3.  NON-REGULATORY MANAGEMENT OF LAGOON AND ESTUARINE FISHERIES

In this section the focus is on some management techniques which can supplement or complement the more classical regulatory management strategies which have been illustrated in the preceding section. Non-regulatory management in the sense used here is the application of various techniques to increase aquatic productivity and hence fishery potential. Firstly, some different applications of hydraulic management of lagoons for fisheries are discussed as a means for significantly improving capture fishery and aquaculture yields. Secondly, some additional non-regulatory management tools such as predator control, stocking , creation of artificial nursery areas, and brush-pile fisheries are examined.

3.1  Hydraulic Management of Lagoon Fisheries

Hydraulic engineering for fisheries is viewed here in as one of the most promising means available for management and development of capture fisheries and aquaculture in lagoons and estuaries. Basically, hydraulic engineering is the means whereby environmental conditions are manipulated through varying freshwater and seawater inputs so as to increase aquatic productivity and hence fishery yield.

The need for hydraulic engineering for fisheries is often brought about by the natural dynamic evolution of coastal lagoon systems -- internal sedimentation, and silting up of connections with the sea by littoral transport -- but increasingly by man-made alterations of the environment which accelerate natural ageing or decrease aquatic productivity in other ways.

Some examples of the kinds of engineering works used in coastal lagoons worldwide to benefit fisheries are portrayed in Table 1, along with some examples of specific lagoons in which hydraulic engineering has been applied, or the need for it identified.

Channelization from the sea to a lagoon or lagoon system is an obvious solution for increasing fishery production in lagoons which would otherwise go completely dry, such as Bardawil in Egypt (Ben-Tuvia, 1979) or periodically go so extremely hypersaline that almost no fish life can be supported such as Laguna Madre de Tamaulipas in Mexico (Hildebrand, 1969). If lagoon-sea connections are not maintained, temperature and hypersalinity can also cause massive fish mortalities such as in Laguna Unare in Venezuela (Mago Leccia, 1965).

In less extreme cases the maintenance of the lagoon connection with the sea can be used to stabilize environmental conditions so that what beforehand was a seasonal fishery can be extended throughout the entire year.

Edwards (1978a) indicates that in Mexico it is generally believed that lagoons of the semi-isolated kind should remain largely so because the fauna is adapted to the brackishwater environment, and therefore artificial connections made with the sea should be furnished with control gates for salinity regulation. Along these same lines in the Laguna Since of 17 000 hm²) in Romania, maintenance of a relatively low salinity within rather narrow limits is advantageous for the fishery to ensure that conditions favouring the economically most important species, mostly freshwater forms, are maintained (Valerian, 1977).

In addition to salinity control, maintenance of the lagoon-sea connection permits the ingress and egress of estuarine dependent, marine transient, and anadromous and catadromous finfishes and crustaceans on which many lagoon and estuarine fisheries are largely dependent, such as those for shrimps, prawns, and mullets worldwide, for Hilsa(shad) in India and for Ethmalosa in West Africa.

Site selection for the channels may be made in such a way that the time and distance for inward migration is considerably reduced as in the Mexican La Joya-Buenavista system from 30 km to 5 km (Huerta Maldonado, 1980) and in the Huizache-Caimanero system (Edwards, 1978a) thereby allowing for an earlier arrival, increased survival, and longer growing period for shrimp. The channel may be designed in such a way as to concentrate fishes or crustaceans for capture during outward migrations (Edwards, 1978a).

Other specific and inter-related purposes of hydraulic engineering in maintaining the lagoon-sea connection for fisheries are for promotion of circulation to carry away domestic, industrial, or agricultural pollutants as is the case with the former in the Lac Tunis (Stirn, 1966) and to increase aquatic productivity as in the nutrient depauperate Panganales Est lagoon system in Madagascar (Collart and Randriamanalina, 1978). Further, such lagoon-sea connections may be also used for improved navigation so that fishery products can be transported from the lagoon by sea to markets or processing plants, and so that supplies can reach isolated fishing communities.

Engineering works within the lagoon also serve a variety of purposes such as providing increased surface area for fishing, and passage of fishes and crustaceans from one water body to another by connecting adjacent lagoons. Interconnecting channels also allow larval fishes and shrimps improved access to nursery areas (see Fig. 4.). Reduction of sedimentation and promotion of circulation are yet other purposes of internal canals (Cervantes Castro, 1980).

