33 - Livestock development and economic productivity in tsetse-infested Africa

Introduction
Economics
Rural development and trypanosomiasis
Conclusion
References

Introduction

Environmental and economic aspects

To tackle economic aspects of trypanosomiasis means to consider the disease as a constraint to rural development in general. We have to face the great variability in trypanosomiasis situations: economic variability in addition to ecological variability. We also are obliged to consider each region as a particular case.

The disease is not just a medical defect to be remedied at whatever price, but it is an environmental obstacle to the increase of national production and revenue which are the two main criteria of development. These two criteria are not necessarily linked.

A basic function of the livestock sector is to provide sufficient animal protein for the human diet. The question arises whether the forced development of national livestock production at whatever cost is the best approach to meet this goal. If we only regard the role of livestock as a source of marketable products in terms of protein, we have to discuss the benefits of imports using comparative cost advantages instead of stressing self-sufficiency. But livestock production in tropical Africa is of much more importance. We should not overlook its eminent role in subsistence providing consumable and easily marketable goods. And we have to regard the potential of animal traction or other intermediate products for rural development.

Trypanosomiasis certainly is not the only constraint to livestock production and rural development. Other human or animal diseases intervene and scarcity of production factors in general may act as development obstacles, even in tsetse-free or reclaimed areas.

Alternative approaches

The nature of trypanosomiasis permits a wide range of methods and measures of control or eradication" A major approach is to fight the vectors by various methods. Another technique is to use drugs to protect livestock against the pathogens in different challenge situations. The introduction of trypanotolerant cattle, or game exploitation are other alternatives. Each of these approaches have many variations, especially the first one and these variations may be combined in integrated programmes depending upon local conditions. Still combinations are rare and campaigns tend to favour a single method.

Evaluation methods: costs, benefits and risks

Responding to trypanosomiasis requires decisions on a political level. Individual producers can hardly meet the problem by themselves. Even private decisions, for example to administer drugs or to purchase typanotolerant cattle by individual producers depend greatly upon the context created by public policy concerning availability and price of various inputs such as drugs and trypanotolerant cattle.

The alternative approaches to the disease have to be treated as separate projects and subjected to economic analysis. Generally we apply the technique of social cost-benefit analysis in order to estimate the monetary benefits of control or eradication campaigns. The cost-benefit flow is calculated and the internal rate of return (IRR) is estimated. The purpose is to reveal the profitability of any anti-trypanosomiasis approach for the society as a whole. All necessary inputs and outputs must be taken into account, either by market prices or shadow prices. The economic evaluation requires the proper accounting of all costs and benefits involved. Besides direct costs and benefits the indirect ones, which are difficult to agree on and to estimate, have to be taken into account.

Attention has to be given to discount rates and duration periods. Strictly speaking, comparisons of results between different approaches against trypanosomiasis can only be made if they are equivalent or transformed in these two areas.

Each cost-benefit evaluation should consider the distribution of the financial burden of campaigns between different economic agents, e.g. state, producers and consumers. The general assumption is that the state takes all the costs and the producers all the benefits. In some cases, for example in the use of drugs, or targets and screens, producers could be obliged to share the costs. Finally we have to take into consideration whether the costs require local or foreign currency. These aspects are rarely considered in existing studies but are extremely important.

Economics

Here we mainly present data from improved production systems such as ranches or stations, one of the reasons being that there are more data available on these than on traditional systems. Data referring to traditional livestock systems are difficult to obtain. Especially the potential benefits of intermediate livestock products are difficult to agree on and further are difficult to assess for cost-benefit analysis. But we have to keep in mind that 95% of the total meat production is provided by the traditional sector.

Costs of tsetse eradication or control

Cost comparisons between different methods of controlling or eradicating tsetse are difficult due to the extreme variability in ecological conditions already mentioned. The available data are not homogeneous, despite the fact that in 1977 the FAO proposed to standardize their presentation during a specialized meeting about the economics of trypanosomiasis. In particular, only current costs are usually counted, thus reducing evaluations to a large but unknown extent. Tsetse eradicated areas are usually considered as being definitely free from trypanosomiasis; almost no notice is taken of maintenance costs to avoid re-invasion by tsetse flies or other biting flies. Comparisons of costs expressed in different currencies at different dates do not have much significance at first sight. To overcome this difficulty and to obtain values in terms of equivalent US dollars (1987) we have used the following method.

To warrant a constant purchase power in US dollars we converted local currencies into US dollars for the year in which the programme occurred (FAO Trade Yearbooks) and connected the results both by price index of manufactured products exported by developed countries (UN) and by effective exchange rate of the US dollar (International Monetary Fund [IMP]). This method is not entirely appropriate for developing countries in general and has even more limitations for Africa due to the great fluctuations in money convertibility. Additionally it ignores the proportions of costs which are in national and foreign currencies.

Despite these restrictions, our approach made it possible to present comparable data.

Insecticides

Tables 1 - 3 show the costs of tsetse eradication per sq. km. reclaimed using insecticides by ground, helicopter or aircraft spraying in several countries of Africa.

Table 1. Costs of tsetse eradication by helicopter spraying in US dollars 1987, per sq. km reclaimed.

