Entomological survey of sand fly vectors and molecular screening for Leishmania parasite in refugee camps in Ethiopia

Abstract

Visceral leishmaniasis is transmitted by infected female sand flies. However, entomological data from refugee camps remained limited. This study aimed to describe sand fly fauna and assess natural Leishmania infection in selected refugee camps in Ethiopia. The study was conducted in four refugee camps. Sand flies were collected using a CDC light trap and sticky trap. Both male and female sand flies were dissected, and morphologically identified. The female sand flies were preserved in 95% ethyl alcohol for Leishmania DNA screening. A total of 2196 sand flies were collected. Phlebotomus rodhaini, the only Phlebotomus species identified, accounted for 1.7% of the collections. Sergentomyia antennatus was the most abundant species (50.8%), followed by S. africanus (15.8%), S. schwetzi (14.3%), S. bedfordi (10.9%), S. clydei (5.2%), S. squamipleuris (1.2%), and S. adleri (0.1%). Species richness was highest in Terkidi camp, whereas overall diversity was greatest in Sherkole. Indoor collections were limited (< 2.3%), confirming predominantly exophilic resting behavior. Termite hills and peridomestic habitats harbored the highest sand fly densities. Polymerase chain reaction screening showed no evidence of Leishmania infection. This first entomological survey conducted in refugee camps in Ethiopia documents a sand fly fauna dominated by outdoor collections, a very low abundance of Ph. rodhaini, and no detectable Leishmania infection. These findings suggest the need for longitudinal monitoring to capture seasonal variation.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37733-2.

Introduction

Visceral leishmaniasis (VL), also known as kala-azar, is a neglected tropical disease caused by protozoan parasites of the genus Leishmania. These parasites are transmitted through the bites of infected female Phlebotomus sand flies¹. The disease is a significant cause of global morbidity and mortality, with East Africa bearing a substantial burden of cases². Among the East African countries, Ethiopia is particularly affected, with six of its regions classified as endemic for VL³. The disease primarily impacts the lowland areas of the country. Annually, Ethiopia records an incidence of 3700–7400 VL cases⁴. Furthermore, an estimated 3.2 million people in Ethiopia live in areas at risk for contracting VL³.

Sand flies of the genus Phlebotomus are confirmed vectors responsible for transmitting leishmaniasis in the Old World⁶. In Ethiopia, the principal vectors for VL are Phlebotomus orientalis, predominantly found in the lowland and highland regions of the northwest, and Phlebotomus martini and Phlebotomus celiae, prevalent in the southern and southwestern areas^3,7–10. The distribution and abundance of these vectors are strongly influenced by ecological factors. Phlebotomus orientalis is closely associated with black cotton clay soils (vertisols) and Acacia seyal woodlands, while Ph. martini and Ph. celiae are typically linked to the presence of termite hills^3,11–13. Conversely, sand fly species within the genus Sergentomyia are generally considered as non-vectors of VL¹⁴. However, the detection of Leishmania DNA in these species suggests the need for further investigation into their potential role in VL parasite transmission. Studies from Kenya¹⁵, West Africa^14,16, and other regions^17,18 have reported the presence of Leishmania DNA in Sergentomyia spp., which often outnumber Phlebotomus populations. Similarly, research in Ethiopia has indicated that Sergentomyia is the predominant genus in VL endemic areas^11,19,20.

A strong association has been documented between civil unrest and the transmission of VL^21,22. The VL infection often arises when immunologically naïve migrants are displaced into VL endemic areas or when Leishmania infected individuals move into new areas and establish additional foci of transmission. In East Africa, the situation is further intensified by limited access to timely diagnosis and effective treatment, HIV co-infection, food insecurity, and malnutrition^4,21 and presence of arid areas with acacias and termite mounds²³. Additionally, a meta-analysis from northern Ethiopia indicated that the pooled prevalence of HIV infection among VL infected individuals was 24% (95%CI: 17–30%)²⁴. Moreover, migration is also documented as a factor to increase the burden of VL⁵.

Ethiopia ranks as the third largest refugee hosting country in Africa, sheltering over one million refugees and asylum seekers, mainly from South Sudan, Sudan, Somalia, and Eritrea. Nearly half of these refugees (48%) are hosted in the Gambella and Benishangul Gumuz Regional States²⁵. Environmental risk assessments for VL transmission have identified these regions as high-risk areas². Many refugees, particularly those from Sudan and South Sudan, originate from VL endemic regions²¹. Previous studies in these areas have revealed a high prevalence of asymptomatic Leishmania infections through serological and molecular methods^22,26,27, while confirmed VL cases continue to be reported annually to the Ethiopian Ministry of Health (MoH unpublished data). Despite these findings, systematic entomological surveillance in refugee settings within the Gambella and Benishangul Gumuz Regional States remains limited. This lack of comprehensive data hampers the development and implementation of evidence-based vector control measures, a critical element of the WHO’s strategy to eliminate VL as a public health issue in East Africa. Ethiopia, together with other East African countries, has committed to achieving this goal by 2030²⁸.