Fig. 4. Hydraulic engineering within lagoon systems as exemplified by the Huizache-Caimanero lagoon complex in Mexico (from Edwards, 1978a)

Diversion of fresh waters to lagoons from nearby rivers or streams is usually accomplished to reduce salinity. For example, Okuda (1965) has proposed an ambitious scheme for diverting Rio Unare waters (Venezuela) to the Laguna Unare to reduce hypersaline conditions during the dry season, and Posewitz (1968) has advocated the use of an upstream reservoir and pumping or gravity feed of fresh water to supply the Unare and adjacent Piritu lagoon with fresh water while eliminating sedimentation from the Rio Unare. Diversion of river waters through seven channels to the Laguna Madre de Tamaulipas (Mexico) has produced favourable conditions for oyster culture there (Sanchez, 1980). In the Huizache-Caimanero system, diversion of river waters has been used to maintain a larger lagoon surface area for a longer duration to benefit survival and growth of shrimp. Circumstantial evidence points to a significant increase in shrimp yield from the system as a result (Edwards, 1978a). Juarez Reyes (1980) mentions the construction of crude temporary dams in three lagoons, presumably across lagoon outlets, to maintain higher water levels for longer periods for the same purpose. As one means to produce better conditions for the reproduction and survival of the eggs and larvae of the anadromous Hilsa in the Hooghly-Matlah estuarine system in India, Gopalakrishnan (1977) suggested the timed release of water from upstream reservoirs.

Finally, the logical extension of hydraulic engineering for fisheries is the creation of new lagoons For example, hydraulic engineering has been used to create the artificial Lagune de Khenis of 350 hm² in a part of the Baie de Monastir in Tunisia. This lagoon supports a capture fishery and is being increasingly developed for aquaculture.

Although the benefits of hydraulic engineering in lagoons for fisheries are obvious, economic data of a cost-benefit nature are difficult to locate. Huerta Maldonado (1980) provides finfish and shrimp yield data which show a notable increase after the construction of a canal from the sea to the La Joya-Buenavista lagoon system and remarks on the comparatively rapid growth and large sizes of the shrimp harvested from the system. After extensive hydraulic management the Huizache-Caimanero lagoon system has a very high average shrimp yield, 126 kg hm^-2 yr^-1 (from Edwards, 1978a). Laguna Madre de Tamaulipas produced about 42 kg hm^-2 yr^-1 of finfishes, crustaceans, and molluscs in 1977 (Martinez Mata, 1980) with only limited hydraulic management whereas historically (Hildebrand, 1969) there were periods when the fishery was severely reduced or non-existent without hydraulic management.

Posewitz (1968) estimated that the engineering works necessary to provide the Piritu and Unare lagoons with fresh water during the dry season would require U.S.$200 000 with $10 000 yr^-1 for maintenance but that the value of the additional fishery yield made possible by the engineering works would be about U.S.$90 000 yr^-1.

Of the information available, Pisanty (1980) provides the most complete picture of the benefits of hydraulic management for the Bardawil lagoon in Egypt. Bardawil, like the Laguna Madre de Tomaulipas, supported a boom or bust fisher from early in this century because of wide variations in salinity. There, without an opening to the sea, the lagoon of 65 000 hm², can dry-up completely (Ben-Tuvia, 1979).

Early in this century with sporadic dredging of channels, or with natural but temporary openings caused by storms, the lagoon produced about 5 kg hm^-2. In the late 1960's and thruogh most of the 1970's the lagoon was brought under well-organized management which included opening two permanent channels to the sea which were dredged every other year, applied research on fishery resources, regulation of the fishery, and an agressive programme of export marketing. A 15 percent tax levied on the gross value of the catch (3.8 to 5.5 million dollars from 1975 to 1978) supported the bi-annual maintenance of the channels (12 percent) and the remainder supported research and administration of the fishery (3 percent). Yield from the lagoon with hydraulic and other management activities was increased to an average of 31 kg hm^-2 yr^-1 and the value of the fishery was increased ten-fold for the latest years of data available, whereas in the past without hydraulic engineering no sustainable fishery existed.

In order to be successful, hydraulic management of lagoons should be based on detailed engineering studies as well as on thorough chemical, physical, and biological surveys. The basic kinds of hydraulic studies which are required are: longshore drift, tides, sedimentation, freshwater input, rainfall, evaporation, occurrence of storms and hurricanes, as well as other physical and chemical investigations including those of surface area, bathymetry, and temporal and spatial distribution of salinity (Cervantes Castro, 1980).