Country	Year	Reference	Cost
Zimbabwe	1983	Hursey and Alsopp	425
Zimbabwe	1986	Hursey et al.	378
Nigeria	1980	Putt et al.	790
Cameroon	around 1980	Cuisance et al.	1923

Most figures range from US$170 to US$790 per sq. km. or expressed as a ratio from 1 to 4.5. Two sources appear to be very different from all others: the one for Cameroon concerning the Adamaoua region (Cuisance et al., 1987) and for Mali (Straw and Kamate, 1981) evaluated, respectively, at US$1900 and US$3400 per sq. km. These differences are due to the duration of the programme for Cameroon (10 years, including annual maintenance and the whole technical staff with expatriate specialists) and for Mali, by taking into account preliminary entomological studies and also the whole technical staff. For these reasons maintenance costs evaluated for Mali from the same source also seem to be extremely high (US$1900 per sq. km. per year).

Table 2. Costs of tsetse eradication by ground spraying in US dollars 1987, per sq. km. reclaimed.

Country	Year	Reference	Cost
Zimbabwe	1983	Hursey and Allsopp	209
Zimbabwe	1984	pers. comm.	235
Zimbabwe	1985	pers. comm.	207
Zimbabwe	1986	Hursey et al.	170
Botswana	1976	FAO. 1977	196
Nigeria	1980	Putt et al.	705
Mali	1981	Shaw and Kamante	3422

Maintenance: US$1914 per year in Mali (Straw and Kamante, 1981) with barriers.

Data for Nigeria (Putt et al., 1980), also including the costs of technical staff, are lower than the previous ones due to the high reclaimed/treated area average ratio.

Finally we can assume that eradication costs by insecticide spraying range from US$170 to US$700 for ground spraying (average 287), from US$380 to US$790 for helicopter spraying (average 530) and from US$180 to US$515 for aircraft spraying (average 405), providing that knowledge of the situation is sufficient to avoid costly preliminary studies and eradication is achieved in a short time (hence excluding data of Cuisance et al., 1987 for Cameroon and Shaw and Kamate, 1981 for Mali).

Table 3. Costs of tsetse eradication by fixed-wing aircraft spraying, in US dollars 1987, per sq. km. reclaimed.

Country	Year	Reference	Cost
Cote d'Ivoire	1978	GTZ/IEMVT	464
Botswana	1976	FAO. 1977	183
Zimbabwe	1984	pers. comm.	515
Zimbabwe	1985	pers. comm.	457

As expected ground spraying seems less expensive than helicopter spraying; the first also reduces the costs in foreign currencies; the share of foreign currencies for ground spraying is estimated to be 20% versus 90% for helicopter spraying. Additionally ground spraying needs less insecticide and thus reduces pollution.

Traps and screens

Table 4 presents the costs of "initial" eradication using traps and/or screens impregnated with insecticides in different situations. In Zimbabwe they are applied against savanna flies and in Burkina Faso against gallery species. In Burkina Faso the method has been associated with the use of the sterile male technique.

Table 4. Costs of tsetse eradication by impregnated traps and screens, in US dollars 1987, per sq. km reclaimed.

Country	Year	Reference	Cost
Burkina Faso	1983	Politzar et al.	39
Zimbabwe	1986	Hursey et al.	345
Cote d'Ivoire	1983	Kupper et al.	54

Maintenance: US$19 per year in Burkina Faso (Politzer et al., In prep.)

Costs are very different in the two tsetse situations, according to the greater surface reclaimed per trap or screen in gallery cases. For Cote d'Ivoire costs are much lower by the use of traps than by insecticide spraying, whereas for Zimbabwe they are roughly the same.

The use of traps represents a satisfactory method because of its cheapness and flexibility; in addition it creates less pollution than insecticides and induces more national added value due to the greater part of inputs of national origin. Its efficiency is comparable with ground spraying but traps cannot be set in rough terrain.

For these reasons traps should be taken into consideration wherever their effectiveness is assumed to be equal or better than that of insecticide spraying. But the use of traps needs available labour and is faced with social constraints that do not occur in other eradication programs. The local population must be aware of the presence and purpose of targets to avoid theft, especially when their participation in maintenance is planned.

Other methods

Sterile insect technique (SIT): This technique has been used since 1980 in Burkina Faso, jointly with insecticide-impregnated traps and screens (Brand!, 1986). The total costs per sq. km. for five years including infrastructure, foreign experts, establishment and maintenance of trap barriers during two years amount to US$626 (1987). It seems expensive but is comparable with most of the figures of insecticide spraying due to the high reclaimed area (mainly riverine glossina). Additionally SIT is the sole non-pollutant control method.

Cattle as "live baits": the death of glossina which had bitten cattle has been reported when cattle were treated against ticks in dipping-tanks (Hursey et al., 1986) or after systemic administration of invermectin (Distelmans et al., 1983). The current dipping costs in Zimbabwe have been reported to be about US$17 per year per sq. km.

Vaccination: The development of a vaccine has been actively investigated, particularly by ILRAD. At present this does not seem possible, but any new research proposals should be supported.

The two latter methods are not yet effective enough to have been used in large-scale eradication schemes. They have not generated economic data that would permit a comparison with the preceding techniques.