To address this gap, we investigated the species composition, relative abundance, habitat preference, and natural infections of Leishmania DNA in sand flies from refugee camps in Gambella and Benishangul Gumuz Regional States, Ethiopia. This study provides essential evidence to guide targeted vector control strategies by enhancing the understanding of sand fly behavior and resting sites, strengthening surveillance systems, and supporting public health efforts aimed at reducing VL transmission among vulnerable populations. Additionally, these findings contribute to better epidemic prevention, more efficient resource allocation, and progress toward Ethiopia’s and WHO’s goals for VL elimination.

Results

Preliminary assessments prior to actual sand fly collection

During the preliminary assessment of sand flies, a total of 474 specimens were collected in Terkidi and Kule refugee camps. However, no sand flies collected in Tsore and Sherkole refugee camps. Morphological examination at the genus level revealed that all samples belonged to the genus Sergentomyia. However, species-level identification was not conducted at this time.

Sand fly species composition and relative abundance across sites

In the current study, a total of 2,196 adult phlebotomine sand flies (1,542 males and 654 females) belonging to eight species in two genera (Phlebotomus and Sergentomyia) were collected using CDC light traps (LTs) and sticky traps (STs). The genus Phlebotomus was represented by a single subgenus (Anaphlebotomus), while the genus Sergentomyia was represented by four subgenera (Grassomyia, Sergentomyia, Sintonius, and Parrotomyia) (Table 1). One species of Phlebotomus and seven species of Sergentomyia were identified from the four refugee camps in Gambella and Benishangul Gumuz Regional States. The majority of the sand flies (85.1%, n = 1,869) were collected from refugee camps in Gambella; specifically, Kule (39.0%, n = 857) and Terkidi (46.1%, n = 1012). In contrast, 14.9% (n = 327) were collected from refugee camps in Benishangul Gumuz Regional State: Sherkole (14.1%, n = 310) and Tsore (0.8%, n = 17). Among the refugee camps, Terkidi accounted for the largest proportion of specimens (46.1%), followed by Kule (39.0%), Sherkole (14.1%), and Tsore (0.8%) (Table 1).

Table 1.

Distribution of sand fly species by refugee camp in Gambella and Benishangul Gumuz regional States of Ethiopia, April-May 2025.

Species (subgenus)	Regional States				Total (%)
	Gambella (n = 1869)		Benishangul Gumuz (n = 327)
	Kule (%)	Terkidi (%)	Sherkole (%)	Tsore (%)
Ph. rodhaini (Anaphlebotomus)	0 (0.0)	26 (2.6)	11 (3.5)	1 (5.6)	38 (1.7)
S. adleri (Sintonius)	0(0.0)	1 (0.1)	0 (0.0)	0 (0.0)	1 (0.1)
S. africanus (Parrotomyia)	78 (9.1)	134 (13.2)	122 (39.4)	11 (64.7)	345 (15.7)
S. antennatus (Sergentomyia)	522 (60.9)	552 (54.5)	41 (13.2)	1 (5.6)	1116 (50.8)
S. bedfordi group (Sergentomyia)	105 (12.3)	84 (8.3)	48 (15.5)	2 (11.8)	239 (10.9)
S. clydei (Sintonius)	84 (9.8)	28 (2.8)	2 (0.6)	1 (5.6)	115 (5.2)
S. schwetzi (Sergentomyia)	59 (6.9)	169 (16.7)	86 (27.7)	1 (5.6)	315 (14.3)
S. squamipleuris (Grassomyia)	9 (1.1)	18 (1.8)	0 (0.0)	0 (0.0)	27 (1.2)
Grand Total	857 (39.0)	1012 (46.1)	310 (14.1)	17 (0.8)	2196 (100.00)

Sergentomyia antennatus was the most abundant species (50.8%, n = 1116) of the total collection, with dominance observed in Kule (60.9%) and Terkidi (54.5%) (p < 0.001). Sergentomyia africanus ranked second in abundance (15.7%, n = 345), with its highest distribution in Sherkole (39.4%) and Tsore (64.7%). Sergentomyia schwetzi (14.3%, n = 315) and Sergentomyia bedfordi (10.9%, n = 239) were also widely distributed, particularly in Terkidi and Sherkole. Less common species included Sergentomyia clydei (5.2%, n = 115), Phlebotomus rodhaini (1.7%, n = 38), and Sergentomyia squamipleuris (1.2%, n = 27), while Sergentomyia adleri was rare, represented by a single specimen (0.1%, n = 1). Species richness was greatest in Terkidi, where all eight species were recorded, whereas Tsore exhibited both low abundance and diversity (Table 1).

