INTEGRATED OVERVIEW OF FISHERIES OF THE BENGUELA CURRENT REGION

 A synthesis commissioned by the United Nations Development Programme (UNDP) as an information source for the Benguela Current Large Marine Ecosystem (BCLME) Programme

 Authors

EXECUTIVE SUMMARY

This Report gives an overview of the major living resources of the Benguela Current Ecosystem, and of the research which has been conducted on them by the three countries bordering the system (Angola, Namibia and South Africa) and foreign partners, as input to management of these resources. It treats the resources on a regional basis, and highlights trans-boundary problems which need to be addressed through a regional rather than a national approach. The report amplifies and updates recent summaries of scientific information on the resources, and adds material on the legal framework and structures under which they are managed in each of the three countries.

The distribution and habitats of the major species exploited by the purse seine, trawl, crustacean, linefish and artisanal fisheries of the region, and of the seal and main seabird populations, are briefly described and discussed with particular reference to questions of stock separation and the connection between apparently different populations. It is pointed out that a number of commercially-important species (e.g. hake, horse mackerel, deep-sea red crab, tuna and, probably to a lesser extent, sardine and anchovy) are distributed or move seasonally across national boundaries, requiring regional compatability in the research and management of these resources.

The life histories of the major exploited species are briefly outlined, with emphasis on temporal and spatial spawning patterns, the dispersal of the early-life stages, migration patterns of recruits and adults, and diet (particularly as it relates to potential competition between species). The broad picture to emerge is that the important offshore resources in the southern Benguela spawn over various parts of the Agulhas Bank, and depend to a greater or lesser extent on the equatorward jet current between Cape Point and Cape Columbine to transport the early life stages to the West Coast, where much of the recruitment takes place. Pelagic recruits return to the Agulhas Bank in the poleward counter current close to the coast. Species which become more demersal with age (e.g. hake and horse mackerel) tend to move into deeper water as they move south in the early stages, resulting in complex movements between the West and South Coasts and more widespread spawning areas. In the northern Benguela, spawning of many of the important species is concentrated in northern Namibia/southern Angola, and is probably influenced by movements of the Angola-Benguela front. Juveniles spawned in this area tend to move south close to the coast and to return northwards farther offshore when older.

The history and current status of the major fisheries of the region are discussed, concentrating on catch trends and changes in the nature of the fisheries. Over the past 30 to 40 years total catches in the south-east Atlantic have declined from a peak of more than 3 million tonnes in 1968 to levels of around 1 million tonnes per year in the 1990s. The most pronounced features in Namibia and South Africa have been the major decline in catches of South African and Namibian sardine in the mid- and late 1960s respectively (followed by a dramatic further decline in Namibia in the mid-1970s), the major decline in the West Coast rock lobster resource (particularly off Namibia) to levels well below those in 1960, and the major reduction in hake and horse mackerel catches off Namibia in the 1990s, due largely to the withdrawal of foreign fishing fleets after Independence in 1990. Off Angola, the most notable feature is the sharp reduction in industrial catches of all the most important species (e.g. sardinellas, horse mackerel and deep-water prawns), due largely to the major reduction in foreign fishing effort from 1985 onwards.

Changes in the abundance and distribution of the major resources, including seals and seabirds, as revealed by acoustic, trawl and aerial surveys and catch-based analytical methods, are discussed. VPA estimates confirm the dramatic decline of Namibian sardine in the late 1960s and the subsequent further sharp decline in the mid-1970s. Acoustic surveys since 1990 show that the population dropped to its lowest ever level in 1995 (co-incident with a major Benguela Niño in the northern Benguela), but that it has recovered somewhat since then. Acoustic surveys of sardine and anchovy in South African waters indicate that there has been a gradual increase in sardine abundance since the mid-1980s (although probably not to pre-collapse levels), and show two cycles in anchovy spawner biomass, with peaks in 1986 and 1991. The distribution of Namibian sardine has moved northwards as the population has declined, an extreme being reached in 1994, when sardine were only found north of the Cunene River, in Angolan waters. Likewise, there has been a major shift in the distribution of sardine in South Africa since the 1960s; the core of the distribution of adults having shifted from the West Coast to the South Coast. Acoustic surveys off Angola indicate that, in the 1990s, the biomass of both sardinella and Cunene horse mackerel has roughly doubled compared to the 1980s. Recent swept-area estimates of hake abundance in both Namibia and South Africa indicate that these populations are relatively stable at present, and there is some evidence that they could be gradually increasing. An interesting trend has been the increase in the abundance of Merluccius paradoxus in Namibian waters since 1992, which may indicate recent expansion or northward displacement of the M. paradoxus population from the South African West Coast. Acoustic and trawl survey estimates of adult horse mackerel in Namibia and South Africa over the past decade do not reveal any pronounced trends apart from a roughly 3-fold increase in the biomass on the South African West Coast since 1991. It is evident from CPUE indices that the abundance of line-caught species in South Africa has declined over the past few decades, to almost zero in some cases. As a result, the distributional ranges of many of these species has contracted, and the magnitude and extent of their migrations has declined. Aerial censuses reveal that the Namibian seal population was roughly halved by the effects of the 1993-1994 extreme anoxia events in the northern Benguela, from which the population now appears to have almost recovered. Aerial censuses have also revealed long-term reductions in the numbers of Cape gannets and Cape cormorants on the West Coast of southern Africa since the 1960s, and continuing declines in numbers of African penguins.

A number of major effects of the environment on the distribution and abundance of commercially important resources in recent years are discussed. The most dramatic of these was the wide-scale advection of low-oxygen water into the northern Benguela from Angola in 1993 and 1994, and the subsequent Benguela Niño of 1995, which appears to have severely affected the Namibian sardine population, and its major predators (particularly seals), and to have directly or indirectly increased mortality of juvenile hake on the Namibian shelf.

The present socio-economic value of the region’s fisheries to Angola, Namibia and South Africa is outlined. Their national importance and the balance between the various sectors (industrial, artisanal, recreational etc.) varies considerably between the three countries. The fisheries sectors in Angola and Namibia rank high in national importance, for local food production in the case of Angola, and in Namibia for exports from the industrial fishery, which are worth over 1,35 billion N$ (approx. 225 million US$) per year at present. In both countries the sector is an important source of employment, which is largely informal in the case of Angola, and formal in Namibia. In South Africa, while the fishing industry is relatively unimportant nationally, exports rival those of Namibia, and the industry is an important source of revenue, food and employment in coastal areas, particularly in the Western Cape Province, which yields 80 - 90% of the total South African marine fish catch. Artisanal fisheries are currently of relatively little importance in South Africa, but the recreational fishery for linefish is large and varied, and directly or indirectly generates revenue and employment opportunities which probably exceed those of the industrial fishery.

The broad national policies, legislation and formal structures for managing marine living resources in Angola, Namibia and South Africa are sketched, and the institutional capacity for research and management (including donor support) in each of the three countries is outlined. Current regional marine research programmes such as BENEFIT are briefly outlined, and other international and regional agreements pertaining to management of the region’s marine resources are listed. Recent and current methods used to assess the size of the major resources in each of the three countries are described in broad detail, and the methods used to manage them discussed.

Major gaps in current knowledge of the region’s resources and threats to their rational management are discussed, with particular attention to trans-boundary issues, which are seen to be notable for hake, horse mackerel, sardine, anchovy and crab stocks. The major general scientific problems are seen as inadequate understanding of stock definition (particularly for shared or migratory stocks), inaccurate or non-existent information on basic biological characteristics (particularly age structure) for many of the harvested species, inadequate absolute estimates of population size and of trends in biomass, the lack of Operational Management Procedures based on population models for many of the resources, and the inability to predict the effects of environmental perturbations on resource dynamics with sufficient confidence for this information to be used in national or regional management. In most cases these problems are particularly severe in Angola and, to a lesser extent, in Namibia. It is noted that the root cause of many of the trans-boundary management problems is the lack of regional agreements and structures for research and management of shared resources, and the shortage of manpower and funds to undertake trans-boundary surveys and other related research activities.

Finally, particular scientific and operational problems in Angola, Namibia and South Africa are discussed on a national basis. In Angola, where resources and their environment have been significantly less studied than elsewhere in the region, there are limited data series for retrospective analyses, large deficiencies in the understanding of fundamental life history characteristics of commercially important species, and no population models on which to evaluate management options. Catch and effort statistics are unreliable, and fisheries regulations are difficult to enforce owing to the extended coastline and small corps of compliance officials. Research capacity is limited because of the small number of people involved and the lack of appropriate local educational facilities. Most importantly, the breakdown of basic services and infrastucture as a result of the protracted civil war, and the severe macroeconomic problems in the country at present seriously inhibit development of local capacity in all spheres. In Namibia, the chief scientific challenges are the development of population models and Operational Management Procedures which would enable alternative harvesting strategies for the major fisheries to be formally evaluated, and the finding of objective ways of including environmental information into management decisions. A major operational constraint is the severe shortage of scientific and particularly technical staff within the Ministry of Fisheries and Marine Resources for the large number of resources which have to be studied, a problem exacerbated by the need for local staff to spend extended periods abroad for further education. In South Africa, the nation’s strong capacity in marine science is being threatened by a reduction in research funding, which has led inter alia to difficulties in maintaining even essential resource-monitoring surveys, and strict curtailment of environmentally-orientated cruises. Loss of senior research and management staff as a result of moves to reduce the size of the Public Service has resulted in a loss of experience and a greater load on remaining staff, which is likely to be aggravated by added responsibilities under the new Marine Living Resources Act.

The Benguela Current Ecosystem can be loosely considered as covering the continental shelf between the Angola-Benguela frontal zone in northern Namibia/southern Angola and the Agulhas retroflection area, typically between 36 and 37 oS (Shannon and O’Toole 1998). As such, it covers the West Coast of South Africa, the entire Namibian coast, and southern Angola (Fig. 1) to an extent depending on the position of the Angola-Benguela front, which moves seasonally typically between 14 and 17 oS.

