Aquatic weeds can be broadly defined as “unwanted and undesirable plants which grow and reproduce in an aquatic environment” (Lawrence, 1966). The problem of weed growth in irrigation and drainage canals in many countries has been reported for many years by various authors (Avault, 1965; Holm et al., 1969; Bates and Hentges, 1976; Reilly, 1984; Brabben and Bolton, 1988) and in many countries including Egypt (Dubbers et al., 1980; Khattab et al., 1981; Huisman, 1983; Van Weerd, 1985; Brabben, 1986; Sadek, 1988), the USA (Legner 1975, 1979; Legner and Fisher, 1980; Legner and Murray, 1981), the USSR (Martishev, 1973), Java and N. America (Shuster, 1952), Sudan (Alabaster, 1981; Redding-Coates and Coates, 1981), Zimbabwe (Fair, 1982; Brabben and Bolton, 1988) and India (Mehta et al., 1976; Petr, 1987) all of which have weed problems which have had or are having a significant effect on the economy.
Weed problems in the tropics are particularly prevalent, since the high year-round temperatures enhance growth rates and shorten the lifespan of plants. As a consequence there is a large accumulation of organic material which, by the release of nutrients, stimulates more growth. Where there is little control over aquatic plant growth their presence very often constitutes a significant nuisance.
Most irrigation and drainage systems are constructed with unlined earth canals, with relatively slow flow velocities to prevent excessive erosion (usually in the region of 0.5 m/s) (Brabben and Bolton, 1988). Drainage canals often contain slow moving or stagnant water. Both the distribution system and the drainage canals are, therefore, ideal places for the growth of aquatic weeds.
The extent of the problem in irrigation systems is partly reflection on the lack of consideration given to potential aquatic weed growth in canals during the design stages of irrigation systems, as well as on poor maintenance of existing systems - often through lack of attention, and financial, mechanical or chemical resources. Some plants have emerged through introduction, others through nutrient enrichment of water bodies and the construction of irrigation systems (Williams, 1980). It is also clear that the problem of aquatic macrophytes has become worse in recent years and is causing increasing financial and environmental problems. In many irrigation schemes maintenance costs account for a large proportion of the budget. In Egypt a seven-year canal maintenance project has recently been commissioned with a loan of US$ 70 million (Brabben and Bolton, 1988).
Since the closure of the High Aswan Dam there have been reports of severe aquatic weed growth in irrigation canals (Holm et al., 1969; Van Zon, 1984; Brabben 1986). The construction of the dam resulted in increased silt deposition in Lake Nasser, and reduced deposition in downstream irrigation canals. This resulted in lower turbidity in the canals, so enhancing weedgrowth through increased light penetration (Van Zon, 1984). The increased use of organic fertilizers has, through eutrophication in canals and drains (Van Zon, 1984) also contributed to this problem (see Section 3.5).
In many cases, whilst tertiary canals are only supposed to contain water during the flooding period, they may hold water for longer periods, thus allowing the establishment of aquatic weeds and disease vectors. Unfortunately, because tertiary and field canals are, in some ways, an ephemeral habitat, they are not suitable for stocking fish for control of nuisance biota (Coates, 1984).
The main aquatic plants, many of which are found throughout irrigation systems, are listed in Table 4.
| Submergent Plants | Floating Plants | Emergent Plants |
|---|---|---|
| Algae sp. | Lemna | Typha |
| Myriophyllum | Azolla | Phragmites |
| Ceratophyllum | Hydrocharis | Carex |
| Naja | Salvinia | Scirpus |
| Hydrilla | Riccia | Cyperus |
| Elodea | Pistia | Acorus |
| Brasenia | Eichhornia | Eleocharis |
| Vallisneria | Limnocharis | |
| Limnophyla | Glyceria | |
| Utricularia | Sagittaria | |
| Potamogeton | Pontederia | |
| Ranunculus | Ludwigia | |
| Polygonum | Persicaria | |
| Sphagnum | Nymphaea | |
| Marsilia | ||
| Jussiaea | ||
| Ipomoea | ||
| Nasturtium |
The main effects of excessive weed growth are in reducing water flow in canals (through physical restriction and by increasing the frictional resistance to flow) (Legner and Murray, 1981; Brabben and Bolton, 1988; Sadek, 1988), providing breeding and feeding habitats for agents of diseases (Dubbers, et al., 1980; Legner and Fisher 1980; Redding-Coates and Coates, 1981) and by often considerable increases in maintenance costs. It is estimated that costs in Egyptian irrigation canals over 10 m wide can exceed US$ 800/km/y (Van Zon, 1986). In many instances the cost of weed removal may be in excess of the increased returns provided by irrigated agriculture (Carruthers and Clarke, 1981). In some canals the increase in resistance may be as high as 70% with an accompanying reduction in carrying capacity of 40% (e.g. Kalabia Canal, Egypt) (Hydraulics research, 1988a).