Species composition relative to collection methods

The distribution of phlebotomine sand flies by collection method is presented in Supplementary File, Table 2. The majority of sand flies (96.0%, n = 2109) were captured using ST (X² = 22.1; p = 0.016), with males representing 70.9% and females 29.1%. Sticky traps, deployed in household compounds, indoor, natural habitats, and termite hill habitats, proved highly effective in capturing all reported sand fly species. Majority of the Ph. rodhaini species (94.7% 36/38) was obtained using ST collection method. In contrast, LTs contributed only 4.0% of the total sand flies collected, with nearly equal proportions of males (52.9%) and females (47.1%). These findings demonstrate the superior efficiency of STs for capturing sand flies in diverse habitats, while LTs provided a more balanced representation of sexes and still collected a wide range of species.

Table 2.

Indoor and outdoor abundance of sand fly species across the four study sites in Gambella and Benishangul Gumuz regional States, Ethiopia, April-May 2025.

Species	Kule		Sherkole		Terkidi		Tsore		Total (%)
	Indoor	Outdoor	Indoor	Outdoor	Indoor	Outdoor	Indoor	Outdoor	Indoor	Outdoor
Ph. rodhaini	0	0	0	11	1	25	1	0	2(5.3)	36(94.7)
S. adleri	0	0	0	0	0	1	0	0	0(0.0)	1(100.0)
S. africanus	8	70	0	122	4	130	1	10	13(3.8)	332(96.2)
S. antennatus	10	512	0	41	12	540	0	1	22(2.0)	1094(98.0)
S. bedfordi	5	100	0	48	2	82	0	2	7(2.9)	232(97.1)
S. clydei	1	83	0	2	0	28	0	1	1(0.9)	114(99.1)
S. schwetzi	2	57	0	86	1	168	0	1	3(1.0)	312(99.0)
S. squamipleuris	0	9	0	0	1	17	0	0	1(3.7)	26(96.3)
Total (%)	26(3.0)	831(97.0)	0(0.0)	310(100.0)	21(2.1)	991(97.9)	2(11.8)	15(88.2)	49(2.2)	2147(97.8)

Indoor and outdoor collections of sand flies

Indoor collections across all sites were notably low, accounting for only 49 (2.2%) specimens out of 2,196 total specimens; however, this result was not statistically significant (p = 0.209). Among the 38 Ph. rodhaini specimens, only 2 (5.3%) were collected from indoor habitats. Sergentomyia africanus 3.8% (n = 13) and S. antennatus 2.0% (n = 22) were among indoor collections, while other species, including S. bedfordi, S. clydei, S. adleri, and S. schwetzi, were rarely found in indoor habitats. This pattern highlights the predominantly outdoor behavior of these species in the study area (Table 2).

Sex ratio and abdominal status of sand flies

The overall male-to-female sex ratio of sand flies collected using LT was 1.1:1, whereas a ratio of 2.4:1 was recorded for sand flies captured using ST. Across all sites and collection methods, male sand flies significantly predominated over females (p < 0.001). Male sand fly predominance was particularly pronounced in Ph. rodhaini (27/11), S. antennatus (888/228), and S. bedfordi (192/47) based on ST collections. In contrast, female-biased ratios were observed in S. schwetzi, S. bedfordi, and S. squamipleuris, while S. clydei showed a more balanced distribution at some sites (Table 3). Overall, STs captured significantly more females than LTs (X² = 13.0; p < 0.001), highlighting their efficiency for sampling host-seeking females. Of the 654 female sand flies examined across all sites, 30 (4.6%) were blood-fed and 75 (11.5%) were gravid. In the current study S. schwetzi, and S. clydei showed notable proportion of blood feed and gravid female sand flies reported. Only one single gravid specimen of Ph. rodhaini was detected in Sherkole (Supplementary Table 1).

Table 3.

Male to female ratio of sand flies among refugee camps by collection methods in Gambella and Benishangul Gumuz regional States, Ethiopia, April-May 2025.

Species	Collection methods
	LT				ST				Total
	Kule	Sherkole	Terkidi	Tsore	Kule	Sherkole	Terkidi	Tsore	Kule	Sherkole	Terkidi	Tsore
	M/F	M/F	M/F	M/F	M/F	M/F	M/F	M/F	M/F	M/F	M/F	M/F
Ph. rodhaini	-	0/1	-	1/0	-	8/2	18/8	-	-	8/3	18/8	1/0
S. adleri	-	-	-	-	-	-	1/0	-	-	-	1/0	-
S. africanus	1/3	8/4	3/4	0/2	22/52	60/50	51/76	5/4	23/55	68/54	57/80	5/6
S. antennatus	11/0	2/0	8/12	-	415/96	30/9	421/111	1/0	426/96	32/9	429/123	1/0
S. bedfordi	2/0	1/4	1/1	-	97/6	37/6	53/29	1/1	99/6	38/10	54/30	1/1
S. clydei	-	0/1	2/0	-	42/42	0/1	14/12	0/1	42/42	0/2	16/12	0/1
S. schwetzi	1/0	2/0	2/6	-	47/11	61/23	97/64	0/1	48/11	63/23	99/70	0/1
S. squamipleuris	-	-	1/3	-	4/5	-	11/3	-	4/5	-	12/6	-
Grand Total	15/3	13/10	17/26	1/2	627/212	196/91	666/303	7/7	642/215	209/101	683/329	8/9

