Quantifying Heterogeneity in Host-Vector Contact: Tsetse (Glossina swynnertoni and G. pallidipes) Host Choice in Serengeti National Park, Tanzania

Abstract

Background: Identifying hosts of blood-feeding insect vectors is crucial in understanding their role in disease transmission. Rhodesian human African trypanosomiasis (rHAT), also known as acute sleeping sickness is caused by Trypanosoma brucei rhodesiense and transmitted by tsetse flies. The disease is commonly associated with wilderness areas of east and southern Africa. Such areas hold a diverse range of species which form communities of hosts for disease maintenance. The relative importance of different wildlife hosts remains unclear. This study quantified tsetse feeding preferences in a wilderness area of great host species richness, Serengeti National Park, Tanzania, assessing tsetse feeding and host density contemporaneously.

Methods: Glossina swynnertoni and G. pallidipes were collected from six study sites. Bloodmeal sources were identified through matching Cytochrome B sequences amplified from bloodmeals from recently fed flies to published sequences. Densities of large mammal species in each site were quantified, and feeding indices calculated to assess the relative selection or avoidance of each host species by tsetse.

Results: The host species most commonly identified in G. swynnertoni bloodmeals, warthog (94/220), buffalo (48/220) and giraffe (46/220), were found at relatively low densities (3-11/km2) and fed on up to 15 times more frequently than expected by their relative density. Wildebeest, zebra, impala and Thomson's gazelle, found at the highest densities, were never identified in bloodmeals. Commonly identified hosts for G. pallidipes were buffalo (26/46), giraffe (9/46) and elephant (5/46).

Conclusions: This study is the first to quantify tsetse host range by molecular analysis of tsetse diet with simultaneous assessment of host density in a wilderness area. Although G. swynnertoni and G. pallidipes can feed on a range of species, they are highly selective. Many host species are rarely fed on, despite being present in areas where tsetse are abundant. These feeding patterns, along with the ability of key host species to maintain and transmit T. b. rhodesiense, drive the epidemiology of rHAT in wilderness areas.

Fig 1. Map of study sites in Serengeti National Park, Tanzania.

The map shows an outline of the protected area boundaries of Serengeti National Park (SNP), Grumeti, Ikorongo and Maswa Game Reserves (GR) and Ngorongoro Conservation Area (NCA), within Northern Tanzania.

Fig 2. Wildlife hosts identified in bloodmeal samples.

The graphs show the percentage of bloodmeals identified per species, out of the total identified samples for (a) Glossina swynnertoni and (b) G. pallidipes. The number of samples identified as each species is shown (n), out of a total of 220 and 46 samples identified for G. swynnertoni and G. pallidipes, respectively. Error bars show 95% binomial confidence intervals.

Fig 3. Feeding indices illustrate selection or avoidance of host species by Glossina swynnertoni and G. pallidipes.

Feeding indices by species and study site on a log scale. Graded colours show the feeding index at each of 6 study sites. Values above 1 indicate a host is selected, values below 1 indicate a host is avoided, by (a) G. swynnertoni and (b) G. pallidipes. Stars indicate significance at p<0.05.