessels (those consistently involved in the fishery) in the June period (when effort had been most constant on the small seamounts) were excluded the pattern was similar to the unstandardized result.

CPUE indices from the Northwest Challenger orange roughy fishery 1992–1993 to 2002–2003

Left panel: Unstandardized for all vessels and full year (Clark 2004). Right panel: Standardized and un-standardized indices for selected core vessels in June (O’Driscoll 2003). 1992 on x axis is the 1991–1992 fishing year.


Stock size appears to have decreased over time, especially the component that resides on the small seamounts where orange roughy spawn or feed. Catch rates have declined less than on Lord Howe, although this appears partly to be a result of the fleet’s spreading effort over a wider area. There has been no management, although New Zealand and Australia have discussed management options for this fishery as well as for the Lord Howe Rise region.

2.3 West Norfolk Ridge

New fishing grounds have recently developed on the West Norfolk Ridge, which runs northwest from the North Island towards New Caledonia. This comprises a chain of ridge peaks and seamount features both within and beyond the New Zealand EEZ.

Up to 15 Australian and New Zealand vessels have been involved in this fishery, which yielded almost 700 t in 2001–2002. Most of this came from three small seamount features. However, it appears to have declined almost as rapidly as it developed, with a decrease in catch to 40 t in 2002–2003. The average catch per tow has decreased from 3 t/tow in 2001–2002 to 0.8 t/tow in 2002–2003.

2.4 South Tasman Rise

The South Tasman Rise is a prominent ridge extending south from Tasmania into the Southern Ocean. It has a series of small peaks near its main summit at about 900 m in depth just outside the Australian 200 mile Fishing Zone.

The orange roughy fishery started in September 1997 and expanded rapidly as more Australian and New Zealand boats entered the fishery. An estimated 3900 t was caught in the 1997–1998 fishing year. Reported catches increased to over 4000 t in 1999–2000. One Belize and three South African flagged vessels fished for a period during the 1999winter, but no other non-Australasian vessels are known to have fished the region. Oreos were previously taken as bycatch in the fishery with over 1000 t caught in both 1997–1998 and 1998–1999. Catches have dropped markedly since then, to around 100–200 t a year, but were greater than the orange roughy catch in 2003–2004.

CPUE analyses have been difficult because the fishery has varied greatly between years in levels of effort and the times of the year when fishing occurred. Average unstandardized catch rates have decreased from over 3 t/tow in 1999–2000, to less than 1 t/tow. Standardized CPUE indices have also decreased substantially (CSIRO, unpublished data).

In contrast to the other fisheries covered here, attempts were made early on to apply a management regime. The fishery has been regulated by a Memorandum of Understanding (MOU) between Australia and New Zealand since early 1998 (Tilzey2000). A precautionary TAC of 2100 t was agreed on for the period 1 March 1998 to 28 February 1999, but the MOU was not renewed for the 1999–2000 fishing season. The TAC was subsequently increased to 2400 t for the 2000–2001 fishing season, before being reduced in 2002–2003 to 1800 t, and subsequently to 800 t and 600 t for 2003–2004 and 2004–2005 respectively (Table 1).

2.5 Louisville Ridge

The Louisville Ridge is a chain of seamount and guyot features extending southeast for over 4000 km from the Kermadec Ridge. It is a “hotspot”chain of more than 60volcanoes, most of which rise to peaks of 200–500 m from the surrounding seafloor at depths around 4000 m. The Ridge is outside the New Zealand EEZ and thus in international waters.

Table 1
Quota levels (under the MOU) and orange roughy catch (t) in the South Tasman Rise fishery

Fishing yearQuota (t)Combined Australian and New Zealand catch
1997–1998None3 930
1998–19992 1001 940
1999–2000None3 620
2000–20012 400830
2001–20022 400170
2002–20031 800110

The fishery started in the 1993–1994 fishing year. Reported catches by New Zealand and Australian vessels rose from about 700 t in that year to over 13 000 t the following year. Since then annual catches have fluctuated between 1000 and 3000 t. Vessels from other countries have fished at times, but catches are not thought to have ever been substantial. There have been strong seasonal trends between years in catch and effort. Initially effort in the fishery was spread over much of the year, but from 1998–1999 onwards effort has been heavily concentrated in June, July and August. The distribution of catches has also varied between years. The fishery initially developed in the central region in 1994–1995, with other grounds quickly developing in the northern region of the Ridge. Southern seamounts also yielded good catch rates from 1995–1996. Over the last three years effort has decreased in the central region and good catch rates have occurred on fewer seamounts. Fishing success on the northern-most seamounts has also been reduced, but fishing has expanded to more features in the southern area (Clark 2004).

Unstandardized CPUE has been examined, based on mean catch per trawl for the total Louisville Ridge area, and the three main regions separately (Clark 1999, Clark and Anderson 2001). Most fishing grounds showed reductions in CPUE from peak values in the first two to three years, but the pattern has differed between the three broad divisions of the Ridge (Figure 5). CPUE on individual seamounts has also been examined and again shows highly variable levels between years.

