K.J.R. MacLennan
Previous issues of World Animal Review have contained articles on African animal trypanosomiasis. This article concentrates more especially on problems of tsetse fly infestation and the techniques used in the control or eradication of the fly. Part I in this issue covers the former aspects and Part II (to be included in the next issue) will cover the latter.
Primarily the rationale for action against tsetse infestation reflects the increasing need to develop the rural economy in the affected areas and the impact and constraints which trypanosomiasis has on the utilization of land resources.
Looked at in general terms the significance of the tsetse-transmitted trypanosomiasis problem derives from the necessity for most African governments to provide an adequate food supply for a rapidly increasing human population that is requiring a better quality diet. Most governments are therefore seeking to increase overall food production and also to increase the availability of animal protein, which in many densely populated areas and particularly in the humid zone is notably deficient. Rapidly increasing human populations are evident in most countries because the more general epidemic diseases have either been controlled with varying effectiveness or may have been eliminated, e.g., smallpox, and remedies are available for many of the debilitating endemic diseases (though often not on an adequate scale). Birth rates are generally high, since infant mortality rates are declining, life expectancy has increased and, although civil strife and war are still, unfortunately, occasional features of grave local import they do not account for significant reductions in overall human population growth. Their economic consequences, however, have been severe in the short term and they are a significant impediment to concerted action against tsetse populations in several localities.
Looking to the future it seems reasonable to take a general 10-year (and preferably 20-year) perspective of land resource requirements in relation to the need to provide a rapidly expanding human population with a diet adequate in quantity and improved in quality. Furthermore, developing economies require an ever-increasing urbanization to serve both industrial and governmental activities and this, in several lightly populated countries (e.g., Zambia, with nearly 30 percent of the population urbanized — MacLennan, 1975) through depletion of traditional food-producing manpower, can give rise to situations where remaining food producers are no longer able to satisfy the needs of the community.
In other circumstances it is the production of commodities such as coffee, rubber, cocoa, palm oil, groundnuts, cotton and sugar for local usage and export that is competing for rural manpower and land resources. Such commodities are often of prime importance to national economies. There is no doubt that the increasing arable utilization of the land resource for some of these objectives is bringing about a diminution in the extent of relatively safe pasturage for dry season transhumance and ruminant production.
For various conservation objectives discussed later it is necessary to take a perspective longer than 20 years of future land requirements. Thus the need for dietary improvement and the facts of population growth combine to exert a multiplying effect on food production targets while cash crop and conservation objectives have also to be accommodated. Since the available land resource is mostly tsetse infested (Ford and Katondo, 1973) the significance of the tsetse problem will escalate sharply rather than diminish with time.
Since historic times tsetse infestation has been a fundamental factor in the interaction between man and his environment, having a basic and deleterious impact on rural and national economies. For a variety of reasons the impact has become more pronounced since the advent of European influences affected both the efficiency of current production and the development of underutilized land resources. Moreover, this factor has had major repercussions on conservation objectives and landuse planning.
The incidence and severity of trypanosomiasis in man and livestock are closely related to the species of Glossina present in the area. The flies can be classified taxonomically in three major groups; the vegetative communities usually, though not exclusively, preferred by the majority of species in each group correspond to the riverine, forest or savanna. It must be emphasized, however, that in highrainfall areas (including the subhumid zone) circumstances may be such that it may be quite usual for riverine tsetse (Baldry, 1964, 1966a and b, 1968, 1969a and b, and Page, 1959) not to be dependent on riparian vegetation and to be found in woodland up to about 6.5 km away while some members of the forest group are exceptional in their independence of forest. However, G. fusca is occasionally present in the subhumid zone in forest relicts in Nigeria and more extensively in the countries west of 3?E (Ford and Katondo, 1973 and U. Spielberg, 1978 — personal communication). G. medicorum is also found extensively in forest relicts of the zone in Ghana, the Ivory Coast and Upper Volta (Ford and Katondo, 1973 and Unesco, 1959). The usual classification of the tsetse subgenera is given in the table, which indicates the main habitat preferences. The geographical distribution has been mapped by the Scientific and Technical Research Commission of the Organization of African Unity (OAU).
Feeding preferences of tsetse flies, trypanosome infection rates, the virulence of transmitted infections and relationships to vegetative habitats vary widely between the subgenera, the species within the subgenus and even within the species, from one locality to another. Recent knowledge of these matters is summarized in articles referred to at the end of the article by MacLennan (1974).
In general terms (there are notable exceptions to practically any general statement about tsetse flies) it can be said that all tsetse communities harbour trypanosomes pathogenic to susceptible livestock. The actual infection rate in livestock at risk varies within very wide limits. Very low levels of infestation by solely riverine species (Jordan, 1961 and Page, 1959) give rise to a trypanosomiasis situation, even in highly susceptible exotic breeds, which can be relatively easily contained by surveillance and treatment (e.g., on the University campus at Ibadan and Yaounde"), whereas heavier infestations by flies of the riverine group give rise to increasingly severe trypanosomiasis problems. In the case of the waterside species G. fuscipes a high density of this fly does not appear to play a significant role in cattle trypanosomiasis transmission (J. Le Roux — personal communication). Flies of the savanna group give rise to severe trypanosomiasis problems in susceptible stock even when present in low, or barely detectable, densities (Jordan, 1961 and Leeflang, 1975). Some members of the forest group (G. fusca and G. medicorum) undoubtedly feed on cattle if they have the opportunity and can very probably be the cause of a serious disease problem (Jordan, 1962 and Jordan, Lee-Jones and Weitz, 1961). Others (G. brevipalpis) can be poor transmitters of infections they harbour (Wilson, Dar and Paris, 1972) and yet others seldom have the opportunity to feed on cattle or might not be attracted to them.
In fundamental contrast all tsetse communities do not harbour trypanosomes pathogenic to man. The pathogenic human trypanosomes have a focal distribution with substantial fluctuation in the actual areas involved over the course of time. Rhodesiantype sleeping sickness (which is not prevalent in the West African region) is transmitted mainly by flies of the savanna group (G. morsitans and pallidipes) and infection is not infrequently derived from wildlife origins (Heisch, McMahon and Manson-Bahr, 1958) with man-to-man transmissions taking place by savanna or riverine tsetse in epidemic situations (Onyango, 1969). It has been shown that domestic livestock harbour the trypanosomes in certain circumstances (Van Hoeve et al., 1967). Gambian-type sleeping sickness (the form of the disease prevalent in West Africa) also perhaps sometimes arises from reservoir hosts other than man — possibly domestic animals (Molyneux, 1973) — but usually man is the source of infections transmitted by flies of the riverine group (notably G. palpalis and close relatives and G. tachinoides). Though it is believed that savanna tsetse do not transmit this disease there is lack of precise information on this point in areas of mixed infestation. However, there is no doubt that riverine tsetse alone are usually responsible for human infections of the Gambian type.
As already noted, susceptible livestock, in contrast to human beings, are at continuous risk wherever tsetse are present, the severity of the problem varying within very wide limits. Some of the mediating factors involving the vector are described above. The picture is also affected by factors involving the animal, notably type, breed, previous experience of infection, therapy, and "stress" factors such as parturition, lactation, inadequate nutrition, and the exertions of traction. The morbidity rate varies within very wide margins, according to the degree of tsetse risk. The clinical manifestation in the animal also varies and can be exceedingly acute, when it is usually rapidly fatal. Frequently it is more chronic in type giving rise to a debilitating illness also often fatal if not treated. Apparent, and occasionally actual, selfcures are sometimes observed in certain livestock and cures can be effected in many animals by therapeutic measures. Animals that have been severely ill (other than for a short period) may, however, suffer permanent damage from the experience; such animals have an impaired productive capability. Losses, therefore, take the form of mortality and depressed productive capability, be it of meat, milk, reproduction or traction, for reasons amply explained by what is known or being learned of the pathology of the disease.
Many livestock projects are basically marginal in financial terms and the occurrence of a trypanosomiasis problem can mean that they easily become unprofitable. The cost of arranging regular surveillance by blood sampling, adherence to a strictly timed and accurately dosed drug regimen required to prevent drug resistance, coupled often with the need to finance periodic activities to reduce the numbers of tsetse present, ensure that projects otherwise successful become loss-makers. In high tsetse-risk areas the disease may become unmanageable but in some locations of low tsetse risk there are several examples of successful production and development, sometimes through the use of trypanotolerant livestock.
Cattle ranch in G. morsitans-infested Guinea savanna woodland. Suppression of regrowth through intensive or improved pasture production. Infested woodland in distance.
Though tsetse flies are not the only vectors of pathogenic trypanosomes to livestock it is tsetse-transmitted trypanosomiasis that is the most serious constraint because of the variety of infections being transmitted and because, during the course of a cycle of development by the trypanosome, the vector itself becomes infected and remains infective for a period of time, in contrast to the ephemeral nature of infectivity in vectors other than Glossina capable of noncyclical transmission. The latter problem, should it develop, is relatively easily contained or even eliminated by simple therapeutic measures. It is a very different matter in the case of tsetse-transmitted trypanosomiasis.
Regarding therapy, the general impression that emerges in the field is one of escalating demand for treatment and, in many areas where antivector measures are either not undertaken or have only been partially successful, of rising infection rates in livestock with a widening gulf occurring between demand for treatment and treatments actually administered. Indeed some countries are failing increasingly to satisfy the demand for treatments, mainly in the traditional livestock sector, but also, significantly, on government and parastatal livestock projects. Reliable numerical data are difficult to obtain because of the immensity and diffuse nature of the trypanosomiasis problem, the inadequacy of techniques for the determination of infection in the individual, and because of the administrative and logistical problems of deploying specially trained junior staff on such activities. Not only is there a gap between what is needed and the treatments officially made available but there is no attempt at control in some countries while in others there is a significant amount of illicit treatment administered. Drug resistance is widespread and probably increasing. The extent and significance of the problem cannot be specified since most countries lack the facilities to study it.
At present most of these constraints, though potent factors in rural development and livestock production, cannot be accurately, or even roughly, quantified in numerical terms. This situation will continue so long as countries lack field facilities to deal with the problems and assemble data on them or, as is sometimes the case, do not adequately deploy and supervise the staff resources which they do have. The fact that the problems are not adequately quantified in numerical terms does not mean that they are not significant.
There are important seasonal variations in the degree of risk to which livestock are exposed in areas that have pronounced variation between wet and dry seasons. In more equable climates the variation is much less. During pronounced dry seasons there is a general regression in the distribution of tsetse particularly if the dry season is also hot. The burning, usual in the savanna lands of Africa, accelerates the diminution in the extent of suitable tsetse habitats.
Pastoralists having susceptible livestock take advantage of these developmerits to make seasonal use of the huge grazing resources present within tsetse-infested savanna lands where infestation is mainly from riverine species or where savanna tsetse density is low or is from flies dispersed from the main tsetse belt. They are, however, unable to utilize successfully, even seasonally, the interior of savanna tsetse belts except where sizeable cultivated enclaves have developed as a result of human settlement. These tend to be used rather briefly so long as the supply of the valuable fodders derived from the farming activities is available. In localities of low risk from riverine tsetse only, pastoralists and some settled cultivators with livestock can be found permanently within the tsetse zone. They can manage and accept the levels of trypanosomiasis that ensue, the pastoralist moving away, if necessary, an option that is not open in areas of higher levels of husbandry and on fixed livestock projects such as ranches.
Kapid regeneration at same location as in the facing photo, following felling and periodic slashing of suckers and seedlings. Vitalization of native grasses favours rapid regeneration of woody species.
