The orange roughy fishery of Namibia: lessons to be learned about managing a developing fishery

Exploration for orange roughy Hoplostethus atlanticus in Namibia started in 1994 and within 12 months several aggregations had been discovered, suggesting the existence of a biomass sufficient to support a viable fishery. At that early stage it was realized that few, if any, recognized management procedures existed for newly developing fisheries, especially with the paucity of data such as existed on Namibian orange roughy. The development of the Namibian orange roughy fishery is reviewed to document the management strategies implemented and how the management of the fishery evolved. The first six years of the fishery are described, including the three-year exploration phase, several years of profitable exploitation, and the severe decline in catch rates. Whether the decline is attributable to fishing mortality or to change in the aggregating behaviour of orange roughy, or both, is not clear. Although many aspects of the precautionary approach were followed, a risk analysis applied and a number of innovative management methods implemented (e.g. incentives to promote exploratory fishing, use of Bayesian statistical methods, implementation of a management plan for long-term total allowable catches), the aggregating biomass declined to between 10 and 50% of virgin levels within the six years. The management methods applied are evaluated in the light of the severe decline in catch rate experienced in 1998 and 1999, so that others may learn from the experience.     

Distribution patterns,stock size and life-history strategies of Cape horse mackerel Trachurus trachurus capensis,based on bottom trawl and acoustic surveys

Research surveys of Cape horse mackerel Trachurus trachurus capensis abundance on the south coast of South Africa are complicated because changes in the species' vertical and horizontal distribution limit the value of stock assessments based a single survey method. Annual bottom trawl surveys conducted in spring provide estimates of the abundance of fish close to the bottom over trawlable grounds. Between 1991 and 1994, hydroacoustic surveys conducted in spring have been used to estimate the pelagic portion of the stock, as well as the portion over untrawlable grounds. These two research datasets, as well as data from purse-seine, mid water and bottom trawl commercial landings, are reviewed to elucidate distribution patterns of horse mackerel and their migratory and spawning strategies. The problems and advantages of bottom trawl and acoustic surveys are discussed in the context of fluctuations in estimates of the size of the stock between 1991 and 1994 and the prevailing environmental conditions. It is concluded that combined acoustic and bottom trawl surveys are the only effective means of surveying horse mackerel, and that effort should be concentrated east of 22°E to assess the spawner stock. It is suggested that research effort directed at improving understanding of exchanges between West Coast (including Namibia) and South Coast population of horse mackerel, as well as of the role of vertical migrations in modulating these exchanges, would be beneficial.     

A review of orange roughy Hoplostethus atlanticus fisheries,estimation methods,biology and stock structure

Orange roughy Hoplostethus atlanticus are unusual fish. They form dense aggregations that have fuelled lucrative fisheries at great depths (600–1 400 m), especially off Namibia, New Zealand and Australia. They are thought to be very long-lived (>100 years, maturity at 22 – 40 years), and to have exceptionally low natural mortality (M = 0.045–0.064 year^?1) and slow growth rates (K = 0.055–0.070 year^?1). In addition, they spawn large eggs and have low fecundity. These factors combine to make orange roughy highly susceptible to overfishing; most stocks are below 30% of pristine levels. Assessments are obtained from indices of catch rate and trawl, acoustic and egg surveys. Acoustic estimates are the most direct, but are confounded by the species' low target strength (?50 to ?53 dB)–attributable to the wax-filled swim bladder. Extracellular wax esters are stored in abundance and comprise mostly mono-unsaturated fatty acids, with low concentrations of the ω-3 fatty acid family. This unusual composition (resultant from the species' diet) ensures neutral buoyancy. Stock separation has been inferred mainly from biological studies, but genetic studies have also found differences among stocks within New Zealand and Australia. Deep-water habitat may be damaged by trawling operations and may take many years to recover, so in some quarters there is a call for a portion of suitable habitat to be set aside for preservation. Although Namibian orange roughy are shallower, smaller and younger than those in other stocks, the Namibian fishery sustained high catches for only a few years before quotas were reduced, from 12 000 to 1 875 tons. Three management lessons are suggested for developing orange roughy fisheries based on the Namibian experience: (1) imposition of catch limits during exploratory fishing; (2) starting the acoustic surveys earlier in the fishery, if possible; (3) greater reliance on trends in catch rate until a survey series has been established.     