Fig. 1 Southern Africa, showing the 200m isobath, considered here to delimit the continental shelf

The Benguela Current is one of the world’s major eastern-boundary current systems, and is rich in pelagic and demersal fish populations, supported by plankton production driven by intense coastal upwelling. These populations have been heavily exploited by man, particularly since the Second World War. Total fish catches in the south-east Atlantic climbed rapidly during the 1950s and 1960s, with the development of hake, sardine, anchovy, horse mackerel and sardinella fisheries (Fig. 2), and a valuable fishery for rock lobster in both Namibia and South Africa. The total annual catch peaked at over 3 million tonnes in 1968, but it subsequently declined to a level of around 2 million tonnes in the 1970s. This was largely attributable to major declines in sardine catches off both Namibia and South Africa, which were only partly compensated by increased (largely foreign) catches of hake and horse mackerel off Namibia. Total annual catches in the region subsequently dropped further to around 1.2 million tonnes in the 1990s, with a further sharp decline in catches of Namibian sardine in the second half of the 1970s, and the cessation of foreign trawling for hake and horse mackerel off Namibia after her independence in 1990. Since the 1960s there has also been a dramatic decrease in rock lobster catches, particularly off Namibia, where catches are now some two orders of magnitude below their peak in the 1960s. It is believed that most of these declines have been due to overfishing, although some of the major fluctuations have probably been influenced to a greater or lesser extent by the large-scale environmental perturbations that have occurred periodically in the system during this period (Shannon and O’Toole 1998).

Fig. 2 Cumulative catches of the principal harvested species in the south-east Atlantic from 1950

This overview gives a brief description of the major living resources of the Benguela Current, and of the attempts that have been made, particularly in recent years, to manage them rationally and sustainably in each of the three countries bordering the system. It treats the fisheries regionally rather than nationally, and builds on material already presented in the regional Benguela Environment Fisheries Interaction and Training (BENEFIT) Programme Science Plan (Shannon and Hampton 1997), with particular emphasis on work carried out since the production of that document. References have largely been kept to recent key publications, and to a number of general articles from the region, wherein further references may be found.

For the purposes of this overview, the Benguela ecosystem has been defined somewhat more widely than its physical limits, to include the extremities of the warm water systems which bound it on both sides. This is necessary, because some of the major resources of the Benguela spend a significant part of their lives outside the system boundaries, and are strongly influenced by interactions between the cool waters of the Benguela and the warm water of the Angola and Agulhas Currents.

2. OCCURENCE AND STOCK IDENTITY

2.1 Small pelagic fish

The major fisheries for small pelagic fish off the west coast of southern Africa are those for sardine Sardinops sagax (also known as pilchard), anchovy Engraulis capensis, juvenile Cape and Cunene horse mackerel (Trachurus trachurus capensis and T. trecae respectively), round herring Etrumeus whiteheadi, and the round and flat sardinella (Sardinella aurita and S. maderensis respectively), which are fished almost exclusively by Angola.

Sardine and anchovy live in temperate waters from southern Angola to KwaZulu-Natal in South Africa, with both species co-existing as quasi-discrete stocks off northern/central Namibia and off the Western Cape (i.e. between Cape Point and the Orange River). The degree of mixing between the southern and the northern populations of these species is unknown, but considering that the populations spawn in different, widely separated areas and are separated by a large, perennial area of cold, upwelled water off Lüderitz, it is probably not significant for management purposes, except in anomalous years. The round herring occurs over a similarly wide latitudinal range, but it appears to be most abundant east of Cape Point, particularly over the Central and Eastern Agulhas Bank. Most of the round herring caught in Namibian waters are juveniles, taken inshore as a small by-catch in the purse seine fishery. Although adults are occasionally taken farther offshore by bottom and midwater trawlers, the adult stock off Namibia is thought to be small compared to that farther south. Interaction between these stocks is probably of little consequence for management.

Juveniles of Cape horse mackerel (i.e. fish < about 20 cm) are most commonly found off the west and south-west coasts of South Africa, and off northern Namibia/southern Angola, south of the Angola/Benguela front. These fish are believed to originate from separate spawning stocks off South Africa's South Coast and northern Namibia respectively. Juvenile Cunene horse mackerel T. trecae are found in subtropical and tropical waters in Angola and (occasionally) off northern Namibia. Their distribution extends from north-west Africa to the Angola/Benguela front, which moves seasonally between about 14 and 17 oS, with an average position at 16 oS (Pereira 1988, Shannon et al. 1987). Because of the highly dynamic nature of the front, it is most likely that the juvenile Cunene horse mackerel found in southern Angola, and very occasionally in northern Namibia, are part of a single population.

Sardinella aurita and S. maderensis are found along the entire Angolan coastline, with the juveniles inshore, predominantly in the north. To the north their distribution extends apparently continuously along the coasts of Congo and Gabon, while to the south S. aurita can extend into northern Namibia in Benguela Niño years. In the north both species undertake extensive spawning migrations along the Angolan coast (S. aurita more so than S. maderensis), making it unlikely that distinctly separate stocks exist in this region. It has however been suggested (Wysokiński 1986) that the Angolan stock of S. maderensis is independent of the stock off the coast of Gabon.

Sardine tend to live within about 50 km of the coast, and are often found close inshore, both in South African and Namibian waters. Anchovy have a similar coastal distribution, but are commonly found more than 100 km offshore on the Agulhas Bank off the Cape South Coast in the spawning season. Round herring are found widely distributed across the shelf both on the West Coast and the Cape South Coast, with a clear increase in size with distance offshore. Juvenile Cape horse mackerel off both South Africa and Namibia, and juvenile Cunene horse mackerel off Angola, are most abundant inshore, generally being found within the 100m isobath throughout the region. S. aurita inhabits the continental shelf and is generally found in calm saline waters (>35 ppt) at temperatures around 24 oC. In contrast, S. maderensis is a coastal, more euryhaline species, also generally found at temperatures above 24 oC, often in the vicinity of river mouths (Luyeye 1995). The two species appear to be roughly equal in abundance, except in the south where S. aurita predominates. Surveys indicate that sardinella density is generally higher in the central (Luanda-Benguela) region than to the north and south (Luyeye 1995).

2.2 Trawled fish

The major species caught by trawl off Namibia and South Africa are the Cape hakes Merluccius capensis and M. paradoxus, which are caught in bottom trawls, and adult Cape horse mackerel, which are mostly caught in midwater trawls off Namibia and in bottom trawls off South Africa as a by-catch in the hake fishery. Other significant by-catch species in the hake fishery in both Namibia and South Africa are monkfish Lophius spp, kingklip Genypterus capensis, snoek Thyrsites atun and the West Coast sole Austroglossus microlepis. In recent years, the monk fishery in Namibia has become increasingly directed, with the West Coast sole as the most important by-catch. On the outer Namibian shelf there is also a valuable deep-water trawl fishery directed at orange roughy Hoplostethus atlanticus and, to a lesser extent, alphonsino Beryx splendens and other deep-water species. Off Angola there is a relatively small bottom trawl fishery for Benguela hake Merluccius polli and M. capensis (in the extreme south), and more important ones in central and northern Angola for demersal species such as Dentex spp. and red pandora Pagellus belloti. The large-eye dentex Dentex macrophthalmus is also taken off northern Namibia in midwater trawls, together with jacopever Helicolenus dactylopterus, another important by-catch species.

The distributions of the three species of hake in the Benguela region are shown in Figure 3. Merluccius polli occurs predominantly in Angolan waters, and is caught on the shelf slope as a by-catch in the prawn fishery and by deep water trawlers in the south, where its distribution overlaps with that of the shallow-water Cape hake M. capensis. The two species of Cape hake are found throughout Namibian and South African waters. M. paradoxus (deep-water hake) occurs in deeper water than M. capensis, although the two species co-occur at intermediate depths (Payne 1989). Typically the former is found in water 150 - 800m deep, mostly at temperatures between 4 and 8oC, whereas the latter occurs from the coast to a water depth of about 380m, in temperatures between 4 and 12 oC. It has been suggested (Payne 1989) that, in South Africa, the relative abundance of the two species is related to the width of the shelf and the steepness of the slope, with M. capensis predominating where the shelf is broad and the slope steep, and vice versa for M. paradoxus. Larger individuals of both species are found at greater depths than smaller fish, and there is little overlap in the distribution of mature fish. M. capensis is the more common species off Namibia, especially in the central region, although M. paradoxus has been increasingly abundant and more widely distributed there in recent years (see following). M. paradoxus predominates off the west coast of South Africa. It is believed that this stock may be the origin of the Namibian M. paradoxus stock. A second population of M. capensis, which for management purposes is treated as a separate stock, exists in the extreme southern Benguela, chiefly over the Agulhas Bank

In the 1970s and 1980s, ICSEAF treated Cape stocks off southern Angola/northern and central Namibia (15 - 25oS), southern Namibia (25 - 30oS), the South African West Coast (30oS - 20oE), and the South African South Coast as separate for management purposes. The species were considered together, as the catch records did not distinguish between them Fig. 3 suggests that the West Coast stocks of both species are probably shared between Namibia and South Africa, although catch patterns between Lüderitz and the Orange River indicate that there may be a measure of separation between the Namibian and South African M. capensis stocks. In contrast, there is some evidence from surveys (e.g. Strømme 1996) and commercial catches that, since 1990, there has been a gradual migration or expansion of M. paradoxus into southern Namibia and farther north, probably from South African waters. This is the only reported evidence of significant longshore movement of any of the three hake species, although it has previously been suggested that such shifts do occur in response to environmental changes such as temperature fluctuations and the movement of de-oxygenated bottom water.

Fig. 3 Distribution of the three hake species in the Benguela ecosystem (from Payne1989)

At least two stocks of Cape horse mackerel exist off southern Africa, viz. off northern Namibia/southern Angola, and off the Western Cape. These stocks were once believed to be genetically separated by the environmental barrier of the Lüderitz upwelling cell, with only a limited interchange between them. However, the finding of large adults on the bottom across the shelf south of Walvis Bay in recent Namibian surveys (E. Klingelhoeffer, NatMIRC, pers. comm.) raises the question of a possible connection between this part of the population and that on the West Coast of South Africa. This is a regional question of major importance for managing this trans-boundary fishery. Opinion is divided on whether the horse mackerel on the Cape West and South Coasts form a single or separate stocks, with catch patterns suggesting the former and genetic studies the latter. There is little information regarding the stock structure of Cunene horse mackerel in Angolan waters, although Sardinha (1996) has suggested on the basis of biological data and distribution patterns that there are separate self-sustaining populations in the north and south of Angola. The hypothesis is currently being tested genetically (Sardinha and Nævdal, submitted).

Orange roughy in Namibia are found mainly over the shelf between about 600m and 1000m depth at bottom temperatures of between 3 and 7 oC (A. Staby, NatMIRC, pers. comm.). The fish tend to be concentrated over hard substrata in a number of small areas, particularly during the spawning season. Alfonsino tend to be more widely distributed over the outer shelf, between about 400 and 700m. The degree to which the distribution of the two species extends into Angolan and South African waters, and the extent of any longshore migrations, is at present unknown.

