Sandfly fauna and ecological analysis of Phlebotomus orientalis and Phlebotomus martini in the lowland foci of visceral leishmaniasis in Somali Regional State, southeast Ethiopia

Araya Gebresilassie, Solomon Yared, Esayas Aklilu

1Department of Zoological Sciences, Addis Ababa University, Addis Ababa, Ethiopia

2Department of Biology, Jigjiga University, Jigjiga, Ethiopia

3Department of Biology, Mada Walabu University, Bale-Robe, Ethiopia

ABSTRACT Objective: To identify the sandfly fauna and analyze ecology of sandfly vector(s) of visceral leishmaniasis in three districts of Somali Regional State, southeast Ethiopia. Methods: Sandflies were collected from four sampling habitats, including indoor, peri-domestic, farm field and mixed forest using light and sticky traps in July 2016, and February and April 2017 in Liben and Dawa zones in the Somali Regional State, southeastern Ethiopia. Results: In total, 4 367 sandfly specimens, belonging to 12 species (three Phlebotomus spp. and nine Sergentomyia spp.) were identified. Phlebotomus (P.) heischi, P. orientalis, and P. martini constituted 45.7%, 31.1%, and 23.1% of the sandfly collection, respectively. There were significant differences in the median number of P. orientalis, and P. martini captured per CDC trap/night between the three sampling districts (P＜0.05). In light trap capture, collection habitats had significant effects on the abundance of P. orientalis, and P. martini (P＜0.05). More median numbers of P. orientalis, and P. martini species were collected in agricultural fields followed by mixed forest and peri-domestic habitats. P. orientalis and P. martini were caught more in outdoor than indoor habitats, suggesting exophilic behaviour. Conclusions: The study demonstrated that the presence of P. orientalis and P. martini are probable vectors of visceral leishmaniasis in this new focus. The findings of our study will improve the understanding of the dynamics of visceral leishmaniasis transmission and will facilitate the implementation of integrated disease control measures based on ecological knowledge of visceral leishmaniasis vector in Liben and Dawa zones and its surrounding regions.

KEYWORDS: Ecology; Phlebotomus orientalis; Phlebotomus martini; Sandfly fauna; Visceral leishmaniasis

1. Introduction

In Ethiopia, visceral leishmaniasis (VL) caused by Leishmania (L.) donovani complex is a public health threat, with approximately 3.2 million people at risk and 3 400-5 000 cases occurring annually[1,2]. This disease is known to be endemic in the semi-arid Humera and Metema plains in the northwest[2,3], and the south-western lowlands, which include Omo and Aba-Roba plains, Woyto, Segen, Dawa and Genale river valleys[3,4]. Moreover, VL cases were reported in the south-eastern semi-arid lowlands of Afder, Liben and Gode zones from the Somali Regional State[5,6].

Like any other vector-borne disease, VL is transmitted by the bite of infected female sandflies of genus Phlebotomus in the Old World and Lutzomyia in the New World[7,8]. In the Old World, proven or suspected sandfly vectors constitute a total of 42 species, of which 26 are implicated in VL transmission caused by L. donovani and L. infantum[9].

Determining the diversity and distribution of sandfly species in different ecologies of Leishmania is an important step in the identification of the proven or suspected vectors. With this in mind, investigations to identify the sandfly fauna and vectors responsible for the transmission of two forms of leishmaniases were undertaken in Ethiopia. This research on sandfly fauna has resulted in identification of 22 Phlebotomus species, representing seven subgenera (Larroussius, Synphlebotomus, Phlebotomus, Paraphlebotomus, Anaphlebotomus, Adlerius, and Parvidens)[10-13]. Specifically, Phlebotomus (P.) orientalis, P. martini, and P. celiae are regarded as the proven vectors of VL in different endemic regions of the country[10,14-16]. In southern Ethiopia, the principal vector of VL is P. martini, which breeds and rests in termite mounds[17]; whereas in north and northwest Ethiopia, P. orientalis is implicated as a vector that is mostly confining to Acacia-Balanites-Zyzphus spp. forest and deeply cracking ‘black cotton clay’ soils[11,18].