Biological and limnological studies in connection with hydraulic engineering in Mexico are usually aimed at providing information for optimum operation of hydraulic structures. These studies relate information on shrimp and fish distribution, movements, and growth to physical and chemical measurements of the lagoon environment, examples of which can be found for various lagoon systems in Gezan Soto (1976), Barrera Huerta (1976), Huerta Maldonado (1980), Martinez Mata (1980), and Sanchez (1980).

Elsewhere, studies such as those by Tesson (1977) and Brethes and Tesson (1978) on the Sebkha Bou Areg in Morocco and by Okuda et al. (1965), Benitez Alvarez, and Gomez (1963), Okuda, Gracai, and Benitez Alvarez (1965), Okuda (1965) and Fukuoka and Gamboa (1973) on the Laguna de Unare in Venezuela provide examples of how short-intensive investigations can point-up hydraulic and other problems in laoogns which suggest the appropriate hydraulic engineering solutions. The investigation reported by Amanieu el at. (1980) on a small semi-artificial lagoon on the Mediterranean coast of France provides an excellent example of how, by detailed biological monitoring of the effects of experimentally controlled seawater input, indications for the optimum strategies for future hydraulic management can be derived.

It seems that much useful information from aquacultural engineering and from the results of coastal aquaculture in small systems could be brought to bear on problems of hydraulic engineering in lagoons, but no study which has attempted to accomplish this on a systematic basis has been found in connection with this review.

3.2  Predator Control

Edwards (1977) based on his field investigations on the Huizache-Caimanero lagoon complex, noted that total mortality of shrimp between post-larval and late juvenile stages is relatively high. Recruitment provides up to about 10 post-larvae m^-2 of lagoon bottom, but subsequent mortality decreases the number of juveniles to about 0.3 m^-2 for most of the lagoon area. In comparison, results of cage experiments indicated that shrimp densities of up to 2.5 m^-2 could be supported without loss of growth. Based on these results and the results of enclosure experiments elsewhere which showed that shrimp harvest rates could be increased by a factor of about 3 in the absence of predators, Edwards (1977) concluded that the establishment of a fishery for the main shrimp predators in lagoons, Galeichthys, Cynoscion and to some extent Callinectes, could reduce shrimp mortality and increase yield. Wing nets set in channels would be appropriate for Galeichthys, and Callinectes could be fished by pot traps. Thus, not only would shrimp production be increased, but a more rational use of the lagoon fishery resources would result by making greater use of finfish and crabs.

3.3  Stocking

Garduño Argueta (1976), in remarking that offshore and lagoon shrimp resources of the Pacific coast of Mexico had surpassed their maximum level of exploitation, points out that one way to realize a greater shrimp yield is through the artificial cultivation of shrimp in hatcheries based on collection of gravid females and culture of larvae for release at post-larval and juvenile stages for stocking in lagoon waters. The same author reports on some pilot work in shrimp rearing but the investigation did not proceed as far as evaluating the contribution of stocked shrimp to fisheries. Edwards (1977) believes that such stocking can be successful only if the hatchery-reared shrimp are first released into enclosures from which predators have been excluded.

The Japanese have engaged in stocking of open but protected inshore waters with Penaeus japonicus since 1968 (Shigueno, 1975). Although the results of these stockings have been difficult to assess, it appears that best results are produced when stocking is accomplished in the inner parts of bays. “Artificial tidelands” have been created as a means to enhance the success of stocking through a better acclimatization to natural waters (Honma, 1980) (Fig.5).

Fig.5. A schematic diagram of an “artificial tide land” for shrimp stocking in Japan (from Honma, 1980).

Intensive stocking of Cyprinus carpio, Sarotherodon niloticus and mullets in the Pangalanes Est lagoon system of Madagascar has been advocated by Collart and Randriamanalina (1978). Some exotic species had been introduced previously, but without notable success except for other Tilapia/Sarotherodon introductions. Mullets occur naturally. These authors note that to increase chances of success, stocking would have to be accomplished repeatedly and in large numbers, and that juveniles would have to be kept in predator-free enclosures until they reached a stockable size.

Texas (U.S.A.) is making an effort towards the stocking of bays (coastal lagoons) with red drum, Sciaenops occellata. This fish is under heavy fishing pressure from both recreational and commercial fishermen and is important to the economy which justifies the stocking effort as an extension of the overall bay fishery management programme. This species has been spawned artificially, and the effect of experimental stocking was to be evaluated. It is noted, however, that there is no advantage to be realized if the stocking and natural production exceed carrying capacity, or if the resource continues to be over-harvested (Hefferson and Kemp, 1980).