The main commercial species of monkfish found in southern African waters are Lophius vomerinus (previously known as Lophius upsicephalus) and Lophius vaillanti. The former is found from northern Namibia to the East Coast of South Africa, but the latter only north of Walvis Bay. Both are demersal species, mainly found at depths of between 150 and 400m. Two separate recruitment areas for L. vomerinus have been located in Namibia: off Walvis Bay and near the Orange River. The relationship between these recruits and L. vomerinus to the south is unknown, as is the extent of any longshore migrations of adults. Nonetheless, it seems reasonable to assume that there is some interaction between the population(s) on the South African West Coast, and that found around the Orange River, in particular.

Demersal fish caught commercially by trawl in Angola can be grouped into species occurring below the thermocline along the continental shelf, and species found above, below and in the thermocline. Sparids are the most important in the former group. particularly the large eye dentex Dentex macrophthalmus, which is fished between 60 and 300m depth between Lobito and as far south as Walvis Bay (Constança 1995). Other important species are the Angola dentex Dentex angolensis and the red pandora Pagellus belloti, which occur across the shelf to depths of around 300m (Bianchi 1986). These three species together usually make up more than half of the demersal catch. In the latter group the most important species is the bigeye grunt Brachideuterus auritus. A significant part of the demersal trawl catch consists of deep-water prawns and to a lesser extent, deep-water red crab, which are discussed in the following Section.

2.3 Crustaceans

The major crustacean fisheries along the west coast of southern Africa are those for the West Coast rock lobster Jasus lalandii off South Africa and Namibia, the red crab Chaceon maritae off northern Namibia and Angola, and for the deep-water rose prawn Parapenaeus longirostris and striped red prawn Aristeus varidens off northern and central Angola.

J. lalandii is a spiny lobster associated with the cool upwelled waters of the Benguela. It occurs in commercially exploitable densities from east of Cape Point to approximately 25oS, and at lower densities beyond its core distribution. Close inshore it is caught by hoopnets deployed from dinghies and by recreational divers, but in deeper waters is harvested by traps.

C. maritae occurs on the slope of the continental shelf from about 27 oS off Namibia, northwards to Angola, Congo and the Ivory Coast. Off Namibia it is found on soft mud substrata at depths of between about 300 and 900m, and is harvested solely by Japanese vessels using traps. Off Angola it is found within a similar depth range, particularly in the southernmost area, and is taken in traps and (occasionally) bottom trawls. It has recently been shown from tagging studies (Le Roux 1997) that adult females migrate from Namibia to Angola, suggesting a single stock in the region, which needs to be managed jointly by Namibia and Angola.

Parapenaeus longirostris is essentially an Atlantic Ocean prawn, found from Portugal to Angola in the east, and from Massachusetts, USA, to French Guiana in the west, whereas Aristeus varidens is an East Atlantic species, found from Rio de Oro (24 oN) to 18 oS (off Namibia). The depth disrtribution of the two species differ. P. longirostris is found on the continental shelf and upper slope, between 50 and 400m depth (López Abellán and de Cárdenas 1990) over sandy bottoms. A. varidens lives on the slope, mainly between 400 and 800m depth, and is strongly associated with muddy bottoms. The size of both species increases with depth, more so in the case of P. longirostris than A. varidens. Angolan-Spanish surveys have revealed that P. longirostris have a more homogeneous distribution than A. varidens, which tends to concentrate in several areas, mainly related to submarine canyons. Concentrations of P. longirostris tend to occur around recruitment areas.

2.4 Line-caught species

The linefisheries of the Benguela Current and adjacent waters exploit a large number of species. They can be broadly classified into a) inshore reef fishes, which are mainly resident on shallow nearshore reefs and have a limited geographic distribution, b) migratory shoaling species, where the adults aggregate and migrate rapidly over large distances, usually as part of annual migratory cycles, and c) offshore large pelagic species such as tuna and billfish, which form large, highly-migratory straddling stocks that cross the borders of many countries, and even oceans.

In Angola, line-caught species belonging to the first two groups make up more than 40% of the total catch by the large inshore artisanal/subsistence fishery which extends along the entire coast. The most important groups caught by line are dentex, croakers (Sciaenidae) and groupers (Serranidae). Catches are highest in the provinces of Benguela, Namibe and Luanda (Anon. 1998a). The recreational fishery for linefish in Angola is underdeveloped and negligible compared to the artisanal/subsistence fishery.

The silver kob Argyrosomus inodorus is the most important of the linefish species caught commercially in Namibia, with roughly equal amounts being taken by commercial fishermen and recreational anglers. Other important angling species in Namibia are the steenbras Lithognathus aureti, the blacktail Diplodus sargus and the galjoen Coracinus capensis.

Off the South African West Coast, the dominant reef fish in the catches is the hottentot Pachymetopon blochii, which has been an important contributor to artisanal/subsistence linefish catches off the West Coast since the early part of the century. It is a highly resident species, which makes it susceptible to localised depletion in heavily fished areas. In contrast, the galjoen Coracinus capensis, which is a major shore-angling species off Namibia and the West Coast of South Africa, is highly migratory. Tagged galjoen have been found to have moved from northern Namibia to east of Cape Point.

Snoek is by far the most important migratory linefish species caught commercially on the West Coast of South Africa, and is important in Namibia as well. It is found along the entire southern African coast from southern Angola to Cape Agulhas, mainly in cool upwelled water, and is a major predator on pelagic fish in the region. Snoek were historically considered to form a single stock extending from Cape Agulhas to northern Namibia, and to migrate seasonally between these regions (e.g. Crawford et al. 1990). However, a recent study of all available evidence (Griffiths, in prep.) suggests that there may be two separate sub-populations - in Namibia and South Africa respectively - with medium-term (of the order of five years) exchange between them in response to environmental events and food availability.

Of the large pelagic species taken in the region, the most important in the southern part is the albacore or longfin tuna Thunnus alalunga, which is currently caught by South African and Namibian pole and line vessels within territorial waters from south of Cape Point to Lüderitz. The species is also exploited by Asian high-seas longliners off both countries. The stock is believed to be part of a single southern Atlantic stock, separated from the southern Indian Ocean stock by the warm water of the Agulhas current. There is also a longline fishery for bigeye tuna Thunnus obesus along the edge of the shelf in both countries, mostly by Asian high-seas vessels. It is not known whether these fish form part of an Indian Ocean, Atlantic Ocean or circumglobal stock. In Angola the most important species taken by local baitboats is the yellowfin tuna Thunnus albacares, while bigeye tuna is the major constituent of the Japanese longline fishery. Yellowfin form part of an Atlantic population, which spawn off Brazil and the Gulf of Guinea, and are most abundant in southern Angola in summer.

Finally, mention must be made of the Cape hake Merluccius capensis, which although primarily a trawl-caught species, is also caught by lines off South Africa's South East Cape, and (to a lesser extent) Namibia, using both hand and hydraulically-hauled longlines. The South African fishery grew out of an experimental longline fishery for kingklip in the 1980s, which severely depleted that stock.

2.5 Seals and seabirds

The Cape fur seal Arctocephalus pusillus pusillus occurs along the southern African coast between Algoa Bay and southern Angola and is harvested in Namibia. Although the harvest is low compared to earlier times (the fishery is centuries old), seals have been included in this overview because they are major top-predators in the Benguela, whose dynamics have been strongly affected by fluctuations in a number of the major fish resources of the region, making them important visible indicators of environmental change. The same is true of resident seabirds such as the Cape gannet Morus capensis, the Cape cormorant Phalacocorax capensis and the African penguin Spheniscus demersus, which breed mainly on nearshore islands and guano platforms off Namibia and South Africa, and feed largely on pelagic fish such as sardine and anchovy. The dynamics of pelagic fish stocks is strongly reflected in changes in abundance, diet and breeding success of the seabirds, to the extent that in South Africa, consideration is being given to using this information directly in the management of the sardine and anchovy fisheries.

3. LIFE HISTORY OF MAJOR RESOURCES

3.1 Small pelagic fish

Sardine and anchovy

Off Namibia, sardine spawn largely within 60 km of the coast in two main areas, one off Walvis Bay and the other farther north, in the mixing zone south of the confluence of the Benguela and the Angola Current systems. In the northern area, peak spawning (mainly by young adults) occurs near the 200m isobath in late summer/autumn in water temperatures between 19 and 21 oC, whereas spawning farther south (mainly by older fish) takes place in summer in cooler water close to upwelling zones. Since the collapse of the sardine stock in the 1970s (Section 4.1), spawning in the south has diminished in importance. The distribution and movement of anchovy off Namibia is similar to that of sardine there, but significant spawning only occurs north of Walvis Bay. The larvae of both species drift south close to the coast, recruiting as 0-group fish into the fishery in the cool upwelling areas near Walvis Bay. This is followed by a return northward migration of juveniles and young adults to the northern mixing area, where they first spawn. In the case of sardine, older fish subsequently return south again to spawn in the Walvis Bay region, although this migration is believed to have decreased in importance since the collapse of the fishery. The behaviour of sardine and anchovy off Namibia is analagous to that found in other upwelling systems, such as those off California, Peru and north-west Africa (Bakun 1995), where both spawning and recruitment occur downstream of the principal upwelling cell (Lüderitz in this case).

In the past, sardine in the southern Benguela have spawned in two areas, one of large adults in cool water west of Cape Columbine, and the other of younger fish in warmer water east of Cape Point, inshore of the Agulhas Current, analagous to the two sardine spawning areas off Namibia. However, since the major stock decline in the mid-1960s, there has been little evidence of spawning on the West Coast, and it seems that the stock is now perpetuated mainly by the younger adults spawning over the Agulhas Bank, particularly the Western Bank, between Cape Point and Cape Agulhas. Anchovy also spawn mainly over the Agulhas Bank, somewhat offshore of the sardine, in the upper mixed layer, in water temperatures ranging between 17 and 19 oC. Peak anchovy spawning is in early to mid-summer, whereas sardine spawn over an extended period, with weak maxima during late winter/early spring and late summer. In contrast to the northern Benguela, spawning occurs mainly upstream of the main upwelling centres of Cape Point and Cape Columbine. Eggs and larvae are transported from the Western Agulhas Bank to the West Coast (Fig. 4) between the 200m and 500m isobaths (Fowler and Boyd 1998), along the offshore boundaries of the upwelling cells by a perennial equatorward jet current, leading to recruitment downstream of the upwelling cells. Bakun (1995) has noted that this is a reversal of the normal spawning and recruitment pattern in other eastern-boundary upwelling areas (e.g. the northern Benguela, California Current and Canary Current), the difference originating from the right-angular configuration of the coastline in the southern Benguela

Fig. 4 Conceptual model of anchovy migration, constructed from acoustic survey data and trends in length in research midwater trawl catches (from Hampton 1992)

Sardine and anchovy larvae transported past Cape Columbine are either carried offshore and lost to the system, or are transported northwards and inshore (possibly assisted by active swimming as they develop), leading to recruitment of both sardine and anchovy inshore along the West Coast at least as far north as the Orange River (Fig. 4). There is evidence that in normal years, the northern limit of the dispersal is the southern edge of the Lüderitz upwelling cell, which causes the larvae either to be carried far offshore, or to be drawn inshore by compensatory flow and returned south by the inshore poleward counter-current. It has been suggested that in years when this cell is abnormally weak (as in 1987), pelagic larvae spawned in Cape waters may be carried past the Lüderitz "barrier" and recruit into the southern Namibian fishery, linking the South African and Namibian pelagic fisheries. This may have occurred in 1987, when an unexpectedly large number of anchovy recruits was caught in southern Namibia, resulting in catches roughly an order of magnitude higher than in the previous and following years.