In recent years, VL has spread to Liben and Dawa zones of Somali Regional State, in south-eastern Ethiopia, areas where it was previously non-endemic (Somali Region Health Bureau, unpublished data)[5]. In this new VL focus, seven cases caused by L. donovani were diagnosed and treated by Medecins Sans Frontieres (MSF), of which most cases occurred within the children of the pastoralist community, suggesting autochthonous transmission. In addition, in the neighbouring Bakool region of Somalia, around 1 671 VL patients were admitted to the Huddur Center of MSF Holland from January 2002 until December 2006[19]. In spite of the public health significance and the recent occurrences of VL in these particular foci, no entomological studies to determine the importance of Phlebotomus species were conducted. Therefore, the current study was carried out to identify the sandfly fauna and bionomics of the suspected vector(s) of VL in Liben and Dawa zones of Somali Regional State, southeast Ethiopia.

2. Materials and methods

2.1. Study areas

Entomological investigations were undertaken in two districts (i.e. Filtu and Deka-Softu) of the Liben zone, and one district (Hudet) of Dawa zone in the Somali Regional State, southeastern Ethiopia (Figure 1) in the months of July 2016, February and April 2017. Filtu district is one of the six districts in Liben zone. In Filtu district, the altitude ranges from 400 m along the rivers to 1 540 m at mount Fiil, east of Filtu town. The town is also located about 715 km southeast of Addis Ababa. Another district in the zone is Deka Softu that is bordered on the west with Fitu district. The centre of Deka Softu is 655 km from Addis Ababa. On the other hand, the third sampling district, Hudet is found in Dawa zone, with a latitude and longitude of 4°45’0” N and 39°13’60” E, respectively. The district is located at an elevation of 912 m above sea level.

Figure 1. Map of the study villages in Liben and Dawa zones of Somali Regional State, southeastern Ethiopia.

The climatic condition of the area is generally ranged from arid to semi-arid, with an annual rainfall of about 200-600 mm per year. The rain season starts at the middle of April and ends at beginning of June. The maximum and minimum rainfall intensities are recorded in April and October, respectively. The maximum temperature varies from 25 ℃ to 40 ℃ in March to the coolest month of July with 19 ℃-26 ℃. The vegetation comprises drought resistant bushes and Acacia spp., Commiphora spp., and Boswellia spp. family trees in pastoral areas. The inhabitants of the study areas are mainly pastoralists, who make their livelihood by raising camel, cattle, goats, and sheep as an essential supplement to the agro-pastoral economy. During the wet seasons, the communities grow maize and sorghum depending on the rain intensity. Squirrels, rodents, whitetailed mongoose (Ichneumia spp.) and foxes are some of the wildlife fauna that are commonly seen in the study villages.

2.2. Sandfly sampling

For different aspects of entomological study, three representative districts, namely Filtu, Deka-Softu, and Hudet were selected. From the three districts, six villages (Mesajid and Bekeka in Filtu; Sora and Takathager in Deka-Softu; Mersha and Luchole in Hudet) were chosen based on VL occurrence. Within the sampling villages, four representatives trapping habitats indoor, peri-domestic, farm field and mixed forest were identified and used for the entire sandfly species collection. Sandflies were trapped for two consecutive nights at each sampling village, constituting a total of 18 collection nights. CDC light traps (CDC-LTs): Sandflies were collected in CDC light traps (LT). LTs were deployed at peri-domestic, agricultural fields, and mixed forest land. Three LTs were set in representative sites of peri-domestic habitats such as cracked walls, stone piles, and animal enclosures. Another three LTs in each village were positioned to sample sandflies in agricultural fields, where they were positioned in open fields, dry riverbeds, and the edge of farmlands. LTs were suspended in mixed forest with the fan 0.4-05 m above the ground level. The traps were deployed 1 h before sunset and collected at dawn the next morning. Afterwards, the sandflies were sorted by sex and genus (Phlebotomus or Sergentomyia spp.), and preserved in 70% ethanol for later species identification.

Sticky traps (ST): White STs of polypropylene sheets, measuring 21 cm width and 29 cm length, coated with sesame oil were used for capturing sandflies. Four STs were installed inside five different houses in each village to intercept and capture any indoor resting (endophilic) sandflies. Similarly, another four STs were randomly installed horizontally on cracked walls, stone piles, and animal enclosures in the peri-domestic biotopes. At the same time, the remaining four STs were deployed in representative sites of agricultural fields, and mixed forest land. Two sets of STs were placed horizontally on the cracks of agriculture fields while the other two sets of STs were hung vertically in a row 30 cm above the ground supported by metal pegs. Each morning, sandflies from STs were removed using forceps and stored in 96% ethyl alcohol in labelled vials for identification.

