Kala-azar in Darfur: Evidence for indigenous transmission in Al-Malha Locality, North Darfur, western Sudan

Background Recent reports showed high numbers of visceral leishmaniasis cases in North Darfur, western Sudan. Due to a lack of previous studies, no information is available on local transmission of the disease in these areas. Therefore, a pilot entomological and epidemiological study was conducted in Al-Malha Locality during the year 2013, to investigate possibility of local transmission and places and times of the year where and when people contract the infection. Methods Kala-azar incidence data were obtained from records of Ministry of Health, North Darfur; Al-Malha rural hospital; and the Federal Ministry of Health, Division of Communicable and Non-communicable Diseases. Sand flies were collected using sticky paper and rodent burrow traps from five different microhabitats during three different phases of the year. Species identification was undertaken using appropriate taxonomic keys. Data were statistically analyzed to determine the distribution of kala-azar among different age groups and between sexes, and to compare the species richness and distribution of different sandfly species between the different microhabitats. Results The most affected age groups with kala-azar during the period 2013–2016 were children between one and five years old and those under one year. Females were found to be more affected than males. A total of 918 sand fly specimens were collected using sticky paper and rodent burrow traps from five microhabitats. Identified specimens belong to 13 species; 5 Phlebotomus and 8 Sergentomyia. Phlebotomus orientalis, the principal vector of visceral leishmaniasis (VL) in Sudan and other East African countries, was found for the first time in the area. No other known vector of VL was found in the collection. The highest number of sand flies was recorded during the summer season (63%), with S. antennata (48%) and S. schwetzi (24.1%) being the most abundant species. Among Phlebotomus species, P. orientalis showed relatively high density (8.6%). A dry seasonal water course (called “Khor”) seems to be the most preferred habitat for most of the sand fly species since most of the collections (41.2%) were made from this site, followed by the rodent burrows. Conclusions The presence of P. orientalis and the high prevalence of VL in infants in the Al-Malha area provide the first evidence for local transmission of the parasite causing kala-azar in Darfur. Transmission is probably occurring during summer near the woodland where a high density of the vector was recorded. As a pre-requisite for designing effective control of VL in North Darfur, large scale entomological and epidemiological studies are recommended.


Background
Leishmaniasis is an important vector-borne disease, endemic in 98 countries, 72 of which are in the third world [1]. About 90% of the estimated annual worldwide cases of visceral leishmaniasis (VL) occur in Brazil, Bangladesh, India, Nepal, Sudan, Southern Sudan, Uganda and Ethiopia [2].
Most leishmaniasis cases are zoonotic, transmitted to humans from animal reservoir hosts by the bite of an infected female sand fly vector [3]. However, in East Africa including Sudan, anthroponotic transmission of the parasite causing VL is thought to occur, especially in villages with a significant presence of A. seyal and/or B. aegyptiaca trees or during epidemic situations [4,5]. Among domestic animals, dogs have been found infected by the parasite, but its role in the epidemiology of VL remains to be determined [6].
In Sudan, leishmaniasis represents a serious health problem. Cutaneous leishmaniasis occurs in a fluctuating pattern in the country mainly in the west, central and northern parts [7]. Visceral leishmaniasis or kala-azar is endemic in savannah areas extending from the Sudanese-Ethiopian border in the east to the banks of the White Nile in the west, and from Kassala in the northeast towards the Blue Nile State in the south, with scattered foci in the Nuba Mountains and Darfur [8].
Recently, the disease has revived in a dormant focus in the White Nile State, Central Sudan which may represent an extension of the disease's geographical range [9].
Darfur is a large region in western Sudan, extending over 190 square miles (about the size of Spain). It is bordered to the northwest by Libya, to the west by Chad and to the southwest by Central African Republic. The region is subdivided into three main federal states (North, West and South Darfur). Due to climate change and political reasons, the region has recently suffered enormously from severe civil war and armed conflicts that resulted in the displacement of millions of people [10].
Darfur is well known to be endemic for cutaneous leishmaniasis [6]. The state was considered to be free of visceral leishmaniasis. However, in the past few years, records of the Ministry of Health (FMoH) in Sudan indicated an alarming incidence of VL in Darfur with about 662 reported cases in the period 2006-2010 to 886 cases in 2011-2016 in North Darfur State. Unfortunately, no publications are available on the disease incidence. Furthermore, it is unknown whether transmission occurs locally or if infection is acquired in other regions of the country, e.g. Gedarif State where Darfurian people go for agriculture.
The current study was conducted in Al-Malha Locality (a VL endemic site) during 2013 to examine the possibility of local transmission of VL through investigation of the sand fly fauna, presence of the sand fly vectors, abundance and seasonal distribution. Additionally, work was done to determine incidence rate and age/sex distribution of VL among the different age groups.