Once inshore, the sardine and anchovy recruits off South Africa move south close inshore along the West Coast in autumn and winter, assisted by the poleward counter-current. They are possibly retained for a period in St Helena Bay before reaching the Agulhas Bank in early summer, towards the end of their first year of life. It is during this period that they are exploited by the purse-seine fleet, which mainly operates from ports on the West Coast. Adults tend to move eastward and (in the case of anchovy) offshore, with increasing age, although there is some return of older fish to the West Coast from the outer edge of the Agulhas Bank, probably assisted by the jet current off the Cape Peninsula.

A portion of the sardine population (mostly fish in their second year of life) undertakes a pronounced and well-documented inshore migration into the shelf waters of KwaZulu-Natal each winter (known locally as the "Sardine Run"), to at least as far up the coast as Durban. This is probably in response to the eastward retreat of warm subtropical water close inshore at that time of the year, and is assisted by equatorward counter-currents inshore of the Agulhas Current. The migration may be analagous to the movement of young Namibian sardine towards the subtropical system boundary in winter. It is most probable (though not yet demonstrated) that the surviving fish return rapidly to the Agulhas Bank later in the year in the strongly flowing Agulhas Current slightly farther offshore.

Analyses of stomach contents up to the end of the 1970s suggested that the diets of sardine and anchovy in the Benguela are similar, with phytoplankton as the main food source. However, laboratory and field studies since then have shown that zooplankton is more important in the diet of both species than was previously believed to be the case. Certainly, juvenile sardine and anchovy, and adult anchovy, feed primarily on zooplankton, although adult sardine appear to utilise more phytoplankton in areas of consistently high phytoplankton abundance. Little is known about the extent to which sardine and anchovy compete for habitat and food in the Benguela ecosystem, but a recent field study on feeding in mixed schools of juveniles, in which a clear size difference was found in the zooplankton taken by the two species (Louw et al. 1998), suggests that direct competition may be limited, at least in the younger stages.

The main spawning area of S. aurita and S. madarensis is thought to be between 5 and 7 oS (Pointe Noire to south of the Congo River), with peak spawning in March-April. There appears to be a seasonal longshore migration pattern for both species, with an equatorward movement towards the spawning grounds during the first part of the year, and a return poleward migration of adults in the second half of the year to central and (in the case of S. aurita) southern Angola. Juvenile S. aurita are encountered over the whole littoral zone from Cape Lopez in Gabon to Baia dos Tigres in southern Angola. Upon reaching a length of 10 – 14 cm, the juveniles leave the littoral zone, and remain for some time in shallower shelf waters before joining the adult stock farther offshore (Troadec and Garcia 1980). Juvenile S. maderensis appear to be largely concentrated inshore north of the Congo River throughout the year (from data in Wysokiński 1986).

Little is known about the diet of sardinellas in Angola, although French studies in the 1970s found that the diet of both S. madarensis and S. aurita in Congolese waters consisted almost entirely of the copepod Calanoides carinatus.

Little is known about the seasonality of round herring spawning in the Benguela, although from ichthyoplankton surveys in the south, it appears that there at least, spawning probably occurs throughout the year, reaching a peak between late winter and early summer. The early life history of round herring spawned on the Agulhas Bank appears to be similar to that of sardine and anchovy spawned there, with the ichthyoplankton being transported to the West Coast by the same jet current, leading to recruitment along the West Coast at the same time of the year as sardine and anchovy recruitment, and movement onto the Agulhas Bank towards the end of the first year of life. The fish appear to move eastwards and offshore with increasing age, inhabiting the entire Agulhas Bank, at least as far east as Port Elizabeth (Roel and Armstrong 1991). As with sardine and anchovy, there is probably some return of older fish to the West Coast from the outer regions of the Bank.

Round herring is a particulate feeder, whose diet in South African waters consists exclusively of zooplankton (mainly copepods, euphausiids and decapods). Juveniles form pelagic shoals in the upper mixed layer, but adults perform pronounced diel vertical migrations, migrating from near the surface at night to near-bottom waters during the day, often passing through a temperature gradient of more than 10 oC.

Cape hakes spawn in midwater throughout the year, with a peak in early summer for both M. capensis and M. paradoxus, and a secondary peak in late summer for M. paradoxus in the southern Benguela. Most M. paradoxus spawning is thought to take place along the edge of the Agulhas Bank, but spawning also occurs over the shelf-break west of St Helena Bay and off central Namibia. In the latter region, M. capensis spawn most frequently between 160 and 250m depth, spawning starting earliest in the shallower waters. The eggs of both species are concentrated around the depth of the thermocline, and the dispersal of M. paradoxus eggs and larvae produced on the South Coast could be similar to that of the pelagic fish spawning there, resulting in young M. paradoxus being plentiful between Cape Columbine and the Orange River. To the north, 0-group M. capensis are particularly abundant off Walvis Bay, which appears to be a nursery area. Juveniles of this species are also plentiful off the Orange River and south to about Cape Columbine, sometimes co-occuring with pelagic fish recruits in winter, on which they feed (particularly on anchovy). As with the pelagic recruits, juvenile hake in the Orange River area move south as they grow older, but unlike them they tend to move offshore as they move south.

There is evidence that off the West Coast of South Africa, juvenile M. paradoxus move inshore in summer and offshore in winter, in response to changing feeding regimes. The adults, which tend to concentrate in depths greater than about 500m in summer and autumn, move inshore in spring to depths of around 300m, and then return offshore, movements which are probably related to both spawning and feeding. Off Namibia, M. capensis follows a similar inshore migration during the spawning season in early summer, followed by an offshore migration in late summer.

Cape hakes feed both close to the bottom and in midwater, and tend to be off the bottom at night, although no clear feeding periodicity in either M. capensis or M. paradoxus has been demonstrated, except in the case of juvenile M. capensis, which off the West Coast of South Africa have been observed to move off the bottom at night to feed on pelagic prey such as juvenile anchovy, and to return to the bottom before dawn (Pillar and Barange 1995). Recent studies (Pillar and Barange 1998) have indicated that M. capensis adults on the South African West Coast also move into midwater at night in response to the vertical migration of their prey, but that they only return to the bottom when satiated, regardless of time of day. This results in aperiodic, asynchronous vertical movements of individuals, depending on food availability and recent feeding activity. This lack of a distinct diel feeding rhythm has also recently been reported by Huse et al. (1998) from studies on the behaviour of M. capensis and M. paradoxus at a location on the central Namibian shelf, although they did find some evidence of increased feeding in the early evening, in common with earlier studies of M. capensis feeding in the same area.

Cape hakes are opportunistic feeders, resulting in considerable seasonal and spatial variability in their diet. On the South African West Coast, young M. capensis and M. paradoxus feed predominantly on planktonic crustaceans (particularly euphausiids), juvenile anchovy (by M. capensis), lightfish Maurolicus muelleri and lanternfish Lampanyctodes hectoris (by M. paradoxus), the diet of both species becoming increasingly piscivorous with age (Punt et al. 1992). Squid, epipelagic fish and, to a lesser extent, mesopelagic fish such as lightfish and lanternfish comprise a significant proportion of the diet of adult M. capensis, but in the larger fish the principal diet items are small M. paradoxus, small M. capensis (to a lesser extent) and other demersal species (Punt et al. 1992). With increasing age, M. paradoxus becomes increasingly cannibalistic on young M. paradoxus, which together with squid, crustaceans and mesopelagic fish constitutes most of the diet of the large mature adults on the Cape West Coast. In contrast, more than 90% of the diet of large M. capensis on the Agulhas Bank consists of pelagic fish (particularly anchovy), horse mackerel and young M. capensis, the size of prey increasing with hake size (Pillar and Wilkinson 1995). Cannibalism on M. capensis appears to be higher than on the West Coast, but the interspecific (hake-on-hake) predation somewhat lower.

The diet of hake in Namibia is similar to that in South Africa (see summary in Gordoa et al. 1995), the chief difference being the greater importance of myctophids (particularly for M. paradoxus), and the significant proportion of gobies Sufflogobius bibartus in the diet of both species. As in South Africa, both species become increasingly piscivorous with age, and hake-on-hake predation becomes increasingly important.

Because of their catholic feeding habits and abundance, hake are extremely important predators in the Benguela. For example, Punt et al. (1992) have estimated that Cape hakes in South African waters could consume as much as 6 million tonnes of food annually. Based on estimates of stock size in Namibia, it would appear that consumption there could be as high.

Cape horse mackerel T. trachurus capensis off the West Coast of southern Africa have broad spawning areas, with most intense spawning in warmer waters immediately west of the shelf-break in both regions. Egg and larva surveys in the 1970s (O’Toole 1977) showed that in Namibia the heaviest spawning occurs in the north between October and March in the mixing zone of warm oceanic water and cool coastal water, and that the timing of spawning is closely linked with the duration and intensity of mixing. Nursery areas exist in both the southern and the northern parts of the ecosystem, adjacent to the spawning grounds but closer inshore, and there are substantial longshore and cross-shelf migrations of both juveniles and adults.

Off Namibia, juvenile Cape horse mackerel live inshore, the smallest fish being found farthest north. Slightly larger individuals appear to migrate south towards Walvis Bay, especially in winter. Maturing fish move offshore and northwards to spawn, the adults generally occuring north of 21 oS.