Comparisons

Though it is hazardous to compare data obtained in different conditions, it seems that generally the most costly methods for short-term eradication are insecticides (mainly by helicopter-spraying) and sterile males and the least costly are traps and screens (whose attractant power is multiplied by 4 to 8 when odour-baited).

A comparison between different methods, based on either field data or prospective evaluations, has recently been made by Brandl (1986) for Burkina Faso. Costs of the SIT increases with the duration period and with the number of glossina species to eradicate. They decrease relative to the size of the eradication area. Costs of traps increase with the period of time and decrease with the size of area. Traps are better than helicopters for periods lower than 15 years, but become more expensive if the duration is longer.

The ranking of costs of the different methods changes depending on the period of time needed to achieve eradication. These comparisons will be recapitulated and illustrated further, including evaluation of drug alternatives.

Economics of livestock production in reclaimed areas

It is still difficult to define the tsetse or trypanosomiasis challenge by a standardized and fully acceptable variable. At present descriptions range from "low" to "high" or "medium", which are imprecise terms to define disease pressure.

As demonstrated most of the cost studies are expressed in terms of "eradication" and not of "control". Eradication seems to be the only incontestable and common criteria. Furthermore, trypanosomiasis challenge is not the only productivity constraint to be taken into account.

Table 5. Cost-Benefit Analysis for Tsetse Eradication.

Country	Situation and cattle breeds	Method of eradication	Date	Reference	Duration (years)	Rate of discount	Results (B/C)
Mali	Zebu, N'Dama and crossbreds*	Ground spraying plus barriers	1981	26	20	12%	0.70 -0.99
Ethiopia	Zebu*	Insecticide ground spraying	1982	13	15	10%	1.44
Ethiopia	Zebu*	Game reduction	1982	13	15	10%	0.98
Nigeria (Sokwa district)	Mainly zebu**	Ground spraying	1980	22	20	12%	2.66 -7.97
Nigeria (Burra district)	Mainly zebu**	Ground spraying	1980	22	20	12%	4.67 -6.51
Cameroon (Adamaoua)	Zebu**	Helicopter "praying	1980	5	12	7.3%	1

*Ed sate analysis
**Ex post analysis

Table 5 shows the results of cost-benefit analyses conducted in three different situations on the basis of two eradication methods (game reduction and insecticide spraying). Game reduction, which is technically difficult and ecologically questionable, gives a benefit/cost ratio (B/CR) below 1.

Among the spraying methods we found values around one for Mali, Ethiopia and Cameroon. Data for Ethiopia derive from an appropriate model and thus are not as significant as the others which were derived from more detailed field studies. For Cameroon the value of one is obtained with a lower discount rate than elsewhere, meaning that the result would have been inferior to one using the same rates as other studies; this is due to the long duration of the programme and to the fact that the method of helicopter spraying is by far the most costly eradication method as seen in Table 1.

Higher figures have been obtained for two districts in Nigeria. This can be explained by local differences: in Nigeria there is high pressure on the land due to the rapidly expanding human population, while the cattle are virtually all highly productive, non-tolerant breeds. Therefore the benefits of tsetse eradication are high and the B/CR is good.

Table 2 showed that in Mali the costs were much higher and the benefits lower than in Nigeria. This is probably due to the lower pressure on land and the presence of some trypanotolerant cattle. Therefore the increase in production after tsetse eradication is not as high as it would be if all cattle were non-tolerant. As expected the B/CR is lower.

These figures (Table 5) show that tsetse eradication may not be economically favourable depending on local situations. The economic results of a campaign depend on the costs caused by the different methods of eradication, but above all on the relative importance of trypanosomiasis as a constraint on livestock and rural development. Unfortunately the method using traps and screens is too new to have any cost-benefit analysis.

In terms of income it has been shown in Northern Cameroon (Cuisance et al., 1987) that the net difference in income per head of cattle, with or without trypanosomiasis, amounts to US$7 (1987). A difference of US$10 (1987) per head of cattle reared warrants profitability at a 15% discount rate during a twenty-year eradication and maintenance programme.

The use of a more detailed herd simulation model covering a ten year period for Burkina Faso (example Brandl, 1985) results in an income difference per head of cattle ranging from US$42 to USS252 (1987) in current values for the ten year period, depending on three benefit scenarios for cattle productivity parameters. These figures give a difference in income per head per year from US$4.2 to US$25.2 (1987), comparable to the Cameroon data. We do not have such data for Zimbabwe. But we assume that the benefits are high, due both to the high productivity of improved breeds found in Zimbabwe and the high quality of the meat produced. Since eradication costs are relatively low (Tables 1 -3), they would result in a good B/CR.

Economics of production in tsetse infested areas

Livestock production in infested areas is possible by the use of chemical protection among non-tolerant cattle, or the use of trypanotolerant breeds which need no treatment except in cases of high trypanosomiasis pressure. Game exploitation by hunting or farming (Gruvel, 1980) may also be a solution that has to be evaluated.

Livestock production with the aid of trypanocidal drugs

Drugs are used in either therapeutic or prophylactic schemes. They are applied to non-tolerant cattle which may reach infested areas during the dry season in pastoral systems. They also are used to maintain on a permanent basis either trypanotolerant breeds (to avoid mortality among young cattle) and sometimes non-trypanotolerant herds in infested areas. The benefits of drugs are easier to calculate because the maintenance of newly reclaimed areas is not necessary. The costs are essentially those of the treatments. Table 6 shows four examples.