2.3. Mounting and identification of sandflies

Sandflies collected during the study were mounted on microscope slides in Hoyer’s medium with their heads separated from thoraces and abdomens. Sandflies were identified to the species level based on the morphology of external genitalia of males, pharynx and antennal features, and spermathecae of females, using different keys[20] and other publications[21].

2.4. Statistical analysis

Statistical analyses were conducted using IBM SPSS statistics, version 20 for Windows (SPSS Inc., Chicago, IL, USA), and Microsoft®Office Excel 2007. Sandfly numbers were checked for normality by Kolmogorov-Smirnov test before data analysis. When trapping data did not conform to the normal distribution, the nonparametric equivalent tests of Kruskal-Wallis and Mann-Whitney-U were applied. Kruskal-Wallis test (K-W) was followed to compare the median numbers of P. orientalis and P. martini collected in the three sampling villages using CDC-LTs. Likewise, Kruskal-Wallis test was used for habitat preference comparisons for CDCLT captures. For multiple comparisons of medians, multiple-Mann-Whitney U-test was used and, P-values were adjusted with the Bonferroni correction to adjust for the inflation of type I errors when several Mann-Whitney tests are performed[22]. Mann-Whitney test (U) was also used for comparing sex ratio of P. orientalis and P. martini in trapping methods. Statistical analyses were considered significant when P＜0.05 unless stated.

3. Results

3.1. Species composition and relative abundance of sandflies

A total of 4 367 sandfly specimens comprising 12 species in two genera (Phlebotomus and Sergentomyia) were collected using LTs and STs during three surveys (Table 1). Overall, 3 and 9 sandfly species in the subgenera Phlebotomus and Sergentomyia were identified, respectively. In the three sampled districts, 1 399, 1 702 and 1 266 specimens of sandfly were identified in Filtu, Deka-Softu and Hudet districts, respectively. In Filtu district, 11 species of sandfly were caught, comprising Phlebotomus heischi, Phlebotomus orientalis, Phlebotomus martini, Sergentomyia squamipleuris, Sergentomyia schwetzi, Sergentomyia bedfordi group, Sergentomyia clydei, Sergentomyia antennata, Sergentomyia adleri, Sergentomyia adami, and Sergentomyia dubia. In Deka-Softu district, however, 8 species of sandflies, belonging to the genus Phlebotomus and Sergentomyia were trapped, which constituted 39.0% of the total collection. Sergentomyia africana was only collected in Hudet district while P. heischi was absent from the faunal list of the district.

Table 1. Relative abundance and fauna of sandflies collected from three districts in Liben and Dawa zones, July 2016, February and April, 2017.

Table 1 depicts the relative abundance of different sandfly species collected from the three sampling districts. Of the 2 301 Phlebotomus specimens, P. heischi constituted 45.7% of sandflies captured, followed by P. orientalis (31.1%) and P. martini (23.2%). Sergentomyia squamipleuris was the most predominant species, accounting for 25.6% and 12.1% of Sergentomyia species and the entire sandfly collection, respectively. The remaining species in descending order were Sergentomyia bedfordi group (8.8%),

Sergentomyia clydei (8.3%), Sergentomyia schwetzi (8.2%), Sergentomyia antennata (7.4%), Sergentomyia adleri (1.8%), Sergentomyia africana (0.4%), Sergentomyia adami (0.3%), and Sergentomyia dubia (0.1%).

3.2. Abundance of Phlebotomus species in the three sampling districts

Table 2 shows the median numbers of P. orientalis and P. martini trapped from the three sampling districts (Filtu, Deka-Softu, and Hudet) using CDC Lts during the three collections periods. There was a significant difference in the median number of P. orientalis captured per CDC trap/night in the three sampling districts [Kruskal-Wallis test (χ2K-W=5.42); df=2; P＜0.05]. Similarly, the sampling districts significantly differed in their median number of P. martini caught in light trap collections [Kruskal-Wallis test (χ2K-W=5.95); df=2; P＜0.05]. Deka-Softu was the most productive sampling village for P. orientalis compared to Filtu and Hudet districts. In contrast, P. martini was trapped in large numbers in the district of Hudet followed by Filtu. The number of P. orientalis and P. martini collected in STs in all sampling habitats were fewer so that no statistically analyses were run.