Off the Western Cape, juveniles also occur inshore and the adults farther offshore. The movements of the fish with increasing age, and particularly the interchange between the West and South Coast populations, are particularly complex, as can be seen from a recent migration model postulated by Barange et al. (1998), reproduced here as Fig. 5. Note that the life history in the first year of life is similar to that of sardine and anchovy, with spawning on the South Coast leading to recruitment on the West Coast in the following year, probably driven by the same transport mechanisms. Recruitment occurs about three months earlier than that of sardine and anchovy, but overlaps with it to some extent, resulting in a by-catch of juvenile horse mackerel in the pelagic fishery. Thereafter the species diverge, with the horse mackerel becoming more demersal and moving offshore, probably ultimately leading to spawning over the shelf-break on the West Coast. Some of these fish however move onshore again in winter in their second year of life, and move onto the Western Agulhas Bank, assisted by the poleward counter-current on the inner shelf. These fish reach maturity at two years of age and move eastwards and offshore with increasing age, leading to spawning across the entire Agulhas Bank, which peaks at different times on the Western and Eastern Banks and which, on the Western Bank at least, appears to be closer inshore in winter than in summer. Barange et al.(1998) suggest that adults on the Eastern Bank move southward and westward to spawn on the Central Bank in spring, a movement which would favour transport of the ichthyoplankton to the West Coast. They also suggest that there is some secondary recruitment directly into the numerous bays of the South Coast from spawning farther offshore, probably aided by the inshore movement of spawners in winter. However, most of these hypotheses are still tentative and need rigorous testing.

Fig. 5 Conceptual model of Cape horse mackerel migration patterns in the southern Benguela, constructed from acoustic and research trawl information (from Barange et al. 1998)

According to Troadec and Garcia (1980), the core spawning area of the Cunene horse mackerel is off Mauritania and Senegal, where spawning peaks from February to June. Although spawning undoubtedly occurs farther south as well, including in Angolan waters, little has been reported about the locality and timing of any such spawning. Little is also known about the seasonal distributional patterns of the different life history stages of the species off Angola, although the fact that at the height of the fishery in the 1960s and 1970s, inshore catches used to show a sharp peak in October/November co-incident with a rise in water temperature, has led to suggestions that there is a seasonal inshore migration in response to changing environmental conditions. This is supported by survey data, which indicate that the proportion of the population in central Angola is highest in summer (Sardinha 1996).

Cape horse mackerel up to the age of two years feed near the surface and are planktivorous. The diet, which consists mainly of copepods, is similar to that of sardine and anchovy, and juveniles up to about 10 cm in length can co-exist in schools with sardine and anchovy. Adults off the west coast of southern Africa are opportunistic feeders on euphausiids (which constitute 95% of their diet in Namibia), polychaete worms, chaetognaths, squid, various crustaceans and fish such as gobies Sufflogobius bibartus, lanternfish and lightfish. Older horse mackerel tend to feed in midwater, and their diet is similar to that of Cape hakes of similar size. Accordingly, there may be interspecific interaction between Cape horse mackerel and Cape hakes, with a decrease in the abundance of the one species benefiting the other, and vice versa. Pillar and Barange (1998) showed that adults on the Cape South Coast feed predominantly on copepods close to the bottom in late afternoon, migrate in synchrony into midwater at dusk, clearly for purposes other than for feeding, and return to near-bottom at dawn. Sardinha (1996) reports that in Angolan waters, horse mackerel of both species undergo a similar diel vertical migration. There have been no comparable studies in Namibian waters, but it would appear from the fact that adult horse mackerel there are caught by midwater trawl throughout the 24-hour period that their vertical migratory behaviour in this region may be different from that to the north and south.

Orange roughy off Namibia have a short spawning period of about a month in July/August, when they spawn in dense concentrations close to the bottom in small areas typically no more than 10 - 100 km2 in extent. They are exceptionally long-lived and slow-growing, possibly only reaching sexual maturity at around 25 years off Namibia, and may have a maximum lifespan of over 100 years. The fish have a low reproductive rate, which together with their aggregating behaviour, makes them highly vulnerable to over-fishing. Alfonsino Berxy splendens are distributed over a wider area and are probably more productive. Little is known about their spawning behaviour or breeding habitat.

Dentex macrophthalmus spawn throughout the year with a peak between October and April. The most intensive spawning occurs between Baia dos Tigres and the Cunene River at depths between 25 and 110m (Kuderskaya 1985). Little is known about migration patterns, except that according to Wysokiński (1986), the stock migrates shorewards in summer and offshore in winter. Even less has been reported on the life histories of the other commercially exploited demersal fish in Angolan waters.

3.3 Crustaceans

West Coast rock lobster

J. lalandii has a well-defined moulting and spawning cycle. Adults moult once per year, the males in spring and the females in late autumn/winter, after which mating takes place. Egg-hatching peaks in October-November and the phyllosoma larvae remain planktonic for a long period, drifting in oceanic sub-gyres until they reach the puerulus (free-swimming) stage and settle. Females reach sexual maturity about five years after settlement, at a greater length in the south than in the north. Maturing males grow faster than females, resulting in the fishery being based largely on males. The adults are generally distributed offshore of the juveniles, except in the north, where the population is constrained close to the coast by low-oxygen water. J. lalandii feed largely on mussels, in particular the ribbed mussel Aulacomya ater, which is abundant in the rocky subtidal zone of the West Coast. In areas of low mussel abundance, the diet consists mainly of echinoderms (sea urchins and starfish), gastropods, bryozoans, polychaetes and seaweeds. The principal predators on J. lalandii are octopus, dogsharks, hagfish, whelks (on injured or weakened animals) and young seals. Cannibalism is known to be prevalent in overcrowded situations, particularly among juveniles.

Chaceon maritae in Namibia appear to spawn throughout the year, judging from the fact that no seasonal cycles in moulting and egg-bearing have been found there (Le Roux 1997). Adult females generally live in shallower water than males, and virtually all egg production and larval release takes place on the shallower part of the continental slope; a pattern which has also been found in Angola. The fact that the migration from Namibia to Angola is almost entirely by females suggests that this is a spawning migration.

Deep-sea prawns

Parapenaeus longirostris spawn throughout the year, with peaks in July and December. Sobrino and de Cardenas (1989) state that females reach sexual maturity at a carapace length of 21.6 mm. According to López Abellán and Garcia-Talavera (1992), there are nursery areas between 8 and 9 oS and between 10 and 11 oS. Eggs are demersal and the larvae planktonic. The larvae enter the post-larval phase at a length of 12 mm. Juveniles are concentrated between depths of 50 and 70m, where recruitment takes place. A sampling programme has shown that there is a major recruitment peak between January and March, and a lesser one in August-September (López Abellán et al. 1993).

Aristeus varidens appear to spawn throughout the year, with a peak in December and probably another in April-May. First maturity is reached at a carapace length of about 25 mm (Sobrino and de Cárdenas).

3.4 Line-caught species

Snoek spawn along the edge of the shelf off the West Coast of South Africa and the western Agulhas Bank, and off southern Namibia, mainly from July to October. There is also some evidence of spawning farther north. It has been suggested that in South African waters the ichthyoplankton is transported by prevailing currents to a primary nursery ground north of Cape Columbine, and to a secondary one east of Danger Point on the South-West Coast (Griffiths, submitted). Juveniles tend to recruit inshore and remain as locally migratory shoals in nearshore nursery areas until they approach maturity, when they join the adult population. Their cross-shelf distribution on the inner shelf is determined by prey availability, and includes a seasonal inshore migration in autumn in response to pelagic fish migration patterns. Adult snoek are found throughout their distributional range, moving offshore and somewhat southwards to spawn. Other than this, there do not appear to be any seasonal trends in longshore movements of adults in South African waters (Griffiths, in prep.). Migrational patterns of juveniles and adults in Namibian waters have not been established with any certainty.

Albacore, the main contributor to the South African and Namibian large pelagic fisheries at present, are believed to migrate across the southern Atlantic to South America, and then northwards to spawn in the tropical central Atlantic. Juveniles occasionally recruit into waters off the Western Cape, but most of the fish caught are large reproductively inactive adults, following and feeding on the rich pelagic prey in the Benguela and Agulhas Current systems.

3.5 Seals and seabirds

The Cape fur seal breeds on small rocky nearshore islands and, most importantly, at six mainland colonies on the Namibian and northern Cape coasts where human access is restricted. Two of these colonies (Kleinsee in the northern Cape and Atlas Bay near Lüderitz) are believed to be the largest mainland seal colonies in the world. The breeding season, during which pupping is followed almost immediately by mating, lasts for 6 to 8 weeks in October/November. Pups are weaned at an age of 8 to 10 months, and thereafter forage widely. While attending their pups, adult cows feed within a few days range of the colonies, but the bulls appear to have separate feeding grounds, probably considerably further offshore. Much of the diet is made up of fish, of which commercial pelagic fish and hake are the most important on the South African West coast, and the bearded goby Sufflogobius bibartus (a non-commercial species), horse mackerel and juvenile hake the most important off Namibia. It has been estimated that seals in the Benguela consume about 1 million tonnes of fish annually, which is of the same order as the total annual fish catch in Namibia and South Africa.

The Cape gannet breeds on islands off southern Namibia and the West and South Coasts of South Africa, usually from September to November. The birds range widely during the non-breeding season, following their prey (largely sardine and anchovy), which they capture by plunge-diving. The Cape cormorant breeds mostly on nearshore islands and guano platforms, but also at islands within estuaries and lagoons, sewage works and on mainland cliffs. The breeding distribution extends from northern Namibia to Algoa Bay. They are generally dependent on large surface schools of fish, which they capture by pursuit-diving, and do not forage as widely as gannets. Anchovy and sardine are the preferred prey species, with bearded goby an important prey in southern Namibia. The breeding range of African penguins extends from Algoa Bay to Sylvia Hill, south of Walvis Bay. They generally breed on islands, although there are a few small mainland rookeries in South Africa and Namibia. Pelagic shoaling fish, particularly sardine and anchovy, are the most important prey, which are caught by deep pursuit-diving. As with the Cape cormorant, the bearded goby has at times been an important prey off Namibia, particularly when sardine have been scarce.

4. HISTORY AND CURRENT STATUS OF THE FISHERIES

Industrial catches of the major species in each country in 1997 are summarised in Table 1. Note that zeroes in parenthesis indicate an assumed zero catch in the absence of data to the contrary. A more general discussion of the history and current status of each sector of the fishery follows.

Table 1. Industrial catches (in tonnes) of major species in each country in 1997, except where indicated otherwise. Assumed zero catches in parenthesis. Data for Angola from IIP. Data for Namibia and South Africa from Stuttaford (1998).