Table 6. Costs of chemical treatments in US dollars, 1987. P = Prophylaxis T = Treatment

Country	Method	Year	Reference	Cost
Botswana	P	1977	FAO. 1977	4.5*
Mali	P	1979	Shaw, Kamante	1.3*
Mali	P	1978	Logan et al.	1.9**
Mali	T	1978	Logan et al.	2.0**

*Per treatment.
**Per head of cattle (zebu-300 kg) per year.

The first figures on Mali and Botswana refer to prophylactic schemes. They are expressed in equivalent US dollars (1987) for one treatment and include all inputs needed for drug administration. The other figures of Mali either in prophylactic or therapeutic schemes are expressed per animal per year. These figures only include the direct costs of drugs. Although the costs of treatment and prophylaxis may differ greatly from place to place, they are comparable when computed in the same conditions. Comparing the costs per animal per year with productivity data from the same source (here zebu cattle in Mali shown in Figure 1) demonstrates that prophylaxis is economically more favourable than treatment in high tsetse-challenge conditions. In addition prophylaxis is used for all cattle at the same time whereas chemotherapy is applied individually, resulting in higher indirect costs.

Figure 1. Average weight gain for two groups of Zebu Maure cattle in Mali from May 1978 to June 1979, treated either by Trypamidium or by Berenil.

In addition data from Kenya shown in Table 7, presented in cost-benefits terms, confirm that prophylactic schemes in cases of permanent risk of trypanosomiasis is preferable to trypanocidal treatments. The B/CRs for both prophylactic schemes are higher than the one for therapeutic drug use. These enormous B/CRs are not comparable with those mentioned for eradication since they concern a specific ranching situation with high trypanosomiasis challenge and no eradication alternative. Unfortunately they have not been presented in monetary terms. But Trail et al. (1984) published detailed data for a comparable ranching situation in Tanzania. They confirmed that in the management situation of the Mkwaja ranch with permanent high risk of trypanosomiasis, non-tolerant Boran cattle husbandry can be successfully maintained by chemoprophylaxis.

Table 7. Estimated cost-benefit analysis for three prophylactic or therapeutic schemes in a ranch population of zebus at trypanosomiasis risk in Kenya.

Drug used	Scheme	Benefit/Cost (US$)
Samorin	Prophylaxis	20-42
Prothidium	Prophylaxis	5-20
Berenil	Treatment	2

Duration: one year
Source: Wilson et al. (1981)

Cuisance et al. (1987) emphasize that after several years of chemoprophylaxis the number of injections per head per year increases from 2-3 to 9-12 due to the increased tolerance of trypanosomes (e.g. Peov ranch in Samorogouan, Burkina Faso). This indicates that the cost of prophylaxis schemes cannot be regarded as constant.

In Mali, Shaw and Kamate (1981) compared different drug application schemes with different eradication methods. This is the sole case where such data are available (Table 8). Zone 1 represents the zone covered by the study, while Kati is a district of the zone. Maintaining trypanocide drugs, Shaw and Kamate (1981) calculated B/CRs superior to 1. Prophylaxis to meet seasonal tsetse challenge during the rainy season and eradication schemes proved to be of no economic interest. These figures differ from those already mentioned but this simply reflects the particular conditions of trypanosomiasis challenge and its seasonal occurrence in zone 1 of Mali. The relative merits of prophylaxis and treatments thus depends on particular conditions.

Table 8. Benefit/cost ratios for different strategies for dealing with trypanosomiasis (Mali, 1981).

Strategy	B/C with inflated benefits	B/C at a low benefit level
Treatment for all zone 1	1.36	1.00
Prophylaxis in the rainy season for all of zone 1	0.66	0.48
Prophylaxis in KATI, treatment elsewhere	0.90	0.66
Prophylaxis of oxen, treatment of other cattle	1.24	0.91
Eradication at high cost	0.70	0.51
Eradication at low cost	0.99	0.72

Source: Shaw and Kamate (1981).

Drug utilization may also face uncertainty in its results. The threat associated with tsetse eradication is re-invasion, while drug use may meet chemical resistance of trypanosomes, especially under deficient management conditions. But these risks are difficult to assess economically.

Comparison between different methods of eradication and the use of drugs

The discounted costs of different methods for controlling trypanosomiasis have been compared by Brandi (1986) for Burkina Faso (Figure 2). Referring to actual costs in Burkina Faso, Brandl simulated a model varying the eradication area, calculation periods and benefit scenarios. In his results he stated that the application of trypanocides in a low tsetse-challenge situation is the most economical way to meet the problem for whatever period ranging from 5 to 20 years. In the case of medium tsetse challenge they are outclassed by traps over a five-year period, by traps and helicopter spraying between five and ten years and by all the methods, including SIT as the most costly for 15 to 20 years. These data emphasize the obvious fact that the more difficult the eradication scheme and the longer it lasts the better the use of drugs appears to be. The time dimension and the level of tsetse challenge are essential concerns in this approach.