Increases in aquatic macrophytes can decrease plankton production (Shuster 1952). For example, Salvinia sp. are known to exclude phytoplankton to the extent that young tilapia will move to more open waters, where plankton is more abundant. This can lead to increased mortality amongst the smaller fish, due to predation (Donnelly, 1969). Death of these plants results in increased amounts of bottom debris, with an associated increase in B.O.D.
The various components of any irrigation scheme will offer suitable habitats to different intermediate hosts of human vector-borne diseases (Holm et al., 1969). The widespread and serious risk attributed to vector-borne diseases has often been associated with poor socio-economic conditions and projects aimed at economic development, such as irrigation schemes and dams, with their associated weed problems. Despite the initial success of chemical control the problem of disease vectors still persists in many countries. However, an increasingly wider application is given to environmental management for vector control. For the last ten years, a joint WHO, FAO and UNEP Panel of Experts on Environmental Management for Vector Control (PEEM) held meetings and coordinated a number of activities relevant to this problem in irrigation schemes. For example, in 1981 the Panel held a technical discussion on the use of fish for mosquito control, which resulted in the publication on herbivorous fish suitable for situations of irrigated agriculture (Haas 1984). New development programmes in Africa, including irrigated agriculture, are based on strict environmental management principles. These are the Cairo Programme for African Cooperation and the Zambesi Action Plan (PEEM Report, 1988).
The construction of irrigation systems often produces large areas where the hydrobiological conditions are vastly different to those of the natural waters (Redding-Coates and Coates, 1981). Formerly fast-flowing waters are reduced to sometimes stagnant or slow-flowing canals, supplying or draining the agricultural land. This change in the environment encourages the growth of aquatic macrophytes, which in turn provide habitats for the water associated vectors of many diseases.
In many cases, it is the extent of aquatic weed cover, which provides a suitable environment for these disease agents (Holm, et al., 1969; Daget, 1976; Coates and Redding, 1981; Petr, 1985). In the Sudan it is estimated that over 90% of the population are affected in some way by the disease bilharzia, transmitted through an intermediate host the snails Bulinus spp and Biomphalaria spp. both of which inhabit vegetation in irrigation canals (Redding-Coates and Coates, 1981).
Presently, preference is given to controlling disease vectors and hosts commonly found in the irrigation and drainage canals through various methods of environmental management, i.e. minimising or eliminating the habitats. The removal of vegetation used as a habitat by agents of disease by the introduction of grass carp, for example, has proven most cost effective (see below).
Mechanical control is costly and requires accessibility to the waterways, and chemical herbicides are often so toxic that the water cannot be used subsequently for fish culture or human consumption or other uses. Fish can play a potential role in this aspect since they are active, feed voraciously and will produce a valuable by-product in the form of protein. The species which has been most widely used in this regard is the grass carp, Ctenopharyngodon idella, although there are other species with potential in this regard (Table 5).
Originally a native of China's river systems the grass carp, has been introduced throughout the world. One of the reasons for its wide dispersal has been its effectiveness as a biological control agent in the control of nuisance aquatic weeds (Hora and Pillay, 1962; Van der Lingen, 1968; Mitzner, 1978; Lembi et al., 1978; Van Zon, 1984; Singit, 1985; Van Weerd, 1985; Whitwell, 1986; Okafor, 1986; Schramm and Jirka, 1986; Wolff, 1988; FFI, 1989) (see Table 6).
