Spatial Predictions of Rhodesian Human African Trypanosomiasis (Sleeping Sickness) Prevalence in Kaberamaido and Dokolo,Two Newly Affected Districts of Uganda

The continued northwards spread of Rhodesian sleeping sickness or Human African Trypanosomiasis (HAT) within Uganda is raising concerns of overlap with the Gambian form of the disease. Disease convergence would result in compromised diagnosis and treatment for HAT. Spatial determinants for HAT are poorly understood across small areas. This study examines the relationships between Rhodesian HAT and several environmental, climatic and social factors in two newly affected districts, Kaberamaido and Dokolo. A one-step logistic regression analysis of HAT prevalence and a two-step logistic regression method permitted separate analysis of both HAT occurrence and HAT prevalence. Both the occurrence and prevalence of HAT were negatively correlated with distance to the closest livestock market in all models. The significance of distance to the closest livestock market strongly indicates that HAT may have been introduced to this previously unaffected area via the movement of infected, untreated livestock from endemic areas. This illustrates the importance of the animal reservoir in disease transmission, and highlights the need for trypanosomiasis control in livestock and the stringent implementation of regulations requiring the treatment of cattle prior to sale at livestock markets to prevent any further spread of Rhodesian HAT within Uganda.     

Sleeping Sickness in Southeastern Uganda: A SystemsApproach

Sleeping sickness continues to be a significant public health burden in southeastern Uganda. Continued spread of the disease into new areas of Uganda highlights our inability to understand and predict the distribution of infection. Multiple factors influence the distribution of sleeping sickness, including climate, land cover, cattle movements, prevention and control activities, and social conflict. We draw on a systems approach to conceptualize and characterize the multiple interacting forces and processes that influence the spatial and temporal dynamics of sleeping sickness in Uganda. This synthesis reveals a complex system of interactions among human and biophysical systems, feedback, and scale dependence. We identify some common analytical modeling approaches relative to our system characterization and identify opportunities for sleeping sickness research and improved understanding of disease dynamics in Uganda.     

Bayesian geostatistical analysis and prediction of Rhodesian human African trypanosomiasis

Background

The persistent spread of Rhodesian human African trypanosomiasis (HAT) in Uganda in recent years has increased concerns of a potential overlap with the Gambian form of the disease. Recent research has aimed to increase the evidence base for targeting control measures by focusing on the environmental and climatic factors that control the spatial distribution of the disease.

Objectives

One recent study used simple logistic regression methods to explore the relationship between prevalence of Rhodesian HAT and several social, environmental and climatic variables in two of the most recently affected districts of Uganda, and suggested the disease had spread into the study area due to the movement of infected, untreated livestock. Here we extend this study to account for spatial autocorrelation, incorporate uncertainty in input data and model parameters and undertake predictive mapping for risk of high HAT prevalence in future.

Materials and Methods

Using a spatial analysis in which a generalised linear geostatistical model is used in a Bayesian framework to account explicitly for spatial autocorrelation and incorporate uncertainty in input data and model parameters we are able to demonstrate a more rigorous analytical approach, potentially resulting in more accurate parameter and significance estimates and increased predictive accuracy, thereby allowing an assessment of the validity of the livestock movement hypothesis given more robust parameter estimation and appropriate assessment of covariate effects.

Results

Analysis strongly supports the theory that Rhodesian HAT was imported to the study area via the movement of untreated, infected livestock from endemic areas. The confounding effect of health care accessibility on the spatial distribution of Rhodesian HAT and the linkages between the disease''s distribution and minimum land surface temperature have also been confirmed via the application of these methods.

Conclusions

Predictive mapping indicates an increased risk of high HAT prevalence in the future in areas surrounding livestock markets, demonstrating the importance of livestock trading for continuing disease spread. Adherence to government policy to treat livestock at the point of sale is essential to prevent the spread of sleeping sickness in Uganda.     

Estimating and Mapping the Population at Risk of Sleeping Sickness

Background

Human African trypanosomiasis (HAT), also known as sleeping sickness, persists as a public health problem in several sub-Saharan countries. Evidence-based, spatially explicit estimates of population at risk are needed to inform planning and implementation of field interventions, monitor disease trends, raise awareness and support advocacy. Comprehensive, geo-referenced epidemiological records from HAT-affected countries were combined with human population layers to map five categories of risk, ranging from “very high” to “very low,” and to estimate the corresponding at-risk population.