Development of acoustic techniques for assessment of orange roughy Hoplostethus atlanticus biomass off Namibia,and of methods for correcting for bias

Orange roughy form dense spawning aggregations in specific small areas in deep water on the Namibian shelf between late June and early August each year. The biomass in three such areas, where most commercial fishing occurs (the Johnies, Frankies and Rix Quota Management Areas, or QMAs) has been assessed acoustically each year since 1997. Acoustic estimates of the aggregated portion of the biomass (the only component that can be assessed reliably using acoustics) were obtained for all three QMAs in 1997 and 1998, but only for Frankies in 1999 owing to increased problems with target identification as the biomass declined. The methodology developed for these surveys, including the equipment used, survey design, target identification, data processing and error analysis are described. Some important biases that should be corrected for when estimating absolute abundance of orange roughy acoustically are addressed. Individual sources of error were quantified as well as possible, and input to an error model that simulated the error process and produced probability density functions of absolute biomass, from which the mean absolute biomass and its standard error could be computed for each survey, effectively correcting for identified sources of bias and quantifying the overall uncertainty. The correction factors ranged from 1.58 to 1.71 and the CVs increased by factors of 1.2–2.1. Target strength uncertainty and negative bias attributable to the dead zone close to the bottom were considered to be the most serious errors. The acoustic estimates indicate a substantial decline in orange roughy biomass in all three QMAs since 1997, in accord with indices from contemporaneous swept-area surveys and the catch rate of the commercial fleet. Acoustic estimates have already been used extensively to manage the resource and are likely to remain important in the future.     

应用贝叶斯方法对有效信息较少的渔业资源进行资源评估-以印度洋长鳍金枪鱼的资源评估为例

众所周知,对有效信息较少的渔业资源进行资源评估面临很大的挑战,而贝叶斯方法在数据数量较少、质量较差的情况下能利用其它种群高质量的数据或已知的先验信息提高资源评估结果的可靠性。由于印度洋长鳍金枪鱼的数据质量较差而数据量有限,长鳍金枪鱼的资源评估结果存在很大的不确定性,为此,本文以印度洋长鳍金枪鱼的资源评估为例,以调查贝叶斯方法在有效信息较少的资源评估中的优势。本文根据不同的先验假设与捕捞数据系列,共构建了8个贝叶斯动态产量模型,以评估长鳍金枪鱼资源。结果表明:(1)分析参数的后验分布能提高捕捞数据系列选择与参数假设的合理性; (2) 利用种群统计学方法为内禀增长率(r)构建有信息先验,能提高资源评估结果的可靠性。与传统方法相比,当基于贝叶斯框架时,能将已知的知识表示为先验信息并能分析参数的后验分布,从而在数据较少或数据质量较差的情况下,能利用各种信息提高参数估计的合理性与资源评估的可靠性。因此,对数据量较少或数据质量较差情况下的渔业资源评估而言,贝叶斯方法非常有效,如本文所示的印度洋长鳍金枪鱼的资源评估。     

Investigation into the declining trend in chokka squid Loligo vulgaris reynaudii Catches made by South African trawlers

The trawl fishery, which targets Cape hake Merluccius spp. and Agulhas sole Austroglossus pectoralis, takes chokka squid Loligo vulgaris reynaudii as by-catch. Catch and effort data from the trawl fishery for the period 1978–1996 are used to obtain annual estimates of catch rate (catch per unit effort cpue) for that period. Examination of the cpue trend shows a sharp decline in the early 1980s and, in order to identify factors that could have influenced that decline, the distribution of fishing effort is investigated both temporally and spatially. There is a possible change in the incidence of squid-directed catches over time, but their overall scarcity could have had only a small impact on the annual cpue trend. Further, using distribution of fishing effort to evaluate the effects of possible changes in fishing patterns, rather than changes in resource abundance, on the trend in trawl cpue, it became clear that there had been a contraction of the trawling grounds and changes in fishing patterns in relation to depth over time. Finally, a general linear model (GLM) is developed to quantify the effect on cpue of factors such as vessel characteristic, depth and position of the drag, season and target species, so obtaining a standardized trawl cpue index of chokka abundance. Analysis of that trend reveals a mean 7.7% annual decline for the period investigated, which should be interpreted as a strong sign of resource decline.     