The genus Glossina
| Fusca group (subgenus Austenina): | Forest tsetse |
| Rain forest species: | G. tabaniformis, G. nigrofusca, G. haningtoni, G. nashi |
| Forest-edge species: | G. fusca, G. medicorum, G. fuscipleuris,
G. schwetzi, G. severini, G. vanhoofi |
| Others: | G. brevipalpis, G. longipennis |
| Palpalis group (subgenus Nemorhina): | Riverine tsetse |
| G. palpalis, G. fuscipes, G. martinii, G. quanzensis, G.
caliginea, G. pallicera, G. tachinoides |
|
| Morsitans group (subgenus Glossina): | Savanna tsetse |
| Savanna: | G. morsitans, G. swynnertoni, G. longipalpis |
| Savanna and thicket: | G. pallidipes |
| Evergreen thicket: | G. austeni |
The most dramatic instability involves the savanna group. This takes the form of longer term fluctuations either in local density within the infested area or of major fluctuations in the margin of the infested area. Over periods of 11 or 12 years it is quite usual for tsetse densities to fluctuate between the very heavy and the barely detectable, a fact that greatly complicates the assessment of risks to which livestock development projects are subject. The territorial expansions in areas of savanna tsetse infestation have been prodigious and are still continuing. When these have moved into unoccupied land resources the immediate consequences have been minimal but where livestock have been present the result has been devastating. Very high mortalities in all classes of mammalian livestock cause the evacuation of the affected area or the impoverishment of the people who remain. Only in countries where Glossina distribution has been adequately mapped and regularly updated can the full extent of such developments be known. Past changes have been described in detail by Ford (1971) — though not everyone would agree with some of his interpretations, the facts of the situations are meticulously described. Brief examples of some of the more important recent events are as follows. Since 1952 the advances of G. morsitans in the central and eastern parts of northern Nigeria have led to the occupation of at least 25 900 km2. Since 1950 in the central part of Cameroon an advance of G. morsitans has occupied about 20 860 km2 and, unless halted, will proceed very probably to occupy a further 9 000 km2 of valuable and extensively utilized land resource. Since 1953 G. morsitans in Zambia has occupied 11 700 km2 in the southwest and lesser advances are taking place at other locations (MacLennan, 1975).
In all these examples the effect on the rural economy has been grievous, resulting in evacuation, in the case of mobile pastoralists, or the impoverishment of those farmers unwilling to move. The impact in southwest Zambia has been particularly serious because the people there lack the mobility of West African pastoralists and depend on cattle sales for the purchase of the foodstuffs that the land cannot provide. Initially this advance moved into an unoccupied land resource that was later converted into a National Park. It proved impossible to contain the advance within the limits of the park, and trypanosomiasis in cattle in the adjacent land resource, which had previously been nil, required 118000 treatments in 1974. However, this was not sufficient to prevent heavy losses among the 37 000 head of cattle most exposed to risk. In the Cameroon example there have been similar consequences with the further complication that greater mobility of the livestock owners has contributed significantly to an everincreasing overstocking and degradation of tsetse-free pasturages. In Nigeria the advances have also occupied very important pastoralist grazing resources and here, as in Cameroon, therapy could not ensure the continued utilization of the fodder resource.
The full extent of advances of the fly in Africa is not known. In addition to the examples mentioned, notable advances are known to have taken place in Botswana, Zimbabwe, Uganda, Tanzania, Ethiopia and very probably in eastern Senegal and western Mali. The largest and most active advance of modern times is probably proceeding in southeast Angola. The problem is not limited to G. morsitans; there have also been some notable advances by G. pallidipes.
Very large scale regression of infestation took place in the early 1890s following on the elimination of natural tsetsefood sources by the rinderpest panzootics. Most of this country has since been reoccupied. Smaller scale local regression of infestation is presently taking place as a result of cultivation and hunting.
Instabilities in the species of trypanosome prevalent in tsetse populations (and in its virulence) are also evident. This is most apparent in the case of T. vivax in G. morsitans deriving its blood meals mainly from cattle. On cattle trek routes (e.g., Dorin to Oyo in Nigeria) infection rates in both cattle and flies can escalate remarkably (Baldry, 1969a and b; Ferguson, 1964; Riordan, 1971; Yesufu and Mshelbwala, 1973). The same thing can happen on cattle ranches and farms as on the Mokwa Ranch and the Shika Farm. Domestic pigs may contract relatively mild infections due to T. congolense or T. brucei until T. simiae is introduced into the relationship, when a fulminating outbreak of fatal trypanosomiasis is the usual consequence, even when tsetse density is exceedingly low. It seems likely that, once introduced by an infected tsetse fly, conditions in piggeries favour rapid mechanical transmission of T. simiae by other haematophagous flies (stomoxys).
A very important trypanosome in fection fluctuation in the case of Glossina is that of the human pathogens. Fortunately these are not often present in tsetse populations and, apart from the inconvenience of bites, human communities, though freely fed upon, live unharmed by tsetse infestations, until such time as the human pathogen is introduced resulting in serious epidemics or continuing endemicity. Serious outbreaks of sleeping sickness can result in the abandonment of valuable land resources.
Searching for G. morsitans with bait, following air spray in Botswana
A balanced utilization must assign parts of the land resource, in varying proportions according to topographic, climatic and edaphic considerations, and community need, to the following usages: (1) conservation of readily degradable areas; (2) conservation of an adequate wildlife resource; (3) conservation of adequate tree cover; (4) provision for forestry plantations; (5) provision for extensive rangelands; (6) provision for village or community pasturage and ranches; and (7) provision for a truly balanced, mixed arable agronomy.
The major objective of increased production is attained principally through item (7) to which items (5) and (6) contribute directly since by a balanced, mixed agronomy is meant food and cash crop production integrated with livestock that consume the byproducts of cultivation and fallows as well as the savanna pasture and browse resources. Items (3) and (4) satisfy vital community needs for fuel, building materials, wild fruits and other items of diet and medicines.
Biconical trap in garden of young mango and banana trees
Tsetse infestations are basically a feature of the wilderness and will persist within infested land resources assigned to objectives (1) to (3) or perhaps invade them if they are not already infested. They will also either persist within or disperse widely into areas assigned to objectives (4) to (7) if these contain the necessary habitat and food sources. Areas devoted to objectives (5) and (6) may not, at the time assigned, harbour detectable infestation but developments within them, i.e., fluctuation of tsetse densities, or favourable environmental or crop situations can cause an escalation of the tsetse problem or else the project can be engulfed by a tsetse advance.
River floodplain habitat of G. morsitans and G. tachinoides in northern Nigeria. Infestations were later removed by ground spraying.
"Settlement schemes", if they result in a sufficiently extensive modification of vegetation and fauna, can result in the areas becoming untenable for resident tsetse populations. Suppression of trypanosomiasis by this means is more easily achieved from infestation with savanna tsetse and when livestock are not involved in the development. For reasons explained earlier, i.e., varying infection rates, factors involving the animal, infrequency of human pathogens in tsetse populations, livestock trypanosomiasis persists as a problem at tsetse densities that do not give rise to a serious sleeping sickness problem.
Some of the riverine tsetse species persist in natural vegetation present in areas of densest human settlement, as, for example, in the rural environment of Tiv country and Ibo land and around Kano city in Nigeria, or in urban areas such as Bida in Nigeria and Bamako in Mali, and cause trypanosomiasis in livestock and periodically in humans (Baldry, 1964,1968, and 1969a and b; Challier, 1973; Touré, 1974).
Wholesale modification of the environment in a manner that renders it uninfestable is impossible since the land resource is not homogeneous in its nature (a variety of land usages have to be accommodated) and because it would be physically, financially, technically and administratively impossible to exert the type of comprehensive control over landuse activity on the scale that would be required.
Viewed in this light it seems that the sound, balanced land utilization that is required to meet the needs of a rapidly expanding human population as described earlier, though it can lead to changes in the Glossina species present and in levels of infestation, will, in many circumstances, ensure the perpetuation of the trypanosomiasis problem except where specific measures are successfully undertaken to eliminate infestation and maintain tsetsefree areas. As mentioned earlier the problem will increase in importance in the future.
Cattle are important converters of the abundant grazing and browse resources of the savanna lands. However, the introduction of cattle into areas apparently free of savanna tsetse infestation has, on frequent occasions, brought to light, through an incidence of trypanosomiasis, the existence of infestations previously present at levels below detection or it has resulted in an escalation in the actual level of infestation or has enabled tsetse colonies to exist where previously this was not possible.
Basically the origin of the cycle of transmission by tsetse flies of trypanosomes pathogenic to livestock is sylvatic (tsetse/wildlife hosts). The relationship of infections in domestic stock to the sylvatic cycle (illustrated in the figure) can be categorized according to whether they are external or internal to it.
Environment hostile for Glossina. Such a situation would exist, for example, on a cattle ranch from which all tsetse-breeding habitats and most of the shaded resting places have been eliminated. Occasional livestock infections can be caused by infected tsetse flies dispersing or being carried into the project area and, before they perish, biting a susceptible animal.
Aerial view of G. morsitans and G. palpalis habitats in Guinea savanna
Aerial view of persisting network of riparian G. palpalis habitats in Guinea savanna zone
In the true savanna zones the risk is relatively light if only riverine species of Glossina are involved. Though dispersal by these species takes place more readily in subhumid and humid areas, tsetse densities and infection rates are sometimes low. If the risk is from G. morsitans it is known from experience that no project is safe from infection if there is a primary focus of these flies located within a radius of about 16 km. Individually dispersed tsetse can be exceedingly difficult to detect, and may not be found even following repeated searches. Such a situation encourages the belief that mechanical transmission by other haematophagous insects is responsible (Kirby, 1963; and Wells, 1972). However, sooner or later tsetse are usually detected, and if the source population is removed trypanosomiasis infections cease (Kirby, 1963). G. longipalpis has somewhat less ability to disperse but is a very effective vector.
If environmental conditions are a little less hostile for Glossina and dispersed flies live for more than five days, they can themselves acquire T. vivax infections derived from infected livestock and, if the dispersed flies are mainly dependent on the livestock as a food source, a progressively more intense transmission cycle of T. vivax develops the longer the flies survive. Thus a small number of flies can cause a high incidence of trypanosomiasis in livestock (MacLennan, 1974; and Leeflang, 1975). Furthermore, with the sylvatic cycle it becomes possible for a fly to acquire drug-experienced infections from livestock and return with them to the sylvatic cycle.
Environment favourable to Glossina. Some cattle-development projects have been sited following repeated tsetse surveys that, though the vegetation seemed suitable, indicated the absence of infestation. On some, either a G. morsitans submorsitans advance has later approached the project, or there was sufficient cover on the project to enable dispersed tsetse flies to survive for a significantly longer period because of the constantly available food source in the form of cattle. Breeding tsetse populations have established themselves where previously this was not possible. As the flies live longer a cyclical proliferation of T. congolense and T. brucei develops in addition to T. vivax. A very intense transmission cycle rapidly escalates and may become so severe that therapeutic and prophylactic agents are powerless to prevent serious losses. This happened on the Shika Stock Farm, the Mokwa Ranch and on the Kontagora settlement scheme in Nigeria. An extreme example of the manner in which infestation and infection can escalate is provided by the slaughter cattle trek route described earlier (Riordan, 1971). In this latter instance it is significant to note that the development took place in an environment not hitherto thought suitable for infestation by G. morsitans submorsitans, and located outside the normal sylvatic associations of this species (Baldry, 1969a and b).
There exist a series of intermediates between the extremes of hostile and favourable environments external to the sylvatic cycle. Ranching projects of an extensive nature will be more vulnerable than those in which bush elimination can be justified economically by increased pasture productivity, a point of special significance for projects located in the southern Guinea and derived savanna zones (Keay, 1953).
Epizootiologically, the return of domestic infections to the sylvatic cycle through the movement of either infected tsetse or livestock is of considerable importance, since the strains of trypanosome are livestockoriented and will have frequently experienced chemotherapy and may have an enhanced level of drug resistance. Drug-resistant strains are readily transmitted by tsetse flies, and can be serially passaged through wildlife hosts by tsetse flies with no diminution in the level of drug resistance (Gray and Roberts, 1971a and b).
The proliferation of habitats favouring savanna tsetse is favoured by the utilization of the grazing resource in areas of fire subclimax. The less severe the annual fires become the greater the grazing intensity, and this favours the densification and proliferation of woody growth. In localities of heavy usage this change progresses remarkably rapidly and provides increasing shelter for tsetse dispersal or for resident satellite tsetse-colonies. A rapidly escalating trypanosomiasis situation can then result and it is only when densification has advanced to the stage that the grass has been suppressed that such localities sometimes become vegetationally unsuited for savanna tsetse. This is an important aspect affecting beef production in the subhumid zone, which favours intensive fodder production from artificial lays etc., and is a point to be considered when making economic evaluations.
Domestic pigs particularly, but also sometimes cattle, that are maintained by family groupings and villages, support satellite colonies of G. tachinoides (Baldry, 1964 and 1968) and sometimes of G. palpalis (Challier, 1973; and Touré, 1974). The situation is particularly prone to develop in the southern Guinea and derived savanna zones (Keay, 1953). Trypanosomes are constantly transmitted to domestic stock by these infestations, which, in some locations, are also involved in troublesome sleeping sickness outbreaks. Habitat associations are de scribed later on in this article under the section on crops.