Results

Approximately 70 million people distributed over a surface of 1.55 million km² are estimated to be at different levels of risk of contracting HAT. Trypanosoma brucei gambiense accounts for 82.2% of the population at risk, the remaining 17.8% being at risk of infection from T. b. rhodesiense. Twenty-one million people live in areas classified as moderate to very high risk, where more than 1 HAT case per 10,000 inhabitants per annum is reported.

Discussion

Updated estimates of the population at risk of sleeping sickness were made, based on quantitative information on the reported cases and the geographic distribution of human population. Due to substantial methodological differences, it is not possible to make direct comparisons with previous figures for at-risk population. By contrast, it will be possible to explore trends in the future. The presented maps of different HAT risk levels will help to develop site-specific strategies for control and surveillance, and to monitor progress achieved by ongoing efforts aimed at the elimination of sleeping sickness.     

Quantifying the Association between Bovine and Human Trypanosomiasis in Newly Affected Sleeping Sickness Areas of Uganda

Background

Uganda has active foci of both chronic and acute HAT with the acute zoonotic form of disease classically considered to be restricted to southeast Uganda, while the focus of the chronic form of HAT was confined to the northwest of the country. Acute HAT has however been migrating from its traditional disease focus, spreading rapidly to new districts, a spread linked to movement of infected cattle following restocking. Cattle act as long-term reservoirs of human infective T. b. rhodesiense showing few signs of morbidity, yet posing a significant risk to human health. It is important to understand the relationship between infected cattle and infected individuals so that an appropriate response can be made to the risk posed to the community from animals infected with human pathogens in a village setting.

Methodology/Principal Findings

This paper examines the relationship between human T. b. rhodesiense infection and human infective and non-human T. brucei s.l. circulating in cattle at village level in Kaberamaido and Dokolo Districts, Uganda. The study was undertaken in villages that had reported a case of sleeping sickness in the six months prior to sample collection and those villages that had never reported a case of sleeping sickness.

Conclusions and Significance

The sleeping sickness status of the villages had a significant effect with higher odds of infection in cattle from case than from non-case villages for T. brucei s.l. (OR: 2.94, 95%CI: 1.38–6.24). Cattle age had a significant effect (p<0.001) on the likelihood of T. brucei s.l. infection within cattle: cattle between 18–36 months (OR: 3.51, 95%CI: 1.63–7.51) and cattle over 36 months (OR: 4.20, 95%CI: 2.08–8.67) had significantly higher odds of T. brucei s. l. infection than cattle under 18 months of age. Furthermore, village human sleeping sickness status had a significant effect (p<0.05) on the detection of T. b. rhodesiense in the village cattle herd, with significantly higher likelihood of T. b. rhodesiense in the village cattle of case villages (OR: 25, 95%CI: 1.2–520.71). Overall a higher than average T. brucei s.l. prevalence (>16.3%) in a village herd over was associated with significantly higher likelihood of T. b. rhodesiense being detected in a herd (OR: 25, 95%CI: 1.2–520.71).     

Clustered or scattered? The impact of habitat quality clustering on establishment and early spread

The match between the environmental conditions of an introduction area and the preferences of an introduced species is the first prerequisite for establishment. Yet, introduction areas are usually landscapes, i.e. heterogeneous sets of habitats that are more or less favourable to the introduced species. Because individuals are able to disperse after their introduction, the quality of the habitat surrounding the introduction site is as critical to the persistence of introduced populations as the quality of the introduction site itself. Moreover, demographic mechanisms such as Allee effects or dispersal mortality can hamper dispersal and affect spread across the landscape, in interaction with the spatial distribution of favourable habitat patches. In this study, we investigate the impact of the spatial distribution of heterogeneous quality habitats on establishment and early spread. First, we simulated introductions in one‐dimensional landscapes for different dispersal rates and either dispersal mortality or Allee effects. The landscapes differed by the distribution of favourable and less favourable habitats, which were either clustered into few large aggregates of the same quality or scattered into multiple smaller ones. Second, we tested the predictions of simulations by performing experimental introductions of hymenopteran parasitoids (Trichogramma chilonis) in ‘clustered’ and ‘scattered’ microcosm landscapes. Results highlighted two impacts of the clustering of favourable habitat: by decreasing the risks of dispersal from the introduction site to unfavourable habitat early during the invasion, it increased establishment success. However, by increasing the distance between favourable habitat patches, it also hindered the subsequent spread of introduced species over larger areas.     