Assessing stocks in data-poor African fisheries: a case study on the white grouper Epinephelus aeneus of Mauritania

The lack of reliable stock assessment for numerous exploited stocks in West Africa often results from poor-quality data, high multi-specificity of captures, and the heterogeneity of exploitation methods. However, many signs of overexploitation exist, particularly for demersal resources, highlighting the urgent need for a more quantitative and comprehensive evaluation of these resources. This study aims to show how, in such a context of poor-quality data and high uncertainty, a multi-method approach for stock assessment can generate a consistent diagnosis of the condition of a resource. As a case study, several methods were combined to assess the stock status of the white grouper Epinephelus aeneus, a flagship species in West Africa that is exploited by industrial and small-scale fisheries in Mauritania. These were estimation of abundance indices using delta generalised linear models; a biomass production model using a pseudo-equilibrium method and including an environmental effect of upwelling intensity; a dynamic biomass production model fitted in a Bayesian framework also including an environmental effect; and an age-structured model based on a modified pseudo-cohort analysis. Sensitivity analyses were performed for most of these assessment methods. Results show that the white grouper stock is highly overexploited due to an excess in the fishing effort estimated at between 30% and 50%, depending on the model used to estimate the effort at maximum sustainable yield.     

Letter to the editor

It is a common observation that the mean length in a fish population progressively decreases as the biomass is reduced by fishing. However, this has not happened with the population of orange roughy (Hoplostethus atlanticus) on the Chatham Rise, New Zealand, even though the biomass is estimated to have declined by almost 80%. Because orange roughy are slow growing it might be expected that mean age (or its proxy, mean otolith weight) would be a more sensitive indicator of biomass decline than mean length. A simulation study shows that this is not true. With equal sample sizes, all three indicators are equally sensitive; on an equal cost basis, mean length is more sensitive than the others. A log transformation increases the sensitivity of age and otolith weight, but it is not clear whether this increase is sufficient to outweigh the much greater costs of measuring age and otolith weight, compared to length. Under the assumption of constant recruitment, it is more likely than not (P = 0.55–0.8) that a change in mean length would have been detected in the Chatham Rise population, given the estimated biomass decline.     

Fisheries management — a New Zealand perspective

The current fisheries management regime in New Zealand involves a system of quota management within which individual quotas are owned by fishermen and are transferable on an open market. The sum of these individual transferable quotas ITQs for a particular species equates to the total allowable catch TAC for that species. The objectives, implementation and consequences of this system are outlined, and its significance for stock assessment for species included in the ITQ system is discussed. A recent need for a TAC reduction in the deep-water fishery for orange roughy Hoplostethus atlanticus is mentioned in the context of the ITQ system. It is suggested that there is potential for economic benefit to the fishing industry in the Benguela system if an ITQ system were to be introduced locally.     

Population abundance and seasonal migration patterns indicated by commercial catch-per-unit-effort of hakes (Merluccius capensis and M. paradoxus) in the northern Benguela Current Large Marine Ecosystem

We developed generalised additive models (GAMs) to estimate standardised time-series of population abundance indices for assessment purposes and to infer ecological and behavioural information on northern Benguela hakes, Merluccius capensis and M. paradoxus, using haul-by-haul commercial trawl catch-rate data as proxies for hake densities. The modelling indicated that individual ship identifiers should be used rather than general vessel characteristics, such as vessel size. The final models explained 79% and 68% of the variability in the commercial catch rates of M. capensis and M. paradoxus, respectively. The spatial density patterns were consistent and confirmed existing knowledge about these species in the northern Benguela system. Furthermore, seasonal migration patterns were described for the first time and were found to correspond to the known spawning areas and seasons for M. capensis and M. paradoxus. Spatial density patterns were validated using the geostatistical modelling results of fisheries-independent trawl survey data. Improved understanding of the relationships between fleet dynamics and fish movement can be achieved by taking into consideration the present catch-rate model and spatial and seasonal distribution maps. We conclude that the yearly standardised CPUE time-series are problematic as proxies for total stock abundance because of spatial coverage issues. Consequently, such CPUE data should not be used for stock-size assessments and fisheries advice concerning northern Benguela hakes until this is solved. We generally recommend the exclusion of standardised CPUE time-series from stock assessments when important and changing parts of the stock distribution cannot be targeted by the fishery, such as due to closed areas or seasons.     

