based on the work of
J.M. Kapetsky (FAO Senior Fishery Resources Officer)
N. MacPherson (FAO Consultant Socio-economist team Leader)
FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS
Rome, 1990
In Ghana the human and material resources available to the Department of Fisheries (DF) for promotion of fish farming are extremely scarce. It is in the interest of the FD to allocate these only in those regions/districts where results are likely to be obtained. The suitability of any particular district for fish farm development depends on the volume of farmed fish considered the minimum justification for a district. In this annex the analysis is postulated on the minimum being 20 hectares of pond surface area1, with an average production of 4 tons/ha and year, or 80 tons for the district, at the end of a five year period. It is of course not essential in an analysis of this kind that the minimum be uniform throughout the country. In a refined analysis the minimum can be specified for each district.
The likelihood that fish farming will be developed in any particular district depends on a number of factors such as: markets for farmed fish, availability of inputs, alternative income opportunities in and outside agriculture, systems of management of natural resources, etc. The previous analysis2 has indicated that the regional distribution of fish farm growth largely depends on cost of transport, and access to markets for both farmed fish and priority fish farm inputs. These factors can be weighed together using the GIS. Once the “combined suitability” of the district for fish farm development has been assessed in this manner, a district, and region-wise, strategy can be developed.
The problem, however, is larger than these economic aspects. Other considerations must be included before a “final” method for ranking districts is developed. Therefore, before analysing below in some detail the methodology for “weighing” the economic data, the overall problem will be addressed.
1 The reasons for selecting 20 ha as a “critical mass” are given in Field Working Paper No. 10.
2 See section xx of main report.
The author is of the opinion that for any location there are four groups of factors which are of about equal importance for a successful fish farming development programme. The four categories are: (i) access to year round water supply3, (ii) access to markets4; (iii) access to priority fish farm inputs5; and, (iv) miscellaneous aspects. The miscellaneous category includes aspects such as: road density, taboos concerning farmed fish consumption, soil conditions, existence of practising fish farmers, extent of farmers' self-management of irrigation schemes.
The author suggests that each of these four categories receive and overall ranking of between 0 and 1. The actual value will depend upon how well the particular district fulfills the ideal conditions. The ideal district, at the very most can obtain a value of 4. Categories (i), which is the access to water, and (iv), the miscellaneous category, will not be discussed further in this annex.
The analysis will distinguish between the two main alternatives for selling the farmed fish: (i) within the district, and (ii) in major urban areas. The latter have been limited, in this excersize, to: Accra-Tema, Sekondi-Takoradi, Kumasi and Tamale.
The size of the market, in five years' time, is a function of the following district information:
| p | population size, |
| pcc | per capita consumption (kgs/person/year) ; |
| t&c | share of tilapia and clarias in per capita consumption; |
| dff | growth in demand for fish; |
“dff” is placed at 4 % per year in districts. In the four urban areas it is placed at 7 % per year6. Map no xx shows the values used for “pcc” and “t&c” by district.
4 Analysed in chapter 5 and annex xx.
The “Maximim Farmed Fish Demand” (MFFD) is expressed in tons. It is calculated as follows:
MFFD = p * pcc * t&c * (1.045 - 1)/1.000
The author stipulates in this analysis that, the projected fish farm output - 80 tons - can not be marketed in the district at the prices used in the analysis, if 80 tons is more than the natural increase in the consumption of tilapia and clarias. The smaller the share of the tilapia and clarias in the projected increase, the better, naturally. In view of these considerations, the weighting might be done as follows:
Wdma = (MFFD - 80)/80
(i) All fish sold within the district
Wdma can be both negative and positive. It is naturally preferable that it be positive. This indicates that the farmed fish will not need to supply all the expected natural growth in demand. However, a share of the increase in markets of 10% is good enough. Therfore, the author suggests that the maximum accepted value for Wdma be 10. To bring the value around to unity it is suggested that Wdma be divided by 10.
If the MFFD = 80, the value of Wdma will be 0. This is not a useful indication, as in fact, in extremis, the production could be disposed of within the district. The author therefore suggests that the formula for the local market, when the market can absorb all or more than all the farmed fish, be as follows:
| Kdm = Wdma/10 + 0.5 | Kdm maximum = 1.5 |
(ii) Some fish sold outside the district
In the case that the district market does not suffice for the increased production, it will be necessary to sell outside the district. In this analysis it has been stipulated that the sales will take place in either of the four urban markets identified above. This situation occurs when Wdma < 0.
| If Wdma < 0, then | Wdma = -1 at maximum. |
The larger the number, in absolute terms, the worse the situation is, as it implies that an increasing proportion of the fish farm output must be sold outside the district. However, in order to facilitate the process of addition, this number will be converted into one that is positive and for which an increasing value reflects an improving situation (analogue to Kdm1). Even if the local market can absorb almost all the product, the prospect is not very favourable and the author proposes that the weight be multiplied by 0.5 to reflect this. The formula will then become as follows:
Kdm2=(1-[Wdma]) * 0.5, thus max Kdm2 = 0.57
In the final summation, either Kdm or Kdm2 must be used, not both.
The available data are:
| up | urban population of nearest “market town”; |
| pcc | per capita consumption of fish (kg/individual/year) in the urban market; |
| t&c | the share of tilapia and clarias in pcc; |
| dff | the annual growth rate in demand for farmed fish; |
| ukm | distance from the district to nearest urban market; |
| mkm | maximum distance that fresh fish in ice can be transported (a function of cost of transport); |
Most of these data are available. The values for “pcc”, “t&c” and “dff” were given in the previous section. The distance of each district centre to the nearest market town is given in Appendix The maximum distance for transporting iced fish in insulated containers has been estimated at 300 kms8
The maximum farmed fish demand in the urban centre (MUFFD) closest to the district (centre) is calculated as follows:
MUFFD = pu * pcc * t&c * (1.075 - 1)/1 000
7 The bracket [] signifies that the absolute value should be used.
8 See Chapter 6 of main report.
The weighting must take into account the share that the district production is likely to occupy in this urban market. The first possibility is that the urban market in fact need not absorb any farmed fish as all of it can be consumed in the district. In that situation the “Urban Market Share” of the farmed fish production will be labelled UMS1. It can be calculated as follows:
UMS1 = (1 + (80 - 80))/MUFFD.