Crisis, what crisis? Control of Rhodesian sleeping sickness

There is an urgent need for cost-effective strategies for the sustainable control of Trypanosoma brucei rhodesiense (Rhodesian) sleeping sickness, which is a fatal zoonotic disease that has caused devastating epidemics during the past century. Sleeping sickness continues to be controlled by crisis management, using active case detection, treatment and vector control - activities that occur only during major epidemics; during the intervening periods, farmers and communities must fend for themselves. There are several methods for assessing the burden of this disease and there is a series of farmer-led methodologies that can be applied to reduce the burden of human and animal trypanosomiases.     

Genetic Diversity and Population Structure of Trypanosoma brucei in Uganda: Implications for the Epidemiology of Sleeping Sickness and Nagana

Background

While Human African Trypanosomiasis (HAT) is in decline on the continent of Africa, the disease still remains a major health problem in Uganda. There are recurrent sporadic outbreaks in the traditionally endemic areas in south-east Uganda, and continued spread to new unaffected areas in central Uganda. We evaluated the evolutionary dynamics underpinning the origin of new foci and the impact of host species on parasite genetic diversity in Uganda. We genotyped 269 Trypanosoma brucei isolates collected from different regions in Uganda and southwestern Kenya at 17 microsatellite loci, and checked for the presence of the SRA gene that confers human infectivity to T. b. rhodesiense.

Results

Both Bayesian clustering methods and Discriminant Analysis of Principal Components partition Trypanosoma brucei isolates obtained from Uganda and southwestern Kenya into three distinct genetic clusters. Clusters 1 and 3 include isolates from central and southern Uganda, while cluster 2 contains mostly isolates from southwestern Kenya. These three clusters are not sorted by subspecies designation (T. b. brucei vs T. b. rhodesiense), host or date of collection. The analyses also show evidence of genetic admixture among the three genetic clusters and long-range dispersal, suggesting recent and possibly on-going gene flow between them.

Conclusions

Our results show that the expansion of the disease to the new foci in central Uganda occurred from the northward spread of T. b. rhodesiense (Tbr). They also confirm the emergence of the human infective strains (Tbr) from non-infective T. b. brucei (Tbb) strains of different genetic backgrounds, and the importance of cattle as Tbr reservoir, as confounders that shape the epidemiology of sleeping sickness in the region.     

Through the Eyes of Children: Perceptions of Environmental Change in Tropical Forests

This study seeks to understand children''s perceptions of their present and future environments in the highly biodiverse and rapidly changing landscapes of Kalimantan, Indonesian Borneo. We analyzed drawings by children (target age 10–15 years) from 22 villages, which show how children perceive the present conditions of forests and wildlife surrounding their villages and how they expect conditions to change over the next 15 years. Analyses of picture elements and their relationships to current landscape variables indicate that children have a sophisticated understanding of their environment and how different environmental factors interact, either positively or negatively. Children appear to have landscape-dependent environmental perceptions, showing awareness of past environmental conditions and many aspects of recent trends, and translating these into predictions for future environmental conditions. The further removed their present landscape is from the originally forested one, the more environmental change they expect in the future, particularly declines in forest cover, rivers, animal diversity and increases in temperature and natural disasters. This suggests that loss of past perceptions and associated “shifting environmental baselines” do not feature strongly among children on Borneo, at least not for the perceptions we investigated here. Our findings that children have negative expectations of their future environmental conditions have important political implications. More than other generations, children have a stake in ensuring that future environmental conditions support their long-term well-being. Understanding what drives environmental views among children, and how they consider trade-offs between economic development and social and environmental change, should inform optimal policies on land use. Our study illuminates part of the complex interplay between perceptions of land cover and land use change. Capturing the views of children through artistic expressions provides a potentially powerful tool to influence public and political opinions, as well as a valuable approach for developing localized education and nature conservation programs.     

Landscape Genetic Structure of a Streamside Tree Species Euptelea pleiospermum (Eupteleaceae): Contrasting Roles of River Valley and Mountain Ridge

We used landscape genetics and statistical models to test how landscape features influence connectivity or create barriers to dispersal for a mountain riparian tree species, Euptelea pleiospermum. Young leaves from 1078 individuals belonging to 36 populations at elevations of 900–2000 m along upper reaches of four rivers were genotyped using eight nuclear microsatellite markers. We found no evidence for the unidirectional dispersal hypothesis in E. pleiospermum within each river. The linear dispersal pattern along each river valley is mostly consistent with the “classical metapopulaton” model. Mountain ridges separating rivers were genetic barriers for this wind-pollinated tree species with anemochorous seeds, whereas river valleys provided important corridors for dispersal. Gene flow among populations along elevational gradients within each river prevails over gene flow among populations at similar elevations but from different rivers. This pattern of gene flow is likely to promote elevational range shifts of plant populations and to hinder local adaptation along elevational gradients. This study provides a paradigm to determine which of the two strategies (migration or adaptation) will be adopted by mountain riparian plants under climate warming.     