Victoria (Australia), too, is contemplating artificial rearing and stocking of finfishes in inshore waters if practical benefits can be thus realized for the enhancement of fisheries (Winstanley, 1981).

3.4  Artificial Nursery Areas

As a management measure for the Pangalanes Est, Collart and Randriamanalina (1978) recommended the creation of artificial spawning areas using various types of local plant materials. These would be located in protected areas of the lagoon, and predators would be removed on a periodic basis using electric fishing.

On a much more sophisticated scale, the Japanese (Honma, 1980) have actively developed a variety of artificial structures to serve as nursery areas for prawns and fishes including breakwaters and man-made weed beds (Fig. 6). Although such structures are economic in Japan, and although their use could contribute to the enhancement of lagoon and estuarine fisheries elsewhere in the world, it will probably be some time before such elaborate engineering for fisheries becomes cost effective in developing countries.

Other non-regulatory management methods

In a review article on fish cropping and production in estuarine waters, Saila (1975) observes that the potential for enhancing fish production in estuarine waters has yet to be fully realized. Among the management measures that he recommends for consideration and additional research and experimentation are controlled artificial enrichment, transplantation, predator control, and controlled breeding.

Fig. 6. An artificial shrimp nursery complex in Japan (from Honma, 1980).

3.5  Brush-park Fisheries

Brush-park fisheries are a traditional form of low-technology aquaculture which is practised in inland and brackish waters in many areas of the world. The brush parks are constructed in a variety of forms and sizes, but basically a brush park consists of an inner core, or concentric circles, of densely packed tree branches or other material surrounded by an outer, more substantial wooden framework which is fished periodically, usually by encirclement (Figs. 7 and 8).

In coastal lagoons the most sophisticated forms of brush-park fisheries occur in Benin. Elsewhere in Africa, brush-park fisheries are found in Nigerian coastal lagoons (FAO, 1969) and in coastal lagoons in Ivory Coast (Kapetsky, pers.obs.), Ghana (Mensah, 1979), Togo (Welcomme, 1971, 1972; Everett, 1976) and Madagascar (Kiener, 1960, 1963; Moulherat and Vincke, 1968). Elsewhere in the world, brush-park fisheries are found in Negumbo lagoon in Sri Lanka (FAO/UN, 1962) and have been recently introduced in some Mexican lagoons (Lizarraga, pers.comm.). In rivers and lakes traditional brush-park fisheries are found in Cameroon (Stauch, 1966), in Sierra Leone (Chaytor, pers.comm.), in Nigeria (Awachie and Ezenwaji, in press; Reed, 1967), Bangladesh (Kapetsky, pers.obs.), Kampuchea (Fily, 1966), China (C.S.F.F.C.E.C., 1972), and Ecuador (Meschkat, 1972).

Brush-park fisheries appear to offer a number of advantages for the management of coastal lagoon fisheries, and therefore their characteristics are examined in detail in the following sections.

3.5.1  Brush-park fisheries in Benin

Buffe (1958), FAO/UNDP (1971), and Welcomme (1971, 1972) have provided detailed information on brush-park fisheries in Benin where they are known collectively as “acadjas”, and Bourgoignie (1972) has studied their historical development over the last two centuries.

There are several basic types of brush-park fisheries in Benin which vary from each other in configuration, construction and fishing characteristics (Figs. 7 and 9). The acadjas are constructed of tree branches with harder woods forming the peripheral structure and with soft wood branches with many ramifications forming the interior of the acadja.

Acadjas are preferentially placed in shallow, quiet waters of no more than 1.5 m depth.

Fishing is accomplished by surrounding the smaller types of acadjas with a net after which all of the branches are removed and the net is then pursed. For the larger brush parks the surrounding net is fished in a step-wise process by gradually moving the net inwards as branches are removed until the fish are concentrated in small areas and can thus be removed with traps, baskets, and hand-nets. After fishing, the used branches are replaced and new ones added, as necessary (Welcomme, 1972).

Yields from the various kinds of acadjas are variable, but high. Buffe (1958) provides data showing yields of 13.6 and 19.2 t hm^-2 for ava-type acadjas for periods varying by only a few days from a full year, and mentions annual yields from other avas at about 10 t hm^-2. Welcomme (1972) provides yield data by type of acadja.