Figure 2. Relative Costs of Different Methods to Control Trypanosomiasis with Reference to the Time period (Sideradougou, Burkina Faso).

Furthermore the simulation is very sensitive to the size of area as shown in Figure 2.1. Enlarging the area to 20,000 km² and considering different tsetse challenge situations, eradication methods perform better than drug application over a ten-year period. Helicopter spraying appears to be the least costly. But again its advantage should be weighted by its environmental effects. Expanding the ten-year period by SIT theoretically becomes less costly than traps, but SIT has not been meant to be used over long periods.

Figure 2.1. Total costs of Different Methods to control Gallery Forest Species and Trypanosomiasis in an Area of 20,000km² (Burkina Faso, 1984).

Figure 3, 3.1 and 3.2 display the benefit-cost data for the previous areas (3,500, 10,000 and 20,000 km²) considering two calculation periods (10 or 20 years) and three productivity scenarios (H1, H2, H3). Figure 3 includes the B/CRs for the use of drugs in "low", "medium" and "high" tsetse-challenge situations. Drugs appear to be only justifiable at a low tsetse challenge.

Figure 3. Benefit-Cost Ratios of Tsetse Control considering different calculation periods and benefit scenarios (H1, H2, H3) in the zone of Sideradougou, Burkina Faso (rate of discount i = 10% control area = 3500 km).

Comparing helicopter and traps, the first is better only for 20 years when the area is 3,500 or 10,000 km² and becomes better whatever the calculation period for 20,000 km² The surface and time effects on B/CRs between the two techniques are perhaps due to the greater marginal costs of traps when the size of the area increases (for the surface effect) and as pointed out above, due to their more regular needs in inputs over time.

Figure 3.1. Benefit-Cost Ratios of Tsetse Control considering different calculation periods and Benefit Scenarios (H1, H2, H3) for an area of 10,000 km² (projected into Burkina Faso conditions; rate of discount i=10%).

Figure 3.2. Benefit-Cost ratios of Tsetse Control considering different calculation periods and Benefit Scenarios (H1, H2, H3) for an area of 20,000 km² (projected into Burkina Faso conditions; rate of discount i=10%).

The SIT appears to be less economical than other methods but it should be weighted by the following criteria. It does not create environmental problems like insecticides do and it does not create acceptability problems like traps do. Additionally the use of a discount rate penalizes the SIT since its expenses are mainly spent before or at the very beginning of a programme. Unlike the other methods, the fixed costs of the SIT should be put on a large regional scale.

As mentioned before programmes generally tend to favour a single method. But combinations of different methods adequate to local conditions, may be successfully implemented. An integrated project based on cattle dipping, trypanocidal drugs and traps and screens has been attempted in Zimbabwe.

Introduction of trypanotolerant livestock

The introduction of trypanotolerant livestock in areas with high tsetse challenge or incidence is often regarded as the only economically relevant approach, especially in areas where the disease has hindered cattle husbandry until today. A lot of work has been done by different research programs and by the ATLN. The data are not always homogeneous, but it is nevertheless possible to point out global outlines and recommendations for the use of trypanotolerant cattle.

It is quite interesting to compare physical productivity data according to different breeds and trypanosomiasis challenge situations. The breeds may be either pure African ones such as East African Zebu, N'Dama and African Shorthorn or crossbreeds between African and European breeds on which research programs have been conducted to improve productivity.

Table 9. Global comparison of productivity of Boran cattle in Kenya and Tanzania and of N'Dama breeds in Western and Central Africa (ranching production systems).

Productivity	Mkwaja Boran high risk chemoprophylaxis	Kenya Boran no risk	N'Dama, middle or high risk, no prophylaxis
Weight of 8-month old calf/100 kg cow/year	29.8	33.90	22.70

Source: Trail et al., 1986.

Table 9 presents the comparison of productivity data between Boran (Zebu) cattle at Mkwaja ranch (Tanzania) with high risk and chemoprophylaxis, in Kenya without risk and N'Dama in middle-or high-risk situations without prophylaxis. All three refer to ranching conditions.

These figures show that the productivity of Boran cattle in high-risk conditions with chemoprophylaxis may be closer to that of Boran cattle without risk than that of N'Dama managed in middle - or high-risk conditions. This fact seems to indicate that Zebu ranching using chemoprophylaxis is preferable wherever it is technically feasible. But we have to balance this result considering the extremely favourable organizational conditions in Mkwaja Ranch (Trail et al., 1984) which are rarely achievable under village conditions. Regarding Table 10 we come to the following conclusions. The productivity of Zebu is the highest among the three breeds studied without risk of trypanosomiasis.

On the other hand they are very sensitive in risk conditions. Their productivity decreases rapidly with increasing tsetse challenge. The productivity of Shorthorn exceeds that of Zebu in middle-risk situations showing that the genetic potential of Shorthorn cattle should not be underestimated. Some' attempts have been made to cross African trypanotolerant cattle such as N'Dama with improved European breeds in order to increase productivity under the constraint of trypanosomiasis (De Rochemonteix, 1984; GTZ, 1982).

