3.1.1 Water
Water is the sine qua non for fish farming in ponds. In comparison, inadequacies of soil texture and permeability can be overcome, albeit at some cost.
Water can come from a number of sources. The most dependable is from agricultural irrigation schemes which have already allocated some of their area for fish farming (Fig. 2). Other sources of water are from rainfall runoff, from streams and from groundwater. The latter is not a recommended source, if through its use the ponds cannot be drained.
Our task was to identify where water would be ample for fish farming other than in irrigation schemes. The approach was to take into account both annual rainfall (Fig. 3) and evaporation during the dry season taken as November to April (Fig. 4). Dry season evaporation ranges from 700 to 1000 mm and annual rainfall from 900 to 2000 mm. Evaporation was subtracted from rainfall and the resulting polygons arranged in 4 categories of effective rainfall which can be interpreted as: ample, adequate, marginal and insufficient (Fig. 5). Where dry season evaporation exceeds annual rainfall by a relatively small amount there are a number of implications for fish farming development: lack of perennial streams as water sources, increased construction costs to build deep ponds to retain water throughout the dry season and danger of incomplete replenishment in dry years.
If the effective rainfall was 300 mm or less, we took this to be a high cost, high risk area for fish farming development. This area corresponds to the parts of the country where it is well known that the availability of water in the dry season is a problem because many dams, dugouts and boreholes have been constructed there.
We allowed for a marginal area of effective rainfall between 300 and 400 mm that could still entail some risk and added costs, but which might be developed, if other criteria are better than satisfactory. If effective rainfall exceeds evaporation by 400 mm, or more, water is not a problem. For example, the “adequate” and “ample” areas correspond to conditions in Ashanti Region where water is not a constraint for fish farming.
3.1.2 Land
“Land” has been analyzed in a number of ways. Ghana possess a large surface that is reserved forest, or is allocated to various kinds of parks and wildlife reserves (Fig. 6). Aquaculture development is not possible in these areas so they have been omitted from consideration. Likewise, for the development of fish farming in ponds, the water surface of Lake Volta and lagoons, has been omitted. In total, there are some 46,000 km2, 19% of Ghana's surface, in these categories. The remainder, about 193,000 km2, the “net area”, at first glance, is the land that can be developed for fish farming. However, soil characteristics are also important for pond construction and operation.
Soil suitability for fish ponds has been analyzed in two ways. For the first, the soils of Ghana as mapped for the FAO-Unesco Soils Map of the World (FAO-Unesco, 1979) were rated according to their suitability for aquaculture taking into account texture, slope and permeability. Four categories were established: good, fair, poor and unsuitable (Fig. 7).
Because the FAO Soils Map of the World is at a large scale, 1:5 million, it is highly likely that there are areas of good or fair soils within soil polygons rated poor or unsatisfactory. These would tend to be the alluvial soils adjacent to rivers and streams. In fact, the distribution of such soils can be seen on detailed soil maps of Northern Ghana (FAO, 1967). Rivers, as mapped by the World Bank in 1986 (Fig. 8), have been used to indicate the locations of such soils. Their area has been conservatively estimated by assuming an alluvial soil width of 1 km. This is relatively close to the average, 0.75 km, for smaller rivers, but considerably less than the 6.8 km width for the White Volta and the 2.4 km width for the Mawli as sampled from maps of Northern Ghana. Thus, for soil polygons rated poor or unsuitable based on the soil map, a check has been made to see if the polygons contain rivers. If so, the area of alluvial soils is estimated as channel length multiplied by an alluvial soil width of 1 km.
These are detailed in the Economics Technical Report (B3), and in Field Working Papers C8 and C9. In brief, the availability of four inputs is evaluated on a district basis: cattle, pig and poultry manure and rice bran (Figs. 9–12). These, taken together, determine whether inputs are sufficient to support the development of 20 ha of fish ponds in each district2. The spatial distribution of inputs availability is mapped in Figure 13 and input indices for individual districts are in Table 1. Similarly, the capability of each district to absorb farmed tilapia and mudfish is estimated. This is based on human population density (Fig. 14), and growth rate, the pattern of fish consumption (Fig. 15) and the supply of tilapia and mudfish (Clarias) If the absorption capability is lacking, transport of the fish to a large urban market is an alternative, if the transport does not prove too costly. The transport cost decision is based on the road distance from each district capital (Fig. 16) to the nearest urban market centre of which there are four: Accra, Sekondi-Takoradi, Kumasi and Tamale. The spatial distribution of market opportunities is shown in Figure 17 and the marketing index of individual districts is in Table 2.
3.3.1 Welfare
People are worse off in some parts of Ghana than in others. Therefore, even though the physical and economic conditions for fish farming are not optimal in some of the worst-off areas, government might wish to stimulate development there to promote welfare. The mission believes that welfare deserves consideration. Welfare has been handled in such a way that by including or excluding it from the models, its effect on can be quantified.
The problem with most of the “quality of life” indicators is that they are outdated, or do not cover the country at district level. Even at regional level not all of the regions are reported individually. The best index available was "Number of Smallholder Households Below Basic Needs Income3 in 1986 (IFAD, 1988) which does cover each of the 10 regions (Fig. 18; Table 3). The values were assigned to categories as quartiles (Table 5).
3.3.2 Extension
“Extension” is a measure of the availability of Fisheries Department Officers in each region. It was calculated as the number of officers devoted to inland fisheries and aquaculture per region divided by the operational pond surface area per region (Fig. 19; Table 4). The index values were arranged in four categories corresponding to quartiles (Table 5).
3.3.3 Agglomeration
Agglomeration measures the positive influence of existing fish farms on the development of new farms. It implicitly takes into account such factors as already developed farming skills, availability of brood stock and fingerlings and markets. The agglomeration index was calculated as the surface area of operating fish farms in each district divided by the district's “net area”. The values were placed in 5 categories (Table 5).
3.3.4 Development
The density of roads, calculated as the length (km) of primary, secondary and tertiary roads (Fig. 16) divided by the area of the district, has been used as a general index of the level of development of each district (Table 5). The values were assigned by quartiles.