Tilapia melanotheron and Chrysichthys nigrodigitatus dominate the acadja fisheries (67 to 95 percent by weight), but open water fisheries are based mainly on other species such as Ethmalosa fimbriata.

Welcomme (1972) has analysed the relationship between yield and the length of time of implantation, the effect on yield of frequency of fishing, and also the relationship between density of branches and acadja yield. The former was analysed for two periods, 1957–59, and 1969–70. The relationships thus derived (Fig. 10) have been interpreted to show that three dynamic factors successively influence yield: immigration of fish into the acadja (colonization), growth and reproduction within the acadja, and possibly emigration from the acadja, while basic yield potential is significantly influenced by density of branches used.

Fig. 7. Diagramatic representation of a godokpono-type brush-park fishery in Benin. (a) Plan; (b) Section. x = hardwood branches; softwood (from Welcomme, 1972)

Fig. 8. A brush, reed, and rice grass-park fishery in China (from C.S.F.F.C.E.C., 1972)

Fig. 9. Various kind and configurations of bursh-park fisheries in Benin (from Welcomme, 1972)

Fig. 10. Yields from brush-park fisheries in Benin as function of time since constuction, (O, 1957-59; x, 1969-70), density of branches, and time between fishings (from Welcomme, 1972)

Composition of the fish population of acadja avas as compared
with that of the open waters of Lake Nokoue (from Welcomme, 1972)

Before physical and economic disruption of acadja fisheries by teredo attack on their branches and by deposit of calcareous deposits limiting epiphytic production caused by the permanent opening of a lagoon/sea channel, the Lake Nokoué-Porto Novo lagoon capture fisheries were perhaps the world's most productive with yield estimates (acadja plus capture fisheries) of more than 1 t hm^-2 prior to 1958 (Buffe, 1958) and about 1.8 t hm^-2 from 1957 to 1958 (Lemasson, 1961)¹ and at least the most productive that have been found in connection with this review. Even with the subsequent decrease of acadjas, lagoon fishery yields there had decreased to only about 357 kg hm^-2 in 1969 (76 kg hm^-2 of this total contributed by acadjas) which is still at a level much higher than the yields obtained in other West African lagoons where acadja fisheries do not exist or are not prevalent.

Since the construction of a cross-channel barrage which reduced lagoon salinities somewhat and hence teredo attack on acadja branches, and because of other favourable economic conditions, acadja fisheries are once again proliferating in the Lac Nokoué/Porto Novo lagoon system (Fig.11). Total acadja yield in 1980–81 in the 15 700 hm² system is estimated at about 2 000 t from about 390 hm² of acadja installations (Welcomme, unpublished), equivalent to a yield of about 127 kg hm^-2 if taken over total lagoon surface area.

3.5.2   Brush-park fisheries in coastal lagoons elsewhere in the world

Elsewhere in West Africa, lagoon brush-park fisheries are known from the Keta lagoon system in Ghana, from the Lagos-Lekki lagoon system in Nigeria, from Lac Togo lagoon in Togo, and from Ebrié lagoon in Ivory Coast; in East Africa, they are used in Madagascar. However, published material providing details on these fisheries is scarce.

In Lac Togo it appears that in the early 1970's there were about 35 hm² of acadjas which yielded about 14 t hm^-2yr^-1 (Togo, 1971), equivalent to about 82 kg hm^-2yr^-1 taken over the total lake surface area of about 6 000 hm².

Information of brush-park fisheries in Nigeria is dated (FAO,1969) but the brush parks were said to be common in Lagos and Lekki lagoons, a lagoon system contiguous with the Benin lagoons. These were constructed of tree branches and palm fronds. Frequency of fishing was evidently at from 2 to 3 month intervals, and the most common fishes in the catch were tilapia and mullet.