When some of the district farmed fish production must be sent outside the district the “Urban Market Share” is labelled UMS2. It will be calculated as follows: UMS2 = (1 + (80 - MFFD))/MUFFD.9
Given the size of the four urban centres, the 80 tons of any particular district, even if it all had to be sent to the same centre, are not likely to occupy a large share of the particular market, except if it all goes to Tamale. That would be problematic. For the other markets, both UMS1 and UMS2 are likely to be quite small in absolute terms. The assignments of Weights to the resulting values is a question of judgement. The author proposes the following:
| UMS1 or UMS2 (in %) | Value of Kkum1 |
| < 1 | 1.0 |
| 1 < 5 | 0.9 |
| 6 < 10 | 0.7 |
| 11 < 20 | 0.5 |
| 21 < 50 | 0.2 |
| 51 < 100 | 0.0 |
The further away they are from the urban market, the worse off the district's fish farmers will be. The maximum affordable distance has been placed at 300 kms. However, there are fixed costs for ice and insulated containers which all farmers must pay irrespective of distance covered. That cost can be expressed in equivalent kg-kms of iced fish transport10, as in the formula below;
Kum2 = 1 - (ukm + 200)/500; but, ukm max is 300
Thus, the minimum value for Kum2 is 0 and the highest is 0.6.
When weighing in the importance of urban markets the combined value (Kum1 + Kum2) should be used.
9 The value 1 is included to avoid the expression becoming 0.
Even if all the additional 80 tons can be disposed of in a particular district, it is useful for the farmers if they also have access to urban markets. Therefore, the weight for the district market should be added to the weight for the urban market. The total market weight will be attained by adding the weights for the district and urban markets, and, in order not to exceed a value of “1”, the value should be divided by the maximum possible.
The maximum possible is:
District market.
Either 1.5 or 0.5, thus 1.5
Urban market:
1.0 + 0.6, thus 1.6
Therefore, the maximum for the market as a whole is 3.1. In order to obtain the Weighting for the Markets (KM) in the overall Weighting of the districts, the following formula should be used:
KM = ((Kdm1 or Kdm2) + Kum1 + Kum2)/3.1
In selecting priority districts for fish farm development, the inputs which are immovable, or costly to move, must be taken into account. Those which are equally accessible in all districts, such as labour, fingerlings, capital and chemical fertilizers, need not be considered. The ones that are difficult to move include: water/land, agro-industrial by-products and some manures. water/land is not discussed in this annex. They are handled by other mission members.
The analysis presupposes that inputs must be found within the district. Therefore, costs of transport are not included.
The analysis has considered “single” feed, or fertilizer, strategies, as feeds and fertilizers (manures) can substitute for each other. The analysis is standardized - that is, the author suggests that the two feeds (rice bran, oil palm fibre) and three manures (cattle, pigs, chicken) all be treated in the same fashion. The procedure outlined below for rice bran, should be followed for oil palm fibre, as well as for manures from cattle, pigs and chickens.
The following data is available:
| cp | crop (rice) production/year in mt |
| kcp | factor indicating the share of “cp” (rice produced) that is useable as feed, in this case in the form of rice bran; |
| mnsf | the minimum quantity of the by-product needed to permit the development of the “minimum” fish farm development programme (20 ha of pond surface area) in the district; |
| p | human population; |
With the above data two relationships will be developed: (i) the share that “mnsf” is of the total feed/manure available in the district (shr1); and, (ii) the relationship between the minimum needed for the average fish farmer (assumed to have 1.000 m2 of ponds) and the average availability of the by-product/manure per household in the district (shr2). The first relationship is important from the point of view of justifying public expenditure associated with fish farm development in a district, and the second is important as it assess the extent to which the population in the district will have access to the use of rice bran without having to obtain it from a number of different suppliers.
The relationships are as follows:
shr1 = (cp * kbp)/mnsf
shr2 = ((cp * kbp)/(p/8)) / (mnsf/200)11
The author proposes the following weighting:
| shr1 | Kin1a | shr2 | Kin1b |
| < 0.02 | 1.0 | > 10 | 1.0 |
| 0.02 < 0.10 | 0.9 | 5 < 10 | 0.95 |
| 0.11 < 0.20 | 0.6 | 2 < 5 | 0.9 |
| 0.21 < 0.50 | 0.3 | 1 < 2 | 0.85 |
| 0.51 < | 0.0 | 0.5 < 1 | 0.8 |
| 0.2 < 0.5 | 0.7 | ||
| 0.1 < 0.2 | 0.3 | ||
| < 0.1 | 0.0 | ||
And, Kin1 = Kin1a + Kin1b
With the above system of assigning weights, each alternative by-product and manure will obtain at best a value of (1.0 + 1.0 =) 2. As five inputs are considered the maximum obtainable for any one district is 10.
However, so far the method proposed does not reflect the fact that it is probably more important for a district to be well endowed with one or two inputs than to have a little bit of everything. Therefore, the author suggests that for any points to be given for inputs12, a minimum of points must be obtained for two. Thus, the district should have at least one input for which the value of Kin is at least 0.9 and one for which the value is at least 0.7.
