It is now nearly a century since systematic research started on tsetse, the principal vector of animal trypanosomiasis or “nagana” and human “sleeping sickness”. This major research effort was the consequence of the outbreak of a traumatic epidemic of sleeping sickness which swept through Uganda and what is now western Kenya in 1901, leading to the depopulation of that region through death and out-migration; the then colonial government appointed a Royal Commission to find a lasting solution to this tropical disease of crucial social and economic importance. importance. In spite of large-scale control campaigns carried out since that time in many parts of Africa—by colonial governments (British, French, Portuguese, and Belgian) in the first half of this century; and since then by specialized agencies of international organizations (the Food and Agricultural Organization of the United Nations, the International Atomic Energy Agency, etc.), and several aid agencies (U.S. Agency for International Development, The Netherlands Government, the Canadian International Development Agency, etc.)—the lasting solution to trypanosomiasis and tsetse vector control which we started seeking some 90 years ago still eludes us.

Trypanosoma congolense and T. simiae have been the two recognized species within the subgenus Nannomonas. T. congolense is pathogenic to cattle and the small ruminants but not to pigs, while T. simiae causesdisease only in plgs. The two species are difficult to distinguish from each other because both share the same developmental cycle in the tsetse and are broadly similar morphologically. However, their isoenzyme and DNA characteristics are quite different. Even more confusing is the fact that T. congolense itself is composed of a number of “strains”, or types; historically the different types of T. congolense were often considered as separate species by different workers.

Now isoenzyme and DNA charaterizatlon show that T. congolense is composed of at least three different types, which are probably equivalent to subspecies. Furthermore, evidence for another species within the subgenus Nannomonas has come recently from isolates from tsetse midguts in The Gambia.

Specific DNA probes have now been produced for the different kinds of Nannomonas. With these, it is now possible to identify the different infections from midgut dissections of wild tsetse. This should now facilitate an elaboratestudy to determine the distribution and prevalence of the different types across Africa, their association with the different species of tsetse, and the significance of each in the causation of disease in domestic livestock. The information gained will also help clarify the taxonomic status of each type within the subgenus.

The application of nucleic acid hybridization techniques in the identification of most protozoan parasites, using species-specific DNA probes, has recently been described by several investigators. Species-specific DNA probes have been employed in the characterization of trypanosome infections in cattle and tsetse from Uganda. Most infections revealed by our DNA probes were mixed. Using these probes, a mixed infection with Trypanosoma brucei, T. vivax and both savannah and Kilifi types of T. congolense was revealed in one cow. This mixed infection could not have been detected by any of the classical parasitological methods. Isolates made from natural field infections, which had been passaged in laboratory animals, were found to consist of homogeneous trypanosome species. This was demonstrated in all of 47 stabilates which were homogeneous infections either of savannah type T. congolense or T. brucei.

The method of sample preparation for DNA probe analysis was modified to suit field conditions. The samples, which were spot-blotted onto nylon filters and either immediately denatured or left undenatured, could be kept at room temperature for 1 month with only a moderate loss of hybridization signal intensities. Although hybridization signals were visible in undenatured samples, those seen with the samples that had been denatured were clearly more intense. This approach eliminates the need for liquid nitrogen and/or an incubator in the field. The simplicity, sensitivity and specificity of this diagnostic technique using species-specific DNA probes, make it an important tool for future studies of the epidemiology of African trypanosomiases.

Trypanosoma brucei stocks (43) isolated from man, cattle and tsetse, all coming from a small area of Lambwe Valley, in Kenya were characterized using isoenzyme analysis. The tsetse isolates revealed the largest number of zymodemes, indicating a lot of heterogeneity among these isolates. Cluster analysis of all the trypanosomes characterized showed that the organisms could be grouped into three distinct trypanosome types; human, cattle and tsetse. The human type was represented in ail the three groups. This was not the case either with the cattle nor the tsetse T. brucei type. T. brucei zymodeme which was found to be predominant during the 1980 outbreak of sleeping sickness was not observed this time round. It is suggested that this zymodeme was either unstable and had changed or it had been eliminated during the insecticides ground spraying operations against G. pallidipes. It is concluded that the heterogeneity observed among the fly isolates is an adaptive mechanism for survival of trypanosomes.

Epidemics of human trypanosomiasis (sleeping sickness) have been occurring in Uganda since the beginning of the present century. The first such epidemic occurred along the shores of Lake Victoria and claimed millions of human lives.