Habitat preferences of the sand flies

Table 4 shows the mean number of sand flies collected for each trap, habitat, and night. Ph. rodhaini, S. africanus, S. antennatus, S. bedfordi, and S. schwetzi showed a strong association with habitat type according to the Kruskal-Wallis H test, indicating a preference for termite hills located in both village and natural habitat environments. However, S. clydei was associated with termite hills in natural habitats only. Additionally, S. bedfordi demonstrated a statistically significant preference for natural habitats, in addition to termite hills in both the village and natural habitats (Table 4).

Table 4.

Mean density with standard error (± SE) of sand fly species per ST per night across different habitats in refugee camps located in Gambella and Benishangul Gumuz regions, Ethiopia, April-May 2025.

Species	Compound	Indoor	Natural Habitat	TH_in village	TH_ Natural Habitat
Ph. rodhaini	0.03 ± 0.02	0.07 ± 0.06	0.28 ± 0.25	0.42 ± 0.20*	0.23 ± 0.08*
S. africanus	0.52 ± 0.16	0.22 ± 0.12	0.59 ± 0.34	3.50 ± 1.96*	2.84 ± 0.97*
S. antennatus	1.09 ± 0.16	0.29 ± 0.12	5.19 ± 0.34	4.00 ± 1.96*	11.59 ± 0.97*
S. bedfordi	0.25 ± 0.61	0.09 ± 0.25	1.28 ± 1.34*	1.21 ± 1.26*	2.25 ± 4.04*
S. clydei	0.10 ± 0.04)	0.01 ± 0.01	0.16 ± 0.12	0.13 ± 0.13	1.61 ± 0.93*
S. schwetzi	0.23 ± 0.81	0.03 ± 0.08	1.16 ± 2.69	0.33 ± 3.30*	2.93 ± 3.48*

*Indicated statistically significant (p < 0.05).

TH natural Habitat= termite hill located in the natural habitat of the study site; TH_in village= termite hills located in the village of the study areas.

Diversity of sand flies among refugee camps

For the sand fly populations in four refugee camps, the Shannon-Wiener diversity index (H’), evenness (E), and species richness (S) were calculated (Table 5). Sherkole had the highest Shannon diversity index (H’=1.430), while Tsore had the lowest (H’=1.200). Compared to Kule, Sherkole had a varied sand fly population (p = 0.001). Conversely, Kule had a greater diversity of sand fly species than Terkidi (H’=0.690 vs. 0.662). On the other hand, Tsore had extremely low total abundance (n = 17), which limits species richness and inflates the influence of small count changes on diversity metrics. The number of species reported in Terkidi (S = 8) was higher than in Kule (S = 6) (p = 0.001).

Table 5.

The Shannon-Wiener diversity index (H’), evenness (E) and richness (S) of the sand fly species from the study areas in Gambella and Benishangul Gumuz regional States, Ethiopia, April-May 2025.

Refugee camp	H’	E	S
Kule	1.237	0.690	6
Sherkole	1.430	0.798	6
Terkidi	1.376	0.662	8
Tsore	1.200	0.670	6

PCR screening of sand flies for the detection of Leishmania DNA

PCR screening was performed on the 11 individually tested female Ph. rodhaini to detect Leishmania DNA. In contrast, female Sergentomyia specimens were pooled by species for PCR-based Leishmania screening as follows: 25 pools of S. africanus (n = 195), 24 pools of S. antennatus (n = 228), 13 pools of S. bedfordi (n = 47), 10 pools of S. clydei (n = 57), 16 pools of S. schwetzi (n = 105), and 4 pools of S. squamipleuris (n = 11), with each pool comprising from 1 to 22 sand flies. None of the pools tested positive for Leishmania, and all individually analyzed Ph. rodhaini specimens were also negative for Leishmania kDNA-based molecular screening (Supplementary Fig. 1).

Discussion

This study provides an entomological assessment of sand fly vectors in refugee camps in the Gambella and Benishangul-Gumuz Regional States of Ethiopia, regions not known to be endemic but at high risk for VL^2,27. The present findings reveal a diverse sand fly fauna dominated by species of the genus Sergentomyia, with clear ecological preferences for outdoor habitats, and importantly, no molecular evidence of Leishmania infection within the sampled sand fly populations.