The Louisville fishery has not been subject to any management measures. Catches, and presumably stock size, have, in general, decreased over time and there are indications that some seamounts have been overfished while other new ones have been found to maintain catches. However, overall the fishery appears more stable than those in the Tasman Sea.


Most fisheries on the high seas around New Zealand have a history of rapid development and then rapid decrease after several years of high catches. With the exception of the South Tasman Rise, none have had any management measures applied. However, despite this lack of management, fisheries on the Lord Howe Rise, northwestern Challenger Plateau and Louisville Ridge have persisted, and average catch rates in recent years have been relatively stable. Catch levels in the West Norfolk Ridge fishery have decreased dramatically in just two years and this may have been because it was based on a small and localized stock.

Stock assessment has been problematic with all these fisheries. There have been no fishery-independent surveys undertaken, as research on stocks within the New Zealand EEZ have taken priority for limited institutional resources. Our knowledge and assessment of changes in the stocks have been based on catches and CPUE. These in turn are affected by the consistency of both the fishing fleet characteristics and fishing practices in the fisheries. Both have typically been highly variable. Fishing outside the EEZ has tended to be undertaken by small New Zealand vessels that had insufficient quota inside the EEZ to fully utilize vessel capacity. This has varied considerably between years and between fishing grounds. A clear example of this is with theLord Howe Rise, where the total datset examined for CPUE analyses by O’Driscoll (2003) comprised about 3400 tows from 53individual vessels. However, the distribution across years, seasons, vessels and smaller fishing areas, was patchy, and only seven vessels met inclusion criteria (based on number of tows and number of years in the fishery); none of these were in the fishery during its early years and, although these vessels accounted for almost 60 percent of all tows in the fishery, combined they carried out fewer than 100 tows in 11 of the 14 years of the fishery. This restricts the usefulness of detailed catch-effort analyses and for the Lord Howe Rise CPUE has not been accepted as a measure of abundance. Data from the Northwest Challenger Plateau and Louisville Ridge are more stable although, even with these fisheries changes in the vessel composition over time and the areas fished between years, they pose difficulties. The lack of reliable abundance estimates or indices for any of the fisheries has restricted any planning for appropriate management.

Unstandardized CPUE (t/tow) by area by year for all months (heavy line) and for the winter period (thin line)


Orange roughy can form dense aggregations for spawning or feeding, which enables high commercial catch rates even as stock size is declining. This makes the species vulnerable to overexploitation. In addition, roughy are slow-growing and long-lived and sustainable exploitation rates are low. Hence recovery from overfishing should take a long time. These characteristics make it important that management occurs from an early stage of any fishery to avoid overfishing. The examples of some orange roughy fisheries in New Zealand, Australia, Namibia and the Indian Ocean that have developed rapidly and then declined to low levels with serial depletion of some fishing grounds (e.g. Branch 2001, Clark et al. 2000) highlight the need for immediate management. In a meta-analysis of New Zealand seamount fisheries, Clark, Bull and Tracy (2001) examined the likely stock sizes of individual seamount-like features and concluded that almost all around New Zealand were of the order of several hundred to 3000 t, with only a few over 10000t. Stocks of this size can be quickly depleted, and long-term sustainable catch levels are only 50–150 t/year.

The Louisville Ridge fishery has continued to sustain a catch of several thousand tonnes, despite concerns when a large amount of effort was being applied to the fishery in the mid-1990s. However, the distance from the nearest country (New Zealand) may have played an important role, as a three to four day steam is a major commitment by a fishing vessel. If catch rates are low, or catches small, such trips can be uneconomic. Catch rates of orange roughy may be affected by heavy trawling breaking up aggregations and often a fishery is more successful when fewer vessels are involved. Although unregulated, the Louisville fishery may have reached such a balance, where the stock’s size, distribution, number of vessels and effort and catch levels have reached an equilibrium.

Several of these factors were considered when the South Tasman Rise fishery developed in 1997. A memorandum of understanding was agreed upon in a short space of time between New Zealand and Australian governments. Issues of catch history, in the first regulation of deepwater fishing activity on the high seas for Australasian vessels, were used to determine a mutually agreed catch division. Compliance and other countries fishing were all problems in the first years and in 1999–2000 the quota was substantially exceeded. There were discussions on research between New Zealand and Australia, and an exploratory research programme was implemented to establish stock structure, and the distribution of the fishery to allow planning for an acoustic abundance survey. However, the survey never happened as catches started to drop and it became clear that the fishery may be relatively small. Quota levels have now been considerably reduced (although they have lagged behind catches) and there is a trigger mechanism that will increase the available catch if aggregations re-appear. The management of this fishery was less than optimal as the initial TAC was too high to be precautionary. However, a lot has been learnt about what is required for effective orange roughy management (Francis and Clark 2005). Central to these factors on the high seas are cooperation between governments, fishers, scientists and managers to limit catches early on until more is known of the stock distribution and until fishery-independent estimates of biomass are made.


Research on the New Zealand fisheries on the high seas has been funded by the New Zealand Ministry of Fisheries. Data on catch and effort by Australian vessels were provided by the Australian Fisheries Management Authority and Bureau of Resource Sciences.