Another outbreak involving about 2500 persons occurred in the same area from Jinja eastwards to the border with Kenya between 1939 and 1945. Since this outbreak, cases continued to be reported within the infected area, though not in epidemic proportions.

Between 1960 and 1971, the two main foci of human trypanosomiasis in Uganda were in south-eastern Uganda (Busoga and Bukedi Districts) and West Nile. In 1971, the infection in south- eastern Uganda spilled north of the usual focus and involved up to 169 persons. Following the control of this small epidemic, surveillance programmes were not instituted because of the prevailing political and economic atmosphere in the country at the time.

Hence, by 1976, the stage was set for another epidemic outbreak of the disease in the country.

This paper presents a detailed review of the present disease situation in the country and a brief review of the past.

The relationships between estimates of tsetse challenge and trypanosome prevalence in trypanotolerant and susceptible cattle were examined at sites of the African Trypanotolerant Livestock Network. Estimates of tsetse challenge were determined as the product of tsetse relative densities, their trypanosome infectionrates and the proportions of feeds taken by them from cattle. Trypanosome prevalence in both trypanotolerant and susceptible breeds of cattle was estimated from monthly examination of groups of ear-tagged cattle. Regression equations between estimates of trypanosome prevalence and tsetse challenge were fitted using an arcsin transformation of trypanosome prevalence and a log10 transformation of tsetse challenge. Highly significant relationships between estimates of tsetse challenge and trypanosome prevalence were found for sites with trypanotolerant cattle (P < 0.0001) and sites with susceptible cattle (P < 0.01). Mean trypanosome prevalence at a given level of tsetse challenge was at least twice as high in susceptible as in trypanotolerant cattle. The slopes of the regression lines for the two groups of cattle were significantly different (P < 0.05). The results confirm the superior ability of trypanotolerant cattle to withstand tsetse challenge in the field. The curves obtained for the two groups of cattle illustrate the necessity in tsetse control campaigns for a major reduction of tsetse populations to take place before significant decreases in trypanosome incidence in livestock can be achieved.

Simple models can be used to study the effects of factors involved in trypanosomiasis transmission, provided rough parameter estimates are available. The models should also be tested against field data; more complex models are needed whichrelate prevalence to the sensitivity of the diagnostic method, and take account of age-dependent changes in host susceptibility.

In an attempt to overcome the problem of catching Glossina austeni Newstead in reasonable numbers, which hitherto has hindered studies of its ecology in Zanzibar, a new tsetse trap (the Chuka trap) was developed. Its sky-blue coloured version proved superior to the white-and grey-coloured versions, the biconical trap, the Epsilon trap and the 3-dimensional “target” respectively, in catching G. austeni. The Chuka trap helped to establish that G. austen is a very low flyer and most trappable in the “forest floor” biotope of Jozani Forest.

A study of population ecology and vectorial capacity of Glossina fuscipes fuscipes Newstead was initiated on Rusinga island and mainland of Mbita Division, South Nyanza District, Kenya. As part of the overall study, investigations into fly distribution and diel activity patterns were carried out using biconical traps. The distribution of flies was found to be confined to a strip of dense vegetation of several metres in width along the lake shore. Trap catches indicated unimodal diel activity patterns for both sexes. Catches were recorded between 0700 and 1900 hr with peaks for females at 0900 hr and for males at 1000 hr. Of the physical factors (solar radiation, relative humidity, temperature and windspeed), solar radiation was found to be the main factor significantly influencing the catches of both males and females during the day. For males Y = −29.46 + 6.30X1 + 0.40X2, (r2 = 0.59, P = 0.0053 and P = 0.0425, where X1 = solar radiation and X2 = relative humidity, respectively). For females, Y = 3.63 + 5.13X1, (r2 = 0.20, P= 0.0960, where X1 = solar radiation).