A total of 2,196 adult phlebotomine sand flies (1,542 males and 654 females) from eight species in two genera (Phlebotomus and Sergentomyia) were collected using LT and ST collection methods. Phlebotomus was represented by one subgenus (Anaphlebotomus), while Sergentomyia included four subgenera: Grassomyia, Sergentomyia, Sintonius, and Parrotomyia. Most sand flies (85%) were collected from Gambella refugee camps (Kule: 357; Terkidi: 1,012), while 15% were from Benishangul Gumuz (Sherkole: 310; Tsore: 17). One Phlebotomus and seven Sergentomyia species were identified from four refugee camps in Gambella and Benishangul Gumuz Regional States. The current cross sectional entomological survey revealed that 98.3% of the collected sand flies belong to the genus Sergentomyia and only 1.7% to the genus Phlebotomus. The report indicated that Sergentomyia outnumbers other species, which is consistent with findings from endemic regions, including northern and southern Ethiopia^11,19,20.

The current study highlights the low abundance of Ph. rodhaini (1.73%), a finding that is consistent with previous reports from VL endemic regions of northern, southern, and eastern Ethiopia⁹. An entomological assessment conducted in the Gambella refugee camps and surrounding areas in 1996 by Hailu et al. (unpublished data); however, this species was not reported. The presence of this sand fly species in the present study may indicate a shift in vector distribution due to climate change or environmental variation since the 1990s. Furthermore, the primary VL vectors reported elsewhere in Ethiopia appear to be absent in these refugee camps. The current finding may reflect the limitation of the current cross sectional study design, as previous research documenting these vectors employed longitudinal and seasonal sampling methods^8,20,29,30. Ph. rodhaini is also distributed across parts of East Africa, including Sudan and South Sudan^31–34. To our knowledge, this entomological survey is the first to document its occurrence in refugee camps within the Gambella and Benishangul Gumuz Regional States. Although reported at low abundance, its presence is epidemiologically relevant, as Ph. rodhaini has historically been implicated as a potential vector for L. donovani in Sudan³⁴. A review by Al-Salam et al. on VL during the South Sudan conflict and its regional implications further highlighted the possible involvement of this species in zoonotic VL (ZVL) transmission in East Africa²¹. While its role in VL transmission has often been underestimated due to its low density^7,35 and preferential feeding on rodents rather than humans, its detection nevertheless signals a potential transmission risk, particularly in settings where primary vectors are rare or absent^31,34. Ecologically, Ph. rodhaini is primarily associated with sylvatic and forest-derived environments, favoring humid, shaded microhabitats with dense vegetation³⁴. It is commonly reported in forested habitats, rock crevices, tree roots, termite mounds, and natural soil cracks, which provide favorable conditions for resting and breeding^8,33,34.

In the present study, seven Sergentomyia species were identified, confirming the wide distribution of this genus across Ethiopia, including VL endemic regions where Phlebotomus species are recognized as the primary vectors of transmission^6,19,20,30. The striking finding was the overwhelming dominance of S. antennatus, which accounted for more than 50% of all collected specimens, especially in the Gambella camps of Kule and Terkidi. Although Sergentomyia species are generally considered zoophilic and of limited significance in VL transmission compared to primary vectors like Ph. orientalis and Ph. martini, their remarkable abundance needs to be further investigated. Large population size could increase human-vector contact rates, underscoring the need to investigate the potential role of specific Sergentomyia species in disease ecology.

The current study demonstrates a significant difference in sampling efficiency between trap types: ST captured significantly more sand flies than the LT. This variation may be partly attributable to the greater total sampling efforts (trap-nights) with STs. Furthermore, the habitats sampled by each trap type differed: LTs were deployed exclusively in indoor habitats and peridomestic (household compound) settings, whereas STs were also placed in outdoor habitats, including termite hills (both within villages and in natural habitats) and mixed forests.

Analysis of species diversity revealed a heterogeneous distribution across refugee camps. Sherkole camp exhibited the highest species diversity and evenness, likely driven by high sand fly densities in termite hills and a more equitable species distribution. Terkidi camp also showed high diversity, with all eight species present; however, the area was dominated by S. antennatus. In contrast, Tsore camp had exceptionally low abundance and diversity this could potentially reflecting suboptimal local ecological conditions for the sand fly spread.

A key entomological finding was exophilic (outdoor resting) behavior of sand flies in the study area. Less than 2.3% of all specimens were collected indoors, a pattern consistent with the reports from other VL endemic regions in Ethiopia, Sudan, and South Sudan^35–37. This has major implications for vector control. While indoor residual spraying (IRS) remains a cornerstone of VL control, our data indicate that effective strategies must also incorporate targeted outdoor interventions. Significantly, the high densities of S. antennatus and S. africanus associated with termite mounds, particularly those near natural habitats and within villages, identify these structures as key ecological niches. The current report is supported by a study conducted in Kenya that showed these species more commonly collected in termite hills³⁸. Environmental management, such as the modification or destruction of termite mounds in peridomestic areas, could be a highly effective strategy to reduce sand fly breeding and resting sites.