¹   In 1959 total yield was about 980 kg hm^-2 of which open water capture fisheries accounted for 715 kg hm^-2 and acadja fisheries 237 kg hm^-2 (FAO/UNDP, 1971)

Fig. 11. Total surface area and distribution of brush-park fisheries in Lac Nokoué, Benin, in 1959, 1970, and 1981 (from FAO, 1971, and Welcomme, unpublished)

In Eastern Ghana in the Keta lagoon area, Mensah (1979) mentions that brush-park fisheries of the circular “godokpono” type (Fig. 7), there called “achidjas” vary in size from 150 to 300 m² and are fished every three to four months. On a visit to the lagoon area in February 1981 (Kapetsky, pers.obs.) brush-park fisheries were observed at several places around the lagoon periphery. These were of two types, the first as described by Mensah (1979) (above), but much smaller in area, from about 7 to 50 m², and common in occurrence. The other type was a smaller version of the “adokpo” (Fig. 9) with a number of small circular brush piles of 1 to 2 m diameter grouped to form a rectangular shape. Based on very limited information from an interview it was gathered that these fisheries have been in use at the Keta area for more than a lifetime, each fishery is individually or family owned, fishing intervals are from one to two months, and the principal species caught is tilapia. Yields from a 3-m diameter (7 m²) “godokpondo” averaged about 15 kg per fishing (about 20 t hm^-2, but varied from considerably less up to about 30 kg per fishing. Teredo attack on the wood used in the “achidja” fisheries had been a problem of annually varying proportions stemming from a time about 13 years previous when the lagoon was directly, but temporarily, connected with the sea.

In the Lagune Ebrié, Ivory Coast, brush-park fisheries also of the godokpondo type had been recently introduced in one area near Abidjan about four years ago by fishermen originally from Benin. In March 1981 (Kapetsky, pers.obs.), these numbered 25, of about 2.5 to 3.5 m in diameter, distributed along about 4 km of a narrow arm of the lagoon. From interviews with five owners, a general idea of the yield and economics of these fisheries was gathered. As in Keta lagoon in Ghana, yields averaged about 15 kg per fishing of a brush pile slightly larger than 3 m in diameter, with one fishing every two months. Yields varied from about 7 kg per fishing to about 40 kg (8 to 42 t hm^-2 per fishing). Tilapia and Lutvanus were said to be the most common fish in the catch. Because of the proximity of these brush-pile fisheries to the lagoon/sea connection, teredo attack on the branches is intense, necessitating complete replacement of branches every 4 to 6 months. The information volunteered by the owners on wood costs and wood replacement rates enabled the calculation of an approximate break-even cost for the average brush-park fishery. The average price of fish paid to fishermen in the area, obtained independently, and the information provided by the fishermen on yield was consistent with the break-even cost, suggesting that the information obtained was fairly reliable. That the brush-park fisheries are economic, even with high wood replacement rates of 200 to 300 percent yr^-1 (compared with a current 36 percent yr^-1 rate in Benin) is also evidenced by their gradual increase in numbers over the last four years.

About one week subsequent to the author's visit to the brush-park fisheries of the Lagune Ebrié, personnel of the Centre de Recherches Océanographiques, Abidjan, observed the fishing of one brush park of the same 25-unit cluster, on 9 March 1981.

The brush park was about 4 m in diameter, and had been last fished 70 days previous. The yield consisted of the following species and quantities of fishes:

The fish yield from this fishing was equivalent to about 17 t hm^-2.

In addition to the fishes tabulated above, Bert (unpublished) notes that a very abundant fauna associated with the branches was also observed including shrimps, crabs, small fishes, and molluscs.

Brush-park fisheries have been recently observed in two other areas of Lagune Ebrié (Durand, pers.comm.) but no details on them are yet available.

Other than from West Africa, the only quantitative information on coastal lagoon brush-park fisheries comes from Negumbo lagoon in Sri Lanka (FAO/UN, 1962), from a single experimental installation in Madagascar (Collart and Randriamanalina, 1978), and from traditional brush-park fisheries in the same country (Moulherat and Vincke, 1968).

In Negumbo lagoon there are two kinds of brush parks, “Mas Athu” and “Katta”. The “Mas Athu” are circular and vary in diameter from 3 to 4.6 m. Construction is of mangrove poles, and branches are placed vertically or slightly inclined in depths of less than 1.5 m depth. Fishing is at intervals of 10 to 15 days. Based on the rough figures given for “Mas Athu” dimensions and catches, yield is estimated herein to be about 7 t hm² per fishing. Species in the catch are mainly Mugil, Siganus, Ambassis, Etroplus, Glossogobius, Eleotris, Epinephelus, eels, prawns, and crabs. More than half of the catch was of undersized young specimens. In the Negumbo lagoon 200–300 of these fisheries were present in the early 1960's.

The “Katta” type of brush-park fishery is larger, of about 13 to 30 m² in area, and is placed in deeper waters than the Mas Athu. The main frame consists of palm trunks set vertically with managrove branches coiled around the lower portions of the trunks. Contrary to the usual fishing practice of surrounding the structures with nets, in the “Katta” brush-park, fishing is done by rod and line. Common species attracted are Lates, Gerres, Ambassis, Scatophagus, and Monodactylus (FAO/UN, 1962). No yields are available from this type of fishery.