Study area
Al-Malha (15°13'40.74"N, 26°15'00.99"E) is a small rural locality, that lies about 215 km north of Al-Fasher town, the capital of North Darfur (Fig. 1). Extending over an area of about 22,000 km 2 , Al-Malha Locality consists of small villages scattered near water and pasture resources, around volcanic mountains. The area is characterized by a long dry season (October-June), followed by a short rainy season (July-September). During the rainy season water drains down seasonal streams known as "Khors" which run through rocky valleys and shallow soil. A cold winter starts around October and ends in December. In March, the area becomes warm and quite dusty. The annual mean maximum temperature varies from 28.3 to 38.1°C, while the annual mean minimum temperature varies from 9.0 to 22.1°C. The average annual rainfall received by the area during 2013 was 30 mm (https:// www.worldweatheronline.com) (Fig. 2).

Kala-azar data collection
Kala-azar incidence data were collected from the records of Al-Malha Rural Hospital, the Ministry of Health, North Darfur and the Federal Ministry of Health, Division of Communicable and Non-communicable Diseases. Only serologically (rK39 RDT) and clinically confirmed VL cases were included in the analysis.

Sandfly collection sites
The following five sites representing different microhabitats in the area were designated for routine sampling of sand flies (Fig. 3a-e):

Sand fly collection and preservation
Sand flies were collected twice per month using Pyrethrum Spray catches (PSC), sticky paper and rodent burrow traps as described by Alexander [11]. At each time point of the study, 5 huts were surveyed by PSC and 10 sticky paper traps were used per site. The traps were fixed overnight from 6:00 pm to 6:00 am. Each trap is made of a white A4 paper sheet, soaked with castor oil on both sides and fixed on sticks held vertically at a height of 20 cm above the ground level. For collections made from rodent burrows, oil soaked papers were rolled and inserted into the burrow. Captured sand flies were removed from the paper traps using fine brushes, washed with distilled water and detergent and preserved in 70% ethanol for identification. Before identification, the head and the abdominal segments of each sand fly specimen were dissected and glass-mounted using locally made Berleze's medium. The specimens were then identified to the species level using keys proposed by Kirk & Lewis [12], Quate [13] and Abonnenc & Minter [14].

Statistical analysis
Data were analyzed using SPSS statistical package (v. 20). Data was not normally distributed. Therefore, the nonparametric Mann-Whitney test was used to compare the median number of VL cases among male and female patients, while the Chi-square test was used to test the distribution of cases among the different age groups.
The Shannon-Weiner (H), evenness (E) and species richness (S) diversity indices of the collected sand flies within different microhabitats were estimated according to Heip et al. [15] using Microsoft Excel sheets (MS Office 2015) as follows: where, Shannon-Weiner index (H) is a quantitative estimate of biological variability as described by Heip et al. [15]. It reflects the number of different species and how evenly the individuals are distributed among these species. The value of the index increases as diversity increases. The relative abundance of each species (p i ) is the proportion of species i relative to the total number of species. Species evenness (E) = H/ln(S) indicates how evenly the individuals in the community are distributed among species. The index ranges between 0 and 1; lower values indicate the presence of dominant species.
Species richness (S) is the total number of different species in the study area.
Species composition was compared between the different collection sites using the Jaccard index of similarity (SC j ) which is calculated using the formula: where, A and B are the richness values of two different collection sites and c is the number of species found in both sites; higher richness indicates higher biodiversity.  were analyzed (Fig. 5). The most affected age groups were children between 1 and 5 years old (125 cases) and those under 1year (71 cases) (χ 2 = 49.94, df= 3 and P < 0.05).