Table 10. Productivity index for trypanotolerant cattle and Zebu in four situations of low, middle or high trypanosomiasis challenge.^a

	N'Dama	Shorthorn	Zebu
Nigeria - no risk/station	47.8	50.3	53.0
Burkina Faso - low risk/village		28.7	20.7
Cote d'Ivoire - low risk/village		17.6	19.5
Central African Republic - low risk/village		26.3	17.8

^a Weight of one-year old calf plus milk equivalent per 100 kg cow weight per year.

Source: Hoste (1987).

Table 11. Productivity rates according to breed and crossbreed in kg of one-year calf/100 kg cow low trypanosomiasis risk. Station data from Avetonou (Togo).

		Male	N'Dama	loc. Breed	N'Dama	75 Crossbreeds	Zebu	Zebu
Genotype	Parents	Female	N'Dama	loc. Breed	50 Crossbreeds	75 Crossbreeds	N'Dama	Zebu
Genotype	Calves		N'Dama	loc. Breed	75 Crossbreeds	Avetonou	F1	Zebu
Index: kg 1 year calf per 100 kg cow/year			31.5	32.9	33.5	34.7	37.3	50.3

Source: GTZ (1982)
50 Crossbreeds = Local breed x German Yellow
75 Crossbreeds = 50 Crossbreed x N'Dama
"Avetonou" = 75 Crossbreed x 75 Crossbreed

Table 11 shows results of crossbreeding between N'Dama and Zebu or N'Dama and different percentages of "German Yellow" compared with N'Dama, local breeds and Zebu. The data are obtained in the research station of Avetonou (Togo) where trypanosomiasis risk is reported to be "low". The Zebu are treated against the disease, which explains why their productivity is the highest. Crossbreeds perform better than pure breeds excluding Zebu, N'Dama being the lowest. However the authors (GTZ, 1982) stated that the differences are not significant in terms of development because of the high sensitivity of improved crossbreeds to changes in environmental conditions. Furthermore N'Dama performs the best when index calculation is changed to consider the increase in weights and the lower losses of calves until they reach the age of two years. This fact reflects a higher survival rate after one year. Taking all into account it appears that the performance of trypanotolerant cattle compared with Zebus is favourable even when there is not trypanosomiasis risk. And in addition they represent the sole possibility of cattle husbandry as soon as the risk increases even a little.

Table 12 demonstrates the annual deficit of income per head caused by trypanosomiasis after modelling the evaluation without treatment administration. The tsetse challenge is considered as being "medium".

Table 12. Evaluation of income and losses due to trypanosomiasis among different cattle breeds in northern Cote d'Ivoire, in equivalent US dollars, 1987.

		Baoule	N'Dama	N'Dama x Baoule	Zebu x Baoule
Annual income per head
	Infected	21.47	22.18	24.39	20.62
	Uninfected	22.17	25.46	27.54	27.74
Mean annual loss per head		0.30	1.95	2.07	3.46

Source: Camus (1981)

Again the crossbreeds including Zebu (here Zebu x Baoule) turn out to be the most profitable breeds when not infected by trypanosomiasis. But we have to say that the generated income per head is very sensitive to infection. In infected herds the highest income per head is reached by N'Dama x Baoule breed. Unfortunately we have no conversion indices to refer to the previous productivity index, but we assume that the order among the different breeds remains the same or even is reinforced by conversion. Again the fact is strengthened that beyond a certain level of trypanosomiasis challenge, only trypanotolerant cattle can be economically raised. The introduction of trypanotolerant breeds has been tried for a long time especially in regions where the occurrence of trypanosomiasis had previously prevented almost all attempts to introduce livestock production (Central African Republic). In these cases trypanotolerant cattle are the only technically feasible approach. Table 13 shows the total costs of introduction of trypanotolerant cattle in US dollars (1987) per head ranging from US$670 to almost US$1400.

Table 13. Costs of the introduction of trypanotolerant cattle in various cases in US dollars 1987 per head of cattle.

Origin	Destination	Breed	Year	Reference	Cost
Cote d'Ivoire	C.A.Rep.	Baoule	1964	Lacrouts et al.	1245
Senegal	Gabon	N'Dama	1985	Shaw	1387
Senegal	Congo	N'Dama	1970	Sarniguet et al.	672*
					775**

*cows
**bulls

Despite these high costs the IRR indicates that ranching (Table 14) of trypanotolerant livestock in high challenge conditions is profitable. The rates of return are not outstanding and the calculation period is long. The IRR for Gabon calculated without expatriate management costs would be even lower if they were included.

There is hardly any data about the introduction of trypanotolerant cattle into traditional farming systems except for the Central African Republic in 1967. A rough calculation resulted in a financial capital appreciation of 11.5% for 20 years with comparable introduction costs to those shown in Table 14.

Table 14. Ex Ante estimated economic IRR for N'Dama cattle ranching.

Country	Management	Date	Reference	IRR	Period (years)
Gabon	Trypanocidal drugs once a year or more	1987	Sauveroche et al.	8.2*	40
Guinea	Treatment if necessary	1980	Sacfinto and VanLancker	6-9	25
Congo	Tsetse eradication and maintenance	1971	Sarniguet et al.	12.7	25

*including expatriate staff

Exploitation of wild fauna

Game reduction has been presented as a means of reducing the tsetse reservoir in order to introduce domestic species. Some authors ask if game exploitation in zones with high risk or difficult environmental conditions would be an appropriate alternative (Gruvel, 1980). Productivity data of wild species seems to bear comparison with those of domestic cattle, but the scarcity of detailed economic studies does not enable us to conclude or discuss their relative advantages. Furthermore wild game does not substitute the social function of cattle.