Branch, T.A. 2001. A review of orange roughy Hoplostethus atlanticus fisheries, estimation methods, biology and stock structure. In: Payne, A.I.L., Pillar, S.C., and Crawford, R.J.M. (Eds). A decade of Namibian Fisheries Science. South African Journal of Marine Science 23: 181–203.

Clark, M.R. 1999. An update and standardized analysis of New Zealand commercial catch and effort information for the orange roughy (Hoplostethus atlanticus) fishery on the Louisville Ridge. NZ Fisheries Assessment Research Document 99/47. 17 pp.

Clark, M.R. 2001. Are deepwater fisheries sustainable? -the example of orange roughy (Hoplostethus atlanticus) in New Zealand. Fisheries Research 1195: 1–13.

Clark, M.R. 2004. Descriptive analysis of orange roughy fisheries in the New Zealand region outside the EEZ: Lord Howe Rise, Northwest Challenger Plateau, West Norfolk Ridge, South Tasman Rise, and Louisville Ridge to the end of the 2002–03 fishing year. New Zealand Fisheries Assessment Report 2004/51. 36 pp.

Clark, M.R. & O.F. Anderson 2001. The Louisville Ridge orange roughy fishery: an update of commercial catch-effort data and CPUE analysis of the fishery to the end of the 1999–2000 fishing year. New Zealand Fisheries Assessment Report 2001/74. 31 pp.

Clark, M.R. & R.L. O’Driscoll 2002. Descriptive analysis of orange roughy fisheries in the Tasman Sea outside the New Zealand EEZ: Lord Howe Rise, Northwest Challenger Plateau, and South Tasman Rise from 1986–87 to the end of the 2000–2001 fishing year. New Zealand Fisheries Assessment Report 2002/59. 26 pp.

Clark, M.R., B. Bull & D.M. Tracey 2001. The estimation of catch levels for new orange roughy fisheries on seamounts: a meta-analysis of seamount data. New Zealand Fisheries Assessment Report 2001/75. 40 pp.

Clark, M.R., O.F. Anderson, R.I.C.C. Francis & D.M. Tracey 2000. The effects of commercial exploitation on orange roughy (Hoplostethus atlanticus) from the continental slope of the Chatham Rise, New Zealand, from 1979 to 1997. Fisheries Research 45(3):217–238.

Francis, R.I.C.C. & M.R. Clark 2005. Sustainability issues for orange roughy fisheries. Bulletin of Marine Science.(76)2: 337–351.

Koslow, J.A., N.J. Bax, C.M. Bulman, R.J. Kloser, A.D.M. Smith & A. Williams 1997. Managing the fishdown of the Australian orange roughy resource. p. 558–562. In Hancock, D.A., D.C. Smith, A. Grant, and J.P. Beumer (Ed). Developing and sustaining world fisheries resources: the state of science and management. CSIRO Australia.

O’Driscoll, R.L. 2003. Catch-per-unit-effort analysis of orange roughy fisheries outside the New Zealand EEZ: Lord Howe Rise and Northwest Challenger Plateau to the end of the 2001–02 fishing year. New Zealand Fisheries Assessment Report 2003/36. 38 pp.

Tilzey, R. 2000. South Tasman Rise trawl fishery. In: Caton, A., McLoughlin, K. (Eds) Fishery status reports 1999: resource assessments of Australian Commonwealth fisheries. BRS, Canberra.

Description of the age-structured production model (ASPM) assessment methodology used for the Namibian orange roughy (Hoplostethus atlanticus) and the South African Prince Edward Islands Patagonian toothfish (Dissostichus eleginoides)

A. Brandão and D.S. Butterworth
Marine Resource Assessment and Management Group (MARAM)
Department of Mathematics and Applied Mathematics
University of Cape Town
Rondebosch 7701, South Africa


Butterworth and Brandão (2005) describe the historical development and assessments of the present status of the Namibian orange roughy (Hoplostethus atlanticus)and the Patagonian toothfish (Dissostichus eliginoides) off South Africa’s sub-Antarctic Prince Edward Islands. This paper sets out the methodologies, based on the age-structured production model (ASPM) approach, which have been used to assess these resources. Common methodology is set out under general headings, within which the differences in the methodology applied for the two stocks are detailed under sub-headings for each.

While the methodology applied for the toothfish is relatively standard, that for the orange roughy incorporates novel features to deal with the “intermittent aggregation”hypothesis that is considered for this resource.


Both the orange roughy and toothfish dynamics are described by the equations.

Ny+1,a+1 = (Ny,a - Cy,a)e -M0 ≤ am-2(2)
Ny+m = (Ny,m - Cy,m)e -M + (Ny,m-1 - Cy,m-1)e -M(3)


Ny,ais the number of fish of age a at the start of year y
Cy,ais the number of fish of age a taken by the fishery in year y
is the Beverton-Holt stock-recruitment relationship described by equation (15) below
is the spawning biomass at the start of year y
Mis the natural mortality rate of fish (assumed to be independent of age) and
mis the maximum age considered (i.e. the “plus group”).

In the interests of simplicity, this approximates the fishery as a pulse fishery at the start of the year. Given that both orange roughy and toothfish are relatively long-lived with low natural mortality, such an approximation would seem justifiable.