In April–May 1986, 180 male and 218 female Glossina morsitans morsitans Westwood, each less than 1 day old, were marked uniquely and released unfed on Redcliff Island, Lake Kariba, Zimbabwe. Flies were recaptured on ox fly-rounds carried out twice daily for 6 months; marks were recorded and the flies released. For females, 56% were recaptured at leastonce; the capture probability for a given 9-day period changed with age, falling from 0.32 after birth to 0.16 by 30 days and rising above 0.5 by 80 days. For males, 57% were recaptured; the probability was 0.21 for age less than 9 days, but > 0.77 for all older flies. The capture probabilities and fly-round catches were used to estimate the change with age in the daily mortality. Female mortality was 6.8% shortly after emergence, < 1% for ages 20–50 days and rose to 5% by 130 days—a pattern similar to that for laboratory reared tsetse. Male daily mortality was 8.3% after emergence, fellto 5.5% by 9 days, then rose continuously to more than 10% by 30 days. The mortality changes are related to age rather than climate. Tsetse caught on ox fly-rounds from a natural population were aged using ovarian dissection and wing-fray analysis. The age distribution (corrected using the capture probabilities from Redcliff) was consistent with the idea that, here too, mortality increased with age.

The development of simple models for describing the population dynamics of species with overlapping generations is briefly described. The paper then presents a minimally complex simulation model for tsetse populations in which seasonally varying density-independent mortality, generally applied only to the adult stage, is combined with density dependence acting on both puparia and adults. Density dependence regulates population size within limits determined by an interaction between the strength of the density-dependent relationships and the variability of the seasonal density independence. Population data sets for Glossina palpalis and G. morsitans in Nigeria, and G. pallidipes in Kenya are adequately described by the model. Relationships discovered between the density-independent mortality and climatic variables for the Lambwe Valley data set allow predictions of population changes in other sites on the basis of their climatic conditions. Attempts to predict changes in the population of G.pallidipes in Somalia highlight the difficulties of this exercise. A tsetse's demographic response to climate may vary from region to region, making extensive predictions from local data sets difficult. There seems, however, no need to make more complex models to describe observed changes in tsetse populations.

This paper describes differences in reproductive and nutritional conditions that develop within 2 weeks of the transfer of wild-caught female Glossina pallidipes Austen to apparently ideal laboratory conditions. In order to investigate the relative sampling efficiency of baited NG2B traps for female G. pallidipes and the flies' changing feeding behaviour on different days of the pregnancy cycle, samples of female G. pallidipes taken directly from the field and those held in the laboratory for a known number of days after larviposition were subjected to both ovarian dissection (including the measurement of the uterine content and the two largest ovarioles) and fat and haematin analysis. Under the laboratory conditions of slightly lower mean temperature (buffered against diurnal fluctuations), lower activity and higher feeding frequency than in the field, the interlarval period was prolonged, the third larval instar was disproportionately prolonged, the ovarioles and larvae in utero showed differential growth rates, pupal size was increased and the fat reserves of post-partum females was depleted. It is concluded that tsetse are very responsive to the precise environmental conditions in which they find themselves and that absolute quantitative measurements made in the laboratory cannot be applied directly to the field situation.

Female sexual receptivity was studied in F1 laboratory-reared Glossina pallidipes Austen, 1903 originating from Nguruman in the Rift Valley of Kenya. Females mated as early as 6 days old but reached peak receptivity at 9–13 days old (60–80%). Females did not immediately mate with more than one male. The mean spermathecal value (MSV) was highest between 7 and 14 days of age, and declined thereafter. The duration of copulation was comparatively short (mean = 24 min), and did not vary significantly with age. High female receptivity corresponded with maturity of ovariole A, which reached maximum length at an age of 10 days. These results are discussed in relation to mass rearing of G. pallidipes.

Phenolic tsetse attractants in host urine were shown to form gradually from proattractants identified as a mixture of glucuronates and sulphates. The breakdown of proattractants is predominantly due to microbial activity. Of the eight bacteria isolated from buffalo urine three have demonstrated varying ability to give rise to free phenols the formation of which was monitored by observing the build-up of 4-cresol by means of gas chromatography. These observations may provide a useful biotechnological model for controlled release of these semiochemicals in the field.

Responses of tsetse Glossina morsitans morsitans and Glossina pallidipes (Diptera: Glossinidae) to host-derived phenolic kairomones were investigated using a wind tunnel. Activation behaviour of G. m. morsitans depended on the mixture of the phenols but not on the dose or ratio of the blend used. Up wind orientation behaviour and actual contact with the odour source depended on both the phenolic mixture and on the ratio of the phenols in the blend. In the present studies, a 4:1 blend of 4-cresol and 3-n-propylphenol, respectively, consistently induced more of the flies to fly upwind and contact the odour source as compared to any other combination of the phenols or even when these two crucial phenols were mixed in a ratio of 9:1; the ratio in which they are found in a typical sample of buffalo urine.