Schematic representation of trypanosomiasis relationships
Livestock entering the territory of the sylvatic cycle. The most important example we have of this is the system of pastoral transhumance used to maintain most of the domestic beef supply in West Africa. As northerly grazing areas dry out and grazing resources are consumed, pastoralists move their livestock southward. The progress is a leisurely one, and much time is spent in consuming the farm byproducts and in manuring the farms of the settled farmers. Where possible, savanna tsetse infestations are avoided, and the climatic influences described earlier moderate the general risk of contracting trypanosomiasis. However, heavy risks cannot always be avoided and some serious trypanosomiasis outbreaks do occur, apart from the lower general incidence of trypanosomiasis resulting from some exposure to the riverine group of Glossina. The tendency is for the pastoralist to escape from areas of heavier risk as soon as possible and, with the advent of sufficient grazing and surface water in northerly locations, the mass movement northward of several million head of cattle is completed in about 10 to 14 days. This accounts for the fact that in the six northern provinces of Nigeria 86 percent of the counted cattle population is located during the rains in the tsetsefree or lightly infested portion amounting to 46 percent of the former Northern Region.
Though it seems fashionable to use adjectives such as "primitive", "wasteful" and "inefficient" to describe this system of husbandry, it can be said that it affords a seasonal respite for overgrazed northerly pastures; makes possible the seasonal exploitation of some tsetseinfested grazing resources; uses up farm byproducts that would otherwise be wasted; and sometimes returns at least some fertility to the cultivated soil. In the present state of knowledge and experience it seems doubtful whether the major part of beef requirements, consisting of some 790 000 head per annum of cattle of Nigerian origin, could be economically and practically provided by any other method.
The subhumid zone is not completely evacuated by pastoralists with zebu stock in the wet season; a small proportion remain. The White Fulani breed in particular appears to possess a degree of trypanosome tolerance that permits allseason utilization of some areas where risk is light and confined to riverine tsetse only. However, this does not indicate that such places are suitable for mixed farming activity by settled cultivators or for higher level husbandry systems.
Regression in areas infested by savanna tsetse is making possible an increased exploitation of the southern Guinea and derived savannas by pastoralists, but adaptations by tsetse populations to domestic circumstances are also taking place. These vegetation zones and the forest zone provide a more favourable environment in which such adaptations can occur.
Isoberlinia woodland infested with G. morsitans in the rains
Another case arises when trypanosometolerant livestock (Pagot, 1974) live within the sylvatic cycle. Three factors seem important in determining whether this will be successful. First, the introduction to risk should be in such a way that the animal has the opportunity to adjust to it or has earlier acquired some tolerance through exposure. Second, the level of husbandry is important since malnutrition and other stress factors may be followed by clinical breakdown. Third, a sustained and very heavy challenge in the laboratory can lead to stunted growth and retarded sexual development (Stephen, 1966). There is some evidence that this also happens in the field though in other locations these cattle are in good bodily condition despite being at quite heavy tsetse risk.
Some examples can be quoted of the relationship between crops and tsetse infestation under various circumstances. The examples are not exhaustive and it seems probable that the full extent of such association has not been revealed by current experience. The developments have usually not been foreseen and the infestations have to be dealt with in retrospect and arise from activities that otherwise are highly desirable.
The occupation of mango groves by G. palpalis and G. tachinoides (Baldry, 1969a and b; Challier, 1973; and Touré, 1974), leading to troublesome sleeping sickness outbreaks and persistent livestock infections, is probably the best known example. The problem can become an urban one. Infestations can also persist in sacred groves, which are of vital significance to the local community. Other actual or potential riverine tsetse habitats can result from guava, cashew, coffee, cocoa, banana and sugar cane plantings. In some areas tree plantings around family groupings harbour G. tachinoides and G. palpalis. Hedging, particularly lantana, although not only this species, harbours peri-domestic G. fuscipes infestations in Nyanza in Kenya and other riverine species in West Africa (Baldry, 1966a and b). G. tachinoides utilizes cocoyam cultivations in the densely occupied land resource of Ibo and in the derived savanna in Nigeria (Baldry, 1968). In this locality, this important vector of sleeping sickness elsewhere is not at present transmitting this disease since the tsetse are feeding mainly on pigs, to which they are transmitting livestock trypanosomes, whereas in Tiv land, in the subhumid zone, sleeping sickness is also transmitted. In some of these examples the tsetse colonies appear to be self-sustaining but in other situations they might be dependent on recruitment from natural populations.
Isoberlinia woodland infested with G. morsitans. Hot season following bush fires (see facing photo).
Another relationship with arable farming is seen in areas where farmers are also cattle users. Family groups are less efficient in producing a marketable surplus if they do not possess effective work oxen for ploughing and for getting the produce to market. It has been stated that possession of work oxen increases the production of the family group sixfold.
Pressure-retaining knapsack sprayers in use on G. tachinoides resting places in a Sudanvegetation zone
Most of the subhumid zone of West and Central Africa is tsetse infested and in this area cultivators are not usually using draught oxen. It has been the usual experience in this area that, if cattlekeeping for draught purposes is encouraged, losses mainly attributable to trypanosomiasis are frequent. In areas of light risk it should be possible to mitigate these losses by treatment but often the risk is a continuous one and the individuals are dispersed over a wide area and therefore difficult to service with treatments as soon as these are required. In such static situations not only does frequent treatment favour the development of drug resistance but anaemia and myocarditis result from trypanosomiasis. Thus heart failure in infected animals (if they possess sufficient strength to be worked) under work stress is not uncommon. The risks are considerable and, generally speaking, it would be wrong to encourage a farmer to get into debt for the purpose of purchasing work oxen in tsetseinfested areas. Tractors are not the complete answer to this problem since, compared with, animals, they cannot use the freely available grazing resources; do not provide byproducts (milk, manure); cannot be eaten; have no financial reserve significance; have as yet no social importance; do not reproduce; and are even more difficult to maintain.
As noted previously, the unhomogeneous nature of the land resource and the necessity to conserve substantial portions of it for particular purposes interact with livestock production, and particularly with the utilization of the abundant grazing and browse resource of the subhumid zone, through the shelter that the conserved resource provides for the continuation of the sylvatic origins of livestock trypanosomiasis.
The most extreme example of this situation is the wildlife reserve, which, so long as it remains infested, provides a permanent source of tsetse flies and pathogenic trypanosomes. It is of interest to note that motivation for the first successful large-area tsetse-eradication operation using insecticides, completed in 1952 in South Africa, arose from the dispersal of G. pallidipes from the Umfolozi, Mkusi and Hluhluwe game reserves and the resulting impact that this had on livestock production in the surrounding land resource of Zululand (Du Toit, 1954). It is probably correct to state that all game reserves and national parks within the subhumid zone are tsetse infested and that the increasing utilization of the surrounding land resources and, in some places, the advancement of tsetse-eradication projects, will give increasing prominence to this situation. This implication for the future has seldom been considered when such reserves have been established.
Tsetse infestations are no protectors of wildlife resources. Wildlife decimation has proceeded and is still proceeding in areas of the densest infestation. Wildlife resources can only be protected through the will of the people and the law of the land. Nor do tsetse infestations preserve the land resource from lowlevel exploitation by cultivators who are not dependent on livestock. They do, however, restrict utilization for any mixed agronomy by cultivators to whom cattle are important and inhibit its utilization by pastoralists. To argue that the continuation of tsetse infestations is essential for the conservation of the land resource is a singularly negative and erroneous position to take up, interfering as it does with the balanced and efficient utilization of the land resource. It is a position not accepted by most national governments. Abuse of the land resource is not a problem peculiar to areas freed of infestation but tsetse eradication does provide a unique opportunity for sound, balanced development.
Rice farm development behind the riparian forest of the river trees
Forest reserves have also in fact provided for the perpetuation of tsetse infestations in Nigeria and these have expanded into more recently conserved areas. Not only, for example, has a very large scale tsetse advance used reserves as stepping-stones, into and through the Anchau Corridor, disrupting the pastoral utilization of a large portion of the subhumid zone in northern Nigeria but, in the Shika area, a similar development gave rise to uncontrollable trypanosomiasis in the higher level management area of the Shika Stock Farm (Kirby, 1963).
Another conservation aspect is involved in statements that have recently been published to the effect that tsetse and trypanosomiasis control have been major factors causing the degradation of the Sahel and that continuation of this activity will result in similar changes in the savanna zone.
However, the fundamental urge in pastoralists is to find new grazing lands. Overutilized areas are avoided as far as possible. Under the transhumance regime pastoralists remain in the savarma grazing resource when it is safe for them to do so. The overutilization of fragile areas such as the Sahel undoubtedly accelerates degradation but, contrary to the statements that have recently had considerable publicity, it is not tsetse and trypanosomiasis control and/or eradication that have resulted in the overexploitation of Sahel grazing land; a multitude of other factors have been responsible. The Sahel grazing land is not used by the large pastoralist community based in the savanna zones. However, the extent of safe savanna grazing resources there is diminishing at a remarkably rapid rate through the pressure to grow food and commercial crops, and to tsetse advances. The consequence is that there is now developing a shortage of safe, common grazing land in the savanna zone. This increases any trend toward pastoral overexploitation and makes less grazing available also for transhumance from the Sahel, making it even more difficult to maintain the seasonal relief that is the fundamental feature of the transhumance system. Thus it can be suggested that, far from contributing to the degradation of the Sahel, tsetse eradication and trypanosomiasis control provide a substantial relief to the overgrazing problem.
Purely pastoral exploitation in the savanna lands of West Africa seems unlikely to lead to changes similar to those that have taken place in the Sahel, which, in the main, is not a fire subclimax. In the fire subclimax of the savannas the heavy utilization of the fodder resource leads in fact to the densification and proliferation of woody species and the area becoming progressively less attractive to pastor alists. Many examples of this kind of succession are to be seen, such changes taking place most rapidly in the subhumid zone. Here, it is cultivation practices that have been responsible for major degradations of the land resource.
The previous sections have described the impact of tsetse infestations on balanced utilization of land resources. These directly or indirectly relate to the need to improve the efficiency of the agronomy in areas of current utilization or to expand the area of utilized land.
Tsetse infestations are far from being the only basic constraint on balanced utilization of land resources. The most important combination of factors is probably where tsetse infestation and river blindness co-exist. In this situation the elimination of either of the vectors alone will not result in long-term, trouble-free development. There are many other factors not related to disease, as well as other diseases (most notably streptotrichosis in zebu cattle in the subhumid zone) that are restraints on rural development. But within the tsetseinfested land resource it is tsetse infestation alone, unless accompanied by river blindness, that is the fundamental constraint in a large area of tropical Africa having the basic capability for balanced development.
It is sometimes said that some of the tsetse-infested land resources are incapable of supporting arable agriculture and/or pastoral usage. But against this it must be pointed out that much of the best-watered and mostfertile land resources are tsetse infested. Similarly it is said that much of the underutilized savanna forage resource of Africa (estimated in total to cover 7 million km2) cannot be utilized with out very great expenditure on infrastructure, such as the provision of water points and possibly not even then. In this case also it is necessary to stress that a very large proportion of good grazing, most notably that of the subhumid zone. could be utilized at once by pastoralists using existing cattle herds and with practically no capital, expenditure other than that required to remove tsetse infestations. It is true that shortages of meat in some areas are due more to offtake/marketing problems than to shortage of stock or grazing. In other very substantial regions of the continent, however, again most notably in the subhumid zones, there is a great human need for more meat, which existing national herds could more effectively supply if available grazing resources were to be freed of tsetse infestation.
Aspects relevant to the economic assessments of the tsetse problem are the economic impacts of tsetse-transmitted trypanosomiasis, the cost and technical effectiveness of counteractivities and the value of the resulting benefits.
Each of the individual items of the complex of effects and interactions described above has an economic connotation. Often this is fundamental, sometimes it may be slight. Some effects are not easily quantified in monetary terms, yet they may have a fundamental economic significance. Not infrequently numerical data are not available or are inaccurate, yet the aspect to which they refer is of major economic significance (as, for example, actual infection rates, mortality rates, production losses and incidence of drug resistance in pastoralist stock presented for treatment) and may be substantial in numerical terms.
Jetty, oho used for bathing and washing clothes, within about 20 metres of the infested mango garden in the photo above. Riparian forest G. palpalis habitats in background.
Garden of young mango trees, papaya, banana, etc., infested with G. palpalis. Horse showing chronic trypanosomiasis.
If equations are to be drawn up that meaningfully place the economic impacts of the problem in relation to the cost of countermeasures, it has to be ensured that all components are accounted for and not simply omitted because their significance has not been noticed or quantified numerically. It seems important that the negative impacts of infestation on balanced land use be adequately assessed. In regard to the problem of trypanosomiasis it is easier to be more precise about the costs of counteractivities than it is to measure the costs resulting from the disease.