The number of fish of age a caught in year y is given by:

C y,a = N y,aSaFy(4)


Fy is the proportion of the resource above age a harvested in year y and

Sa is the commercial selectivity at age a.

The mass-at-age is given by the combination of a von Bertalanffy growth equation for length-at-age l(a) defined by constants l, κ and and a relationship relating length to mass. Note that here l refers to standard length:

ℓ(a) = ℓ [1 - e -K(a-to)](5)
W a = Cℓ(a)d(6)


wa is the mass of a fish at age a.

The total catch by mass in year y is given by:


which can be re-written as:



Orange roughy

The commercial fishing selectivity, Sa, for orange roughy is assumed to be knife-edge and given by:



ar is the age at recruitment to the fishery, assumed equal to the age at maturity (am) for the Namibian orange roughy populations.


The commercial fishing selectivity (assumed to vary from year to year), Sy,ar, for toothfish is assumed to be described by a logistic curve, modified by a decreasing selectivity for fish older than age ac. This is given by:



a50,yis the age-at-50% selectivity (in years) in year y, where
The are assumed to be normally distributed with mean zero and standard deviation σ a50
δdefines the steepness of the ascending section of the selectivity curve (in years-1)
ωydefines the steepness of the descending section of the selectivity curve for fish older than age ac in year y, where and are assumed to be normally distributed with mean zero and standard deviation .

In cases where equation (8) yields a value of Fy > 1 for a future year, i.e. the available biomass is less than the proposed catch for that year, Fy is restricted to 0.9, and the actual catch considered to be taken will be less than the proposed catch. This procedure makes no adjustment to the exploitation rate (Sy,a Fy) of other ages. To avoid the unnecessary reduction of catches from ages where the TAC could have been taken if the selectivity for those ages had been increased, the following procedure is adopted for the toothfish fishery (CCSBT 2003):


The modified selectivity is denoted by ,



so that




Now Fy is not bounded at one, but g(Sy,a Fy) ≤ 1 hence Cy,a = g(Sy,aFy) Ny,a ≤ Ny,a as required.


The spawning biomass in year y is given by:



fa is the proportion of fish of age a that are mature (assumed here to be knife-edge at age am).

The number of recruits at the start of year y is assumed to relate to the spawning biomass at the start of year y by the Beverton-Holt stock-recruitment relationship (assuming deterministic recruitment):


The values of the parameters α and β can be calculated given the initial spawning biomass Ksp and the steepness of the curve h, using equations (16) – (20) below. If the initial (and pristine) recruitment is R0 = R(Ksp), then steepness is the recruitment (as a fraction of R0) that results when spawning biomass is 20% of its pristine level, i.e.:

hR0 = R(Ksp)(16)

from which it can be shown that:


Rearranging equation (17) gives:


and solving equation (15) for α gives:


In the absence of exploitation, the population is assumed to be in equilibrium. Therefore R0 is equal to the loss in numbers due to natural mortality when Bsp = Ksp, and hence:





Given a value for the pre-exploitation spawning biomass (Ksp) of a fish stock and the assumption that the initial age structure is at equilibrium, it follows that:


which can be solved for Ro.

The initial numbers at each age a for the trajectory calculations corresponding to the deterministic equilibrium, are given by:


Numbers-at-age for subsequent years are then computed by means of equations (1) -(4) and (7) -(15) under the series of annual catches given.

The model estimate of the exploitable component of the biomass for orange roughy is given by:


In the case of the toothfish assessment, the model estimate of the exploitable component of the biomass is adjusted for the fact that some estimated selectivities never reach one. This adjustment is given by:


where n’ is the number of the summed terms (15-5+1).


The age-structured production model is fitted to all available indices of abundance to estimate model parameters. The likelihood function is calculated assuming that each observed abundance index is lognormally distributed about its expected value:



is the ith abundance index for year y (for orange roughy these consist of the GLM standardized CPUE, the acoustic and the research swept-area abundance indices whereas for toothfish only CPUE are available)

is the corresponding model estimate


is the model estimate of exploitable biomass of the resource for year y and
qiis the catchability coefficient for the ith abundance index and
εyis normally distributed with mean zero and standard deviation .

Orange roughy

Under the “catch-induced”hypothesis for the decline in indices of abundance described in Butterworth and Brandão (2005), the negative of the penalized log-likelihood (ignoring constants), which is minimized in the fitting procedure (results given in Figure 6 of Butterworth and Brandão (2005) consists of the combination of the contribution of each abundance index plus the contribution of the penalty placed on the natural mortality parameter as well as the contribution of the penalty applied to the multiplicative bias of the acoustic abundance series (both these penalties are frequentist analogs of Bayesian priors for these quantities). The (penalized) negative log-likelihood function minimized for each aggregation is thus given by:



qAC is the remaining multiplicative bias of the acoustic abundance series whose maximum likelihood estimate is given by:

qSA is the ‘catchability coefficient’for the research swept-area abundance indices, whose maximum likelihood estimate is given by:

qCPUE is the catchability coefficient for the standardized commercial CPUE abundance indices, whose maximum likelihood estimate is given by:

is the standard deviation of the penalty function applied to qAC, which is input; its value is the CV of the distribution of the product of the systematic bias factor distributions applied to the acoustic abundance indices
qestis the mean of the penalty function applied to qAC, whose value is taken to be equal to one as the distribution of the bias factors for the acoustic estimate have been defined in such a way that the corrected acoustic estimate is intended to be an unbiased estimate of abundance
Mis the natural mortality rate
Mesis the mean of the penalty function applied to M (i.e. the prior distribution mean), which is input
is the standard deviation of the log acoustic abundance estimate for year y, which is input and is given by


is the CV of the sampling error distribution and
is the CV of the distribution of the product of the random bias factor distributions taken to apply to the acoustic abundance indices,
is the standard deviation of the log research swept-area abundance index for year y, which is input and is given by the sampling CV of the research swept-area index of relative abundance,
σCPUEis the standard deviation of the standardized CPUE series (assuming homoscedasticity of residuals), whose maximum likelihood estimate is given by


is the acoustic series estimate for year y
is the research swept-area series index for year y
is the standardized CPUE series index for year y and
nCPUEis the number of data points in the standardized CPUE abundance series.

The estimable parameters of this model are qAC, qSC, qCPUE, KSP, σCPUE and M.

Under the “intermittent aggregation”hypothesis (Butterworth and Brandão 2005) estimable multiplicative bias factor Xy is included in the model so that the various terms in equation (26) become


where i represents the type of abundance index in the likelihood; for example, i = AC, when dealing with the acoustic abundance index and so on. This x factor allows for the possibility that not all the orange roughy belonging to an aggregation are present at that site each year.

The results of the acoustic survey carried out in 2002 for the Frankies ground (closed to commercial fishing since 1999) show an index of abundance for 2002 that is in the region of the 1997 estimate (Brandão and Butterworth 2003) indicating that the low indices of abundance observed in years subsequent to 1997 cannot be interpreted as purely a consequence of fishing down of the population, but instead reflect that the extent of aggregation of the stock that occurs from year to year is variable. A penalty function is applied for the proportion of stock present (xy) to reflect an assumption that these proportions follow a beta distribution, which is restricted to the pertinent [0,1] range by construction. Thus the following term is added to the negative of the log‑likelihood function given in equation (26) (in which the terms corresponding the the abundance indices are modified to reflect equation (29)):


N   is the total number of years considered in the assessment
η   is a parameter of the beta distribution, with η > 0
φ   is a parameter of the beta distribution, with φ > 0.

Clearly the data cannot uniquely determine both η and φ, as this model would then not preclude an interpretation of the data that reflects an enormous resource little of which aggregates each year. Thus for precautionary management purposes, these two parameters were constrained to yield a proportion present in 1997 of at least 80% for the Frankies aggregation, and the remaining degree of freedom is used to match the spread of the distribution of annual proportional aggregation evident from the abundance indices. Only Frankies provides sufficient data contrast to be able to effect such estimation, so that the values of η and φ obtained there were taken to apply to the other aggregations.


The age-structured production model is fitted to the GLM standardized CPUE of toothfish to estimate model parameters. The likelihood is calculated assuming that the observed CPUE abundance index is lognormally distributed about its expected value as indicated by equation (25).

The negative log-likelihood function (ignoring constants), which is minimized in the fitting procedure (results given in Figure 10 of Butterworth and Brandão (2005)), consists of the combination of the contribution of the CPUE abundance index plus the contribution of the penalty placed on the parameters a50,y and ωy of the selectivity function given by equation (10). The negative log-likelihood function is thus given by:


The estimable parameters of this model are and σCPUE. The values of σa50 and σω were fixed at 1.0 and 0.3respectively. Closed form estimates for qCPUE and σCPUE are given by equations (27) and (28) respectively.


The model above provides estimates of the catch-at-age (Cy,a) by number taken by the fishery each year from equation (4). These in turn can be converted into proportions of the catch of age a:


Using the von Bertalanffy growth equation (5), these proportions-at-age can then be converted to proportions-at-length -here under the assumption that the distribution of length-at-age remains constant over time:



Aa,ℓ is the proportion of fish of age a that fall in length group ℓ. Note that there-fore:

for all ages a(34)

The A matrix has been calculated here under the assumption that length-at-age is normally distributed about a mean given by the von Bertalanffy equation, i.e.:



N* is a normal distribution truncated at ± 3 standard deviations (to avoid negative values) and

θ(a) is the standard deviation of length-at-age a, which is modelled here to be proportional to the expected length at age a, i.e.:

θ(a) = τ ℓ {1-e-k(a-t0)}(36)

with τ a parameter estimated in the model fitting process.

Note that since the model of the population’s dynamics is based upon a one-year time step, the value of τ and hence the θ(a)'s ’s estimate will reflect not only the real variability of length-at-age, but also the “spread”that arises from the fact that fish in the same annual cohort are not all spawned at exactly the same time and that catching takes place throughout the year so that there are differences in the age (in terms of fractions of a year) of fish allocated to the same cohort.