Annual landings of the major species exploited in the purse seine fisheries of the region from the 1950s to the present are shown in Figs. 6, 7 and 8 for Angola, Namibia and South Africa respectively. They are discussed by species below.

Fig. 6 Purse-seine catches of sardinella (Sardinella maderensis and S. aurita combined) and (predominantly) juvenile Cunene horse mackerel off Angola since 1956

Fig. 7 Purse-seine catches of sardine, anchovy and juvenile Cape horse mackerel off Namibia since 1950

Fig. 8 Purse- seine catches of sardine, anchovy, round herring and Cape horse mackerel off South Africa since 1950

Industrial catches of sardine in Angolan waters from the 1950s to the present, which are reduced to meal and oil, have fluctuated widely from almost zero to a maximum of 146 000 tonnes in 1957, depending on the state of the Namibian stock and the degree to which it has extended into Angolan waters. The fish were initially caught by a large local fleet of small vessels, but after Independence in 1975 were mainly taken by larger, foreign-owned purse-seiners which began fishing on sardine and sardinellas off Angola in 1976, processing their catches to a large extent at sea. Since 1994, when the adundance of sardine in Namibian waters has been particularly low, a number of Namibian vessels have fished under license for sardine in southern Angola, but only in 1995, when 47 000 tonnes were caught by such vessels, have these catches been significant.

In Namibia, annual catches of sardine (mainly adult fish, taken for both canning and reduction to meal and oil) rose rapidly from levels of around 200 000 tonnes to a maximum of nearly 1.4 million tonnes in 1968, whereafter there was a sharp decline to below 300 000 tonnes in 1971, followed by a slight increase in catches for a few years and a precipitous collapse in 1977 and 1978. Since then, annual catches have rarely exceeded 50 000 tonnes, reaching an all-time low of a little over 1 000 tonnes in 1996. It is most likely that these collapses were largely due to over-fishing (especially in the late 1960’s when in addition to the Walvis Bay fleet there were two factory vessels operating outside territorial waters), perhaps exacerbated by a number of years of poor recruitment as a result of adverse environmental conditions. A change from sardine nets to anchovy nets in the late 1960s, which would have placed greater pressure on the recruits, may also have been a contributing factor. With the decline of the stock in the 1970s the fishery moved increasingly northwards, and the fleet changed from small, predominantly wooden-hulled vessels to larger steel-hulled refrigerated-seawater vessels capable of returning the fish from northern Namibia in a condition suitable for canning.

South African catches of adult sardine for canning, and of both adults and juveniles for fishmeal and oil, rose from around 100 000 tonnes per annum in the early 1950s to a peak of around 400 000 tonnes per annum in the early 1960s. Thereafter, catches declined sharply to well below 100 000 tonnes per annum, a level which has only recently been regained with the steady growth of the stock since the mid-1980s. Prior to the collapse, the fishery was based on large adults off St Helena Bay, caught predominantly in winter. Thereafter it shifted to younger adults farther south, caught throughout most of the year, and recruits close inshore on the West Coast in autumn and winter as a by-catch in the anchovy fishery, a pattern which still persists. The fleet changed from 32-mm mesh sardine nets to 13 mm anchovy nets in the mid 1960s.

In the 1970s and 1980s, catches of South African and Namibian anchovy, which were not targeted in South Africa prior to 1964 and were hardly caught in Namibia before 1966, have been less variable than those of sardine, fluctuating under quota control around a level of around 300 000 tonnes in South Africa and 200 000 tonnes in Namibia. An exception was the pronounced peak in 1987 and 1988 in South Africa and in 1987 in Namibia, when catches were roughly double the average. The peak in South Africa is thought to have been due to exceptionally good recruitment, at least partly associated with enhanced transport of eggs and larvae to the West Coast in those years, while in Namibia, the peak was probably caused by an anomalous influx of anchovy recruits from the Cape stock, driven by the same environmental factors which caused the good recruitment in the south that year. Annual catches in South Africa in the 1990s have varied between 40 000 and 347 000 tonnes, but the Namibian catches have averaged less than 50 000 tonnes per annum during this period, with a decline to virtually zero in 1996 and 1997. Recent surveys confirm that the anchovy stock in Namibian waters is severely depleted at present.

The purse-seine fishery for predominantly juvenile T. trecae off Angola started in the 1950s. Catches rose in fluctating fashion from an average of slightly under 100 000 tonnes per annum between 1956 and 1965 to 261 000 tonnes in 1972. Yields then decreased with the cessation of Portuguese rule in Angola, but a record catch of at least 380 000 tonnes was recorded in 1978, by which time distant-water foreign midwater and bottom trawlers, targeting more on adults, had entered the fishery. In the early 1980s, the proportion of the catch taken by purse-seine varied between 15 and 47%, the remainder being taken by midwater (47 - 76%) and bottom (5 - 18%) trawl. Catches declined sharply during this period, and by 1984 only 55 000 tonnes of T. trecae in total were landed. Angolan catches since then have fluctuated between about 25 000 tonnes in 1985 and 1991 and 130 000 tonnes in 1996. A small part of the catch (approximately 2 000 tonnes in 1997 – Anon. 1998a) is taken by small-scale purse-seiners operating close inshore in what is essentially an artisanal fishery.

In Namibia, horse mackerel were not recorded in purse-seine landings until 1971 when, following the first collapse of the sardine fishery, 140 000 tonnes were caught. Since then there have been sporadic catches in excess of 100 000 tonnes per year, with an average of 59 000 tonnes and a maximum of 116 000 tonnes in 1992. The fish are utilised entirely for meal and oil.

Purse-seine catches of horse mackerel in South Africa, which were once second only to sardine in the landings, peaked at 118 000 tonnes in 1954, and thereafter declined steadily to the early 1970s. Since then horse mackerel have only been a minor constituent of the pelagic fishery, with annual catches never having exceeded 10 000 tonnes. This decline is thought to have been due to a decline in the parent stock, probably aggravated by increased fishing pressure on the younger juveniles with the introduction of anchovy nets in the mid-1960s.

Annual catches of the two sardinella species off Angola by the industrial fishery up to the mid-1970s fluctuated between about 50 000 and 150 000 tonnes. With the introduction of the large purse-seiners after Independence, annual catches rose to around 300 000 tonnes, followed by a steady decline from the mid-1980s to 1992 when catches stabilised at around 50 000 tonnes per year. In 1994 there were 36 vessels operating in this fishery, of which 26 were purse-seiners and 10 pelagic trawlers. The catch is used for fishmeal because the fish is too bony for canning. A small amount of sardinella (around 2 000 tonnes in 1997 – Anon. 1998a) is also caught by artisanal fishermen in small-scale purse- seine operations close inshore for local sale and consumption.

Round herring have been a minor constituent of the South African purse-seine fishery since the late 1950s with annual catches in the 1980s and 1990s running at a level of about 50 000 tonnes, peaking at 76 000 tonnes in 1995. The juvenile fish are caught as a by-catch in the fishery for juvenile anchovy and sardine inshore on the West Coast, and the adults in targeted fishing between Cape Columbine and Cape Point, somewhat farther offshore, particularly during the first three months of the year. There have also been sporadic attempts to exploit the large adults in the Algoa Bay region by both purse-seine and midwater trawl. Some of the catch has been canned and the rest reduced to meal. The round herring resource in South African waters is believed to be under-utilised at present, and attempts at greater exploitation have been encouraged. In Namibia, around 1 000 tonnes of juvenile round herring are usually taken each year by the purse-seine fleet for reduction, although catches as high as 14 000 tonnes (in 1996) have been recorded. The potential for canning the larger fish has been investigated, but they have generally been found to be too soft for this purpose.

Annual catches of Cape hakes (M. capensis and M. paradoxus combined) in Namibian and South African waters by local and foreign fleets since 1950 are shown in Fig. 9. Although the demersal fishery began around the turn of the century, catches prior to 1950 seldom exceeded 50 000 tonnes per annum, with most fishing effort being in South African waters. The Namibian fishery started in the late 1950s. In the early 1960s there was an explosive increase in effort and landings throughout the Benguela, with the arrival of foreign trawling fleets, and by 1972, the annual hake catch in the south-east Atlantic exceeded 1.1 million tonnes. Subsequently, catch rates and landings of hake declined sharply, and conservation measures were introduced, including the declaration of a 200-mile fishing zone by South Africa in 1977. Since then hake catches in South African waters have remained relatively stable at just over 140 000 tonnes per year. Off Namibia, hake catches between 1973 and Independence in 1990 averaged 500 – 600 000 tonnes annually, mainly taken by foreign fleets. At Independence, strict conservation measures were introduced, including the exclusion of foreign vessels. The hake catch is now taken exclusively by Namibian-registered vessels and the annual local catch has risen from 55 000 tonnes at Independence to around 120 000 tonnes over the period 1996 – 1998. Catches of M. capensis and M. polli in Angola are of a lower order, amounting to less than 1 000 tons per year in recent years.

Fig. 9 Catches of hakes off Namibia and South Africa by foreign and local fleets since 1950

Adult horse mackerel are the main target for midwater trawlers operating in Namibia and Angola, the Namibian fishery being the largest by volume in that country. Trawl catches of the two species rose from under 50 000 tonnes per annum in the early 1960s (when horse mackerel were only trawled in Namibia), to between 600 000 and 700 000 tonnes per annum from 1982 to 1984 (Fig. 10), by which time most of the catch was being taken by foreign (mainly former Soviet bloc) vessels. The fish were largely frozen and shipped back to Europe for human consumption. Trawl catches of horse mackerel (mainly T. trachurus capensis) in Namibia since Independence in 1990, when Namibia took control of the fishery, have fluctuated around 350 000 tonnes per annum, with a decline to between 200 000 and 250 000 tonnes per annum in recent years. The number of midwater trawlers in the fishery now is less than half that at Independence. The fleet is largely made up of ageing ex-Soviet bloc vessels, about half of which are now registered in Namibia, but which are mostly operated by foreign crew. The major part of the catch is frozen and transhipped to reefer vessels for export as a relatively low-value product to West Africa, but a small amount is now being smoked or dried-salted ashore for export to African countries.

Fig. 10 Catches of (predominantly) Cape horse mackerel by midwater trawl and purse-seine off Namibia since 1961

Purse-seine and trawl catches of Cunene horse mackerel in Angola were summarised in Section 4.1. Compared to these catches, which have generally been between 50 000 and 100 00 tonnes per annum over the past decade, catches of Cape horse mackerel in Angolan waters in the past decade have been insignificant.