Comparison of the responses of the two species indicate that significantly fewer G. m. morsitans fly upwind and contact the odour source on stimulation with a blend of 4-cresol and 3-n-propylphenol as compared to G. pallidipes. This may explain why phenols only produce small improvements in the catches of G. m. morsitans as compared to G. pallidipes.

The simple wind tunnel developed, not only provides a useful tool for rapidly screening candidate attractants, but also a means to investigate how various compounds alone or in various combinations affect tsetse behaviour.

Plotting the movement of air through typical tsetse habitats in Zimbabwe revealed that it is liable to change direction by over 90° within a few metres. In winds of < 1 m/sec, there was a negative correlation between windspeed and the wind's tendency to meander, whether in thick bush or out in the open. And at the mean windspeed of wet season mopane woodland (0.3 m/sec), the wind changed direction by c. 15°/sec. Accordingly, host odour does not move downwind in simple trajectories that tsetse flies could easily follow. Even 5 m from a source, odour (modelled with smoke) approached a notional tsetse fly for a quarter of the time from >90° away from the true source direction. Also, air turbulence at common tsetse resting sites generates much nonsense information about the “true” direction of any odour-bearing wind. The suggested answer to how tsetse manage to find distant, invisible hosts, in spite of this confusing information from the wind, is that they progress by a biased random walk which is the outcome of their upwind anemotactic responses to odour-bearing wind and the proportion of the time for which the wind does blow in the “correct” direction. Computer simulation shows that this might work in principle.

Tsetse colonized either for laboratory studies or for release in S. I. T. programmes are assumed to be healthy and geneticallyw similar to flies in natural populations. However, insect colonies are subjected to many of the same evolutionary forces that influence genetic changes in natural populations, i.e. drift, selection, hitch-hiking, mutations, assortative mating and immigration. The influence of these on genetic structure of tsetse fly colonies is outlined, and examples are presented from several species. There is little or no evidence for adaptation during the early phases of laboratory colonization of five species of tsetse. A model is presented indicating that with as little as a 5% fitness difference between males, some colonies have existed long enough to have undergone significant changes in the relative numbers of males having “standard” and “enhanced” fitness. Slight changes in heterozygosity of colonized flies is documented by comparing colonies and field-collected flies and by comparisons within colonies over periods of several generations or years. An example of hitch-hiking is illustrated with the closely linked genes Sr (sex ratio) and Est-X in Glossina morsitans submorsitans. A possible interaction between alleles at these loci is discussed. A summary is presented of polyacrylamide gel electrophoretic methods for monitoring 16 polymorphic loci distributed among the X chromosome and autosomes of tsetse.

Tsetse flies have been eradicated from an area of 48,000 km2 in Zimbabwe since 1981. The main approach has been pesticidal control, involving ground spraying, aerial spraying, cattle dipping and artificial bait techniques, in some circumstances in combination. This represents a change in tactic from the pre-independence period when control relied solely on ground spraying. Factors which have influenced choice of technique are discussed in the paper, including scale of trypanosomiasis problem, technical limitations of the technique, cost-effectiveness and environmental considerations.

Ground spraying, the mainstay of tsetse control operations in Zimbabwe prior to 1980, is becoming increasingly unfavourable because of high logistical requirements and environmental reasons, but remains an effective technique under a wide range of circumstances. Aerial spraying has proved very effective for management of crisis situations but is comparatively expensive, capital intensive in hilly terrain and less effective against Glossina pallidipes than G. morsitans. Vector control by applying insecticide (deltamethrin) to cattle appears technically feasible at a very competitive cost. This is a most promising technique in areas where there are already cattle and few wild hosts.

Control of tsetse using odour-baited and insecticide-treated screens has proven technically feasible in a wide range of terrain, and appears economically competitive with ground spraying. Both technical and financial performance of the technique are still being improved. Bait technology offers scope for the new approaches to the control of the fly and defence of cleared areas from reinvasion.

Pyramidal traps (8000) have been set up against Glossina fuscipes fuscipes to control Trypanosoma rhodesiense in Busoga. In order to understand the pattern of tsetse/human contact, the collection of data on the density, spatial distribution and mobility of flies has been related to the dynamics of human trypanosomiasis in Busoga. As for T. gambiense, few reasonable correlations between global fly densities and incidence of the disease have been found. However, field observation with a recording time of seasonal variations of movements of people and animals associated with particular habitats, remains the key to understanding the transmission of the disease.