The use of racial, ethnic, and ancestral categories in human genetics research

The global dispersal of anatomically modern humans over the past 100,000 years has produced patterns of phenotypic variation that have exerted—and continue to exert—powerful influences on the lives of individuals and the experiences of groups. The recency of our common ancestry and continued gene flow among populations have resulted in less genetic differentiation among geographically distributed human populations than is observed in many other mammalian species. Nevertheless, differences in appearance have contributed to the development of ideas about “race” and “ethnicity” that often include the belief that significant inherited differences distinguish humans. The use of racial, ethnic, and ancestral categories in genetics research can imply that group differences arise directly through differing allele frequencies, with little influence from socially mediated mechanisms. At the same time, careful investigations of the biological, environmental, social, and psychological attributes associated with these categories will be an essential component of cross-disciplinary research into the origins, prevention, and treatment of common diseases, including those diseases that differ in prevalence among groups.     

Density dependence slows invader spread in fragmented landscapes

Patchiness is a defining characteristic of most natural and anthropogenic habitats, yet much of our understanding of how invasions spread has come from models of spatially homogeneous environments. Except for populations with Allee effects, an invader's growth rate when rare and dispersal determine its spread velocity; intraspecific competition has little to no influence. How this result might change with landscape patchiness, however, is poorly understood. We used simulation models and their analytical approximations to explore the effect of density dependence on the spread of annual plant invaders moving through heterogeneous landscapes with gaps in suitable habitat. We found that landscape patchiness and discrete invader population size interacted to generate a strong role for density dependence. Intraspecific competition greatly slowed the spread of invasions through patchy landscapes by regulating how rapidly a population could produce enough seeds to surpass habitat gaps. Populations with continuously varying density showed no such effect of density dependence. We adapted a stochastic dispersal model to approximate spread when gap sizes were small relative to the mean dispersal distance and a Markov chain approximation for landscapes with large gaps. Our work suggests that ecologists must consider reproduction at both low and high densities when predicting invader spread.     

Mechanistic modelling of animal dispersal offers new insights into range expansion dynamics across fragmented landscapes

Understanding and predicting the dynamics of range expansion is a major topic in ecology both for invasive species extending their ranges into non‐native regions and for species shifting their natural distributions as a consequence of climate change. In an increasingly modified landscape, a key question is ‘how do populations spread across patchy landscapes?‘ Dispersal is a central process in range expansion and while there is a considerable theory on how the shape of a dispersal kernel influences the rate of spread, we know much less about the relationships between emigration, movement and settlement rules, and invasion rates. Here, we use a simple, single species individual‐based model that explicitly simulates animal dispersal to establish how density‐dependent emigration and settlement rules interact with landscape characteristics to determine spread rates. We show that depending on the dispersal behaviour and on the risk of mortality in the matrix, increasing the number of patches does not necessarily maximise the spread rate. This is due to two effects: first, individuals dispersing at the expanding front are likely to exhibit lower net‐displacement as they typically do not travel far before finding a patch; secondly, with increasing availability of high quality habitat, density‐dependence in emigration and settlement can decrease the number of emigrants and their net‐displacement. The rate of spread is ultimately determined by the balance between net travelled distance, the dispersal mortality and the number of dispersing individuals, which in turn depend on the interaction between the landscape and the species’ dispersal behaviour. These results highlight that predicting spread rates in heterogeneous landscapes is a complex task and requires better understanding of the rules that individuals use in emigration, transfer and settlement decisions.     