The negative result from the most sensitive kDNA PCR assay, applied to both Sergentomyia and individual Ph. rodhaini specimens, was a significant and valuable finding. These findings suggest several non-exclusive possibilities in the context of the refugee camps’ asymptomatic Leishmania infection distribution²². First, parasite circulation may have been at an extremely low, undetectable level during the sampling period, which may not have coincided with seasonal transmission peaks. Second, the zoonotic transmission cycle may involve animal reservoirs that are not effectively bitten by the sampled sand flies. Finally, known VL vectors might exist at such low densities that they were not captured in sufficient numbers to include an infected specimen. The absence of infection provides crucial baseline data, indicating a currently low transmission intensity within the local vector population, which is essential for monitoring future epidemiological trends. This finding contrasts with reports from other endemic regions where Sergentomyia and Phlebotomus spp. have tested positive for Leishmania DNA. For instance, in Iran, S. dentata (4/48, 8.33%) and S. sintoni (2/4, 50%) tested positive for Leishmania major¹⁸, while in Portugal, S. minuta was reported to have tested positive for Leishmania DNA (2/1867)¹⁷. This discrepancy may be partly explained by the absence of these specific Sergentomyia species in our study area. Furthermore, studies have challenged the paradigm of exclusive Phlebotomus transmission, implicating S. squamipleuris in Kenya as a potential vector¹⁵ and detecting L. donovani in Ph. rodhaini from Sudan³⁴. The absence of Leishmania DNA in Ph. rodhaini in the current study could therefore be explained by several factors, including ecological differences between sites, such as variations in host feeding preferences, local transmission intensity between sites, or differences in sand fly microbiota. Additionally, the predominance of asymptomatic human infection in our study areas²² may result in lower parasitemia, further reducing the probability of parasite detection in sand flies.

The strength of this study lies in its rigorous methodological design. The parallel use of LTs and STs across diverse habitats enabled a comprehensive assessment of sand fly ecology in previously under-surveilled refugee camps in Ethiopia. The application of highly sensitive PCR assays provides reliable data on Leishmania infection status within the vector populations. Furthermore, focusing on a vulnerable population and inclusion of ecological diversity indices generate crucial baseline data essential for informing targeted public health interventions against VL. On the other hand, the study also has several limitations that should be considered when interpreting these findings. First, while the study encompassed four refugee camps, its results may not be fully generalizable to all refugee camps in Ethiopia. Second, the cross sectional design provides only a temporal snapshot of sand fly ecology and infection, which may have missed seasonal variations in vector density and transmission peaks. Third, the specific laboratory analyses were not performed, limiting insights into host feeding preferences and potential reservoir dynamics. Additionally, the reliance on morphological identification alone, without complementary molecular confirmation, e.g., Cytochrome Oxidase I (COI) mitochondrial gene sequencing, may affect species-level certainty. To address these gaps and build upon this work, future studies should prioritize longitudinal sampling to capture seasonal dynamics. Laboratory analysis should include performing blood meal analysis to improve our understanding of host preferences and reservoir dynamics. Cytochrome Oxidase I (COI) mitochondrial gene sequencing should complement morphological species identification and resolve cryptic species complexes.

Conclusion

Our study showed a detailed picture of the sand fly fauna in southwestern and northwestern Ethiopian refugee camps for the first time. The vector community was characterized by high diversity, a strong predominance of outdoor-resting Sergentomyia species, and a notable absence of Leishmania DNA in the tested sand fly specimens at the time of the study. Integrated vector management has been implemented in these areas that includes environmental management, specifically targeting termite mounds in peridomestic areas. It is recommended to establish longitudinal entomological surveillance across different seasons to better capture temporal variations in sand fly density and Leishmania infection rates. Future studies should employ a larger sample size for Phlebotomus species and consider screening blood-fed females for host preferences to identify potential reservoir hosts involved in the transmission cycle. It is also desirable to investigate potential animal reservoirs in and around the camps to understand the complete transmission cycle.

Materials and methods

Study areas

The study was conducted in four refugee camps of Terkidi and Kule (Itang district, Gambella Regional State in southwestern Ethiopia) and Sherkole and Tsore (Homosha district, Benishangul-Gumuz Regional State in northwestern Ethiopia). These regions border South Sudan and Sudan. As of 31 May 2025, Gambella hosted 395,203 refugees, and Benishangul-Gumuz hosted 110,606 refugees²⁵. These camps were randomly selected among a total of seven and five refugee camps in Gambella and Benishangul Gumuz Regional States, respectively^39,40 (Fig. 1).

Fig. 1.

Map showing sand fly collection sites in refugee camps in Gambella and Benishangul Gumuz regions, Ethiopia [Kule Camp: represented by a white triangle; Terkidi Camp: represented by a red circle; Sherkole Camp: represented by a yellow circle; Tsore Camp: represented by a green diamond]. The sand fly collection sites are represented as orange dots where; A: Sherkole, B: Tsore, C: Kule and D: Terkidi refugee camp sand fly collection sites. The maps were built using the free and open source QGIS software version 3.36.3 (QGIS Development Team (2024). QGIS Geographic Information System, version 3.36.3. Open-Source Geospatial Foundation Project. https://qgis.org) and shapefiles were obtained from the free and open-source site https://data.humdata.org.