It has widely been found to be much cheaper than either chemical or mechanical control (Avault 1965), and was effective in the removal of habitats for two bilharzia-.carrying species of snail (Van Schayck, 1985, 1986). Van Zon, et al. (1978) estimated that weed control in Egypt using biological methods, would cost 75% less than mechanical means, and further experiments by Khattab, et al. (1981) suggested that a combination of biological (the introduction of grass carp) and mechanical methods was a more effective means of control than mechanical alone.
| Common name | Scientific Name | Distribution | Notes | References |
|---|---|---|---|---|
| VORACIOUS HERBIVORES | ||||
| Grass carp | Ctenopharyngodon idella | Widely distributed Origin China | Consumers of a variety of plant food including terres- trial (leaves and grasses) and aquatic plants | Ling, 1967 |
| Tilapia | T. rendalli | Africa | Effective grazers on sub- merged macrophytes - also feed on plankton, benthic animals and detritus | Spaturu, 1978 |
| T. zilli | Africa | |||
| Wuchang bream | Megalobrama amblycephala | China | Consumes higher plants | Coche, 1980 |
| Tawes | Puntius goniotus | South East Asia | Feeds particularly on fila- mentous algae; also terres- trial plants | Hora and Pillay, 1962 |
| Giant gourami | Osphronemus goramy | Asia | Feeds mainly on plant leaves | Hora and Pillay, 1962 |
| OMNIVORES | ||||
| Tilapias | Oreochromus mossambicus | Widely distributed | Efficient consumers of vege- tation but may prefer peri- phyton attached to larger macrophytes | Lahser, 1967 |
| Oreochromus niloticus | Widely distributed | Efficient consumers filamentous algae - but less efficiently than other tilapias | Avault, et al., 1968 | |
| Sepat Siam (snakeskin gourami) | Trichogaster pectoralis | |||
| Goldfish | Carassius auratus | Readily consumes filamentous algae, e.g. Pithophora sp. | Ruskin and Shipley, 1976 | |
| Catla, rohu, mrigal | Catla catla, Labeo rohita, Cirrhina mrigala | |||
| Common carp | Cyprinus carpio | Feeds on plants only if other food not available. Controls weeds by sediment disturbance | Hora and Pillay, 1962 | |
| Milkfish | Chanos chanos | Feeds mainly on decaying | Hora and Pillay, plant material. Large fish also consume fresh filamen- tous algae | 1962 |
| POTENTIAL | ||||
| Silver dollar fish | Metynnis roosevelti and Mylossoma atgenteun | South America | Aggressive schooling fish Stocked at 1 200 2 500/ha Rapidly removes vegetation by biting off at base and consuming later | Ruskin and Shipley, 1976 |
| Tilapia | Tilapia guineensis | West Africa | Grows in estuarine condi- tions when fed leaves from higher plants | Coche, 1983 |
| Tambaqui | Colossoma bidens | Amazon region South America | High commercial value. Grows to large size, hardy to culture conditions. Eats fruit in particular | Ruskin and Shipley, 1976 |
| Pirapitinga | Mylossoma bidens | |||
| Cachama sp. | Venezuela | FAO, 1983 | ||
| Brycon chagrensis | Panama | July, 1981 | ||
| Pearl spot | Etroplus suratensis | Southern India and Sri Lanka | De Silva and Perrera, 1983 De Silva, et al., 1984 | |
| Distichodus engycephalus | West Africa | Afinowi and Ezenwa, 1982 | ||
| Distichodus brevipinis | Efficient herbivores - can be pests in rice-growing areas | |||
| Distichodus rostratus | ||||
| Species | Place of and reason for introduction | Result | Reference |
|---|---|---|---|
| C. idella | Malaysia, to clear 1.8ha pond 375 fish stocked | 22t cleared in 110 days | Hickling (1960) |
| USSR, to clear Kara Kum Canal | Planned flow reduced by weeds loss of 20,000ha irrigated cotton Decreased weeds | Edwards (1980) | |
| Arkansas, to clear 20000ha public lakes | After 15 years no problem | Ruskin and Shipley (1976) | |
| T. rendalli & T. zillii | 2–10 ha reservoirs in Kenya | Total removal after 2–5yrs | Van der Lingen (1968) |
| T. zillii | Canals of Imperial Valley S. California 2500 fish/ha stocked | Complete removal | Ruskin and Shipley (1976) |
| P. gonionotus | E.Java. Indonesia irrigation dams | Cleared 284 ha reservoir in 8 months | Shuster (1952) |