Table 2. Median numbers (IQR) of Phlebotomus orientalis and Phlebotomus martini collected by CDC light traps from three different sampling districts.

3.3. Habitat preferences of P. orientalis and P. martini

A significant difference was recorded in the median density of P. orientalis among the three sampling habitats using CDC Lts [Table 3; Kruskal-Wallis test (χ2K-W=11.94); df=2; P＜0.05]. Higher median number of P. orientalis per CDC light/night was caught in agricultural fields [3.17 (0.43) per trap/night] than mixed forest habitats [1.46 (0.55) per trap/night], and peri-domestic [0.18 (0.12) per trap/night]. Similarly, habitat type had significant effects on the density of P. martini in light trap collections [Kruskal-Wallis test (χ2K-W=6.91); df=2; P＜0.05]. The median density of P. martini collected per CDC light traps per night in peri-domestic, mixed forest, and agricultural fields were 0.40 (0.11), 1.47 (0.45), and 3.85 (0.23) per trap/night, respectively.

Table 3. Median numbers (IQR) of Phlebotomus orientalis and Phlebotomus martini collected in CDC light traps/trap/night from different sampling habitats.

3.4. Indoor and outdoor abundance of P. orientalis and P. martini

In total, 30 and 33 specimens of P. orientalis and P. martini were captured on sticky traps deployed outdoor, respectively (Table 4). However, sticky traps placed indoors were not productive in collecting a single specimen of P. orientalis and P. martini.

Table 4. Number of Phlebotomus orientalis and Phlebotomus martini captured by sticky traps indoor and outdoor

3.5. Sex ratio

Sex ratios (males: females) for different sandfly species showed that males caught by all methods combined was slightly higher than that of females (2 220 males: 2 147 females), with an overall sex ratio of 1.03:1 (Table 5). For P. orientalis, the sex ratio in LTs was 0.72:1, which was not statistically different [U(1)=156.50, Z=-0.18, P＞0.05]. However, a relatively high ratio of male versus female (2.53:1) on the sticky traps was observed. P. martini had the sex ratios slightly in favour of males with 1.17:1 and 2.00:1 in light and sticky traps, respectively; while the former was statistically not significant (Mann Whitney U-test, P＞0.05).

Table 5. Sex ratio of sandfly species collected from different habitats using CDC light traps and sticky traps in Filtu, Deka-Suftu and Hudet districts.

4. Discussion

A cross-sectional entomological investigation was carried out in three districts of Somali Regional State, southeast Ethiopia. This study showed that 12 species of sandflies were distributed in this region and that 3 were Phlebotomus and 9 were Sergentomyia. To the best of our knowledge, this is the first report of sandfly fauna in this region of Ethiopia. Overall, the sandfly fauna in the present study area has distinctive Afrotropical elements, with essentially the same species found elsewhere in East Africa[8,23]. Three species of Phlebotomus were identified in the study areas, where P. heischi was the dominant, constituting 24.1% (1051/4367) of total sandfly captures. Nevertheless, the role played by this species in the epidemiology of the VL is yet unclear.

The two sandfly species, P. orientalis and P. martini, are the proven vectors of VL in East Africa[23]. P. orientalis is a known vector of L. donovani in Sudan, South Sudan[24,25], southwestern Ethiopia[14], and the northern Ethiopia[15,16]. P. martini has been implicated as the principal vector of VL in southern Ethiopia[10] and Kenya[26,27]. The dominance of P. orientalis and P. martini in our study was previously reported in various parts of Ethiopia[11,28,29]. These sandfly species apparently play a significant role in the epidemiology of VL in this focus area.

In addition to Phlebotomus spp., 9 species of Sergentomyia were collected and identified. The most abundant species was Sergentomyia squamiplueris, while the other four (Sergentomyia bedfordi group, Sergentomyia clydei, Sergentomyia schwetzi, and Sergentomyia antennata) were also common; the remaining four species were rarely trapped. In general, the sandfly fauna encountered in the present study is consistent with previous reports in other parts of Ethiopia[11,12,28-30].