In purely economic terms trypanosomiasis infections can be of little account where tsetse densities are low, where livestock infections arise from flies with low infection rates, not primarily oriented on livestock and where facilities for diagnosis and treatment are readily to hand. This is exemplified in the forest zone in places where the risk from G. palpalis, and where savanna species (G. longipalpis) and forest species (particularly G. fusca and G. medicorum and possibly some others), are not present. This is the case in some areas of relatively high human population density.
In these circumstances cattle, exotic to the tsetse zone (zebu and temperate breeds) can be maintained and trypanosomiasis can be one of the lesser and more readily treated problems to be countered. Facilities for the diagnosis and treatment of trypanosomiasis must, however, be available. Usually they are provided from the general background of services and supervision that such projects require for reasons other than trypanosomiasis. The availability and cost of the facilities have to be considered in the economic assessment. Though characteristic of specific locations in the forest zone this situation may occasionally be encountered in the savannas, but will be inherently less stable and there will, therefore, be the need;' for a greater risk component to be included in the economic assessment. Furthermore, no risks can be assessed without repeated surveys being carried out by a specialized tsetse field unit under the control of a specialist entomologist. Both the availability and cost of this requireihent have also to be considered.
The other extreme relating to the disease exists in areas of moderate or high densities of savanna tsetse where the incidence of the disease, despite the availability of therapeutic agents, is such that continuous production from susceptible livestock, even from low-level husbandry systems, is impossible. The severity of the economic impact in relation to production from susceptible livestock is complete, as exemplified in areas overtaken by tsetse advance and described earlier. This situation exists over much of the subhumid zone.
Trypanotolerant West African shorthorn (Muturu). Dense G. morsitans infestations in vicinity (Nigeria, Katska type).
Intermediate situations exist where treatments, accompanied when necessary by vectorsuppression activity, or periodic withdrawal of stock from risk (as in transhumance), can make livestock production possible. A point to be considered is that large numbers of pastoralists can be serviced relatively easily at fixed treatment points, which they readily attend. It is much more costly and difficult to service similarly a scattered community of smallholders or ranches. In practice it has on occasions proved impossible to provide the degree of service necessary to prevent losses, resulting in livestock projects becoming uneconomic. Small projects are more readily serviced in the physical sense but at higher cost per animal unit and they are not likely to contribute greatly in alleviating national beef deficits, though any step in the direction of the development of a balanced, mixed agronomy is to be encouraged. Higher level husbandry projects can be more vulnerable than basic pastoralism.
Glossina derives its economic significance solely from its capability to transmit trypanosomiasis to man and livestock. The economic impact is directly related to the risk of contracting trypanosomiasis and, as previously noted, there is a wide range in the degree of trypanosome risk in a variety of circumstances. Essentially, economic aspects concern: the necessity to know which tsetse species are present and their density, what are trypanosome infection rates in the flies, and what is the relationship of the infestation to man and livestock; and what is the cost of the antitsetse activity required.
An essential component, is the availability and cost of a specially trained field organization, adequately serviced with transport and travel allowances under the control of a specialist entomologist performing repeated surveys and compiling records of tsetse distribution over the course of time. Single surveys can reveal a degree of the economic risk but cannot fully assess the instabilities discussed previously. Livestock projects that are developed in the subhumid zone in the absence of adequate information on Glossina run a high risk of failure. Unless the circumstances are sufficiently favourable in regard to the Glossina risk they cannot be protected so as to ensure adequate levels of production at a cost that can be justified in economic terms. In more extreme situations, adequate protection is not feasible in purely technical terms. Additionally there is the cost of arranging trypanosomiasis surveillance and treatment in livestock (and possibly humans) and the actual cost of antitsetse activities that may be required.
Anti-tsetse objectives are discussed in Part II of this article. There are very important differences in the economics of control compared with largearea eradication (where the latter objective is technically feasible). Examples can be quoted from Botswana, Nigeria and Zambia and from the first such operation, ever completed in South Africa (Du Toit, 1954).
In Botswana, taking the annual cost of containing G. morsitans infestation and excluding any other costs, the annual expenditure following a successful eradication becomes substantially less by the third year. On generous assumptions the maintenance of the eradication situation settles down at half the annual cost of control. The major expenditure on the eradication is recouped by the fifth year (Negrin and MacLennan, 1977).
It must be emphasized, however, that for this expenditure the area of the eradication operation would be 3.7 times that of the area over which control is currently exercised and that the value of the land resource freed of infestation has not been included in the economic equation. The continuing expenses of diagnosis and treating livestock trypanosomiasis inherent in the control situation, the constraints on balanced land utilization and the effects of human trypanosomiasis have also not been taken into consideration.
Trypanotolerant N'Dama of Fonta type (Guinea)
Trypanotolerant N'Dama (Hamitic longhorn) work ox. Dense G. morsitans infestations in vicinity (Gambian type).
The other significant aspect is that it is impossible technically to develop an extermination operation of smaller size because of the necessity of reaching a defendable perimeter. The fact that half the area reclaimed would not be put to "productive" use but would be assigned to conservation objectives does not detract significantly from the economic advantage of effecting eradication though the inclusion of this area is essential in reaching that objective. A more extreme example of this aspect is provided by the Zululand operation and the current problem in southwest Zambia.
In the Zambia example (MacLennan, 1975) a G. morsitans advance overtook the southwest corner of the country with the consequences described earlier. The cumulative cost of vectorcontrol activities, which proved far from effective in providing complete protection to the livestock, over a six-year period of containment, was more than double the estimated cost of eliminating the entire infestation (on the assumption that current airspray techniques achieve extermination). The annual cost of maintaining the eradicated area free from reinvasion on the vulnerable perimeter would be less than the present running cost alone on the "holding lines" of game and stock fences supported by clearings and game eviction. Again, the value of the land resources available for production and the benefit of eliminating other constraints inseparable from infestation are not taken into consideration.
In Nigeria anti-tsetse operations developed from the need first, to control very serious sleeping sickness epidemics (most active in the Sudan and subhumid zones) and later, to deal with the consequences of G. morsitans advances on the pastoral utilization of the land resource in several areas. The presence of the disease resulted in the abandonment of valuable land resources, much of it the more fertile and better watered portions (as, for example, the alluvial floodplains). From the vectorcontrol activities it proved possible to develop substantial extermination areas within a longterm programme of eradication. By 1977, 194 532 km2 in a programme covering 259 000 km2 had been completed. The removal of disruptive effects on the rural economy, particularly on pastoral and arable utilization and on conservation activities (especially those in connection with forestry and the reservation of pasture resources) had very important economic connotations.
It has frequently been observed that, in development, the cost of eliminating tsetse infestations (where this is technically feasible) is a small part of the total cost of development. The most recent example, quoted from Nigeria (Jordan et al., 1977), indicates that, in a proposed grazing development, the tsetse-eradication component amounts to 1 to 2 percent of total investment or 9 to 17 percent of investment (less the cost of livestock). However, in an alternative "low key" proposal for the same location, eradication costs would amount to 26-42 percent of infrastructure costs. It has to be stressed that the calculation is dependent on the project falling within a large extermination area. At the proposed location it cannot be suggested that periodic tsetse suppression (as opposed to eradication) would ensure an economically viable project or that an eradication "enclave" could be maintained for a reasonable cost. In the control situation it is unlikely that each suppression operation would cost significantly less to execute than a single effective eradication. Recoveries in G. morsitans populations following air-spraying in Botswana, which fell short of eradication, were slow but were rapid in areas where livestock were located and this caused a significant incidence of trypanosomiasis within 12 months of spraying. It is doubtful also whether repeated ground-spraying of large areas, without achieving eradication, could be accepted over the years (assuming largely unaltered tsetse habitats within the project and primary G. morsitans infestations within a radius of about 16 km), because of the cost, environmental aspects relating to persistent insecticides in current use, the transmission capability of lowdensity G. morsitans populations and their orientation on cattle.
Viewing the tsetse-eradication programme in Nigeria it is evident that it has eliminated considerable direct and indirect problems deriving from trypanosomiasis. For example, tsetse have probably been removed from most of the epidemic sleeping-sickness zone, large G. morsitans advances (which had severe impacts on utilization of the grazing resources) have been halted or reversed and, in addition, a large area, hitherto unusable. or only partially utilizable, has been freed of infestation.
In effect this has provided some compensation for the loss of grazing resources resulting from population growth and food/cash crop production. In the absence of a "land budget" it is not known what is now happening or what will be required in the future, but even by 1973 it appeared that the transhumance system was under severe stress and that a major reason for this was the progressive loss of safe grazing areas and dry season "survival" grazing areas.
Though much of the tsetse eradication area was on occupied land resource substantial areas of new grazing have been made available at a cost in 1973 of about Naira 200 per 2.59 km2 (US$ 117 per km2). This represents the actual operational cost of one eradication unit for one year achieving eradication over 16 000 km2 and including all unit personnel and labour charges, allowances, all materials, depreciation on all transport and equipment, and assuming that the proportion of the area included in discriminative spraying was 10 percent. Since that time there has been a severe inflation and, in the subhumid zone, discriminative spraying has increased possibly to 18 percent with a proportionate increase in cost and a reduction in the rate of annual progress. The equivalent cost today, in intermediate situations in the subhumid zone, could be estimated to be about 2.5 times the 1973 costs, bringing the cost up to US$ 293 per km2 or something near US$ 3 per hectare. This would still seem a cheap way to create usable grazing. Over much of the eradication area this has been the only development charge, since the livestock already existed and was simply re-accommodated. Specific infrastructural support that did not already exist has been minimal or nil.
Trypanotolerant Gambian N'Dama showing some zebu blood
Trypanotolerant Gambian N'Dama. Old female with chronic trypanosomiasis
It appears that, in Nigerian conditions, the most important follow-up activity is to secure the appropriate portions of the land resource for conservation and grazing objectives based on land use, capability assessments and on community need. If the situation is not stabilized in this manner it is only a question of time before problems that have developed in tsetse-free areas are seen again in the eradication areas.
Traditional methods of land use will, however, have to continue to be the main method of land utilization. Not only do higher level developments have to be technically feasible but, on the scale required to make a significant impact, they require a very large investment. They also have to be fostered by a body of technical expertise that knows exactly what it is about and the activities have to be acceptable and enforceable in the administrative and political senses. It seems that, in actual production terms and in relation to the needs of the larger human populations, production by traditional methods will have to provide what is required during the lengthy and difficult transition to more advanced practices where these are appropriate.
The problems of ensuring sound and balanced land utilization are common to tsetsefree and to infested locations and are certainly not confined to Africa. Tsetse eradication provides scope and space for the attainment of objectives that were not previously available in the Sudan and subhumid zones. Toooftenrepeated dicta to the effect that tsetse eradication activity should only proceed where the whole land resource can be put to productive use immediately and that settlement of the land resource will prevent reinfestation have little meaning in the West and Central African context. The former is inconsistent with balanced land use and in meeting future population needs and the second has no basis in fact in this region. ■
References
Baldry, D.A.T. 1964. Observations on a close association between Glossina tachinoides and domestic pigs near Nsukka, Eastern Nigeria: ecology and trypanosome infection rates. Ann. trop. Med. Parasit., 58: 3234.
Baldry, D.A.T. 1966a. Lantana camara as a breeding site for Glossina tachinoides Westwood in south-eastern Nigeria. In 11th meeting ISCTR, Nairobi, 1966. OAU/STRC Publication No. 100: 91-93.
Baldry, D.A.T. 1966b. On the distribution of Glossina tachinoides in West Africa. In 11th meeting ISCTR, Nairobi, 1966. OAU/STRC Publication No. 100: 95-109.
Baldry, D.A.T. 1968. Observations on peridomestic behaviour and resting sites of Glossina tachinoides near Nsukka, East-Central State, Nigeria. Bull. ent. Res., 59: 585-593.
Baldry, D.A.T. 1969a. Distribution and trypanosome infection rates of Glossina morsitans sub morsitans along a trade cattle route in south-western Nigeria. Bull. ent. Res., 58: 537-548.
Baldry, D.A.T. 1969b. Variations in the ecology of Glossina spp. with special reference to Nigeria populations of G. tachinoides. Bull. Wld Hlth Orgn, 40: 859-869.
Challier, A. 1973. Ecologie de Glossina palpalis gambiensis en savane d'Afrique occidentale. Mémoires ORSTOM, 64:274.