| Giant gourami O. goramy | India irrigation | Control of aquatic weeds | Hora and Pillay (1962) |
| Tilapias | Hawaiian sugar cane plantation irrigation canals, 75000 fry at 5–10cm | Cost of herbicides $5000 reduced to $3000 with fish control. No regrowth | Little & Muir (1987) |
| Average individual weight (g) | Stocking Density kg/ha | Average no./ha |
|---|---|---|
| 10 – 15 | 50 – 60 | 5000 |
| 20 | 60 – 90 | 4000 |
| 30 | 60 – 120 | 3000 |
| 100 | 120 – 150 | 1500 |
| 200 | 180 – 250 | 1000 |
| >300 | >200 – 300 | 500 – 850 |
In general grass carp are not specialised feeders and have been shown to consume over 170 different species of aquatic plants. Mitzner (1978) and Zonneveld and Van Zon (1985) concluded that little shift in the qualitative structure of plant species in a water body will be evident after stocking with grass carp, although the biomass of vegetation would decrease. The less preferred species with spiny or hard tissues are not consumed until later in the season after supplies of other plants have declined. Smaller fish prefer different species of plant compared to the larger sizes, hence, over a period of time, the control of plants in a water body will be fairly comprehensive.
The differences in food preferences seen by some authors are a result of the size of the fish used or the ambient temperature. Smaller fish with smaller mouths prefer softer tissues of the submerged vegetation, and the growing roots of water hyacinth. Thus it would be preferable to stock smaller grass carp with water hyacinth since the larger sizes will only feed on the leaves and not the growing points (Zonneveld and Van Zon, 1985). Mitzner (1978) found that grass carp of approximately 380 g have a preference for Najas and Potamogeton sp.
Low temperatures seem to affect the food preference of grass carp, the softer more succulent species of submerged vegetation (e.g. Elodea spp., Hydrilla verticillata, Myriophyllum and Potamogeton spp.) being consumed when temperatures are below 12– 15° C. This should be a consideration when using grass carp for weed control in temperate waters.
In Egypt the problem was in prevention of fishermen taking too many of the stock from irrigation canals and drains (Nour, pers. comm.). The fishermen now realise that the alternative to the grass carp would be chemical control which would have a detrimental effect on the natural fishery. In Michigan the main concerns about aquaculture in canals and the introduction of exotic species are the reduction in water quality, the spread of disease and the effects of escapees on the species used for sport fishing (Whelan, pers. comm.).
Experiments on the bio-energetics of grass carp have shown them to have a low metabolic expenditure in relation to other species, such as the common carp. This enhances their potential for fisheries and aquaculture, as more of the assimilated energy is available for growth (Huisman, 1979; Zonneveld and Van Zon, 1985). Grass carp also has low food conversion efficiencies, as its physiology is not totally adapted to its herbivorous diet, with the result that it has to consume large quantities of plant matter to satisfy its daily metabolic requirements (Van Zon, 1984). Hickling (1966) found that 48 g of vegetation was required to produce 1 g in weight gain. Whilst this is not normally an attractive feature in the aquaculture or fisheries context, it is highly advantageous in the species role as a weed control agent.
Large-scale stocking of grass carp in irrigation and drainage canals in Egypt proved to be successful at controlling weeds and increasing the protein production in Egyptian waterways (Van Zon, 1984). Fish were stocked at 20–30 g in weight, at densities of 60–120 kg/ha. Stocking larger fish has the disadvantage that stocking densities have to be increased to attain the same level of weed control. It was found in Egypt that fish of 20–30 g stocked at 50–60 kg/ha would control weeds to the same extent as 200 g fish at stock densities of 200 kg/ha (Table 7) (Van Zon 1984). However, where natural predators are likely to be a problem, it may be necessary to stock with larger fish.