The abundance of P. orientalis and P. martini differed among the three study districts. For P. orientalis, Deka-Softu was the most abundant sampling district while Filtu and Hudet were the least. The vast area of Deka-Softu and its environs is consisted of a vertosol soil type. This soil type is characterized by extensive cracking during dry season that contributes to a favourable habitat for resting/breeding of P. orientalis[18]. This factor has contributed to the moderate abundance of P. orientalis in the district. In contrast, P. martini was trapped in large numbers in the districts of Hudet, followed by Filtu. The ecological conditions in Hudet are mainly characterized by the presence of red soils, consisting of large number of pinnacletype termite mounds. Many species of sandflies rest in the airshafts of termite hills, and P. martini, which occupies old shaded termite mounds, is the vector of “termite hill kala-azar” in Kenya[27] and southern Ethiopia[17]. These may explain the predominant abundance of P. martini in this district.

With respect to microhabitats, agricultural fields were the most productive microhabitat for the Phlebotomus spp. compared to other habitats as shown by light trap captures. Higher numbers of P. orientalis and P. martini were trapped in agricultural fields than other habitats, which could be related to the dominance of black cotton soil (vertosolic). Higher humidity in combination with stable temperatures maintained throughout dry season, and the availability of organic matter that provides food for immature development in the deeper layers of cracked vertisol, creating a suitable microclimate for the P. orientalis[18,31]. In addition, the presence of some vegetation in the farm field could provide shade and source of sugar for adult populations[31]. Earlier studies in eastern Sudan and northern Ethiopia also showed that P. orientalis is mainly associated to forest area with large expanses of vertisol[11,31] and rarely associated with human dwellings.

In Hudet, agricultural fields consist of large numbers of pinnacletype termite mounds that provide ideal breeding and resting sites for the three East African species sandflies of the subgenus Synphlebotomus (P. martini, P. celiae and P. vansomerenae) and many other sandflies in southern Ethiopia and different parts of Kenya[17,26,33]. Moreover, various species of animals, particularly rodents, mongooses, and reptiles (snakes and lizards) live in the ventilation shafts of termite mounds, some of which (rodents and mongooses) are suspected as reservoir hosts of leishmaniasis[34]. As a result, the termite hill ecology requires more researches to fully understand the significance in the epidemiology of VL.

Attempts to sample indoor resting sandfly species in general and Phlebotomus spp. (P. orientalis and P. martini) in particular yielded fewer captures. In the same way, trapping endophilic Phlebotomus species using pyrethrum spray catches did not yield any specimens (result not indicated). This observation suggests that both sandfly species have a propensity towards exophilic (outdoor resting) behaviour, a finding which is in agreement with earlier studies in some districts in northern Ethiopia[11,30]. However, this finding is unlike that of Lambert et al.[35] where 79% of the total collected P. orientalis by light traps were from inside human dwellings in eastern Sudan. From epidemiological point of view, this observation has important significance in planning VL control measures based on conventional vector control methods, such as indoor residual spraying.

In our study, P. orientalis and P. martini had slightly higher proportion of male to female both in CDC light traps and sticky traps. Similar trends were also noted for other sandfly species [11,29]. Given a normal sex ratio of 1, then skewed ratios in favour of males might be related to the fact that the traps were placed near emergence sites, where males are generally abundant.

In conclusion, our data showed the occurrence of 12 species of sandflies, belonging to two genera: Phlebotomus and Sergentomyia in the southeast VL focus of Ethiopia. P. orientalis and P. martini reported herein are the two principal vectors of VL, which could probably contribute to the VL transmission in this new VL focus. The study also revealed that P. orientalis and P. martini exhibit exophilic behaviour. This knowledge is important in the selection and application of effective vector control interventions. Programs of the exophilic behaviour of P. orientalis and P. martini should be designed and taken into account. Moreover, further studies to determine the infection rates and host preference analysis of P. orientalis and P. martini should be conducted in the current study area and other VL endemic foci in the Somali Regional State.

Conflict of interest statement

The authors declare they have no conflict of interest.

Acknowledgements

The authors would like to acknowledge Jigjiga University for providing financial support to the study. We are also indebted to the people of the three study areas for their kind co-operation. Our thanks also goes to Mr. Bisrat Tesfay for his kind support in designing the study area map.

Authors’ contributions

AG and SY conceived the investigation. AG, SY, and EA performed the field work and identified the sandflies. AG, SY, and EA analyzed the data and prepared the draft manuscript. All authors read and approved the final version of the manuscript.