The model is fitted by adding the following term to the negative log-likelihood of equation(31):



is the proportion by number of the catch in year yin length group ℓ, and
σlenhas a closed form maximum likelihood estimate given by:

Equation (37) makes the assumption that proportions-at-length data are log-normally distributed about their model-predicted values. The associated variance is taken to be inversely proportional to py,ℓ to reduce the contributions from expected small proportions, which will correspond to small observed sample sizes. This adjustment (originally suggested to us by A.E. Punt) is of the form to be expected if a Poisson-like sampling variability component makes a major contribution to the overall variance. Given that overall sample sizes for length distribution data differ quite appreciably from year-to-year, subsequent refinements of this approach may need to adjust the variance assumed for equation (37) to take this into account.

The wlen weighting factor may be set at a value less than 1 to reduce the contribution of the catch-at-length data to the overall negative log-likelihood compared to that of the CPUE data in equation (31). The reason that this factor is introduced is that the data for a given year frequently show evidence of strong positive correlation, and so are not as informative as the independence assumption underlying the form of equation (37) would otherwise suggest.

In the practical application of equation (37), length observations were grouped by 2 cm intervals, with minus - and plus-groups specified below 54 and above 138 cm respectively, to ensure values in excess of about 2% for these cells.


Brandão, A. and D.S. Butterworth 2003. Stock assessment of Namibian orange roughy populations using an age-structured production model and all available indices of abundance from 1994 to2002, and making allowance for annually variable aggregation of the stocks. Unpublished report. National Marine Information and Research Centre, Namibia. DWFWG/Wkshop/Mar03/doc3: 39pp.

Butterworth, D.S. and A. Brandão 2005. Experiences in Southern Africa in the management of deep-sea fisheries. In R. Shotton (Ed), Deep Sea 2003: Conference on the Governance and Management of Deep-sea Fisheries. Queenstown, New Zealand, 1–5 December 2003. FAO Fisheries Proceedings. No. 3/1. Rome, FAO. 2005. Proceedings of the Deep Sea Conference: 226–234.

CCSBT (Commission for the Conservation of the Southern Bluefin Tuna) 2003. Report of the Second Meeting of the Management Procedure Workshop. Queenstown, New Zealand, March 2003.

High seas resource management: some discussion of the Madagascar ridge, Western Indian Ocean

M. Mayekiso, A. Naidoo and R. Leslie
Department of Environmental Affairs & Tourism
Private Bag X2, Roggebaai, Cape Town, South Africa

“Conscious that the problems of ocean space are closely inter-related and need to be considered as a whole”, Preamble to the 1982 Convention of the Law of the Sea.


The negotiation of the 1982 United Nations Convention on the Law of the Sea (LOSC) is a triumph of perseverance. The fact that this convention exists, that it has spawned so many crucial legal instruments and that the Convention itself has been signed by more than 150 states is evidence that the majority of states acknowledges that the oceans must be managed globally. Fundamental to the LOSC is the freedom of the high seas and an obligation to cooperate in the management of high seas. The need to cooperate on the high seas has resulted in the establishment of several Regional Fisheries Organizations.


South Africa has developed fisheries in recreational, artisanal, subsistence, small-scale commercial and heavy-industrialized fisheries. The National Department of Environmental Affairs & Tourism advises the Minister on the allocation of some 683000 tonnes and 166 500 tonnes of total allowable catch (TAC) in the small pelagic and demersal hake trawl fisheries. In 2003 these two fisheries were the two largest industrialized fisheries in South Africa with 112 rights holders in the small pelagic fishery and 53 rights holders in the demersal hake trawl fishery. With capacity such as this it was inevitable that the industry would begin to explore resources outside the national EEZ, as is their right. In 1996 the Minister issued South African companies with permits to fish the EEZs of the Prince Edward Islands and South Africa’s Antarctic Territories. Within a year these companies progressed to fishing other CCAMLR areas. Expeditions to explore South East Atlantic and Western Indian oceans for catches of orange roughy (Hoplostethus atlanticus), alfonsino (Beryx splendens) and oreo dory and large pelagics like tuna began to increase. Of course South Africa’s history of highs-seas fishing does predate the 1990s as rock lobster fisheries off Tristan and Gough Islands and tuna fishing off Vema Seamount were established in the early 1980s.

Currently South Africa does not exploit South East Atlantic high-seas stocks of orange roughy, alfonsino and oreo dory. The industry, like those of other nations, is more successful in the Western Indian Ocean on the Madagascar Ridge (Figure 1). This harvesting is undertaken in the absence of management by a regional fisheries organization.

In attempting to find new and more economically efficient reserves of orange roughy, a South African company, Irvin and Johnson, undertook exploratory trawls on a number of seamounts on the Madagascar Ridge, during January and February 2000. The trawls were undertaken between the Ndomed and Gallieni Facture Zones (Figure 2). Material from a total of 42 trawls was sampled. Additional material was supplied from other trips to the Melville Bank area. The following results describe the work of Leslie, Compagno and Hulley (2001). Figures 3–12 show some of the unusual fishes taken from this area.