Although smaller fish can be stocked at lower densities they will not feed on wooded stemmed plants such as Phragmites and Typha sp. However since over 200 species of plant are readily consumed by these fish throughout their life-span and the fact that their small size dissuades fishermen from removing them, there is a distinct benefit to stocking at smaller sizes. Earlier experiments using larger sizes (260 g) indicated that the loss from human predation tended to decrease the effectiveness of the grass carp stocking programmme for weed control (Van Zon, 1984).
The use of grass carp for weed control has the added benefit of enhancing the endemic fish populations, by increasing the rate of nutrient recycling in the system (Van Zon, 1984). They are also reported to have fast growth rates with 350 g fish increasing in weight to 1 571 g in 80 days in lakes in the USA (Mitzner, 1978), and 48.7 mm fish increasing in size to 186 mm in six months (Colle et al., 1978).
An economic analysis of the system showed biological weed control to be considerably cheaper than conventional methods (Tables 8, 9, 10, 11). Chemical and mechanical control tend to create a shift in the vegetation to the species less susceptible to the chemicals used (Table 12). This would create ever increasing costs of control and long-term planning is necessary (Van der Bleik et al., 1982). Grass carp do not have this effect, their use resulting in an overall decrease in numbers.
Costs in the USA for chemical control were found to be many times higher than for using grass carp. In Florida it was estimated that 15 000 hectares of Hydrilla were treated with chemicals at a cost of US$ 9.1 million whereas the cost of grass carp stocked at 35 fish/ha would have cost US$ 1.71 million (1977 prices) (Haller 1979). In California, the cost of maintaining weed free canals with chemicals was estimated to be in excess of US$ 150 000 a year, but with the introduction of grass carp into the canals the costs have fallen to just US$ 15 000 per annum (Fish Farming International, 1988). In addition to the economic benefits, the use of grass carp is a longer lasting measure.
| Canal Width | Drain width | |||||
|---|---|---|---|---|---|---|
| Methods of Control | >10m | 5–10m | <5m | >10m | 5–10m | <5m |
| Mechanical only | 1700 | 1140 | 540 | 1150 | 865 | 540 |
| Chemical +Mechanical | 1175 | 925 | 410 | 925 | 670 | 410 |
| Grass Carp +Mechanical* | 7851 | 5402 | 3 | 7851 | 5402 | 3 |
Notes
* After second year of stocking
1. Estimated average width 15m.
2. Estimated average width 7.5m
3. Not suited to stocking fish
| Annual costs | US$/ha | |
|---|---|---|
| restocking of 750 fingerlings @ 30g | 190 | |
| estimating price of fencing (per km 5yr deprec) | 25 | |
| estimated additional maintenance | 175 | |
| guarding/cleaning | 300 | |
| Total | 690 | |
| Annual revenues | ||
| 600kg grass carp | 1050 | |
| 500kg endemic fish species | 625 | |
| Total | 1675 | |
| Canals | Drains | |||
|---|---|---|---|---|
| Costs | Revenuesb | Costs | Revenuesb | |
| Mechanical control only | 1700 | 550 | 1150 | 550 |
| Chemical+additional mechanical control | 1175 | 550 | 925 | 550 |
| Grass carp+additional mechanical controla | 785 | 2500 | 785 | 2500 |
a second and further years after stocking
b conventional yields of endemic species: up to 300kg/ha/yr.After introduction of grass carp: up to 500 kg/ha/yr.Harvestable grass carp production : at least 600 kg/ha/yr.