Rural development and trypanosomiasis

Before concluding and after having evaluated different methods of responding to trypanosomiasis, we focus on possible interactions between tsetse-infested and tsetse-free areas.

Do tsetse-infested areas constitute an obstacle for the development of the tsetse-free areas?

We all know -the threat of tsetse re-invasion of formerly cleared areas, especially when cattle or wild life move across the border between free and infested areas. This may particularly occur when transhumant cattle are moving between dry season and wet season pastures. Tsetse-infested areas are also responsible for increased population densities in tsetse-free areas and degradation of arable land due to overpressure and overgrazing with all the associated environmental problems. The question therefore arises whether these problems justify eradication campaigns.

In general if an area is in danger of tsetse re-invasion, not just random contacts, the question arises whether the population density and intensity of land use is appropriate enough to destroy the tsetse habitats permanently. Otherwise an eradication attempt would only enlarge the total endangered area. Enlarging the area to release population pressure has never been a creative solution. The improvement of farming systems and an increase of intensity are more economical.

We stated before that to tackle trypanosomiasis by any of the numerous direct and indirect methods may not be obviously justifiable economically in each case. But the first step before any decision is taken has to be a careful and complete analysis of all alternatives and benefits to rural development (FAO, 1987). Tsetse-infested areas can definitely constitute a potential for economic development, as some have already proved. Only studies dealing with the development problem as a whole can fully answer the question.

Conclusion

There is an urgent need to introduce economic concerns at each stage of reflection or research about trypanosomiasis.

Economic data are not only rare, but also lack homogeneity to allow comparisons between the costs and benefits involved in trypanosomiasis programs. Already as early as 1977 the FAO had proposed a standardized presentation for the costs of an eradication campaign and introduced a framework for provisional studies concerning land utilization. But these have rarely been taken into consideration and the need for comparable studies still remains. As methodological issues have been mentioned in many former meetings we do not want to repeat them once more. There is a definite need for global agreement about a standardized economic framework, for example that suggested by FAO, to obtain the homogeneity in the field economics, which the ATLN has achieved in the zootechnical field.

In the context of scarcity of public funds, it seems particularly relevant to itemize local or foreign currencies among the expenses, as this is a major criterion in economic decisions. It also seems necessary for economic purposes to define more precisely what is known as "trypanosomiasis challenge" by a continuous, standardized index with ecological and epidemiological components (abundance of tsetse, their infestation and probability of cattle being infected). This index should be related to productivity measurements and thus enable us to finally evaluate ex ante the benefits of any strategy. Such an index would make more precise the selection of a method or a combination of methods replacing the rather empirical practice which prevails at this time. ILCA's current research will provide fundamental background data to define and establish this index.

In addition, training programmes which already exist must be extended to enable African authorities to manage or participate in the research or studies which must be carried out and in any public and large-scale decisions made on the disease.

To perpetuate the success of any programmes, particularly those which aim at avoiding tsetse re-invasion or resistance of trypanosomes to insecticides, depends mainly on the participation of all social and economic groups of the rural population. Advertising campaigns must not be neglected.

Finally work has to be done to agree on and to estimate indirect costs and benefits in order to focus precisely on other alternatives to combatting trypanosomiasis. The great potential of livestock production for rural development is reflected in its important role in subsistence production, its income function and its potential for generating intermediate products such as manure and animal traction. It definitely should not be evaluated only as a means of providing protein of high quality.

References

Brandl, F.E. 1985. The use of a herd simulation model for the estimation of direct economic benefits of tsetse control. Application to the pastoral zone of Sideradougou, Burkina Faso. Rev. Elev. Med. vet. Pays trop. 38: 364-370.

Brandl, F.E. 1986. Beurteilung verschiedener Verfahren zur Bekampfung von Trypanosomiasis und Tsetsefliegen. Schlussbericht einer begleitenden Evaluierung des Projektes Biologische Tsetsebekampfung, Burkina Faso. GTZ GmbH Eschborn/Taunus.

Camus, E. 1981. Evaluation economique des pertes provoquees par la trypanosome sur quatre types genetiques bovine dans le nord de la Cote-d'Ivoire. Rev. Elev. Med. vet. Pays crop. 34: 297-300.

Cuisance, D., G. Tacher, N. Clair, Ndoki and Muller. 1987. Republique du Cameroun. Lutte contre les glossines: bilan et perspectives. BDPA, Paris.

De Rochemonteix, J. 1984. Cinq annees d'experimentation sur le croisement du betail N'Dama x Pie Rouge de l'Est (N'Damance). SODEPRA, Abidjan. Cote d'Ivoire.

Distelmans, W., F. d'Haseleer and J. Mortelmans. 1983. Efficacy of systemic administration of invermectin against tsetse flies. (Efficacite de l' administration systemique d'invermectine contre les glossines). Ann. Soc. beige Med. trop. 63: 119-125.

FAO. 1977. Consultation FAO/PNUD sur les repercussions economiques de la trypanosomiase. FAO, Rome.