Incidence of kala-azar in Al-Malha Locality
Females were more affected by VL than males, with a total number of infected females of 186 and 116 males during the period 2013-2016 which was significantly higher than that of males (Mann-Whitney U-test: U = 6374.50, Z = -6.019, P = P < 0.0001)
Sergentomyia antennata was the predominant species represented by (47.9%) of the total collections, followed by S. schwetzi (19.9%) and S. clydei (14.5%). Among Phlebotomus species, P. orientalis was the most abundant species (8%). Table 1 summarizes the sex ratio, total number, relative abundance and total number of flies per m 2 for each of the collected sand fly species.
Seasonal fluctuations of sand fly species from January to October 2013 The majority of sand flies (63.2 %) were collected during the summer season (March-June), with a higher abundance rate of Sergentomyia species (88.1%) than Phlebotomus species (11.9 %). Eighteen percent of collections were recorded in early dry season (January-February) and another 18% during the winter season (October-December). No collections were made during the rainy season (July-September) due to the inaccessibility of the collection sites which became completely covered by flood from different Khors ( Table 2).
Phlebotomus orientalis appeared in early dry season in the woodland and Khor habitats and reached its peak abundance in late summer. Significant fluctuations of the abundance of P. orientalis were observed in Khor, woodland, and rodent burrows.
Phlebotomus papatasi appeared at different habitats in small numbers in the summer season, while extremely small numbers of other Phlebotomus species (P. duboscqi and P. bergeroti) were rarely encountered in the collection (Fig. 7).

Sand fly prevalence in the different collection microhabitats
A clear variation in the distribution of the collected sand fly species across the different microhabitats was observed,  with 41.2% of the total collections being made from the Khor, followed by the rodent burrows (35.1%) and the woodland (20.9%) ( Table 3). Although the woodland microhabitat showed less collection percentages and species richness (9 species) than the Khor and rodent burrows, this site resulted in the highest diversity indices (Table 4). This may be due to the dominance of S. antennata in the Khor (52.9%) and rodent burrows (47.2%). Other species showed different relative abundance at the different sites. For example, S. schwetzi was most abundant in rodent burrows (35.5%), S. clydei was most abundant in Khor (23%), while P. orientalis was most abundant in the woodland (26%) than in other sites. It seems that the eroded hill habitat was not preferred by most species since only 4 species were collected from this site (2.4%). The village showed the lowest density (0.44%) with only 4 specimens (2 S. antennata and 2 S. clydei) collected from this site using knockdown method ( Table 3).
The Jaccard index of similarity (SC j ) showed a high species similarity between the Khor and rodent burrows (84.6%) and between the Khor and woodland (75%).