| First year: | stocking of 3000 grass carp of 30g | 750 |
| estimated price of fencing per km (depreciation 5 years) | 25 | |
| estimated additional conventional maintenance (per km 10–15%) | 175 | |
| guarding/cleaning (1 labourer/2km) | 300 | |
| Total Costs | 1250 | |
| Yield possible | 2375 | |
| Fourth year: | stocking of 750 grass carp of 30g | 190 |
| estimated price of fencing (per km depreciation 5 yrs) | 25 | |
| estimated added conventional maintenance (per km 10–15%) | 175 | |
| guarding/cleaning (1 labourer/2km) | 300 | |
| Total Costs | 690 | |
| Yield* | 1735 |
| Type of control | Mechanical | Herbicidesa | Grass Carp | ||
|---|---|---|---|---|---|
| Biota | Persistent | Non persistent | |||
| Macroflora | |||||
| biomass | 4 | 5 | 4 | 5 | |
| diversity | 3 | 5 | 5 | 3 | |
| Phytoplankton | |||||
| biomass | 6 | 4 | 4 | 2 | |
| diversity | 3/6 | 5 | 4 | 3/2 | |
| Zooplankton | |||||
| biomass | 6 | 2 | 4 | 2 | |
| diversity | 3/6 | 5 | 5 | 3 | |
| Macro-invertebrates | |||||
| biomass | 6 | 5 | 4 | 2 | |
| diversity | 3/6 | 5 | 5 | 3 | |
| Fish | |||||
| biomass | 6 | 6 | 6 | 1 | |
| diversity | 3/6 | 6 | 6 | 3 | |
Notes :
1 - more than ]
2 - temporarily more than ]
3 - same as ] -in undisturbed situation
4 - temporarily less than ]
5 - less than ]
6 - no or not much reliable data ]
a - Non-persistent means a half lifetime of the activecompound of less than 14 days
Since the first publicity of the potential of the grass carp for weed control, the level of interest and research work has increased considerably. Experiments using grass carp to control aquatic weeds were performed in the irrigation and drainage canals in south Florida (Schramm and Jirka, 1986). Effective control of Hydrilla verticillata was obtained at stocking densities of 110 fish (average weight 0.6kg) per hectare of water. The poor survival of the fish in this study was related to low water levels in the canals during irrigation and drainage procedures, herbicides and pesticides in drainage water, and low oxygen levels in areas of dense weed growth.
Grass carp have been used extensively in the Netherlands for control of weeds in irrigation and drainage canals as an alternative to chemical and mechanical control (Van Zon, 1984). A number of limitations to the use of grass carp have been found there. These concern the quality of the water, the physical nature of the waterbody and the nature of other uses being made of the waterway. The major limitations are discussed in Zonneveld and Van Zon (1985) and are summarized below.
The possibility of escape or migration to other waterbodies must be considered. This would have the effect of lowering the stocking density and the effectiveness of weed control in the immediate area. Use of fencing to control stock movements may not be practical, since it will reduce waterflow in the canals. Schramm and Jirka (1986) found that movement of grass carp increased with rising water levels, resulting in an “upstream” migration, only partially controlled by barriers which themselves slowed the water flow. The same authors suggested that efficient capture methods and controlled distribution would enable more effective use of the fish.
Low dissolved oxygen levels (<4 mg/1), or fluctuating dissolved oxygen levels decrease the feeding rate of grass carp substantially (Shireman et al., 1977). High levels of weed growth may cause oxygen fluctuations by day and by night, and, therefore, it has been suggested (Aliyev, 1976; Sutton et al., 1979; Van Zon, 1980) that grass carp be used as a secondary method of weed control, after major weed removal has been carried out. Schramm and Jirka (1986) also found that fish mortality increased in areas of dense vegetation due to the creation of anaerobic conditions. Smaller fish were found to be more tolerant of low oxygen levels. Oxygen thresholds of 0.32–0.60 mg/O2/1 were found for grass carp of 2–3 g (Opuszynski, 1967), and it was proposed that this tolerance relates to the natural environmental conditions of waters used by juvenile fish (areas of dense vegetation) for protection.
Although grass carp effectively control weeds at temperatures of 15–30°C (Kilambi and Robison, 1979), they tend to avoid cooler temperatures, and their feeding preferences change at lower temperatures, tending to be less generalized.
Grass carp are very tolerant of disturbance, but tend to only remove plants in these areas when all others have been eaten. (Van Zon, 1977). They also prefer water depths of over 30 cm (Janichen, 1978), which tends to restrict their use for weed control in the very shallow field canals, unless fencing is used to prevent movement to deeper areas.
In addition, the large volumes of faecal material produced by grass carp may lead to changes in the water quality (Hickling, 1966). Changes in the turbidity of the water and the concentration of potassium were found in ponds stocked with grass carp used for weed control, but not at levels which were considered to cause ecological problems (Lembi et al., 1978; Avault et al., 1968). Such problems, however, should be of little consequence in irrigation canals.