Itang is a special district that lies in the low-lying floodplain area of Gambella, characterized by a flat to gently undulating landscape. The district is bordered to the south and southeast by the Anuak Zone, to the west by the Nuer Zone, to the north by the Oromia region, and to the northwest by South Sudan, with the Alwero River demarcating part of its southern boundary. The altitude of this district ranges from 350 to 480 m above sea level⁴¹. The district’s average annual temperature and total rainfall are 29 °C and 1000 mm, respectively. Temperatures remain high year-round with peaks in March–May. Vegetation is dominated by savanna woodland, Acacia seyal, and extensive grasslands⁴². Moreover, there are scattered distributions of small-sized termite hills in the village and natural habitats of the refugee camps.

The Homosha district is found in the eastern parts of the Benishangul Gumuz Regional State, which is bordered by Kurmuk in the northwest, Menge in the northeast, and Assosa in the south. The study area receives 588-1,549 mm of rainfall annually, with maximum temperatures of 21.4–31.5 °C during the dry season and minimum temperatures of 7.4–17.6 °C. The soil is deep reddish-brown sandy clay, and the landscape is largely covered by natural vegetation, including shrub grassland, dense woodland, bamboo thickets, and open grassland⁴³. Termite mounds are common in the district, particularly in woodland and grassland habitats, but rarely distributed in the refugee camp.

Collection of sand flies

The collection of adult sand flies was conducted by employing two standard techniques: LTs and STs, as detailed elsewhere^6,10,30. Each method was deployed individually, without the presence of the other methods.

Before initiating the main sand fly collection, preliminary field assessments were conducted in order to assess the abundance, distribution, and diversity of sand flies in potential collection sites. The first assessment took place during the fourth week of May 2024 in refugee camps in Gambella Regional State. LTs were deployed indoors and in household compounds. We used 32 trap-nights (4 LT per night for 8 nights). Whereas, for ST, we deployed four traps for each of the five collection habitats (indoors, household compounds, termite mounds within villages, termite mounds in natural habitats, and natural habitats) per night. Sticky traps were deployed for eight nights each in all refugee camps, accumulating to a total number of trap nights of 160 (four ST in each of five habitats over eight nights). A second assessment was conducted in Benishangul-Gumuz Regional State in August 2024 using the same trapping procedures; however, no sand flies were captured during this assessment.

CDC light trap (LT)

Two fixed sampling points (indoor and household compounds) were established per refugee camp. A single LT was deployed at each point per night, positioned 30–50 cm above ground level and operating from dusk to dawn. A total of 5 LTs were deployed per night, amounting to a cumulative total of 40 trap-nights. This LT sampling was performed over two consecutive nights across all collection sites in the refugee camps of the Gambella (April 6–10, 2025) and Benishangul Gumuz (May 5–9, 2025) Regional States. Specimens were traced and aspirated each morning from the collection LT, inactivated using a combination of cold and mechanical vibration, and subsequently preserved in 95% ethyl alcohol for future morphological and molecular analysis (Fig. 2).

Fig. 2.

Field sampling methods and habitat characterization for sand fly collection in selected refugee camps in Gambella and Benishangul Gumuz Regional States, 2025: (A & B: Landscape views of savanna woodland and shrubland vegetation in Kule and Terkidi refugee camps respectively; C & D: Natural habitat presentation with mixed vegetation plots mainly bamboo tree in Sherkole and Tsore refugee camps respectively; E: Sand fly collection in Kule camp natural habitats; F: Depicted that sand fly collection in the periphery of the compound in Terkidi refugee camps and G: Indicating sand fly collection in Kule refugee camp compound using LT).

Sticky traps (STs)

White transparent polypropylene sheets (A4 size) coated with a thin layer of castor oil were used as STs for sand fly collection. In each survey site, four traps were deployed in five habitat types, including indoors, household compounds, termite mounds within villages, termite mounds in natural habitats, and natural habitats with mixed forest cover. A total of 48 STs were deployed per night, amounting to a cumulative total of 384 trap-nights. Traps were operated from dusk to dawn. Captured sand flies were collected the following morning using fine-tipped forceps, transferred into labeled vials containing 95% ethyl alcohol, and preserved for subsequent morphological identification (Fig. 2).

Mounting and identification of sand flies

All collected sand flies were sex segregated based on the morphology of their reproductive organs observed under a stereomicroscope. Female sand flies were dissected by carefully removing the head and the last three abdominal segments using fine dissecting pins. Whereas, the male sand fly’s genitalia were detached from the rest of its body parts. These parts were mounted on properly labeled glass slides in a drop of gum chloral mounting medium with the head ventral-side up under a single coverslip. Prepared slides were allowed to clear and dry at room temperature for one to two weeks before examination under a compound optical microscope for species identification. The remaining body parts (thorax and anterior abdominal segments) were preserved in sterile 1.5 mL Eppendorf tubes containing 95% ethyl alcohol, labeled identically to the corresponding slide, and stored for subsequent DNA extraction (Fig. 3). Species identification was carried out using well established taxonomic keys⁴⁴ and additionally⁴⁵ with diagnostic morphological features, including the cibarium, cibarial teeth, pharynx, and spermatheca.