Du Torr, R. 1954. Trypanosomiasis in Zululand and the control of tsetse flies by chemical means. Onderstepoort J. vet. Res., 26: 317-387.
Ferguson, W.W. 1964. A survey of disease incidence in slaughter cattle in Northern Nigeria. Bull. Epiz. Dis. Afr., 12: 331-338.
Ford, J. 1971. The role of the trypanosomiases in African ecology. Oxford, Clarendon Press.
Ford, J. & Katondo, K.M. 1973. The distribution of tsetse flies (Glossina) in Africa 1973. Nairobi, OAU/STRC Bureau of Animal Resources.
Gray, A.R. & Roberts, C.J. 1971a. The cyclical transmission of strains of Trypanosoma congolense and T. vivax resistant to normal therapeutic doses of trypanocidal drugs. Parasitology, 63: 67-89.
Gray, A.R. & Roberts, C.J. 1971b. The stability of resistance to diminazene aceturate and quinapypamine sulphate in a strain of Trypanosoma vivax during cyclical transmission through antelope. Parasitology, 63: 163-168.
Heisch, R.B., McMahon, J.P. & Manson-Bahr, P.E.C. 1958. The isolation of Trypanosoma rhodesiense from a Bushbuck. Brit. med. J., 2: 1203-1204.
Jordan, A.M. 1961. An assessment of the economic importance of tsetse species of Southern Nigeria and the Southern Cameroon based on their trypanosome infection rate and ecology. Bull. ent. Res., 52: 431-441.
Jordan, A.M. 1962. The ecology of the Fusca Group tsetse flies in Southern Nigeria. Bull. ent. Res., 53: 355-385.
Jordan, A.M., Lee-Jones, F. & Weitz, B. 1961. The natural hosts of tsetse flies in the forest belt of Nigeria and Southern Cameroon. Ann. trop. Med. Parasit., 55 167-179.
Jordan, A.M., McIntyre, W.I., Le Roux, J. G. & Negrin, M. 1977. Economic assessment of present operations for the control of animal trypanosomiasis in Nigeria. (Report to FAO Animal Production and Health Division)
Keay, R.W.J. 1953. An outline of Nigerian vegetation. 2nd ed. Lagos, Government printer.
Kirby, W.W. 1963. A review of the trypanosomiasis problem on the Shika stock farm Northern Nigeria from 1929 to the present day. Bull. epiz. Dis. Afr., 11: 391-401.
Leeflang, P. 1975. The predominance of Trypanosoma vivax infections of cattle at a distance from savannah tsetse concentration. Trop. Anim. Hlth Prod., 7: 201-204.
MacLennan, K.J.R. 1974. The epizootiology of tsetse transmitted trypanosomiasis in relation to livestock development and control measures. Actes du Colloque, IEMVT.
MacLennan, K.J.R. 1975. A review of operations against tsetse flies and trypanosomiasis in Zambia and recommendations for the future. Report to the Director of Veterinary and Tsetse Control Services, Lusaka, Zambia. 121 p.
Molyneux, D.H. 1973. Animal reservoirs and Gambian trypanosomiasis. Annls Soc. beige Méd. trop., 53: 605-618.
Negrin, M. & MacLennan, K.J.R. 1977. Economic assessment of present operations for the control of animal trypanosomiasis in Botswana. (Report to FAO Animal Production and Health Division)
Onyango, R. 1969. New concepts in the epidemiology of Rhodesian sleeping sickness. Bull. Wld Hlth Orgn, 41: 815-823.
Page, W.A. 1959. The ecology of G. palpalis in Southern Nigeria. Bull. ent. Res., 30: 617-631.
Pagot, J. 1974. Les races trypanotolerantes. Actes du Colloque, IEMVT.
Riordan, K. 1971. Trypanosome infection rates in Glossina morsitans submorsitans in South-Western Nigeria. In OAU/STRC Publication No. 105: 183-186. Lagos, ISCTR.
Stephen, L.E. 1966. Observations on the resistance of West African N'dama and Zebu cattle to trypanosomiasis following challenge by wild Glossina morsitans from an early age. Ann. trop. Med. Parasit., 60: 230-246.
Touré, S.M. 1974. Notes sur quelques particularités dans l'habitat de Glossina palpalis gambiensis Vanderplank observées au Sénégal. Rev. élev. Méd. vét. Pays trop., 27: 81-94.
Unesco. 1959. Vegetation map of Africa. London, Oxford University Press.
Van Hoeve, K., Onyango, R.J., Harley, J.M.B. & De Raadt, P. 1967. The cyclical transmission of Trypanosoma rhodesiense isolated from cattle to a man, a cow and a sheep. Trans. roy. Soc. trop. Med. Hyg., 61: 684-687.
Wells, E.A. 1972. The importance of mechanical transmission in the epidemiology of Nagana: a review. Trop. Anim. Hlth Prod., 4: 74-88.
Wilson, A.J., Dar, F.K. & Paris, J. 1972. A study of the transmission of salivarian trypanosomes isolated from wild tsetse flies. Trop. Anim. Hlth Prod., 4: 14-22.
Yesufu, H.M. & Mshelbwala, A.R. 1973. Trypanosomiasis survey in cattle and tsetse flies along a trade route in Southwestern Nigeria. Ann. trop. Med. Parasit., 67: 307-312.
K.J.R. MacLennan
Part I of this article, in World Animal Review No. 36, dealt in general terms with the problems of tsetse infestations and their effects on rural economies. This part discusses the general principles involved and the techniques in use in the control or eradication of tsetse infestation.
Successful anti-tsetse activities have included methods such as hand-catching and trapping (Principe, G. palpalis). game eviction or destruction (Uganda, Zimbabwe, G. morsitans, G. pallidipes), various kinds of tree-clearing and the use of insecticides, or combinations of some of these techniques. Game eviction has been the cheapest method of suppressing savanna tsetse populations (Wooff, 1968), but is repugnant to most thinking people and is now difficult to administer. It is unlikely to succeed in West Africa where savanna tsetse populations thrive on very low game densities and it is not effective against the riverine species. In practical terms most of the anti-tsetse activity in the savanna zones has been by tree-clearing and insecticides either alone, or, more usually now, in combination.
Because of the expense of clearing and the considerable maintenance commitment resulting from rapid regenerative growth, clearing activities in recent years have tended to be minimal, although, before modern insecticides became available, clearing operations against riverine tsetse were a major factor in the suppression of the sleeping sickness epidemics in Nigeria in the 1930s (G. palpalis and G. tachinoides). Though current anti-tsetse activity is mainly conducted through the use of insecticides, clearing still has a place in some sleeping sickness situations, as a means of protection on limited sectors of the perimeter of eradication projects (usually in combination with insecticides) and on the more intensive livestock projects.
Ground spray unit assembled for action in Nigeria. Such a unit can reclaim about 1 600 km2 (1 000 sq. miles) of G. morsitans and G. tachinoides infested land resource in one dry season operation.
A variety of clearing objectives are involved, ranging from sheer clearing to selective methods tailored co the habitat and behaviour preferences of particular tsetse species and to climate. However, as with insecticides, selective methods become more difficult to devise and execute in the more humid environments because the habitat relationships of the flies become progressively more diffuse; the flies are more invasive and less dependent on particular components of the tree cover. Selective techniques have proved most successful in the more stressful Sudan and northern Guinea zones (Keay, 1953; Davies, 1964; MacLennan, 1967 and MacLennan and Na'Isa, 1971).
Sheer clearings have an important place on limited portions of the perimeter of insecticidal tsetse eradication areas, which are particularly vulnerable to reinvasion. Mainly because of expense it is not usually feasible now to think in terms of clearings more than 1.6 km (one mile) wide. These are usually supported by periodic insecticidal applications on each side or by buffer zones on the infested side from which infestations are periodically eliminated by insecticidal applications. Sheer clearings can also be important on less-extensive livestock projects where intensive fodder production or pasture improvement leads to the suppression or elimination of woody growth. If only an absolute minimum of shade trees remain, potential tsetse resting places under these can be periodically treated with persisting insecticides and this can ensure that tsetse levels are kept sufficiently low for treatment/prophylaxis to successfully control livestock trypanosomiasis. The economic feasibility of this situation is finely balanced and successful implementation is probably limited to localities of higher quality soil types and favourable rainfall. Furthermore, livestock have to be kept at a reasonable distance from any infested perimeter woodland/forest, or some kind of insecticidal buffer zone created, or else the adjacent area used for arable production rather than for grazing.
The actual technique of clearing that is appropriate depends very much on the circumstances. Selective objectives and the maintenance of the treated area have to be met by hand methods. In practice the performance of motorized saws has been disappointing. In some places large-area sheer clearings have been created by chain-dozing the woodland. The technique may be applicable to the savanna of West Africa though there appear to be no good examples of this. The big problem is again the control of regenerative or successional growth of woody species. Arboricides have performed disappointingly in the subhumid zone though possibly it is worth persevering to discover how to use these agents more effectively. The dozing, windrowing and root-ploughing of woodland or savanna followed by utilization for intensive forage production and intensive grazing possibly offers the best prospect of success provided rainfall and soil characteristics are suitable. The areas that can be handled, however, will be relatively small in relation to overall livestock production needs and the magnitude of tsetse infestation. Sheer barrier clearings are usually done by hand while in riparian associations regrowth is eliminated by piling cut branches on the stumps and eventually burning these, having protected them in the meantime from inadvertent or premature burn and theft of the cut material.
Ground spray of G. morsitans dry season resting places in the Guinea vegetation zone
Principles and methods of use. A va�riety of principles and methods is involved.
Since tsetse flies do not feed on vegetation, and wildlife food sources cannot be systematically dosed, lethal doses of insecticide are acquired by the flies either from the places where they rest, or the surrounding atmosphere while they are in flight or at rest. Methods may be classified according to the duration of lethality of the insecticide emission (transient, or non-residual, as opposed to persistent) and according to the method of dispensing the emission — for example, from the ground or by aerial application. In practice, persistent emissions are applied both by ground spray (usually by knapsack equipment, either hand-pumped, pressure-retaining sprayers or motorized mist-blowers) and by air spray (helicopter) while non-residual emissions are usually dispensed from fixed-wing aircraft or, occasionally, by helicopter. A wide variety of options in equipment and material is available, suited to different circumstances. The situation has been comprehensively reviewed in "Insecticides and application equipment for tsetse control" (FAO, 1976b).
Ground spraying. In recent practice ground spraying has been performed by the relatively precise placement of persisting deposits by teams of carefully supervised knapsack sprayers (Davies, 1964; MacLennan, 1967).
A single placement of insecticide on preferred tsetse resting places is used; the insecticide remains lethal to any adult tsetse that rests on it for a duration at least as long as the pupal period in the ground. An adequate safety margin is required to allow for the fact that in practice all resting places cannot be covered. The deposit is most effective when a lethal dose is picked up and adheres to the feet of the fly after a single contact, but lethal doses can also be acquired through an accumulation of sub-lethal contacts resulting from the daily movements of the flies to a variety of resting places.
Persistence for about eight weeks should be sufficient. However, in locations where reinfestation pressure is continuous (control situations) and in buffer zones protecting eradication areas, the duration of persistence should be as long as practicable (Lycklama à Nijeholt, 1965; Mac-Lennan, 1967).
To achieve these degrees of persistence relatively heavy dose rates are necessary but these are applied in a restricted manner developed from a detailed knowledge of the ecology of the target Glossina species in the particular environment. The insecticide must be placed where flies will contact it when resting, that is, in the preferred resting places, otherwise it is wasted. Thus blanket applications are ruled out mainly because of the difficulty in distributing such a weight of insecticide, its cost and the impact on non-target organisms.
Two degrees of limitation are practised: first, "discrimination" or dispersal only within tree cover of components known to be frequented by the flies, and secondly, "selection" or limitation within such localities to that stratum of the vegetation and places within it on which flies customarily perch. In certain circumstances the localities are quite specialized and easily recognized once the basic studies of resting behaviour have been carried out. The technology is based on ecological studies initiated by Nash in the early 1930s and continued since then (MacLennan, 1967; MacLennan and Cook, 1972; Scholtz, Spielberger and Ali, 1976). It is most effective when discrimination is of the order of 10-18 percent of the infested area. When the degree of discrimination is 10 percent, a dose rate of 60 g/ha overall in the sub-Sudan zone has achieved extermination, the local application rate being 600 g/ha. Overall dose rates increase with the proportion of the area included in the discriminative application.