The most obvious advantage to the use of grass carp as a biological control agent, in preference to chemical or mechanical methods, is the direct increase in food production within the water body for very little capital expenditure. The increased nutrient turnover created by the fish can encourage the growth of local fish species, resulting in increases in production of 100 – 200 kg/ha in Egypt (Bailey, 1978; Shireman and Smith, 1981). Part of the stock can be removed annually, up to a figure of 500 kg/ha, since as part of the management programme 25% of the number of fish originally stocked are added each year. Management aspects of biological control programmes are concerned with removal of a certain percentage of the fish, since overpopulation can lead to high turbidity and bank erosion. Numbers can be estimated by mark and recapture techniques (Robson and Regier, 1964, 1971). Where there is also a natural fishery operating the numbers caught must be regulated. There are several methods which can be used, including:
The regulation of the fishery - this is not a particularly effective method with artisanal fishermen
Operation of a closed period when weeds are most dense - this method would also benefit the endemic fish species.
Regulate the fishing gear used - this would be useful only if one size were the most popular for the market. In many countries there is also a market for the smaller sized fish.
Undersized fish returned to the water.
Other species which have been found to reduce or control certain weeds (see Tables 5 and 6) include the common carp, Cyprinus carpio, which stirs up the substrate sufficiently to prevent the growth of submerged weed through shading (a consequence of the increased turbidity). This occurred when carp were stocked at a minimum density of 400 per hectare. The Israeli strain of common carp was found to control the alga Pithophora, and other monofilament algae, when stocked in limited numbers (55 adults per hectare). Tilapia species Oreochromis mossambicus, O. niloticus and Tilapia melanopleura are also known to control Pithophora, and other species of filamentous algae and submerged weed (Lawrence 1966).
Further species used for weed control include Metynnia sp. (Holm et al., 1969) and Tilapia zillii (Petr 1987). The latter species was used to control submerged species of plant in Arizona and California. However the tilapia were also subject to heavy predation by bluegill and largemouth bass (Petr 1987). In contrast the use of grass carp for the same purpose was implicated in the decline in the stocks of bluegill and largemouth bass. The reason for the decline may have been due to the increased turbidity and the invasion of breeding sites by grass carp (Forester and Lawrence, 1978). Thus the balance between effective control of weeds and the effects on the natural populations must be maintained carefully.
In the Nile irrigation scheme in Egypt, water hyacinth, Eichhornia crassipes, is collected from the canals and used to replace approximately 30% of the leaf protein concentrate in the food fed to common carp. This reduces the amount of fish meal required and the proportion in the diet increased during times of food shortage. Farm wastes are also being used as supplementary feed for fish, e.g. plant debris after harvesting of crops etc. (Nour, pers. comm.).
Lemna minor was found to be a preferred food for Nile tilapia in cages in Indonesia, in place of Hydrilla verticillata and Chara spp. (Rifai, 1979, 1980). Another aquatic plant, Ipomoea aquatica, is produced throughout Asia primarily as a food for human consumption, although the remainder after the tips are removed is used as fish feed. Little and Muir (1987) report the use of Azolla pinnata with the blue green algae, Anabaena azolla, as a food for Nile tilapia, although experiments by Pullin and Almazin (1983) suggest that the growth rates of fish fed on this plant are poor, due to its high fibre and low carbohydrate content. Productivities of the more widely occurring aquatic weeds are shown in Table 13.
Aquatic weed growth in irrigation canals, and other waterways, is a significant problem in many parts of the world, and can constitute a substantial drain on the agro-economy, of developing countries in particular. Alternative methods of weed control are being sought as the traditional methods (chemical and manual/mechanical control) become increasingly expensive or impractical.
Biological agents of control are increasingly being seen as a feasible solution to the problem. The level of interest and the research effort into the use of fish (particularly the grass carp) to control excessive aquatic weed growth in irrigation canals has steadily gained ground in recent years, and there are now several studies which show that not only can fish be successfully and economically used to control weed growth, but that such schemes also result in an improved supply of protein to the rural communities.
In contrast to their generally accepted image as nuisance weeds, aquatic plants can also yield an economic crop. This is especially true when their growth is integrated with other farming activities, such as fish production, where they can provide a cheap source of feed or fertilizer.
