The primary goal of this study was to evaluate and map the transmission patterns of sleeping sickness prevailing in the Bipindi HAT focus during two contrasting periods in order to provide knowledge that will help to develop a cost-effective control strategy. Within this scope, entomological and parasitological parameters were recorded and compared in three villages of the Bipindi sleeping sickness focus.
During the entomological surveys, G. p. palpalis was the most trapped tsetse species. A significantly higher density (ADT) of tsetse flies was found in July but the highest infection rates were observed in March. Among the 15 successfully identified blood meals, 10 (66.7%) originated from humans and 5 (33.3%) from pigs. The highest transmission risk index was observed in the Lambi village; the evaluation of the TRI at the trap level shows that the risk was higher in July and near perennial water sources.
The presence of G. p. palpalis, G. pallicera and G. nigrofusca in the Bipindi sleeping sickness focus corroborates previous observations in the same focus [3, 21]. In this study as well as in other studies carried out in the Bipindi sleeping sickness focus [19, 22, 23], the most prevalent tsetse species was G. p. palpalis (> 98.0%). Previous studies  using the “Vavoua” tsetse fly trap, which has been proven to be very efficient on a large variety of tsetse fly species , also revealed a high proportion of G. p. palpalis in the same environment where pyramidal traps were set up. These results show that G. p. palpalis was the most trapped tsetse species whatever the type of trap used in most sleeping sickness foci of the forest region of Cameroon. They also show that the higher proportion of G. p. palpalis observed in the present study cannot be a consequence of the trapping technique or to the type of trap used. Although this species has a high power of adaptation, the high prevalence observed is a reflection of the effect of human action on the biotopes where the traps were set [12, 21].
The significantly higher ADT observed in July corroborates results obtained in previous investigations in the Bipindi focus  as well as in the Campo sleeping sickness focus in the forest region of the southern Cameroon . The relative abundance of tsetse flies showed high correlation with factors such as climate, vegetation or the contact frequency between tsetse flies and their vertebrate hosts . It was also shown that the type of soil, human activities, the breeding sites of tsetse flies as well as the distribution of mammals might significantly influence the abundance of tsetse flies in a given biotope [25, 26]. In our study, the traps were set in the same position during both trapping periods to allow the measurement of the direct or indirect impact of climate (temperatures and rainfalls) variations on the abundance of flies at the same capture sites. Therefore, the changes occurring in the environment or in the host behaviour throughout the trapping period are driven by the temporal variations of the climate. Indeed, the climatic conditions prevailing in July (beginning of the little dry season) and March (end of the major dry season) were probably highly influenced by the cumulative effect of the climatic conditions prevailing during the months or seasons preceding each trapping period, as shown in the ombrothermic diagram [see Additional file 1: Figure S1]. Subsequently, some water sources tended to dry up after the dry months and to be regenerated after rains. However, permanent water sources were maintained throughout the year, thus favouring the development and maintenance of a micro-habitat for tsetse flies. The significant drop in ADT (from 3.05 in July to 0.91 in March) in Ebiminbang may result from the environment around each trap. In fact, most streams in this village are not permanent during the year. Therefore, almost all traps in Ebiminbang were set up around temporary water sources that dried up in March (Table 1, Figure 1a and 1b). On the contrary, in Memel I and Lambi, most of the traps were set across the Miguili and Mougue streams (these permanent streams are the tributaries of the Lokoundje river) respectively. A noteworthy observation in this study is that fly densities were particularly elevated where host activities, in and/or around large and perennial water sources, were generated. Consequently, the ADT were almost constant or not significantly affected from one trapping period to another. These results confirm clearly that the water availability, along with the associated activities of hosts (bathing, swimming, laundry), may have a real impact on tsetse fly density and consequently on the transmission of sleeping sickness.
The significantly higher proportion of teneral flies observed in July compared to the value in March might be due to some environmental conditions (climate, vegetation …), which favour the completion of the fly life-cycle . During the dry months, the high temperature does not favour the development of pupae because dry soil does not enable tsetse flies to put their larvae in appropriate conditions or in environments where they can easily develop. Ideally, precipitation should be moderate, an excess of rain can drain the soil and lead to the destruction of pupae buried by tsetse flies . As expected, the little rainy season preceding the sampling in July was favourable to the development of tsetse flies. Furthermore, the higher proportion of teneral flies recorded in July reflects the higher transmission risk index observed during this trapping period because teneral flies are highly susceptible to infections during their first blood meal .
The higher proportion of tsetse flies harbouring a blood meal in March compared to July could also be explained by the climatic conditions. Though tsetse flies can find favourable climatic conditions and diverse hosts everywhere in the forest zone, the temperature in the dry period is high and vertebrate hosts move close to permanent water sources and shade where they can get water or rest. In such biotopes, tsetse flies can easily get their blood meals ; this is in accordance with the higher proportion of blood meals observed in March. Despite the identification of human and pig blood meals in tsetse flies of the Bipindi sleeping sickness focus, most blood meals were not identified. These unidentified blood meals were probably from wild animals because the most common domestic animals found in this focus were used as a reference during the identification of blood meals . This hypothesis is strengthened by previous results in the Bipindi HAT focus where tsetse blood meals originating from different wild animal species including Python sebae, Tragelaphus spekeii, Trionyx and Kinixys were reported [3, 29].
The high risk for HAT transmission observed in July (Figure 1a and 1b) is not in line with results obtained by Grébaut and colleagues  who found a similar level of the transmission risk during the dry and the rainy seasons. The transmission risk indices found by these authors were 2.03 in May (rainy season) and 2.09 in February (dry season), despite the difference in the relative density of tsetse flies during the two periods (2.3 in May and 1.7 in February). The discrepancy between our results and those of Grébaut et al.  could be explained by the sampling period. Grébaut et al.  sampled tsetse flies in May (in the middle of the little rainy season) and in February (at the end of the major dry season), whereas our surveys were conducted at the beginning of the little dry season and at the end of the major dry season. Another explanation of this discrepancy will be the variation in the rainfall over the years as observed in March [see Additional file 1: Figure S1]. The comparison of the transmission risk indices between villages shows that the Lambi village had the highest level of the transmission risk, whatever the trapping period. This result is in line with those obtained during medical surveys where most of the patients previously diagnosed were from this village. The high transmission risk index observed in Lambi could also be explained by some environmental factors such as permanent water sources and stable vegetation cover which favour a high level of sleeping sickness transmission. In Ebiminbang, however, most of the rivers dried up during the dry seasons. This may explain the drop of the ADT as well as the low value of the transmission risk index observed in this village.