The method depends for its success entirely on the acquisition of a sufficiently detailed knowledge of the resting preferences of the tsetse species involved in the particular environment of the project (MacLennan, 1967; MacLennan and Cook, 1972; Scholtz,
Spielberger and Ali, 1976). Preferences vary very greatly from species to species and place to place (for example, that which is effective in the drier savannas does not work in the moister zones and that which is required for the elimination of savanna tsetse is very different to that required for riverine species).
Having decided upon the criteria for discrimination and selection and on the dosage rate, one usually applies the insecticide in linear swaths, but only to the vegetational features decided upon. The swath meanders with the vegetational feature and the width varies with local preferences and circumstances but may consist of four sprayers spaced at intervals of about 10 metres advancing in line and following the features decided upon
A single swath may follow both sides of minor drainages and vegetational interzones, particularly ecotones between savanna woodland and tree savanna, or one swath may be needed on each side of larger drainages. All drainages should be followed to their origins. For riparian forest ribbons the swath may be split between the outer and the stream-side margins of the forest and one swath may be needed for each bank. When dealing with savanna species a sharp look-out has to be kept for smaller habitat features such as small tree and thicket clumps associated with termitaria and footslope ecotones that are visible from the swath line and sprayers should be detached to deal with them. If substantial blocks of woodland are not penetrated even by minor drainages or tracks followed by spray groups, swaths should be sprayed at intervals of about 150 metres.
The groups spraying the swaths consist of labourers operating the pumps. Two men may be allocated to each pump and they alternate between actual spraying and carrying insecticide supplies for the man spraying. The groups are directed by trained field staff of the Control Assistant or Scout category, these in turn being supervised by Field Assistants and the whole operation in the field being under the direction of a Control Officer. The terminology varies from country to country but the chain of direction is similar in most places. The Control Officer is very dependent upon adequate maps on which the operation is planned and daily progress is recorded. Usually special maps have to be prepared for this purpose.
To gain access it is necessary to prepare in advance a network of rough motorable tracks and to cut access paths into denser vegetational features. The operation also requires an adequate number of lorries, pick-ups and tractors with trailers to transport personnel, insecticide and water.
The organizational or logistical commitments involved in the treatment of large areas are substantial but an operation unit of manageable size has been able to deal with 1 600 km2 of G. morsitarts infestation in the sub-Sudan zone in Nigeria in one operational season. At that time Nigeria had four such units operational.
As a means of eradicating both G. morsitans and riverine tsetse infestations this approach has been highly successful in the Sudan and sub-Sudan vegetation zones but, although successful, it has proved more costly and more difficult to apply effectively in the northern Guinea zone. Much less is known of the relevant aspects of tsetse ecology in more humid locations and there are as yet no indications that similar approaches can be developed for large-area eradication in the southern Guinea or moister areas. Not only are relationships between the tsetse population and the vegetative environment much more diffuse and the tsetse more invasive (Molyneux, Baldry and Fairhurst, 1979) than in areas of greater climatic stress for Glossina, but the persistence performance of insecticide deposits is much lower in moister environments. This leads to the use of heavier deposits more generally applied and to sharply rising costs. Even between the Sudan and northern Guinea vegetation zones there exist very important differences in the resting behaviour of G. morsitans and in the persistence behaviour of insecticides.
Though further field study may reveal how this method could be applied to the southern Guinea zone and more humid locations for the elimination of tsetse populations, the outlook is not at present encouraging, though clearly the approach is appropriate in control rather than eradication.
The method is technically and financially efficient at levels of discrimination of 10 percent but if more than about 18 percent of the area has to be covered, costs escalate in proportion and there is greater concern about the environmental side-effects of the weight of insecticide being dispensed over wide areas. The annual rate of progress of an operational unit diminishes in proportion to any increase in the percentage of the area included in the discriminative spray.
Helicopter applications. Since the basis of the ground spray technique is the application of insecticide along linear swaths oriented to preferred tsetse habitats it is evident that a similar arrangement could probably be achieved by helicopters and that this would reduce the formidable organizational problems inseparable from large-area ground spray operations. Such a procedure would speed up the rate of reclamation of land resources from tsetse infestation. The concept has been made financially feasible through the development of ultra-low-volume (ulv) insecticide formulations and spray equipment suitable for use on helicopters, reducing very considerably the weight of the payload, which would be prohibitive if conventional insecticide formulations were used (Lee, 1977; Spielberger, Na'Isa and Ardurcahim, 1977).
Small helicopters are used (e.g., the Bell G4A) which follow closely the same localities as those visited by ground spray units. The helicopter flies at a ground speed of 32-40 km per hour (20-25 mph) and 24-32 km per hour (15-20 mph) over forest so that there is an adequate downdraught to achieve a dispersal of insecticide within and below the tree canopy. The machines fly as close as possible to the canopy (1-2 m) and operate from a large number of small temporary clearings made for landing and spaced to reduce non-spray ferrying time to the absolute minimum. The effective swath width in this case is about 20 metres.
The emission equipment consists of six small, electrically operated and electronically controlled rotary disc atomizers mounted on a 3.2-m transverse boom under the rotor downdraught. Droplet size (determined by the rate of flow and viscosity of the insecticide and the rotation speed of the atomizers) is of crucial importance. Droplets must be sufficiently large to give a deposit with residual activity yet sufficiently numerous to give adequate coverage without being so light that they do not strike. Droplet size is also critical with regard to canopy penetration. Because of this, and the relatively small volumes of insecticide being dispensed, the metering of flow rates and atomizer speeds and the physical characteristics (viscosity, volatility) of the insecticide are of crucial significance to success. Adequate control is achieved through sophisticated electronic control systems and specially designed insecticide formulations. Ground speed becomes a crucial factor also in area insecticide dosage rates. Following any alteration to the several variables it is essential to monitor the droplet spectrum to determine that it conforms to specifications. For this purpose it is as well to seek independent specialist assistance at, or near, the commencement of the spray season.
Motorized knapsack mistblowers in use against infestation in dense habitats
Livestock at Matyoro Lakes in northern Nigeria, formerly heavily infested with G. morsitans, G. tachinoides and G. palpalis. Freed of infestation by knapsack spraying of DDT and dieldrin
The technique is very sensitive to meteorological conditions. After the earth is warmed by the sun a condition of turbulence develops in the lower air, which can carry small droplets upward and disperse them. Ideally a condition of temperature inversion should exist (ground temperatures lower than those at say 7.0 m). Because operations cannot be conducted in darkness, spray activity is effectively restricted to short periods around dawn and dusk. This can be extended briefly by increasing the insecticide flow rate from 4 to 5 litres per hectare, and the droplet sizes by reducing the atomizer rotation speeds. Atomizer speeds of 7 000 rpm will give the generally preferred droplet with volume median diameter of 150 (p.m, while reducing the rpm to 6 500 gives the larger droplet of 170 [pm needed when less favourable meteorological conditions develop. This increases droplet recovery at ground level by 25 percent. However, once turbulence becomes anything but minimal, or if wind speeds in excess of 7 m per second are encountered, spraying can become ineffective. It is emphasized that these figures relate to a particular insecticide formulation and that different criteria will apply to other formulations which will have different physical characteristics.
Helicopter in action returning from spray sortie
One of six electronically controlled spinning-disc atomizers mounted on helicopter spray boom
Organizational aspects of deploying and maintaining helicopters in the bush are formidable but, generally speaking, the requirements for labour and transport are much less than for ground spray of a similar area although the cost of helicopter hire more than offsets the saving on labour and transport. A two-machine spray unit can handle about 3 625 km2 of infested area if the degree of discriminative spray is 10 percent, provided that the unit is smoothly serviced with supplies of insecticide, fuel and spare parts.
The helicopter spray unit requires the same degree of support in regard to pre- and post-spray tsetse surveys, map making and operational planning as does ground spraying and, in practice, it has usually been necessary to protect helicopter sprayed areas by ground spray methods at points where there are particular perimeter re-invasion threats.
In comparison to ground spray, higher local and overall dosages, are required since the element of selection within the swaths, discriminately sprayed, is minimal. As a result of this the impact on non-target organisms is substantially greater. Local dose rates in the Guinea zones are of the order of 800-1000 g/ha. If the degree of discrimination is 10 percent this results in overall dose rates of 80-100 g/ha of infested area. This will be proportionately higher at higher percentages of discriminate spray.
Both ground spray and helicopter applications become progressively less cost effective as the percentage of the area included in discriminative spray rises and, possibly at about 18 percent, is near the upper limit for both methods. At higher rates both cost and environmental aspects give cause for increasing concern. However, where G. morsitans advances are in progress or rapid action is required to reclaim an infested land resource there is no proven option, at present, other than helicopter spraying, which can be applied in the Guinea zones in cases where there is insufficient time to train, build up and equip the large organizations required for ground spraying.
Helicopter spray techniques have eliminated G. morsitans, G. tachinoides and G. palpalis infestation from substantial areas (Spielberger, Na'Isa and Ardurrahim, 1977). G. morsitans foci have sometimes persisted and have required follow-up applications that have usually proved successful. In practice the method has worked well for the extermination of riverine tsetse infestations in the northern Guinea, sub-Sudan and Sudan zones. Similar limitations on ground spraying exist in more humid areas.
Helicopter applications of non-persistent aerosols, though relatively costly, are feasible technically and could be considered appropriate to certain circumstances (van Wettere et al., 1978).
Fixed-wing aircraft applications. Sequential emissions having a transient effect are blanketed at intervals in time and space over the whole of a project area (MacLennan, 1967). The flies acquire the lethal dose from the surrounding air, not from the surfaces on which they rest, as is the case when using deposits with residual activity. All, or nearly all, adults are eliminated following each application.
The applications have to be carefully timed to ensure that no flies that emerge from the ground following an application live long enough to deposit a viable puparium, and they must be repeated at this interval until the flies are beyond the end of the pupal period in the ground, by which time all puparia deposited before the spray cycles started have hatched. A knowledge of the timing of the reproductive cycle is, therefore, crucial for success. Both the period to first larval birth and the pupal period in the ground vary significantly according to prevailing temperature, becoming shorter at higher temperatures. Usually about five or six applications are required at intervals of about 10 to 19 days for a period of about two months. The mean periods can be calculated from average shade temperature to an accuracy of 10 percent (Phelps and Burrows, 1969) from tables compiled by Mulligan and Potts (Glasgow, 1970).
Small, light aircraft are used, of a kind that can take off fully loaded from bush landing-strips located as close as possible to the operational area.
Project areas are preferably rectangular in shape, with the aircraft flight lines oriented transversely to the prevailing wind. The aircraft flies at a speed of about 155 knots, as close as possible to the tree canopy — or about 7.0 m. As in helicopter spraying, a degree of skill and understanding is required of pilots that is of a much higher order than that required for ordinary agricultural crop spraying.
The project is covered progressively by parallel flight lines spaced at intervals of 300 m (sometimes less). The flight lines are controlled at each end from the ground, from cut lines made before the spraying commences and pegged off at the selected flight-line interval by personnel equipped with two-way radios and spotlamps. This personnel is responsible for calling a repeated flight on lines going wider than the acceptable minimum.
Very accurate navigational equipment is required to control both direction and distance. This operates either on the radar-doppler principle or from global, very low frequency radio beacons. If this equipment is sufficiently accurate with regard to direction and distance, a single-ended marking can be adopted, though some reservation has to be expressed regarding a complete lack of monitoring at one end of the flight lines. In some circumstances it is feasible to work from a single, central cut line, but again similar reservations apply.
Meteorological conditions are even more critical in aircraft applications than with helicopter spraying. Normally spraying is restricted to a short period of about one hour around dusk and dawn, while temperature inversion is still present, but, in flatter terrain, the general efficiency of the operation can be greatly increased by night flying, using a special beam-light fitted to the nose of the aircraft. The harmattan wind can be a problem, as in the case of helicopter spraying.
If a single aircraft only is available the operation becomes very vulnerable to any delays resulting from equipment failure. There is little scope for flexibility in timings and if there is slippage beyond the period to first larval birth the whole series of cycles has to start again from scratch.
There is a distinct limitation on flying over broken terrain. Since aircraft have to fly as close as possible to the tree canopy, the operation is best carried out in flatter types of country; it becomes hazardous in more broken country and quite impracticable in hilly country.
The aircraft usually flies at a speed of about 155 knots, emitting insecticide concentrate at the rate of 4-5 litres per minute (depending on the dosage required and the concentration of the active ingredient). The formulation is emitted from a single wind-driven rotary atomizer spinning at about 8 000-12 000 rpm (depending on the physical characteristics of the formulation) to obtain a droplet emission of 30 [pm volume median diameter.