Position of Melville Bank area in relation to South Africa
Location of exploratory trawls

Specimens of Etmopterus cf priceps (Etmopteridae) collected on the Madagascar Ridge closely resemble E. princeps, which is known only from the North Atlantic


Specimens from the Madagascar Ridge of Etmopterus brachyurus (Etmopteridae) and South Africa resemble E. brachyurus, which is known from Japan, the Philippines and Australia


Proscymnodon plunketi (Somniosidae). Catches of this species show a significant range extension for this species, which was known only from New Zealand and Australia.


Somniosus antarcticus (Somniosidae), widely distributed in the Southern oceans but not previously recorded from the Melville Banks region.


Parmaturus macmillani (Scyliorhinidae). This species was regarded as endemic to New Zealand. These records constitute a significant range extension.


Sladenia sp. (Lophiidae), an undescribed anglerfish fish species. The first record of this genus in the Indian Ocean.


Apisturus cf investigatoris (Scyliorhinidae). Close to or perhaps identical to A. investigatoris, known only from the type locality in the Andaman sea


Centrosymnus coelolepis (Somniosidae). Widely distributed in deep waters of the Mediterranean Sea and the Atlantic (Grand Banks to South Africa) and Pacific (Japan to New Zealand and Australia) Oceans


FIGURE 11–12
An undescribed catshark, Apristurus sp. (Scyllorhlnldae) and chimaera, chimaera sp. (Chimaaridae)



The results of this analysis show that the chondrichthyan bycatch from the Melville Banks area had unexpected strong links to deep-water fauna from the orange roughy grounds off Australia and New Zealand rather than to the geographically-closer shelf off Africa. This could be a reflection of the relatively poor knowledge of the deep fauna off the shelf of the east coast of Africa. Alternatively it could suggest that there is an ancient common fauna on the Indo-Pacific sea mounts. The results of this analysis, although restricted to the chondrichthyan and teleost fauna, make it clear that there is a vast reserve of knowledge on high seas biodiversity that must be tapped.


These results present fascinating research challenges and open up novel avenues for biodiversity research. Such topics for research have not as yet been explored and it will be unfortunate to lose possible clues before these opportunities are explored.

All nations will agree that the protection of high seas biodiversity is necessary; the question that must be answered is where that responsibility lies. Can Regional and International Fisheries Organisations function in this role? In the Atlantic Ocean, fisheries for tuna and tuna-like species are managed by ICCAT. The newly established South East Atlantic Fisheries Organisation will assume control of fisheries for other resources in the South East Atlantic on the high seas, like the Tristan Bluefish (Hyperoglyphe antarctica). The picture in the Indian Ocean is rather different. Although management of tuna and tuna-like species falls under the gambit of the Indian Ocean Tuna Commission (IOTC), this body has yet to develop into an authoritative body, and there is no organisation responsible for other species. It is in this managerial void that the demersal longline and trawl fisheries on Walters Shoal and the Madagascar Ridge have developed in the western Indian Ocean. Evidence of the increase in high seas fishing in the southern African region can be observed though the number of international longline vessels that have requested permits to make port calls to South African ports (see Table1).

Table 1
Increase in high seas fishing

YearNo. of countriesNo. of vesselsNo. of calls

Note the more than 100 percent increase from 2001 to 2002. The apparent drop in 2003 is because the data are only for the first five months of that year. The figures for these first five months are already over 50 percent of the 2002 figures and will probably exceed the 2002 figures.

Regional and international fisheries organizations are attractive means to consider as custodians of high seas management because there currently exists a host of RFOs and similar commissions. Regional (RFOs) and international Fisheries Organizations generally face a number of challenges. There exists inadequate scientific knowledge for decision making, often due to data constraints. This affects their ability to make informed management decisions and to adopt appropriate measures for conservation and management. Even if only management measures are adopted, members can opt out from measures decided by the RFO, which results in voluntary compliance. This is compounded by a general inability to adequately enforce or monitor compliance. Further, non-members may undermine efforts by RFOs as they are not bound in any manner to abide by RFO measures. Finally, the management and allocation of fishing rights for migratory and straddling stocks generally underscore the agenda of RFOs. Ecosystem conservation and high-seas biodiversity is not a priority.

All of these limitations can be addressed through the funding of practical solutions. However the solutions and their funding must be a result of the political will and the adopting of appropriate National and International Laws. Political will remains fundamental to success (Pearce 2002).

We must seize the opportunity that the Final Plan of Implementation of the World Summit on Sustainable Development (WSSD) offers us in providing a statement of will to conserve high-seas biodiversity: The WSSD Plan of Implementation calls upon states to “maintain the productivity and biodiversity of important and vulnerable marine and coastal areas, including in areas within and beyond national jurisdiction”(para 31a). Paragraph 31 must be acknowledged as a commitment from heads of states not only to coastal, but also marine, ecosystem conservation.

Ideally we would want scientifically sound and rigorously tested scenarios on which to develop management plans for high seas biodiversity. However, while developing these and encouraging the international political will to fund such research there must be a parallel process of beginning to identify unique and vulnerable habitat types for conservation and protection. Marine protected areas offer a management solution for protection and conservation that is less dependent of detailed assessments.