Genetic assessment of environmental features that influence deer dispersal: implications for prion-infected populations

The landscape can influence host dispersal and density, which in turn, affect infectious disease transmission, spread, and persistence. Understanding how the landscape influences wildlife dispersal and pathogen epidemiology can enhance the efficacy of disease management in natural populations. We applied landscape genetics to examine relationships among landscape variables, dispersal of white-tailed deer hosts and transmission/spread of chronic wasting disease (CWD), a fatal prion encephalopathy. Our focus was on quantifying movements and population structure of host deer in infected areas as a means of predicting the spread of this pathology and promoting its adaptive management. We analyzed microsatellite genotypes of CWD-infected and uninfected deer from two disease foci (Southern Wisconsin, Northern Illinois). We quantified gene flow and population structure using F _ST, assignment tests, and spatial autocorrelation analyses. Gene flow estimates were then contrasted against a suite of landscape variables that potentially mediate deer dispersal. Forest fragmentation and grassland connectivity promoted deer movements while rivers, agricultural fields and large urbanized areas impeded movement. Landscape variables, deer dispersal, and disease transmission covaried significantly and positively in our analyses. Habitats with elevated host gene flow supported the concept of dispersal-mediated CWD transmission by reflecting a concomitant, rapid CWD expansion. Large, interrelated social groups isolated by movement barriers overlapped disease foci, suggesting that philopatry exacerbated CWD transmission. Our results promote adaptive management of CWD by predicting patterns of its spread and identifying habitats at risk for invasion. Further, our landscape genetics approach underscores the significance of topography and host behavior in wildlife disease transmission.     

Permissible Home Range Estimation (PHRE) in Restricted Habitats: A New Algorithm and an Evaluation for Sea Otters

Parametric and nonparametric kernel methods dominate studies of animal home ranges and space use. Most existing methods are unable to incorporate information about the underlying physical environment, leading to poor performance in excluding areas that are not used. Using radio-telemetry data from sea otters, we developed and evaluated a new algorithm for estimating home ranges (hereafter Permissible Home Range Estimation, or “PHRE”) that reflects habitat suitability. We began by transforming sighting locations into relevant landscape features (for sea otters, coastal position and distance from shore). Then, we generated a bivariate kernel probability density function in landscape space and back-transformed this to geographic space in order to define a permissible home range. Compared to two commonly used home range estimation methods, kernel densities and local convex hulls, PHRE better excluded unused areas and required a smaller sample size. Our PHRE method is applicable to species whose ranges are restricted by complex physical boundaries or environmental gradients and will improve understanding of habitat-use requirements and, ultimately, aid in conservation efforts.     

Sex-Biased Dispersal of a Frog (Odorrana schmackeri) Is Affected by Patch Isolation and Resource Limitation in a Fragmented Landscape

Sex-biased dispersal is widespread in the animal kingdom and is affected by numerous factors including mating system, social factors and environmental conditions. Unlike birds and mammals, there is no common trend in amphibians and explaining the direction and degree of sex-biased dispersal in species-specific cases is difficult. We conducted a study on dispersal of the Chinese piebald odorous frog (Odorrana schmackeri) in a fragmented landscape associated with dam construction. Ten microsatellite loci were used to analyze 382 samples sourced from 14 fragmented ‘islands’. Assignment tests indicated a significant pattern of female-biased dispersal on one island with inconsistencies in the strength and direction of this pattern between nearby islands. The effects of four island attributes and two potential impact factors on the pattern of sex-biased dispersal were examined. We found that the extent of isolation from the mainland and the number of breeding sites both showed a negative correlation with female biased dispersal, such that the closer an island is to the mainland the more likely it is to display female biased dispersal, and the more breeding sites on an island the more male immigrants. Based on these results, we conclude that geographic isolation and limited breeding resources are the most likely explanation for the patterns of dispersal observed in this fragmented population of amphibians.     

Evidence of climate‐driven regime shifts in the Aegean Sea’s demersal resources: A study spanning six decades

Climate change (CC) can alter the configuration of marine ecosystems; however, ecosystem response and resilience to change are usually case‐specific. The effect of CC on the demersal resources of the Aegean Sea (east Mediterranean Sea) was investigated during the past six decades applying a combination of multivariate analysis, non‐additive modeling and the Integrated Resilience Assessment (IRA) framework. We focused on the study of: (i) the biological “system” complex, using proxies of biomass (landings per unit of capacity) for 12 demersal taxa, and (ii) the environmental “stressor” complex, described by 12 abiotic variables. Pronounced changes have occurred in both the environmental and biological system over the studied period. The majority of the environmental stressors exhibited strikingly increasing trends (temperature, salinity, primary production indices) with values started exceeding the global historical means during late 1980s‐early 1990s. It is suggested that the biological system exhibited a discontinuous response to CC, with two apparently climate‐induced regime shifts occurring in the past 25 years. There is evidence for two‐fold bifurcations and four tipping points in the system, forming a folded stability landscape with three basins of attraction. The shape of the stability landscape for the Aegean Sea''s biological system suggests that while the initial state (1966–1991) was rather resilient to CC, absorbing two environmental step‐changes, this was not the case for the two subsequent ones (intermediate: 1992–2002; recent: 2003–2016). Given the current trajectory of environmental change, it is highly unlikely that the biological system will ever return to its pre‐1990s state, as it is entering areas of unprecedented climatic conditions and there is some evidence that the system may be even shifting toward a new state. Our approach and findings may be relevant to other marine areas of the Mediterranean and beyond, undergoing climate‐driven regime shifts, and can assist to their adaptive management.     