Geostatistical modelling of the spatial life history of post-larval deepwater hake Merluccius paradoxus in the Benguela Current Large Marine Ecosystem

Optimal and sustainable management of fish resources cannot be ensured without a thorough understanding of the migration patterns and population (demographic stock) structure. Recent studies suggest that these aspects of the economically and ecologically important deepwater hake Merluccius paradoxus are not reflected in the current assessment and management practices for the Benguela Current Large Marine Ecosystem. In this study, we compiled data from multiple demersal trawl surveys from the entire distribution area and applied state-of-the-art geostatistical population modelling (GeoPop) to estimate growth rate, mortality, and spatial and temporal distribution patterns of M. paradoxus. The data and the model enabled us to follow temporal and spatial changes in the distribution and infer movements from the recruitment/nursery areas, through the juvenile phase and the adults’ migration to the spawning areas outside/upstream of the nursery areas. The results indicated one primary recruitment/nursery area on the west coast of South Africa and a secondary less-productive recruitment/nursery area on the south coast near Port Elizabeth. Juveniles initially migrated away from the main recruitment area, followed by natal homing by larger individuals. This pattern was highly consistent through the time-series of the study. This perception of a, primarily, panmictic population that performs transboundary migrations between Namibia and South Africa corresponds largely to the hypothesis and data plots given in recent studies. We recommend that fisheries assessment, advice and management take into consideration these aspects of the distribution and population (stock) structure of M. paradoxus.     

Hauraki Gulf snapper fishery, 1972 and 1973: Some evidence for a declining catch‐rate

The commercial fishery for snapper Chrysophrys auratus (Bloch and Schneider) in the Hauraki Gulf is one of New Zealand's principal marine fisheries. The snapper catch from this area declined by 22% in 1972, and by a further 4% in 1973, following 4 y of high catches by Danish seiners.

Some of the decline can be attributed to economic factors within the industry, and adverse winter weather. However, the 1972 decline in Danish seine catches is an acceleration of an existing trend. It probably results from heavy exploitation of an accumulated stock, and from movement of a few strong year classes past the age at which they have contributed most to the catch. The slight recovery in the 1973 Danish seine catch resulted only from increased effort; catch per day at sea continued to fall. An improvement in catch and catch per day is unlikely for some years at least.

Snapper catches by other methods (trawl, long line, and set net) also declined during 1972, but the available data on fishing effort do not allow a detailed analysis. The trawl catch in the gulf fell further in 1973 as fishing effort was diverted to other grounds, but the catch by lines and nets improved slightly.     

Impacts of trawling on benthic macro-fauna and -flora of the Spencer Gulf prawn fishing grounds

The overall effects of trawling on benthic habitats and their assemblages are dependent on the distribution and intensity of trawl effort. The benthic habitats of the Spencer Gulf prawn trawling grounds are subjected to known variable levels of trawling disturbance recorded from fisher's logbooks. These habitats have not been quantitatively investigated. The aim of the study was firstly to characterise the macro-faunal and -floral assemblages and secondly, to comparatively assess trawl impact by testing the null hypothesis of no differences between five sites exposed to different intensity of trawl effort. The distribution and abundance of benthic macro-fauna and -flora were studied at two sampling resolutions by using beam trawl sampling (10,000 m²) and underwater stereophotography (4.5 m²) at five sites with different levels of trawl disturbance (effort). The results showed that the Spencer Gulf prawn trawling grounds are characterised by sandy sediments with a low content of silt and clay, with the exception of one site with very fine gravel. Biomass, abundance and cover of macro-fauna and -flora were generally low throughout, but with large differences among sites. Biomass, abundance and cover were found to be negatively correlated to both trawl hours from 1994–1998 and during the period of study. ANOSIM and SIMPER analyses using biomass, abundance and percentage cover as variables showed significant differences between sites with eight species or taxonomic groups contributing more than 10% to the observed similarity within sites. The two northern sites were dominated by sponges and the bearded mussel, Trichomya hirsutus, and the southern hammer oyster, Malleus meridianus. Other species that contributed to the similarity within sites were the ascidian, Polycarpa viridis, mobile epifauna (the blue swimmer crab, Portunus pelagicus, and the western king prawn, Penaeus (Melicertus) latisulcatus) and demersal fish species (Degens leatherjacket, Thamnaconus degeni, and the Melbourne silver belly, Parequula melbournensis). Simpson's Index of Diversity (1 − λ) and species richness in the beam trawl samples were fairly even across all sites, but a significant effect of site was found in the stereophotographic transect samples. It was concluded that prawn trawling has a strong influence on the structure of the benthic habitats of the Spencer Gulf prawn fishing grounds highlighted by the magnitude of change and the strength of the gradient of effects. Confounding factors linked to the biophysical attributes of the sites are believed to be minor.     

