6. RESULTS

In this section we highlight the implications for fish farming development that are brought out by considering each criterion separately. This provides a foundation for interpretation of the combined criteria in the integrated models.

6.1 Land and Water

Land and Water opportunities are clearly best in the southern part of the country with districts having a high KL index clustered in the SW (Fig. 20). Twenty-five percent of the districts have relatively high values as shown by the statistical frequency distribution of KL (Fig. 22) and by Table 6. There are 28 districts in the fourth quartile; however, nearly all of them have index values of 1. In contrast, nearly one-half of the districts have indices below 0.25.

There is relatively little soil area rated “good” (Table 6). Thus, relatively large areas of soils rated “suitable” contribute to high values of KL where water is sufficient.

None of the fourth quartile districts contain irrigated areas (Table 6). As would be expected, districts with irrigation schemes have low land-water availability indices because of the lack of water. However, for fish farming development, it would be a mistake to ignore these districts because of their low KL indices. Thus, how these districts rate with regard to other criteria should be used to decide on development opportunities.

The implication for development is that about one-fourth of the districts are outstanding from the land-water viewpoint, but that the majority are deficient according to the criteria we have used. Some of those which are deficient index-wise, have irrigation schemes in which fish farming development should be stimulated.

6.1.1 Inputs and Markets

Indices of inputs (KI, Fig. 13) and of market (KM, Fig. 17) do not have the same spatial distribution. This is shown by a correlation coefficient of only 0.36. Most of the inputs are in the northern half of the country while the markets have a less distinctive pattern, but with a concentration in the south central and central west. The highest inputs index is 0.75 and most districts have indices between 0.3 and 0.7 (Fig. 23; Table 1).

There are only two high marketing indices. As would be expected, they are the main population centres of Accra and Kumasi. The statistical distribution of the marketing index is similar to that of inputs, with most values distributed around 0.5 (Fig. 23).

The interpretation is that there are relatively few districts that offer good overall economic prospects for fish farming in that a high availability of inputs and markets occur together. The implication is that farmed fish would have to be transported from “fish farming districts” to “fish marketing districts”, thereby adding costs to the final product.

6.1.2 Other Factors

Agglomeration-development-extension, as the “Other Factors” index (KF) are highest in the SE with scattering of districts with higher values in the central W. Index values were relatively high for only a few districts (Table 5; Fig. 24). There were only three in the fourth quartile and only about 15% had values exceeding 0.5.

This suggests that there are few districts that offer significant opportunities for development due to the potential influence of “other factors”.

6.2 The Integrated Models

The four models provide different pictures of development opportunities (Figs. 25–28; Tables 7–10)⁵. However, as would be expected from the results from individual criteria, it is the S and SW that are favourable for development according to all of the models. With the basic model (K1), the best fish farming opportunities are clustered in Ashanti Region with 10 of 11 high-scoring districts there and the other in Volta Region (Fig. 25; Table 7). As compared with the basic model (K1), not taking welfare into account (Model K2) clearly lowers scores for much of Northern Region and Upper East and Upper West. At the same time development opportunities are enhanced in Brong Ahafo (4 high-scoring districts), Eastern (5), Volta (2) and Western (1). Ashanati is reduced to three high-scoring districts.

⁵ For Figures 25–28 the values generated for each district by the models have been cast into four categories which can be interpreted as “good” (3), “fair” (2), “poor” (1) and “unsuitable” (0) in terms of opportunities for fish farming development. The accompanying tables, 7–10, show individual values for each district according to the same categories.

Including welfare (Model K1) shows the districts in which fish farming could be developed, but at a higher cost to government and with a greater risk commercially. Also including welfare, but emphasizing the economic aspects (Model K3), highlights the districts where development can go ahead with a better chance of commercial success given that other criteria are satisfied reasonably well. Nine districts fall into this category of which 5 are in Ashanti, and of which there are one each in Volta, Brong Ahafo, Northern and Greater Accra (Fig 27; Table 9). Emphasizing inputs and markets, Model K3, places part of the north into a better light because that area was relatively well off for inputs.

Taking out agglomeration (Model K4) puts the focus on districts that are high-scoring for other criteria, but where fish farming may be little practiced at present. It is in these latter districts that extension would have to be relatively intense to stimulate fish farming. Fifteen districts are in the high-scoring category for model K4 (Fig. 28). Of these, there are 5 each in Brong Ahafo and Eastern, 2 in Ashanti and one each in Volta, Western and Central (Table 10).

Looking at the frequency distributions of the normalized results of the four models (Fig. 29) gives an overall view of relative development opportunities. Districts with normalized index values higher than 1.25, of which there are 20 each for Models K1 and K2, 15 for Model K3 and 19 for Model K4, fall into a natural grouping of districts with the best opportunities for fish farming development.

In order to narrow down the selection still further, the districts were arranged according to the number of times that the highest scoring districts appear in each of the four models. These can be considered as the best all around disticts for fish farming development (Table 11).

There were three districts, Atwima, Amansie W and Ho that had high scores in each of the four models and two districts, Kumasi and Sunyani, that had high scores in 3 of the 4 models. There were 11 other districts that scored highly in 2 of the 4 models and an additional 10 districts that scored high for one of the models.

Table 11. Occurrence of high scores (value=3) among the four models.