In the Pangalanes Est lagoon system in Madagascar, a kind of “mini” brush-park fishery is employed called “Vovomora” consisting of a central core of loosely packed ferns surrounded by vertical stakes (Fig. 12). The small version of the vovomora is from 50 to 60 cm in diameter and placed near the shore in depths of less than 80 cm. In certain areas the vovomora reach 1.5 in diameter and are planted in correspondingly deeper water.

The vovomora are fished at 3 to 4-day intervals, and the catch consists of Macrobrachium, and young of Gobius giuris, other small gobids and atherinids. Yield ranges from a minimum of about 2 t hm^-2 per fishing to about 8 t hm^-2 with yields up to 20 t hm^-2 per fishing during the cold season as calculated herein based on approximate data provided by Moulherat and Vincke (1968). For the Pangalanes lagoon system the 2 217 vovomora provided an estimated 57.8 t of fish and shrimp in 1967 Taken over the total surface area of the lagoon system, the vovomora contributed about 5.2 kg hm^-2 equivalent to 15 percent of the total fish, shrimp and crab yield of the system in 1967.

Collart and Randriamanalina (1978) report on an experimental brush-park fishery of 250 m² using branches of local trees implanted in the Pangalanes Est lagoon system in Madagascar. After slightly more than a year, the brush park was fished. The yield was equivalent to 252 kg hm^-2. The species in the yield (percent by weight) were Tilapia rendalli (26.6 percent), Ptychchromis oligacanthus (21.9 percent), Cyprinus carpio (23.0 percent), Gobius giuris (11.3 percent), Paratilapia polleni (8.9 percent) and Sarotherodon mossambicus) (8.3 percent).

3.5.3   Some perspectives on the contribution of brush parks to the management of coastal lagoon fisheries

Brush-park fisheries appear to offer a number of biological and economic advantages for the management of coastal lagoon fisheries. Among these are relatively high yield per unit area, the low-level of technology required, the high intensity of labour employed, a potential for increase in the biological productivity of the lagoon system as a whole through nutrient input from wood, and a positive effect on adjacent capture fishery yields. At the same time the dissemination of brush-park fishery “technology” could also create a number of biological and socio-economic problems which might include actual or perceived competition with capture fisheries for space and resources, interference with navigation, eutrophication due to wood nutrient input, increased silting or sedimentation rates due to physical interference with currents and water flows, and local deforestation.

Fig. 12. Fishing of a “Vovomora” in Madagascar (adapted from Kiener, 1960)

Yield

The high yield per unit of area of brush-park fisheries in a wide variety of locations has been demonstrated in the previous sections. However, the frequency of fishing and quantity of wood or other material utilized may influence yield greatly. From the viewpoint of fishery management it is the frequency of fishing which determines whether the brush-park fishery contributes to, or at least does not compete with, adjacent capture fisheries (Welcomme, 1972). From the information available it seems that brush-park fisheries as used in Ghana, Ivory Coast, Nigeria, Sri Lanka and Madagascar, are operated as refuge traps rather than as structures which contribute significantly to aquatic production, and thus, these may compete somewhat with adjacent capture fisheries by attracting fishes away from open-water capture fisheries to the brush parks.

Intensity of Labour

The relatively high intensity of labour required to construct, fish and maintain brush-park fisheries may be an advantage where the economic situation dictates that the quantity of employment generated is a prime consideration in fishery management, especially when compared with the relatively small quantity of labour required for high-intensity mechanized aquaculture. The large quantities of branches required can also contribute to the local economy by fostering a wood growing, harvesting and transporting industry.

Level of Technology

The level of technology required for construction of brush-park fisheries is low relative to some types of “modern” intensive aquaculture, and the fishing skills required are no more complex than for many kinds of capture fishing. This translates to minimal personnel costs to governments for extension services to assist with advice on brush-pile fishery construction and operation when compared with the costs of maintaining such services for other forms of intensive aquaculture.