The fishery for T. trachurus capensis in South Africa changed from a purse-seine operation for juveniles on the West Coast in the 1950s and early 1960s to a bottom-trawl fishery for adults on the South Coast thereafter, caught by both local and foreign (mainly Japanese) trawlers. Catches rose from about 10 000 tonnes in 1964 to a peak of nearly 100 000 tonnes in 1978 (Fig.11), and have averaged about 30 000 tonnes per annum over the past two decades. Between 1967 and 1975, horse mackerel contributed some 40% of the landings of the inshore demersal fishery on the South Coast, but since the gradual phasing out of the foreign fishery from 1982, this proportion has declined to around 20%. In the 1990s a targeted midwater trawl fishery for horse mackerel developed, mostly on the Eastern Agulhas Bank. Catches have been relatively low (average just under 8 000 tonnes per annum between 1990 and 1997) due to operational and marketing constraints, but there is concern that this fishery has the potential to threaten the resource. Horse mackerel catches in bottom trawls on the West Coast in the past decade have never exceeded 5 000 tonnes per year.

Fig. 11 Cape horse mackerel catches in the South African purse-seine (West coast) and trawl (South Coast) fisheries since 1950

The Namibian monkfish fishery was initially a by-catch fishery, but in recent years it has developed into a target fishery in response to increasing market demand. At Independence in 1990 the fishery changed from an international to a local fishery. Since then annual catches (mainly by small wetfish trawlers) have risen from 1 500 tonnes (the lowest value in the past decade) to around 10 000 tonnes, comparable to peak levels reached by the international fishery in the 1980s. The value of the product is high, making this fishery an important contributor to the Namibian economy (Olsen 1997). The by-catch of monkfish in the South African demersal trawl fishery on the West and South Coasts has been relatively constant over the past two decades, at around 5 000 tonnes per year.

In Angola, roughly 10 000 tons of demersal fish have been caught annually in trawls in recent years, mainly dentex spp., croakers (Scianidae), grunters (Pomadasydiae) and groupers (Serranidae). In Namibia the most important other species caught in bottom trawls are kingklip and sole, catches of which have averaged 1 275 and 340 tonnes respectively per year since 1990. The average combined catch of dentex and jacopever in midwater trawls in Namibia since 1990 is 3 750 tonnes (Boyer et al. 1998). In South Africa the two most important other trawled species are snoek and kingklip. Between 1990 and 1995 catches of these two species averaged 10 900 and 2 200 tonnes per year respectively.

Exploratory fishing for deep-water trawl species, particularly orange roughy, began in Namibia in 1994. Small catches were made in 1995, but in the following season over 12 000 tonnes of roughy were caught (by bottom trawl) making this the second-largest roughy fishery in the world. The alfonsino catch during these two years was nearly 3 000 tonnes. Indications are that the 1997/98 catch of both species will be at the same level. At present only four roughy aggregations are being targeted, although there is some exploratory fishing for further aggregations. Five vessels are currently operating, in joint ventures between foreign and Namibian companies. The roughy and alfonsino catch is entirely exported, mainly in the form of high-value frozen fillets for the USA and Japanese markets respectively.

There has been a fishery for West Coast rock lobster since the early part of the century, and for a long period the fishery was the world’s largest for any Jasus species. During the 1940s and 1950s South African catches were relatively stable at around 9 000 tonnes (whole mass) per annum, but from the mid 1960s to the present catch levels have declined under quota control to around 1 500 tonnes per annum, with particularly sharp declines in the late 1960s, early 1980s and early 1990s, the most recent one being attributed at least in part to a sharp decline in the growth rate of individuals. In Namibia, where the resource was clearly over-exploited, annual catches have declined even more dramatically, from a peak of nearly 9 000 tonnes in 1966 to about 3 000 tonnes in the early 1970s, to half this level a decade later, and to a few hundred tonnes in recent years.

Deep-sea red crab

The red crab fishery in Namibia started in 1973. Catches rose to a peak of about 10 000 tonnes in 1983, after which annual landings declined steadily to 2 676 tonnes in 1991, clearly as a result of over-fishing. Catches have fluctuated around this level since then. In only two of the years since the introduction of TACs in 1989 has the TAC been reached. The entire catch is taken by Japanese vessels fishing with traps. Off Angola, a few hundred tonnes of red crab are taken by trawl each year as a by-catch of the deep-water prawn fishery. Since 1986 there has also been a directed crab fishery by a single Japanese vessel, which has caught a few thousand tonnes per year on average.

The deep-water prawn fishery off the west coast of southern Africa (mainly for rose and striped prawns) began in 1966, when a single Spanish trawler started operating off Angola. By 1972, a fleet of 54 foreign (mainly Spanish) trawlers was operating off Angola, and catches had risen from about 1 000 tonnes per annum to over 8 000 tonnes per annum. The catch peaked at over 12 000 tonnes in 1973, whereafter there was a sharp reduction in Spanish operations, ending in complete withdrawal from the fishery by 1977. The Spanish fleet was replaced by a Cuban fleet, which continued to fish until 1979, the total catch peaking at 11 400 tonnes in 1975. Since then catches have declined, with annual landings of the two species combined over the past decade fluctuating between about 3 500 and 7 000 tonnes per year (Fig. 12) peaking in 1997. In all but two of the years, landings of rose prawn have been higher than those of the striped prawn. At present, there are 44 vessels in the fishery, of which 22 are fishing under an agreement with the European Union, and the remainder are national.

Fig. 12 Annual landings of deep-water rose prawn and striped prawn in Angola since 1988 (from Sardinha) 1998)

The total annual catch of tuna in Angolan waters between 1993 and 1995 reported to ICCAT varied between 291 and 910 tonnes, with yellowfin tuna (caught by local baitboats and longliners) and big-eye tuna (caught by foreign longliners) making up most of the catch .

Foreign longliners caught tuna in Namibian waters under South African licence prior to Independence in 1990. The Namibian-controlled tuna fishery started in 1991, and since then an average of 2 330 tonnes of tuna (predominantly southern albacore) have been taken per year by a fleet of about 30 local and foreign-owned pole or line vessels. The foreign longline tuna fishery, which targets bigeye tuna for the high-value sashimi market, started in 1993. Catches have varied between 52 tonnes in 1996 and 1 005 tonnes in 1994. Experimental catches of swordfish Xiphias gladius taken by surface longlining have been low (50 tonnes in three years), but are regarded as encouraging given the low level of effort. An average of 770 tonnes of snoek has been taken per year by commercial vessels since Independence, mostly by handline (Boyer et al. 1998). Catches of silver kob from commercial linefish vessels have fluctuated over the past three decades around an average of about 500 tonnes per year. It is estimated that at present the recreational catch of silver kob is roughly equal to the commercial catch (Kirchner 1998).

The commercial linefishery in South Africa developed rapidly after the Second World War, reaching its peak in the late 1960s, when about 20 000 tonnes on average were caught per year. Thereafter, catches declined despite increasing effort. The average annual catch of all species excluding tuna between 1993 and 1997 was 11 700 tonnes, of which just over half was caught west of Cape Point. Snoek have consistently contributed over half the catch since the start of the fishery. The average catch over the past five years was 6 650 tonnes, of which more than half was landed at St Helena Bay and Yzerfontein on the West Coast. The second-most important species in terms of landings was hake, with average landings of just under 1 000 tonnes per annum, caught mostly on the South Coast. Other important species include yellowtail Seriola lalandi, kob spp. (particularly Argyrosomus inordus) and geelbek Atractoscion aequidens, which together contribute almost as much as snoek to the annual landings.

There is a large and varied recreational fishery for linefish species in South African waters, particularly on the South and East Coasts. Off the West Coast, the species most commonly taken from boats are tuna, snoek and yellowtail, while galjoen, hottentot, silver kob and white stumpnose Rhabdosargus globiceps are the bony fish most commonly caught by shore-anglers.

The baitboat (pole) fishery for tuna in South African waters started in 1980, and by 1990 about 10 000 tonnes of tuna (predominantly southern albacore) were being caught per year. Catches subsequently declined to about half this (6 571 tonnes on average between 1993 and 1997), but reached about 8 000 tonnes again in 1998. For the past two decades there has also been a longline fishery for large bigeye and yellowfin tuna by Japanese and Taiwanese vessels fishing under license, catches of the latter species being confined to the warmer waters east of Cape Point. In 1997, experimental pelagic longline permits were issued to South African fishermen, primarily to target high-grade tuna for fresh export. The vessels have made good tuna catches along the edge of the continental shelf, and have also caught substantial quantities of swordfish (probably nearly 1000 tonnes in 1998), which has a high market value and is likely to be exploited more heavily in the next few years.

By far the most important artisanal fisheries in the region are in Angola, where they contribute some 30% of the total landings of fish and shellfish. Declared landings from this sector in 1997 amounted to nearly 30 000 tonnes (Anon. 1998a). In 1995 there were more than 23 000 registered artisanal fishermen, landing their catches at some 105 controlled landing places, predominantly in the central provinces of Luanda, Benguela and Kwanza Sul (Delgado and Kingombo 1998). The total number of boats in 1995 was estimated at 4 677, ranging in size and sophistication from unmotorised canoes (about 25% of the total) to small wooden boats 5 - 6.5 m long with or without motors ("chatas"), which make up about 70% of the total, to 8 – 12 m vessels ("catrongas") with inboard motors and some preservation facilities (Delgado and Kingombo 1998). Larger "traineiras" (8 – 25m semi-industrial deckboats with inboard engines) are also involved in small-scale fisheries along the coast. The most common fishing gears are gillnets, longlines, and beach- and boat-operated seine nets and traps. The main species caught are pelagic fish such as sardine and sardinellas, horse mackerel, mullet, bigeye grunter and small tuna, and demersal fish such as bigeye dentex, croakers and groupers. Fresh fish is sold in urban areas close to the landing sites, or because of a lack of freezing facilities, is dried or smoked for sale in the rural areas.

Artisanal/subsistence fisheries off the West Coast of South Africa (particularly for mullet Liza richardsoni, which is now caught mainly by beach-seine and driftnets), have been in existence for centuries, but are of negligible value compared to the nation’s major industrial fisheries. In Namibia there is virtually no artisanal/subsistence fishing, because of the uninhabited desert coastline.

Marine recreational fishing from skiboats and larger gamefishing boats, and by rock-and-surf anglers and spearfishermen is an important leisure activity in South Africa, and directly or indirectly generates a significant amount of employment and revenue (see Section 7). The sector has grown rapidly since the Second World War, to the extent that surveys indicate that approximately 15% of coastal residents now fish in the sea on a regular basis. As the fishery has grown, there has been a tendency along the entire coast for catch rates to decline, and for endemic inshore reef species (such as galjoen and white stumpnose in the Western Cape) to be replaced by cartilaginous fish in the rock-and-surf catches.