Discussion
Studies on sand fly fauna and their seasonal distribution are crucial for predicting transmission of leishmaniasis, and for the successful implementation of control programs [16]. This study provides baseline data on sand fly ecology, distribution, and their expected role as vectors of leishmaniasis in Al-Malha Locality, North Darfur, West Sudan. The analyzed numbers of kala-azar cases in Al-Malha Locality during 2013-2016 indicates a slightly higher incidence of disease among females than males in most age groups. The data also provided strong evidence for the indigenous nature of the kala-azar transmission in Al-Malha Locality since the most affected age groups were found to be children between one and five years, and those under one year. Studies on other parts of western Sudan reported a similar finding [17]. Phlebotomus orientalis, the vector of VL is known to be more abundant in Acacia-Balanites woodland and rarely associated with human dwellings [13,18]. It seems that the presence of the village at the edge of the woodland makes the villagers more vulnerable to peridomestic transmission of the disease especially during the hot seasons of the year when most people sleep outdoors [19].
More information is needed about the source of infection, parasite strain and the expected reservoir hosts. According to the kala-azar risk map based on predicted vector distribution [20], western Sudan includes some of In eastern Sudan, several studies provided evidence for the existence of zoonotic cycle of L. donovani in the woodlands of Dinder National Park [21,22], with the possibility of the Egyptian mongoose (Hyrpestes ichneumon) acting as a sylvatic reservoir host [23]. Transmission has also been demonstrated to take place in the village habitat, although it is more likely that in these habitats transmission is probably anthroponotic [24].
Although western Sudan has been endemic for cutaneous leishmaniasis for a long time, no records are available on the epidemiology of the disease or its transmission cycle in the region. It is noteworthy that P. papatasi, the main vector of L. major in Sudan and North Africa, and P. duboscqi the main vector in West Africa, were collected in this study. The study area is rich with burrows of rodents and other small mammals that may be suitable reservoir hosts of Leishmania parasites. Therefore, detailed studies on the role of these vectors in the transmission and maintenance of CL in Darfur are recommended.
Thirteen sand fly species were recorded in this study, indicating the fauna richness of the area which can be attributed to the diversity of microhabitats and vegetation in the collection sites. Moreover, availability of rodents and other small animal burrows may provide suitable breeding and resting sites for many sand fly species.
In Sudan and other parts of the old world, most sand fly species including P. orientalis, show seasonal abundance with an increasing abundance during the dry season (January-May), reaching the peak in the early rainy season (April) and drastically drop in numbers during the wet season (August-November) [13,18,25]. Unfortunately, the authors were unable to access the study area during the rainy season and made collections so as to confirm such observations on seasonality of sand fly species in this far part of Sudan.
The abundance of P. orientalis in the woodland was also reported by Quate [13] and Elnaiem & Osman [26]. They confirmed that P. orientalis thrives in habitats dominated by Acacia seyal/Balanites aegyptiaca. In a subsequent publication, the same research group showed a high abundance of VL and active transmission of VL in the village and presented evidence for inter-annual fluctuation of vector density [5]. No sand flies were collected from indoor knockdown collections. For P. orientalis, this result agrees with previous reports which indicated that the vector rarely rests at indoor sites [18]. However, for other species the results are surprising  since Sergentomyia spp. and P. papatasi are known to prefer indoor resting sites [27]. The preference of S. schwetzi in the rodent burrows may indicate a close association between this sand fly species and rodents as a preferred feeding host. Moreover, the relatively more humid microclimatic conditions within these burrows present a more suitable ecological niche for this sand fly species. This has also been observed long time ago by Kirk & Lewis [28]. The predominance of S. antennata throughout the year is in accordance to the findings of Quate [13] in Polich area who mentioned that S. antennata is a nonseasonal species.
The overall sex ratio of most of the collected sand fly species in this study showed that males were more abundant than females, especially for P. orientalis. This finding corroborates the findings of Doha & Samy [29] who also observed a high male/female sex ratio for P. orientalis in Al-Baha, Saudi Arabia. This biased sex ratio is probably due to difference in behavior of either sexes or bias introduced by collection methods [30]. It may also be that females are not attracted to non-baited oil sticky traps [31].

Conclusions
Results obtained in this study are essential to understand the abundance, and seasonality of sand fly species in Al-Malha Locality which represent an essential pre-requisite of the control of leishmaniasis through the control of its sand fly vectors. However, many factors interfere with the fulfillment of the objectives of the study. This includes, for example, the refusal of some Al-Malha residents on the usage of light traps thinking that it may attract the fighting groups. Also, data on seasonality of sand fly species was missing due the inaccessibility of the collection sites during the flood season, in addition to difficulties in obtaining detailed records of kala-azar incidences which add some limitations to interpretation of transmission data. Largescale entomological surveillance in North Darfur and surrounding areas is highly needed to elucidate the spatial and temporal distribution of sand fly vectors and design effective control policy.