Contribution to Increased Aquatic Productivity in the Lagoon System as a Whole

There are no hard data to show that nutrient input to a lagoon system through the use of large quantities of wood in brush-park fisheries contributes to increased aquatic production and hence to increased adjacent capture fishery yield. However, the use in Lac Nokoué in Benin of about 30 t of wood hm^-2 of brush-park installation in 312 hm² of brush parks with a 36 percent yr^-1 annual wood replacement rate (Welcomme, unpubl.data from 1981) is equivalent to an annual wood input (wet weight) of about 3 370 t, or 225 kg of wood hm^-2 if taken over total lake surface area. When nutrient input from wood was compared (crudely) with nutrient input from the two principal freshwater rivers flowing into the system, the former appeared to be inconsequential. Thus, nutrient inputs from wood used for brush-park fisheries are insignificant in quantity in comparison with the introduction of these nutrients to the lagoon by river inflow for the Lac Nokoué situation. Thus, by the same reasoning, eutrophication caused by wood nutrient input does not appear to be an environmental danger.

Competition for Space and for Fishery Resources

Welcomme (1972) provides quantitative data which indicate that the species composition of brush-park fishery catches is different from that of adjacent capture fisheries. Therefore, at least in the Lac Nokoué instance, competition between capture and brush-park fisheries was minimal. Welcomme (1972) also hypothesizes that brush-park fisheries may benefit adjacent capture fisheries quantitatively to some extent if left unfished for long periods, as fish populations build up through reproduction and growth and as some of the individuals disperse to open waters or colonize newly built or recently fished adjacent brush parks.

How the sites and sizes of brush-park fisheries are allocated traditionally among the families and cooperatives of fishermen of open waters of lagoons in Benin and elsewhere is unknown, but an important consideration for management purposes in the dissemination of this technology is that the capture fishermen understand that, if operated properly, brush parks present no threat to their livelihood. Evidently, due to misunderstandings and other sociological or economic pressures, once successful brush-park fisheries in Lac Togo were completely destroyed when it was attempted to rapidly increase their expanse there from about 35 hm² to 250 hm² (Togo, 1971; Everett, 1976). Further, brush-park fisheries introduced to Lac Ahémé, Benin, in the late 1960's although initially successful, have now disappeared due to socio-economic and political problems (Welcomme, pers.comm.).

Disadvantages of Brush-park Fisheries

The greatest disadvantage of brush-park fisheries is the large quantity of branches required to establish and maintain the fisheries. This can be disadvantageous in a number of ways -- economically in the cost of purchasing, harvesting, transporting, and installing the wood for initial construction, or for periodic replacement, and with regard to the latter, especially in relation to the intensity of teredo attack and the incidence of other marine infestation.

Environmentally, the large quantities of branches required could contribute to local deforestation and through this, environmental degradation such as erosion, increased sediment loads and increased turbidity which could ultimately adversely affect natural fishery potential. One means to counter this potential difficulty is to establish tree plantations to supply brush-park fisheries, much the same way that tree plantations have been established adjacent to some African cities to supply firewood.

Another potential disadvantage of brush-park fisheries is that they may contribute to the rapid ageing of a lagoon through promoting increased sedimentation rates and hence shallowing of lagoon waters. Although Texier et al. (1980) mention that brush-park fisheries may be a factor in sedimentation in Lac Nokoué, Texier (pers.comm.) does not believe that their influence on this process is great.

Interference of brush parks with navigation in Lac Nokoué, where they are relatively dense, is much more akin to inconvenience than to obstruction at present (Kapetsky, pers.obs.). If necessary, channels could be set aside and bouyed to facilitate navigation through brush-park areas.

Brush-park Fisheries: Summary

Brush-park fisheries offer an attractive, non-regulatory means for management of lagoon fisheries. Among the advantages are high yield, simple technology, high intensity of labour, and low management expense on the part of government. Disadvantages are that brush parks might be perceived by fishermen as competing with already established capture fisheries for fishery resources or for space (or might be operated in such a way that they actually do compete), that large quantities of wood are required for initial construction and for periodic maintenance, and that brush parks might foster some environmental degradation.

The suitability of brush-park fisheries for dissemination as a management technique should be decided based on intensive studies of traditional allocation of fishing rights, fishery biology, fishery and forestry economics, environmental considerations such as possibilities for environmental degradation, and ecological aspects such as the presence of teredos and fouling organisms.

From an economic and environmental viewpoint brush-park fisheries could have a wider potential for dissemination if a cheap, readily available substitute substrate could be found to completely or partially replace the use of tree branches and other natural materials. Such artificial structures are already available as used for artificial reefs in marine waters (Fig. 13), but at this time might not be cost-competitive with natural materials except where the incidence of teredo attack is very high.