Dynamics of the pelagic component of Namibian hake stocks

A study was undertaken to investigate the spatial, diurnal and interannual dynamics of the pelagic component of Namibian hake stocks. Data were derived from trawl surveys conducted on board the R.V. Dr Fridtjof Nansen during the period 1991–1999. In addition to trawl surveys, acoustic methods were applied concurrently to quantify fish biomass off the bottom (above the headline). The pelagic component of hake was variable in both time and space – it was denser in the north and at depths of 200–500 m. Pelagic densities also correlated well with the density of adult fish (≥30 cm) at the bottom, but not with that of young hake (>18 and <30 cm) at the bottom. Pelagic densities were lowest at midday and highest at night, early morning and late afternoon densities being intermediate between the two. Incorporation of acoustic methods allows for at least partial account to be taken of diurnal variation in determining catchability. The fraction of the biomass off the bottom was not the same each year, thus affecting assessment and management advice for the stock, even if the survey results are treated only as relative indices of abundance. More field studies are needed to investigate the size range and species of hake active in vertical migration.     

Surviving off junk: low-energy prey dominates the diet of African penguins Spheniscus demersus at Mercury Island,Namibia, between 1996 and 2009

The diet of African penguins Spheniscus demersus in Namibia consisted mainly of sardine Sardinops sagax in the 1950s. Since the collapse of pelagic fish stocks in the 1970s, birds fed mainly on bearded (pelagic) goby Sufflogobius bibarbatus, a low-energy prey species. We present diet data for African penguins breeding at Mercury Island, the largest colony for this species in Namibia, between 1996 and 2009. Bearded goby was the main prey item throughout the study period, both in terms of frequency of occurrence (67.8%; SD 31.2) and in terms of mass (59.2%; SD 31.5). Diet composition varied throughout the year as well as between years; birds occasionally fed on a variety of fish species other than bearded goby. In Namibia, poor prey abundance is considered as a major factor contributing to the decline of penguin numbers after the collapse of the sardine stocks. However, bearded goby appears to be relatively abundant along Namibia's southern coast and low prey quality rather than low abundance appears to be a key factor influencing population dynamics of African penguins and other marine top predators in southern Namibia.     

Towards a catchability constant for trawl surveys of Namibian hake

A trawl catches only a portion of the fish in its path. The term catchability therefore refers to the fraction of the available fish caught. A method was developed and tested to establish catchability constants for trawl surveys of Namibian hake Merluccius spp. A catchability constant can be expressed as a simple relationship between hake area densities calculated from trawl catches and acoustic biomass estimates. Initial values were on an order of magnitude of 0.8, meaning that the catch takes 80% of the hake available to the trawl. The methodology depends on careful area selection (flat bottom, homogenous fish distribution), following the same trawl lane during subsequent hauls in an area, and thorough acoustic post-processing. A pronounced and repetitive pattern in catchability within surface daylight hours was found. Early morning and early afternoon catches were low, and the best catches of the day were made around noon, a result that may influence stock assessments based on trawl data, because morning and afternoon data would under-represent actual abundance.     