Fig. 3.

A workflow diagram depicting the process from field collection of sand flies to molecular screening for Leishmania DNA.

DNA extraction and molecular screening of Leishmania DNA

We individually screened the only Phlebotomus species encountered during our study, Ph. rodhaini, while the Sergentomyia genus were pooled in sets of 1–22 female sand fly specimens based on species, trap type, collection site and habitat. The individually prepared and pooled female specimens were soaked in DNA shield solution with beads and homogenized twice at high speed (6.5 m/s for 3 min) using a FastPrep-24 bead beating homogenizer (MP Biomedicals), followed by centrifugation at 21,000 g for 1 min. From the clarified homogenate, 180 µl was transferred into a 1.5 ml tube for DNA extraction. DNA was isolated using the DNeasy Blood & Tissue kit (QIAGEN) following the standard manufacturer’s protocol for isolation of insect genomic DNA (Fig. 3).

A real-time PCR (rtPCR) assay targeting the minicircle kinetoplast DNA (kDNA) was used to screen Leishmania DNA from sand fly specimens. The primers employed were kDNA-CMF (5′-CTTTTCTGGTCCTCCGGGTAGG-3′) and kDNA-CMR (5′-CACCCGG CCCTAT TTTACA CC AA-3′). The assay was performed as previously described²². In brief, each reaction was set up in a final volume of 25 µl containing 1× HotStarTaq Master Mix (Qiagen, Venlo, The Netherlands), 0.6 µM of each primer, 0.4 µM of the probe (Integrated DNA Technologies, Leuven, Belgium), 0.1 mg/ml bovine serum albumin (Roche, Vilvoorde, Belgium), and 5 µl of DNA template. Amplification was carried out on a QuantStudio 5 real-time PCR system (Thermo Fisher, Cat. No. A28568). Each run included genomic DNA from a Leishmania-positive culture as the positive control, a non-template control (NTC) as the negative control, and one used extraction control (ExC).

Data analysis

Data were entered using Microsoft Excel 2016, simply calculating the relative proportion from the overall sand fly collected⁶ and exported to SPSS software version 20.0 (SPSS Inc., Chicago, IL, USA) for further analysis. Sand fly density was calculated as the number of sand flies collected per trap per night using LT and ST to compare differences between habitats. The normality of the data was assessed using the Shapiro–Wilk test. When the data did not meet the assumption of normality, the non-parametric Kruskal–Wallis test was employed. This test was used to compare the mean number of sand fly species collected across different sampling habitats using LTs and STs. The chi-square (χ²) test was applied to evaluate associations between categorical variables. We calculated the Shannon-Wiener (H) diversity index in PAST v4.03 for each study site using the parameters of the proportion in which each species was collected. Shannon-Wiener index: H=−∑si = 1(Pi)ln(Pi), where S is the number of species and Pi is the proportion of the total samples belonging to the i^th species. H expresses the differences in the diversity of the sand fly fauna between refugee camps. Evenness: E = H/ln(S), where H is the value of Shannon-Wiener, and E expresses how evenly the individuals in the refugee camps are distributed over the different species. Species richness: S = the number of species in each refugee camp. The permutation test was employed to assess the statistical significance of differences in Shannon diversity indexes between study sites⁴⁶. Statistical significance was determined at p < 0.05.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CDC LT

Centers for Diseases control and prevention light trap

CPD

Compounds that contains households

DNA

Deoxyribonucleic acid

MoH

Ministry of health of Ethiopia

NH

Natural habitat

PCR

Polymerase chain reaction

ST

Sticky trap

TH

Termite hill

VL

Visceral Leishmaniasis

Author contributions

HB, AA, GT, BE and EA conceived and designed the study. All authors were involved in proposal writing and participated in field coordination, data collection, supervision and overall implementation of the study. HB analyzed the data and drafted the manuscript. All authors read and approved the final manuscript.

Funding

This study was financially supported by grants from Addis Ababa University.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare that they have no competing interests.

Ethics approval and consent to participate

Consent was sought from each refugee camp coordinators including protection, security and health coordinators) and informed consent was also sought from head of households from where sand flies were collected in their compound and/or indoor. The study obtained ethical clearance from the Aklilu Lemma Institute of Health Research-Institutional Research Ethics Review Committee (ALIHR-IRERC) prior to data collection (Ref. No.: ALIPB IRERC/112/2015/23). A permission letter was obtained from Federal Democratic Republic of Ethiopia Refugees and Returnees Service (FDRE RRS) and subsequently from RRS regional office and camp coordination offices.

Consent for publication

Not applicable.