As in helicopter spraying the specifications are critical and careful monitoring of relevant aspects is essential to success. The technology of the emissions has been progressively developed since the early 1950s and is now highly refined, though there is still something to be learned regarding the effects of volatility and drift on optimum flightline intervals. A particularly important feature of the method is the very low dosage of insecticide used: it is of the order of 6-12 g/ha of project area for each spray cycle or 36-72 g/ha in total if six applications are made. The current use of aircraft, both fixed and rotary wing, for anti-tsetse work has recently been reviewed in detail by Lee (1977).
The result of aircraft applications on tsetse populations has been a rapid decline to zero, or something near it, in most instances followed, in some places, by a remarkably slow recovery, the population remaining reduced by about 90 percent for a period of at least three years. Unfortunately, there are no areas of a sufficiently large size that have remained completely free of infestation for long enough for it to be categorically stated that eradication of G. morsitans has been achieved by this technique, and there are certainly several locations where eradication has not been achieved. However, there is still scope for much further study on the entomological aspects of spraying by aircraft and a real prospect that further investigation of the survival and drift phenomena will ensure that this method regularly achieves extermination, although this is only possible under suitable conditions of terrain and climate. An additional impact could be produced by reducing the inter-spray cycle intervals to about 10 days, which, recent studies indicate, would ensure that no insemination takes place after the operation commences. There is no experience of the effectiveness of the technique over forest habitats and it would be wrong at this stage to dismiss the possibility of success. The point is of considerable importance since, at present, it is the only approach that can be utilized in these locations that is capable of striking and entering into the moister parts of the sub-humid zone.
However, regarding the current state of knowledge it must be said that the technique has been developed in environments differing in most respects markedly from the moister savannas, and has been used only against G. morsitans centralis. The as yet limited experience of the technique against G. morsitans submorsitans and G. palpalis of the sub-humid zone has been termed "encouraging" but it is still too early to know its actual potential for control or eradication of these species. However, since discriminatory techniques, which are known to achieve extermination, are reaching the limits of their capability in the more humid zone, it is essential to proceed with further development of the fixed-wing spraying technique.
Compounds and formulations. For ground spraying, only ddt and dieldrin have proved fully effective in the field over the years. The former is often used as a wettable powder containing 75 percent active ingredient mixed with water to give a suspension containing 2.5 to 3.75 percent of active ingredient. ddt has the merit of being exceedingly low in its toxicity for humans and it is possibly as a result of the prolonged biological activity at the locations to which it has been very selectively applied that tsetse eradication operations in the Sudan zone have proved so successful. Persistence of activity against Glossina can be demonstrated in situ for periods well in excess of 12 months (Lycklama a Nijeholt, 1965; and MacLennan, 1967). However, this degree of persistence is obtainable only if it is applied to bark in the dry stage in the Sudan zone and drier parts of the sub-humid zone. Performance falls off rapidly in moister environments where it may fail to achieve results even at much heavier dosages.
Light aircraft being prepared for fixed aerosol spraying in Botswana
Wind-driven spinning atomizer. One only is required on wing trailing edge inboard of engine.
In more-humid locations, ground spray operations have used dieldrin in emulsion form at strengths of 2-4 percent derived from 20-25 percent emulsion concentrate. The actual nature of the formulation of the concentrate has been shown to be of importance and, in some countries, a special 18-percent formulation is in use.
In denser habitats, such as swamp forest and thicket, it is sometimes necessary to use motorized knapsack mist-blowers in order to get an adequate dispersal of insecticide within the infested habitat (Koeman et al., 1971). For this purpose a dieldrin emulsion is usually used at a final concentration of 2-4 percent active ingredient derived from 20-25 percent emulsion concentrate.
In helicopter spraying of residual deposits, special ulv formulations of diedrin (18 percent) and endosulphan (25 percent) are used, designed particularly to meet the requirements of anti-tsetse work. It is quite clear that failure will result unless criteria relating to droplet deposits (such as size, volatility, crystalline or surface availability on leaves or bark) are met.
For fixed-wing spraying of aerosols, endosulphan concentrates, sometimes diluted with a suitable solvent, are used at concentrations of 18-35 percent depending on the dosage rate required.
Over the course of the years all potential substitute compounds for these have been tried in the field. In spite of encouraging activity demonstrated in laboratory tests, none (with the exception of some of the new synthetic pyrethroids) has shown any capability of replacing ddt, dieldrin and endosulphan. Certainly, as regards biological effectiveness, there does at last seem to be some prospect of substituting the only compounds that have proved effective over the years with synthetic pyrethroids.
Direct environmental side-effects. These result from the effects of the insecticide emissions on organisms other than Glossina and can be immediate or longer term.
The restricted nature of ground spray means that the immediate impact is mainly restricted to organisms that occupy the places to which the insecticide is applied — which, ideally, may be 3-10 percent of the total project area. The fauna that occupies the same niche is affected, particularly insects and insectivores. Animals that are not insectivorous show little if any change and those, including insectivores, that frequent other niches, such as the canopy, are little affected. Some aquatic creatures do suffer an immediate impact. However, the majority of species in the area are not affected, and those which are mostly re-establish themselves from neighbouring unsprayed localities; the rate at which this happens depends on the invasive capability of the animal (FAO, 1977a; and Koeman et al, 1971). Two bird species in particular are slow to re-establish, namely the snowy crowned robinchat and the bluebreasted kingfisher. The latter is migratory and its permanent absence can be attributed as much to habitat alteration resulting from changed land use as to the direct effects of the pesticide.
Though it has been searched for, it has not proved possible in Nigeria to find any significant accumulation problem with ddt or dieldrin of a magnitude that would indicate that these compounds are having a significant, deleterious impact in those tsetse eradication areas where the desired result is achieved mainly through a single application of insecticide. It is different in control situations, where application may be required annually, but the number and extent of these are much more limited. It must also be borne in mind that side-effects will be greater in magnitude as the degree of restriction is relaxed, a point favouring fixed-wing application in more humid areas.
In the drier part of the sub-humid zone (the Matyoro Lakes in the sub-Sudan zone) ddt and dieldrin applied in 1961 as a restricted ground spray in a complex of woodland, forest and swamp forest eradicated G. morsitans and G. palpalis and possibly also G. tachinoides (MacLennan and Aitcheson, 1963). Following an investigation in 1970 of immediate and shortterm impacts in a neighbouring area after dieldrin applications, the 1961 area was revisited. All bird species noted to have been affected by the 1961 spray were present and fish were exceedingly numerous (and good to eat) in the small lakes of the system, which have only a very limited seasonal outflow.
The lack of an appreciable accumulation problem can be ascribed to the fact that, in eradication operations of this nature, the result is usually achieved by a single application and to chemical degradation resulting from high ultra-violet values and other features of the tropical environment
The immediate environmental side-effects of helicopter applications are substantially less specific since, although the discrimination is similar to that in ground spraying, the same degree of selection is not, and there is a significant deposition in canopy and aquatic habitats. The immediate impact on non-target organisms is more severe; some heavy mortalities result and some species are slower to reestablish (Koeman et al., 1978). Longer-term accumulation problems are probably not very different to those following ground spraying. The direct impact is mitigated by having ground-sprayed blocks within the helicopter-sprayed area, or adjacent to it, by having unsprayed habitats adjacent to the area and by being careful not to include the whole of any particular type of habitat in the area sprayed (FAO, 1977a). Furthermore, the toxicity ranges of dieldrin and endosulphan differ: for example, bees are numerous in endosulphan areas 12 months after spraying while fish are less affected by dieldrin applications. Some reduction in the undesirable impact can be achieved by alternating blocks sprayed by either of these two compounds.
Fixed-wing applications of endosulphan have been studied over a period of three years in the Okavango delta in Botswana. No mortalities directly attributable to the applications have been observed in any species other than Glossina, Temporary accumulations of endosulphan, below levels of biological significance, have been observed (Russell-Smith, 1977 - personal communication; and Wood and Turner, 1975) in some species of fish and there is also some indication of transient behavioural responses in some species of fish and ants. The lack of side-effects is possibly not surprising when it is realized that dosages are of the order of 6-12 g/ha per application, much of which does not reach the ground. Glossina is particularly sensitive to this compound dispensed in this manner and it is again emphasized that there is no observable mortality of the other numerous Diptera present in sprayed areas, nor of any other animal studied so far.
Usually those engaging in anti-tsetse activity probably rank with the more responsible users of insecticides. Seldom, if ever, have they resorted to large-scale and repeated dispensations of the type that have caused trouble in the USA (Carson, 1961) and in Europe and they have shown no complacency about the continued use of ddt and dieldrin. An overall, balanced judgement is required (Provost, 1972). It is still necessary in some circumstances to use ddt for medical and agricultural reasons and it is not unusual in tropical agriculture to find application rates of 1 000 g/ha, sometimes repeated as many as five times in one growing season and repeated year after year, and to find that compounds such as ddt are a constituent of the applications. Application rates and usage practice in anti-tsetse work have been very different.
The basic objective of large-area anti-tsetse activity is often to facilitate land usage. If this objective is achieved, there is an unavoidable loss of wilderness habitats where these are not specifically conserved. The faunal changes that result from such a development are often of a much greater degree than those arising directly from the application of the insecticides and are of a more permanent nature.
Anti-tsetse techniques are still being improved and new ones developed and at last there is a real prospect of alternatives to the chlorinated hydrocarbons becoming available.
The most promising of the longeracting pyrethroids (decamethrin) has been found to be highly toxic to some fresh-water crustaceans but preliminary experiences show that these are again numerous after 12 months (Spielberger, 1978 - personal communication).
West African N'Dama exhibiting high degree of tolerance to many West African strains of trypanosomes
West African Dwarf shorthorn (Muturu) cattle. These exhibit a high degree of tolerance to local strains of trypanosomes.
There is a considerable diversity in the type of action against Glossina that will be appropriate under the different circumstances within the sub-humid zone. This will range from urgent ad hoc medical and /or veterinary measures (mainly diagnostic and therapeutic and sometimes augmented by vector control activity) required to contain the immediate consequences of a tsetse infestation affecting established livestock production or resulting in a sleeping sickness epidemic, to longer-term actions designed to eliminate infestation from very substantial areas so that planned development of the land resource can proceed without the serious complications that persisting tsetse infestations can cause. Though continental-scale eradication is not feasible, more local eliminations are feasible in appropriate circumstances and have been achieved in the past.
Anti-tsetse activities have two very different aspirations, namely control or local eradication.
In local eradication, a vector-free status is achieved and maintained throughout a large project area. The area is chosen so that tsetse elimination can proceed in a series of annual phases to embrace the entire local tsetse-belt or to take the project perimeter to locations that can be defended for an acceptable expenditure of funds and resources in relation to the area being protected. In control, activity is focal, related to specific sleeping sickness or livestock requirements and, since tsetse-free islands dispersed in a sea of infestation are relatively difficult to maintain and defend, a continuing, permanent commitment develops over a large area. This commitment involves vector and disease surveillance and countermeasures as appropriate. When disease episodes occur, the correct course of action is not always easy to formulate since it is often not readily determinable if the episode is due to a tsetse advance, to an occult infestation, drug resistance or maladministration of medicaments. The continuous deployment of scarce technical resources and the substantial supervision commitment are both difficult and expensive to execute and will continue permanently.
In the eradication project the supervisory commitment within the area becomes progressively less until it either takes place at increasing intervals or becomes oriented to the vulnerable sector of the perimeter, if this does not extend to the local limit of infestation. The technical expertise can then be deployed in extension of the fly-free area rather than becoming increasingly embroiled in the supervision of scattered control commitments and the elucidation of perplexing disease outbreaks.
However, the local eradication option can only be exercised where this is feasible in technical and financial terms. In this connection it has to be strongly emphasized that the possibility of formulating an eradication plan cannot even be perceived until a detailed knowledge of tsetse distribution has been acquired and a local field capability exists that is able to do this, collect basic information on habitat relationships and execute, or provide, field support for the anti-tsetse operation. In the absence of this field capability not only do possibilities of action remain obscured but, even if they should be known to exist, they cannot be executed in spite of the availability of financial and technical assistance that could be drawn upon.
It is essential to have an anti-tsetse technique that has been proved to achieve extermination of the particular species in that environment and this is usually developed from control activities or pilot projects. In the absence of a national field capability this experience also cannot be acquired nor can the organization be trained.