FAO. 1987. Evaluation du programme d'action pour la lutte contre la trypanosomiase africaine et la mise en valeur des zones en cause. FAO, Rome.

Gruvel, J. 1980. Glossines, Betail domestique et faune sauvage, une conciliation est-elle possible? Acta Zoolog. Patholog. Antwerp. 75: 29-48.

GTZ. 1982. Trypanotolerance - recherche et application; Seminaire de Lome et Avetonou (Togo).

GTZ/IEMVT. 1979. Evaluation des projets du lutte contre les glossines et les trypanosomes.

Habtemariam, T., R.E. Howitt, R. Ruffanner and H-P. Riemann. 1983. The benefit-cost analysis of alternative strategies for the control of Bovine trypanosomiasis in Ethiopia. Prev. Vet. Med. 1: 157-168.

Hoste, Ch. M. 1987. Elevage et trypanosomiase animale Africaine. These de doctoral d'Etat es sciences naturelles. Universite Pierre et Marie Curie, Paris.

Hursey, B.S. and R. Allsopp. 1982. The eradication of tsetse flies (Glossina spp.) from Western Zimbabwe by integrated aerial and ground spraying. Tsetse and Trypanosomiasis Control Branch, Department of Veterinary Services, Zimbabwe.

Hursey, B.S., G.W. Whittingham and V. Chadenga. 1986. The integration of insecticidal techniques for the control and eradication of Glossina morsitans in North East Zimbabwe. OAU/IBAR, Nairobi.

Kupper, W. A. Manno, A. Douati and S. Koulibali. 1984. Impact des pieges biconiques impregnes sur les populations de Glossina palpalis gambiensis et Glossina tachinoides: Resultant d'une campagne de lutte a grande echelle contre la trypanosomose animale au Nord de la Cote-d'Ivoire. Rev. Elev. Med. vet. Pays trop. 37: 176-185.

Lacrouts, M., J. Sarniguet and F. Tyc. 1967. Le cheptel bovin de la Republique Centrafricaine. Production - Commercialisation - Perspectives d'avenir. Secretariat d'Etat aux Affaires Etrangeres, charge de la Cooperation. Paris.

Logan, L.L., J.T. Goodwin and S. Tembleys. 1979. Observations on maintaining zebu maure cattle in a tsetse infested area of Mali. 16e reunion, Yaounde (Cameroun). OUA/CSITRC, 3 241-250.

Molyneux, D.H. 1982. Trypanosomes, trypanosomiasis and tsetse control: impact on wildlife and its conservation. Symp. Zool. Soc. London. 50: 29-55.

Politzar, H., D. Cuisance, I. Tamboura and Ch. Merot. Cout de l'emploi de barrieres de pieges et d'ecrans insecticides pour la protection de la zone pastorale d'accueil de Sideradougou (Burkina Faso). A paraitre.

Putt, S.N.H., A.P.N. Shaw, R.W. Matthewman, D.M. Bourn, M. Underwood, A.D. James, M.J. Hallam and P.R. Ellis. 1980. The social and economic implications of trypanosomiasis control. A study of its impact on livestock production and rural development on northern Nigeria. University of Reading, Study No. 25.

Sarniguet, J. et al. 1971. Projet de ranch d'elevage de bovine dans la plaine de Dihesse, Sedes, Paris.

Sauveroche, B. 1987. Le ranch de la Nyonga (Gabon). Introduction de l'elevage bovin dans un pays d'Afrique Equatoriale. Memoire de Dess, IEMVT, Maisons-Alfort.

Shaw, A. 1986. Economic aspects of the trade and multiplication of trypanotolerant cattle. In: FAO, programme de lutte contre la trypanosomiase animale africaine. Compte rendu de reunions a Bamako (1985).

Shaw, A.P.M. and C. Kamate. 1981. An economic evaluation of the trypanosomiasis problem in the zone one. Chemonics, International Consulting Division, Washington, USA.

Socfinto and van Lancker. 1980. Programme d'Agro-Industrie et d'elevage en R.P.R. de Guinee. III: Programme d'elevage. Bruxelles, Belgique.

Tacher, G. and D. Planchenault. 1981. Le ranch de selection de Madina Diassa: evaluation ex post et perspectives d'avenir. IEMVT, Maisons-Alfort (France).

Trail, J.C.M., M. Murray, R. Sones, J.M.C. Jibbo, J. Durkin and D. Light. 1984. Chemoprophylaxis: Boran cattle can be productive under high tsetse challenge. Rev. Elev. Med. vet. Pays trop. 37: 270-275.

Trail, J.C.M., K. Sones, J.M.C. Jibbo, J. Durkin, D.E. Light and M. Murray. 1986. Productivite de Bovins Boran Proteges par chimoprophylaxie contre la Trypanosomiase. Rapport de recherche no. 9: Addis Abeba, Ethiopie: CIPEA.

Wilson, A.J., A.R. Njogu, G. Gatuta, P. Mgutus, H. Alushula and R. Dolan. 1981. An economic study on the use of chemotherapy to control trypanosomiasis in cattle on Galana ranch, Kenya. OAU/CSRTC, 112: 306-317. 17th meeting, Arusha (Tanzania).