Recreational fishing in Namibia is pursued by rock-and-surf anglers, who have access to about 20% of the coastline, and skiboat fishermen operating from Walvis Bay, Swakopmund and Lüderitz. The sector has grown rapidly in recent years, and now attracts many visitors to the coast, particularly from Namibia and South Africa, generating considerable employment and revenue in the coastal towns (see Section 7). With the recent increase in fishing pressure, the number of fish caught per angler has declined significantly.

5. CHANGES IN ABUNDANCE AND DISTRIBUTION OF MAJOR RESOURCES

Virtual Population Analysis (VPA) estimates of sardine biomass in Namibia between 1952 and 1988 (Fig. 13) confirm the marked decline of the stock in the second half of the 1960s, and the collapse to well below 1 million tonnes in the mid 1970s. Catch patterns have clearly indicated a northward shift in the core distribution since the collapse of the fishery in the 1970s, possibly as a result of the depletion of the southern spawning population and the cessation of associated migrations. Acoustic survey estimates of sardine on the Namibian/southern Angolan shelf since 1990 (Fig. 14) indicate that the adult stock is still well below pre-collapse levels, with an all-time lowest estimate of only a few thousand tonnes in the summer of 1995/1996, co-incident with the Benguela Niño at that time (Section 6.1). The surveys also indicated a further northward shift in distribution in 1994, when the majority of the biomass was found off southern Angola. Survey estimates of anchovy and juvenile round herring combined (they are not distinguished in the surveys) have dropped from levels of around 200 000 tonnes between 1990 and 1993 to below 100 000 tonnes since then. The decline is reflected in the anchovy landings, which have been negligible since 1996.

Fig. 13 VPA estimates of sardine biomass in Namibian waters from 1952 to 1988 (from Boyer 1996)

Fig. 14 Estimates of sardine biomass in Namibia/southern Angola since 1990, obtained from acoustic surveys (updated from Boyer 1996). * = Angola not surveyed.

In South Africa, acoustic surveys of sardine over the past 14 years have revealed a gradual increase in spawner biomass from below 50 000 tonnes in 1984 to around 600 000 tonnes at present (Fig. 15). The same surveys (supported by concomitant egg-production surveys between 1983 and 1990) have revealed two major peaks in anchovy spawner biomass in South African waters during this period, with pronounced troughs in 1989/1990 and in 1996 (Fig. 15). Acoustic surveys of sardine and anchovy recruitment off South Africa show the same general pattern (Fig. 16), except in 1995, when strong anchovy recruitment was not matched by a high spawning biomass in that or the previous year. Of particular note is the high degree of correlation between the sardine and anchovy recruitment estimates throughout the time-series, which suggests that, in South Africa at least, recruitment of these two species is controlled largely by common environmental influences on the early life history stages, and not primarily by spawning biomass.

Fig. 15 Acoustic estimates of anchovy and sardine spawner biomass off South Africa since 1984 (from Barange et al. in press)

The South African surveys have also revealed significant shifts in sardine and anchovy distribution in the Southern Benguela since the early 1980s, the most prominent of which was a sudden large increase in the number of anchovy spawners on the West Coast in 1986, followed by a progressive reduction to the more normal low levels over the next three years (Hampton 1992). A significant eastward shift in both the anchovy and the sardine spawning populations has also been recorded in the past decade, from 1990 in the case of anchovy, and from 1988 onwards for sardine (Barange et al. in press).

Fig. 16 Acoustic estimates of the number and biomass of anchovy and sardine recruits off South Africa since 1995 (from Barange et al. in press)

Dr Fridtjof Nansen acoustic survey estimates of the biomass of sardinella (S. aurita and S. maderensis combined) in Angolan waters between 1984 and 1997 are shown in Fig. 17, which suggests that sardinella biomass has increased from levels of around 200 000 tonnes in the 1980s to more than double this in the 1990s. (This increase was probably at least partly due to the withdrawal of a major part of the distant-water foreign fleet in the late 1980s). It should be noted that at least some of the survey estimates are possibly negatively biased because of the difficulty in assessing sardinella acoustically when the fish are on or close to the surface, which frequently occurs.

Fig. 17 Estimates of sardinella (S. maderensis and S. aurita) biomass in Angola since 1985, from Dr Fridtjof Nansen acoustic surveys (Data from IIP)

According to acoustic survey data, round herring spawner biomass in South Africa appears to have remained relatively constant at around 1 million tonnes since the first estimate was made in 1986. No pronounced shifts in the distribution of adults have been noted from the survey data, although since the surveys have not always encompassed the distributional range of the population, the possibility of such shifts cannot be excluded.

Survey estimates of M. capensis and M. paradoxus abundance off South Africa and Namibia over the past decade, derived from swept-area demersal surveys (supplemented by acoustic estimates of fish above the trawl in the Namibian surveys), are shown in Fig. 18. It is notable that there has been a marked increase in the abundance of M. paradoxus off Namibia since 1992, which is confirmed by a marked increase in the proportion of M. paradoxus in the Namibian hake catches in recent years (E. Voges, NatMIRC, pers. comm.). This increase may indicate northward displacement or expansion of the stock from South Africa, or alternatively, a shoreward displacement in response to changes in the oxygen content of bottom waters. Fig. 16 indicates that in Namibia, the component of the population targeted by the fishery (i.e. M. capensis >35 cm long) grew between Independence in 1990 and 1992, but thereafter declined for the next four years. The most recent survey results indicate exceptionally good recruitment in 1997 (Boyer et al. 1998), with a subsequent increase in fishable biomass to the highest level since Independence. Off South Africa, the survey results suggest an increase in M. paradoxus biomass on the West Coast over the past decade, but no clear trend in the biomass of M. capensis. The low estimate for M. paradoxus in 1992 supports the hypothesis of a longshore displacement of the Cape stock into Namibian waters in that year, but given the uncertainty in the estimates, this apparent connection should be viewed with caution.

Fig. 18 Research survey estimates of the biomass of Cape hakes (Merluccius capensis and M. paradoxus) off the West Coast of South Africa and off Namibia (data from M&CM and NatMIRC respectively)

Fig. 19 (from Anon. 1998b) shows trends in exploitable and total biomass of hake on the West Coast of South Africa from 1917 to the present, estimated from an age-structured production model. All indications are that the stock has been relatively stable over the past two decades, with signs of a gradual increase in recent years.

Fig. 19 Estimates of the exploitable and total biomass of Cape hakes in South African waters since 1917 from an age-structured production model

Acoustic estimates of horse mackerel (mainly T. trecae) biomass in Angolan waters between 1985 and 1997, obtained from Dr Fridtjof Nansen surveys, are shown in Fig. 20. There appear to be two distinct periods: from 1985 to about 1995, when T. trecae biomass fluctuated around a level of about 200 000 tonnes and capensis biomass was occasionally of the same order, and from 1995 to the present, when T. trecae levels have been about twice as high, and T. t. capensis biomass very low. In all surveys the abundance was highest in central and southern Angola, except in 1996 (a year of anomalously warm conditions along the whole Angolan coast) when a third of the biomass was found in the north, between Luanda and Cabinda. The level of these estimates compared to the catches made at the height of the Angolan horse mackerel fishery (over 380 000 tonnes in 1978) is clear indication of the decline of the resource, which is supported by the more than 10-fold reduction in annual catches which has occurred since then (Section 4.1).

Fig. 20 Acoustic survey estimates of Cunene and Cape horse mackerel biomass in Angolan waters since 1985. Data from IIP (NB: check this Fig!)

Using a VPA of catch data from the former Soviet bloc fleet, Vaske et al. (1989) estimated that the biomass of Trachurus spp. in ICSEAF Divisions 1.3, 1.4 and 1.5 (i.e. between southern Angola and Cape Agulhas) between 1973 and 1987 fluctuated between about 1.5 and 2.3 million tonnes. This estimate must however be treated with caution because of, inter alia, the unreliability of the age-length keys used in ageing the fish. Since 1990, the biomass of adults and juveniles combined off Namibia and southern Angola has been estimated by annual acoustic surveys, predominantly from Dr Fridtjof Nansen (Fig. 21). The estimates generally fall between 1 and 2 million tonnes, with a maximum of 2.1 million tonnes in 1992. In an absolute sense, these estimates are questionable because of the lack of verification of the target strength expression used. For example, use of a target strength expression for T. trachurus capensis developed in South Africa (Barange and Hampton 1994) would lower the estimates by a factor of approximately 3. Furthermore, some of the differences between the surveys have been attributed to variations in the proportion of the population which is acoustically detectable rather than to fluctuations in the biomass (Anon. 1998c) – a problem which has also been experienced in acoustic surveys of horse mackerel in South Africa (see below).

Fig. 21 Acoustic survey estimates of (predominantly) Cape horse mackerel biomass in Namibian waters since 1989. Data from NatMIRC

Bottom trawl survey estimates of T. trachurus capensis biomass on the South African West Coast between 1985 and the present have increased from less than 10 000 tonnes in 1986 to between 50 000 and 300 000 tonnes since 1991. Equivalent estimates on the South Coast over this period for the 0-200m depth stratum have fluctuated between 100 000 and 580 000 tonnes, with an average of about 300 000 tonnes. There is however evidence from acoustic surveys and other acoustic observations that the South Coast trawl estimates are negatively biased because of the unavailability of part of the population to the bottom trawl at certain times and in certain areas (particularly on the outer shelf and over the shelf-break).

Deep-water fish

From a swept area analysis of commercial catches in 1996 and 1997, Branch (1996) and Branch and Roberts (1998) estimated the total abundance of orange roughy on the Namibian shelf at that time at over 200 000 tonnes. No biomass estimates considered reliable have yet been obtained for alfonsino. An acoustic survey of the three most important roughy grounds from Dr Fridtjof Nansen in 1997 gave an estimate of 147 000 tonnes after correction for the major sources of bias. A major difficulty in assessing the total roughy population by either trawl-based or acoustic methods is the lack of knowledge regarding the proportion of the population outside the known aggregations.

West coast rock lobster

The Schaefer and Fox production models used to estimate the MSY for the Namibian rock lobster resource (Section 8.4) have given values of 4 300 and 3 100 tonnes respectively for the period 1958 - 1991. If the very variable data from the 1960s and the TAC-restricted data from the early 1990s are ignored, these estimates drop to 1 900 and 1 800 tonnes respectively. Recent modeling estimates put the stock at around