The facts of Glossina distribution and biology make possible the attainment of an eradication objective over very substantial infested areas of the savanna grazing resources where distribution is not homogeneous. The objective has been attained in several countries, and experience in Nigeria indicates that the objectives can be attained, though not so readily, in the drier parts of the sub-humid zone. The techniques have been developed after many years of field investigation and there is no reason to believe that, after similar expenditure of effort on investigation and technological development (materials, equipment, methods) eradication areas will not be progressively extended, though at present, as has been explained previously, existing, proven technologies are near the limits of their capabilities.
Where eradication objectives are unattainable at present, control situations have to be accepted. In terms of production, that which is feasible in the case of susceptible livestock by treatment or prophylaxis is determined by the degree of risk. As has been explained, higher-level production is usually not feasible except in areas of low risk.
Treatment and prophylaxis. This is appropriate in conditions of low tsetse risk but, in areas of high risk, effective production cannot be ensured by such means unless accompanied by vector suppression (Finelle, 1976).
The difficulties arise from the fact that diagnostic techniques in the individual require a degree of simple laboratory support, that techniques are not completely reliable, and that they require the handling and bleeding of animals. The diagnosis of a trypanosomiasis problem is more readily made on a herd basis and the whole herd dealt with as appropriate.
Drug resistance can be an economically significant problem under static management systems dependent for success on a specific level of production. Repeated treatment in the presence of infestation, particularly if it is livestock-oriented, favours the development of a resistance problem. However, there is still some scope in situations of low tsetse risk, through the alternation of appropriate trypanocides, of reducing the impact of resistance even though, by now, crossresistance to several trypanocides is widespread. It seems that the unstable nature of the resistance phenomenon might soon lead to greater degrees of crossresistance between the currently available trypanocides. No new compounds are available to counter this, most of those available can be significantly toxic to one class of animal or another and there is no generally suitable trypanocide for T. simiae infections in pigs (even if there were one it would usually not be possible to administer it in time to save the lives of the animals).
It has been the usual experience that block treatment of entire groups of cattle at appropriate intervals gives better results than waiting for an individual positive diagnosis or until a percentage of the group becomes positive (this latter procedure should not be practised). Surveillance for infection of the entire group on a fixed management system is essential at regular intervals. The treatment of individual trypanosomiasis cases as they arise works in practice only in situations where tsetse risk is minimal, accurate diagnosis is swift and effective treatment follows promptly. The practical difficulty and cost of doing this, as, for example, on extended range, or on a ranch, or in servicing a dispersed group of smallholders with work-oxen, should not be underestimated, nor should the degree of technical expertise required to handle both animal health and the entomological aspects. Pastoral transhumants can accept greater risks and lower levels of production than is practicable under more static management systems. Properly controlled, closely supervised, mass curative treatment and strategically timed prophylaxis have done much to mitigate annual losses from trypanosomiasis and to increase the dry season utilization of infested areas. However, since drug resistance is of widespread occurrence, this practice has to be used with great care and also to ensure that treatment is not just encouraging owners to linger in high-risk areas. Thus, it can be appreciated that the careful supervision and management in the "control" situations require levels of expertise in the animal health and entomological aspects that are either not readily available or can only be deployed in relatively limited areas. Indeed, breakdowns resulting in serious disease outbreaks are not unusual even on very high level management projects. It is also evident that the needs of transhumants are not being fully met in a number of countries for a variety of reasons. These are some of the realities that have to be faced when deciding whether to proceed with the establishment of tsetse eradication areas where these are technically feasible or a control commitment where eradication is not feasible.
Though the difficulties involved have been emphasized, the trypanosomiasis problem can be adequately countered provided the risks have been correctly assessed as being of an acceptable degree and the animal health and entomological expertise is available to carry out the required countermeasures against the disease and, when necessary, against the vectors.
West African shorthorn in Ghana. These exhibit a high degree of tolerance to local strains of trypanosomes.
Tolerant and non-susceptible livestock. Stock that are exotic to the tsetse zone in origin are particularly susceptible to trypanosomiasis but, within the zone, there are breeds of cattle (Pagot, 1974), pigs, sheep and goats, horses and donkeys that exhibit degrees of tolerance of infection. As with treatments/prophylaxis the degree to which the phenomenon can be utilized to achieve effective production is related very significantly to the severity of the risk from Glossina. Constant exposure to high-risk situations and the intervention of "stress" periods (Stephen, 1966) can lead to clinical breakdown and sometimes death. Furthermore, it is usually the case that, until they have become infected and the immune defences activated, these animals are about as vulnerable to infection as any other animal (Roberts and Gray, 1973).
Trypanotolerant cattle, once adjusted to the local trypanosomes, produce meat effectively where fully susceptible animals would die. Daily weight increase and time to maturity of selected types under good management and negligible or unspecified degrees of tsetse risk can be impressive (Pagot, Coulomb and Petit, 1972; and Roberts and Gray, 1973). However, the working and milking capabilities of the cattle are usually low and animals reach optimum average slaughter weights at a size and time when the average zebu is near the rate of peak daily weight gain. There is still much to be learned about the production and reproductive capability of these animals (Stephen. 1966) under conditions of moderate and high tsetse risk, and there is no straight comparison available of the production per unit area from closed herds of zebu and tolerant breeds under comparable conditions in which Glossina is absent. All this information is essential to any assessment of the extent to, which the use of these animals can mitigate the trypanosomiasis problem or rather whether it would be better to pursue actions that make possible production from susceptible stock. There is much to be learned but though there is clearly an important potential for the appropriate use of tolerant animals under certain circumstances, it cannot be suggested that they provide a complete solution in situations of high trypanosomiasis risk. Their ability to cover ground is less than that of the zebu, though they appear to be much more resistant to streptothricosis. In any case these animals are in short supply.
There appear to be no livestock capable of producing meat that are not susceptible to tsetse-transmitted trypanosomiasis, other than poultry, fish and, in practice, rabbits.
Housing. Under zero-grazing systems susceptible livestock can be isolated from risk by keeping them in fly-proof accommodation when this is practical and economically feasible (as might be the case in pig production units and certain dairy units).
Sterile male release. The method is still under development and has the important advantage that it becomes more efficient, in the biological sense, at lower tsetse densities and has no detrimental direct side-effects on other organisms. However, efficient breeding colonies of each of the Glossina species to be eliminated have to be established and, though real progress is being made on colonization techniques, the individuals to be released are still costly to produce. Natural males have to be outnumbered 3:1 by sterile males, which can cost about US$ 1 to produce. Natural savanna tsetse populations have first to be reduced by non-persistent insecticide applications. In the sub-humid zone it is possible that the method may be applied more readily to the riverine species. However, regardless of the species, sterilized males have to be released regularly throughout the area for a period of several months, which, in some circumstances, could prove a major obstacle in areas of low human population density. This aspect, however, could prove no more difficult than the deployment of ground spray teams. A considerable logistic support, as in ground spraying, would, of course, be required.
Growth-regulating hormones. Compounds are available that have been shown to affect, for example, maturation of the puparium. Until such time as a powerful attractant is available for use on Glossina there exists an insurmountable problem in applying the compound.
At tract ants. There is as yet no powerful, long-range attractant though some promising advances in developing one are being made. The availability of such an agent would greatly improve the efficiency of survey and detection methods, would make possible the sterilization of natural populations rather than having to produce individuals from colonies and would make possible other channels of attack such as the use of hormones. In the present state of knowledge the availability of an effective attractant that could be used in the field would contribute more than anything else to our ability to deal effectively with the tsetse problem. ■
References
Carson, R. 1961. Silent spring. Boston, Houghton-Mifflin. 308 p.
Davies, H. 1964. The eradication of tsetse in the Chad river system of northern Nigeria. J. appl. Ecol., 2: 387-403.
FAO. 1977a. The environmental impact of tsetse control operations. Rome. FAO Animal Production and Health Paper No. 7.
FAO. 1977b. Insecticides and application equipment for tsetse control. Rome. FAO Animal Production and Health Paper No. 3.
Finelle, P. 1976. Chemotherapy and chemoprophylaxis of African trypanosomiasis. Rome, FAO. Report AGA: TRYP/WP/76/28.
Glasgow, J.P. 1970. In The African trypanosomiases, ed by Mulligan and Potts, p. 352-353. London, Allen and Unwin.
Keay, R.W.J. 1953. An outline of Nigerian vegetation. 2nd ed. Lagos, Government Printers.
Koeman, J.H., Rijksen, H.D., Smies, M., Na'Isa, B.K. & MacLennan, K.J.R. 1971. Faunal changes in a swamp habitat in Nigeria sprayed with insecticide to exterminate Glossina. Netherlands J. Zool., 21: 434-463.
Koeman, J.H., Den Boer, W.M.J., Feith, A.F., De Jongh, H.H. & Spliethoff, P.C. 1978. Three years' observation on side effects of helicopter applications of insecticides used to exterminate Glossina species in Niger. Environ. Pollut., 15: 31-59.
Lee, C. 1977. New developments in tsetse fly control using aircraft. Agric. Aviat., 18: 6-17.
Lycklama À Nueholt, P. 1965. Persistence of dieldrin deposits on bark applied in various concentrations. Document, 12th meeting of the International Scientific Council for Trypanosomiasis Research and Control. Bangui, Organization of African Unity. Publication No. 102.
MacLennan, K.J.R. 1967. Recent advances in techniques for tsetse-fly control. Bull. Wld Hlth Orgn, 37: 615-628.
MacLennan, K.J.R. & AITCHESON, P. 1963. Simultaneous control of three species of Glossina by the selective application of insecticides. Bull. ent. Res., 54: 199-211.
MacLennan, KJ.R. & Na'Isa, B.K. 1971. The current status and future prospects regarding tsetse extermination in Nigeria. Lagos, Organization of African Unity. Publication No. 105, p. 303-309.
MacLennan, K.J.R. & Cook, A.G. 1972. The resting behaviour of Glossina morsitans sub morsitans in the northern Guinea vegetation zone in relation to control using insecticides. Entomologist, 105: 144-152.
Molyneux, D.H., Baldry, D.A.T. & Fair-hurst, C. 1979. Tsetse movements in wind fields: possible epidemiological and entomological implications for trypanosomiasis and its control. Acta tropica, 36: 53-65.
Pagot, J. 1974. Les races trypanotolerantes. Actes du Colloque, IEMVT, Maisons-Alfort.
Pagot, J., Coulomb, J. & Petit, J.P. 1972. Revue et situation actuelle de I'emploi des races trypanotolerantes. WHO/FAO African Trypanosomiasis Information Service. TRYP/INF/74.57.
Phelps, R.J. & Burrows, P.M. 1969. Prediction of pupal period of G. morsitans orientalis under field conditions. /. appl. Ecol, 6: 323-337.
Provost, M.W. 1972. Environmental hazards in the control of disease vectors. Env. Entomol., 1: 333-339.
Roberts, C.J. & Gray, A.R. 1973. Studies on trypanosome resistant cattle, Parts I and II. Trop. Anim. Hlth Prod., 5: 211-233.
Scholtz, E., Spielberger, U. & Ali, J. 1976. The night resting sites of the tsetse fly Glossina palpalis palpalis (Robineau-Desvoidy) (Diptera: Glossinidae) in northern Nigeria. Bull. ent. Res., 66: 443-452.
Spielberger, U., Na'Isa, B.K. & Ardurrahim, U. 1977. Tsetse (Diptera: Glossinidae) eradication by aerial (helicopter) spraying of persistent insecticides in Nigeria. Bull. ent. Res., 67: 590-598.
Stephen, L.E. 1966. Observations on the resistance of West African Ndama and Zebu cattle to trypanosomiasis following challenge by wild Glossina morsitans from an early age. Ann. trop. Med. Parasit., 60: 230-246.
Wettere, P. van, Baldry, D.A.T., Molyneux D.H., Clarket, J.H., Lee, C.W. & Parker, J.D. 1978. The experimental application of insecticides from a helicopter for the control of riverine populations of Glossina tachinoides in West Africa. IV. Evaluation of insecticides applied as aerosols. PANS, 24: 433-446.
Wood, A.B. & Turner, A. 1975. Environmental effects of the endosulphan spraying programme against tsetse flies in Botswana. London, Tropical Products Institute.
Wooff, W.R. 1968. The eradication of the tsetse Glossina morsitans and Glossina pallidipes by hunting. Document, 12th meeting of the International Scientific Council for the Eradication of Trypanosomiasis. Bangui, Organization of African Unity. Publication No. 102.
The author's address is Balchraggan, 2 Craigrory, N, Kessock, Inverness IVI IXB, United Kingdom.
The author's address is Balchraggan, 2 Craigroiy, N. Kessock, Inverness IV1 1XB, United Kingdom.