Evolution in a Changing Environment

We propose a simple model for genetic adaptation to a changing environment, describing a fitness landscape characterized by two maxima. One is associated with “specialist” individuals that are adapted to the environment; this maximum moves over time as the environment changes. The other maximum is static, and represents “generalist” individuals not affected by environmental changes. The rest of the landscape is occupied by “maladapted” individuals. Our analysis considers the evolution of these three subpopulations. Our main result is that, in presence of a sufficiently stable environmental feature, as in the case of an unchanging aspect of a physical habitat, specialists can dominate the population. By contrast, rapidly changing environmental features, such as language or cultural habits, are a moving target for the genes; here, generalists dominate, because the best evolutionary strategy is to adopt neutral alleles not specialized for any specific environment. The model we propose is based on simple assumptions about evolutionary dynamics and describes all possible scenarios in a non-trivial phase diagram. The approach provides a general framework to address such fundamental issues as the Baldwin effect, the biological basis for language, or the ecological consequences of a rapid climate change.     

La Trypanosomose Humaine Africaine dans l’espace ivoiro-burkinabé : optimisation des stratégies de surveillance épidémiologique

The objective of this paper was to describe recent data from Burkina Faso and Côte d’Ivoire on Human African Trypanosomosis medical monitoring in order to (i) update the disease situation in these two countries that have been sharing important migratory, economic and epidemiological links for more than a century and (ii) to define the future strategic plans to achieve the goal of a sustainable control/elimination process. Results of active and passive surveillance indicate that all sleeping sickness patients diagnosed these last years in Burkina Faso were imported cases from Côte d’Ivoire. Nevertheless the re-introduction of the parasite is effective and the risk of a resumption of transmission exists. In Côte d’Ivoire, few cases are still diagnosed in several historical foci and the fear exists that the disease could reemerge in these foci or spread to other areas. In order to achieve a sustainable elimination of sleeping sickness in these two countries, control entities have to adapt their strategy to the different epidemiological contexts. At the exception of specific cases, the current disease prevalence no longer justifies the use of expensive medical surveys by exhaustive screening of the population. New disease control strategies, based on the exchange of epidemiological information between the two countries and integrated to the regular national health systems are required to target priority intervention areas. Follow-up in time of both treated patients and serological suspects that are potential asymptomatic carriers of parasite is also important. In parallel, researchers need to better characterize the respective roles of the human and animal reservoir in the maintenance of transmission and evaluate the different control strategies taken by National Control Programs in term of cost/effectiveness to help optimize them.     

A global sensitivity analysis for African sleeping sickness

African sleeping sickness is a parasitic disease transmitted through the bites of tsetse flies of the genus Glossina. We constructed mechanistic models for the basic reproduction number, R0, of Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense, respectively the causative agents of West and East African human sleeping sickness. We present global sensitivity analyses of these models that rank the importance of the biological parameters that may explain variation in R0, using parameter ranges based on literature, field data and expertize out of Uganda. For West African sleeping sickness, our results indicate that the proportion of bloodmeals taken from humans by Glossina fuscipes fuscipes is the most important factor, suggesting that differences in the exposure of humans to tsetse are fundamental to the distribution of T. b. gambiense. The second ranked parameter for T. b. gambiense and the highest ranked for T. b. rhodesiense was the proportion of Glossina refractory to infection. This finding underlines the possible implications of recent work showing that nutritionally stressed tsetse are more susceptible to trypanosome infection, and provides broad support for control strategies in development that are aimed at increasing refractoriness in tsetse flies. We note though that for T. b. rhodesiense the population parameters for tsetse - species composition, survival and abundance - were ranked almost as highly as the proportion refractory, and that the model assumed regular treatment of livestock with trypanocides as an established practice in the areas of Uganda experiencing East African sleeping sickness.