Sumaila et al. (2000) make a convincing argument for the use of marine protected areas (MPAs) as a management tool for marine and coastal ecosystems. The effects on fishing on the ecosystems are largely not factored into current management measures. Total allowable catches and total allowable effort measures are dependent on accurate stock assessments and assessments of fishing capacity, which are difficult to achieve even when both scientific and legal infrastructure and capacity exist. Most management measures tend to focus on the species being exploited and are therefore not sensitive to symptoms like fishing down the foodweb as described by the landmark paper of Pauly et al. (1998). Myers and Worm (2003) claim that fishing pressure has reduced predatory fish levels of the world’s oceans by 90 percent. This must surely impact ecosystem functioning and hence there is a need to rebuild entire community structure.

In addition, we are increasingly beginning to appreciate the destruction caused by trawling gear. One may argue that the gear has a relatively spatially confined footprint, however this footprint may cover a significant portion of a unique habitat. The trawlable fishing areas of sea mounts (top and moderately steep sides) may in fact be different ecotones from the steeper slopes and gullies and deeper regions between sea mounts.


If the slow changes to predator-prey relationships and to the structure of the interactions within high seas areas that become increasingly significant with time are to be avoided, then marine protected areas on the highs seas must become a reality.

The size and number of MPAs required is a complex question. The success of an MPA is a function of its size related to the life history and mobility of the species which are being protected (Attwood and Bennet 1995). It is commonly accepted that 20 percent of the unfished amount of spawner biomass is required for sustainable fisheries. This percentage is also suggested as the minimum of a total habitat type that should be protected to ensure conservation of that ecosystem type (Plan Development Team 1990). However, it is unlikely that one reserve will serve the needs of a region and a system of a network of MPAs will be required (NRC 2001, ATEGH 2003).

The more protection offered to a habitat type, the greater insurance against any system failure or localised disaster. More protection can be offered through a combination of increasing the scope of a single MPA or increasing the number of MPAs. The balance between increased conservation or protection and the perceived limiting of economic gain from high seas stocks will be severely debated. Agardy (2000) makes the argument that protected areas often translate into better economic returns to the fishery through improved fish yields. In addition protected areas allow for a control site to test management options that ultimately serve the needs of the fishery.

The Committee on the Evaluation, Design and Monitoring of Marine Reserves and Protected Areas, of the United Sates list the possible benefits of marine protected areas in seven categories (NRC 2001):

  1. conservation of biodiversity and habitat
  2. fishery management
  3. scientific knowledge
  4. educational opportunities
  5. enhancement of recreational activities and tourism
  6. sustainable environmental benefits and
  7. protection of cultural heritage

If high seas MPAs are to be a reality and supported by sufficient political will, then meetings such as Deep Sea 2003 must begin to identify mechanisms to unlock benefits from these seven categories.

Finally, if we accept that legally defined MPAs on the high seas are necessary then the most challenging question remains, who has the authority to declare these MPAs or how do we mandate any authority to declare MPAs? Are the RFOs and other commissions on migratory and straddling stocks suitable to be mandated bodies, should research collaborations like ICES take the lead or is there a need to begin thinking about an International World Oceans Commission? As for where this will be negotiated, the success of United Nations with regards to the LOSC and the ensuing instruments, makes such an organization an ideal candidate for these tasks.


Agardy, T. 2000. Effects of fisheries on marine ecosystems: a conservationist’s perspective. ICES Journal of Marine Science 57:752–760

AHTEG REPORT 2003. UNEP/CBD/SBSTTA/8/INF/11: Technical advice on the establishment and management of a national system of marine and coastal protected areas. Paper prepared by the Ad Hoc Technical Expert Group on Marine and Coastal Protected Areas.

Attwood, C.G. & B.A. Bennet 1995. Modelling the effect of marine reserves on the recreational shore-fishery of the south-western Cape, South Africa. South African Journal of Marine Science, 16: 227–240.

Leslie, R.W., L.J.V. Compagno & P.A. Hulley 2001. New records of deep-sea fish from the Madagascar Ridge. Poster presented at the Sixth Indo-Pacific Fish Conference, Durban South Africa 2001.

Myers, R.A. & B. Worm2003. Rapid worldwide depletion of predatory fish communities. Nature, 423, 280–283.

NRC 2001. Marine Protected Areas. Tools for sustaining ocean ecosystems. National Research Council (NRC), Committee on the Evaluation, Design, and Monitoring of Marine Reserves and Protected Areas in the United States. National Academy Press, Washington. USA. 272 pp.

Pauly, D., V. Christensen, J. Dalsgaard, R. Froese & F.J. Torres 1998. Fishing down the foodweb. Science, 279: 860–863.

Pearce, J. 2002. The future of fisheries - marine protected areas - a new way forward or another management glitch? Marine Pollution Bulletin 44: 89–91.

Plan Development Team 1990. The potential of marine fishery reserves for reef fish management in the U.S. Southern Atlantic. NOAA Technical Memorandum, NMFS-SEFC-261. 40pp.

Sumaila, U.R, S. Guenette, J. Alder & R. Chuenpagdee 2000. Addressing ecosystem effects of fishing using marine protected areas. ICES Journal of Marine Science 57:752–760.

Previous Page Top of Page Next Page