A Bayesian state-space model for mixed-stock migrations,with application to Northeast Atlantic mackerel Scomber scombrus

Management of fisheries that exploit mixed-stock populations relies on assumptions made concerning stock structure and mixing in different areas. To address the problems of accounting for uncertainty when formulating scientific advice for the management of highly migratory fish stocks, management decisions need to be based upon assessment models that represent plausible alternative hypotheses for stock structure and migration patterns of the exploited populations. We present a multi-stock, multi-fleet, multi-area, seasonally structured Bayesian state-space model in which different stocks spawn in spatially different areas and the mixing of these stocks is explicitly accounted for in the absence of sufficient tagging data with which to estimate migration rates. The model is applied to the Northeast Atlantic mackerel Scomber scombrus population, accounting for the annual spawning-feeding-overwintering migration patterns of the three spawning components, together with uncertainty in the extent to which the southern component migrates north to feed and overwinter, and consequently the extent to which it mixes with the other components and is subject to exploitation. The model allows the effect of exploitation on the individual components to be assessed, and the results suggest that the fishing mortality of southern spawning adults was insensitive to the extent to which they migrated north.     

Juvenile life history of NE Atlantic orange roughy from otolith stable isotopes

Otoliths of pre-recruit orange roughy (Hoplostethus atlanticus) were sampled from different deep-water habitats (‘flat’ and ‘hill’) and areas (north and south) on the Porcupine Bank. Age-based profiles for stable isotopes of carbon (δ¹³C) and oxygen (δ¹⁸O) in these otoliths were developed, which provide a fish life-history record of water depth and metabolic activity. These profiles were consistent among all individuals implying endogenous (ontogenic) influences on the pattern. The data indicate that post-larval orange roughy are mesopelagic active foragers, early juvenile fish move into a low energy deep-demersal phase, and older pre-recruit orange roughy assume the habitat depth and metabolic rate typical of adults. Comparison of otolith stable isotope profiles among areas and habitats on the Porcupine Bank suggest that juvenile orange roughy from South ‘hill’ and ‘flat’ habitats experience differing temperature and metabolic status at certain life stages. This may reflect oceanographic and ecological divergence between the two environments and suggests fine-scale population structure that may reduce resilience to exploitation.     

The effects of spatial targeting of fishing effort on the distribution of the Norway lobster,Nephrops norvegicus,on the Farn Deeps grounds,northeast England

Nephrops norvegicus is an essentially sedentary species of lobster that forms the basis of valuable fisheries in the northwest Atlantic and western Mediterranean. Fishers exploiting a sedentary stock are likely to visit the most profitable (highest catch rate) areas first. Such spatial targeting of fishing effort is likely to have important consequences for stock monitoring and assessment. We used underwater television surveys of Nephrops burrow densities on the Farn Deeps grounds, northeast England, to describe changes in abundance and distribution between the beginning and end of a winter fishing season. Above a threshold of c. 0.6 burrows m^–2, overwinter depletion increased with burrow density, consistent with fishing effort being targeted at the highest densities. A simple simulation model showed that this pattern of mortality is an expected consequence of spatially targeted fishing behaviour. The model also predicted that there is decreased spatial variability in density after fishing. An overall decrease in variability was not evident from the survey data, but geostatistical analysis indicated that there was “flattening” of the density profile along a north‐south axis, consistent with the dominant direction of commercial trawling. We concluded that Nephrops fishers are able to find and exploit the highest densities of their target species. A potential consequence is that catch per unit effort (CPUE) data used to monitor trends in this stock potentially could mask declines in stock abundance. CPUE might be more effective if analysed at finer spatial scales, but this is not currently possible. In the absence of these fine scale commercial data, fishery‐independent surveys (e.g., underwater television) are an important source of information on trends in stock abundance.     

Fisheries for orange roughy (Hoplostethus atlanticus) on seamounts in New Zealand

Major commercial fisheries for orange roughy (Hoplostethus atlanticus) occur on seamount features, which are widely distributed throughout the New Zealand region. When the fishery developed in the late 1970s to early 1980s, it occurred mainly on flat bottom, but over time has become more focused on seamounts. In the 1995–1996 fishing year, it is estimated that about 70 % of the catch of orange roughy within the New Zealand EEZ was taken from seamounts. Seamounts on the Chatham Rise have been fished for over ten years. Examination of commercial catch and effort data show strong declines in catch rates over time, and a pattern of serial depletion of seamount populations, with the fishery moving progressively eastwards to unfished seamounts along the southern margins of the Rise. Catch rates on seamounts in other regions of New Zealand have also generally shown a similar pattern of rapid decline. There is growing concern over the impact of trawling on seamounts, and the effects this can have on the benthic habitat and fauna, and the long-term sustainability of associated commercial fisheries.