Knowledge about diurnal resting sites of sand flies is scanty and often anecdotal. In this study, we explored a part natural – part agricultural oasis in Neot Hakikar, Israel, looking for sand fly resting sites. To achieve this, we developed a new type of emergence trap. Sixteen types of microhabitats were examined and in seven of these, we also investigated the rodent burrows. We found that Phlebotomus papatasi showed clear preferences for resting sites characterized by vegetation cover, type of vegetation, and the presence of a mulch layer. In habitats with bare soil and little shade, few or no resting sand flies were found outside rodent burrows. Apart from the trunks of date trees, most resting P. papatasi were found in disturbed habitats, especially in large piles of organic waste and in a plowed field. Though catches from rodent burrow exits were always higher than from the nearby ground, it is safe to assume that the few burrows in this vast oasis do not play an important role for breeding and resting of P. papatasi. It also appears that disturbing the natural environment further increases the already considerable sand fly population.
The sand fly Phlebotomus papatasi (Scopoli), 1786 is widely distributed throughout the Mediterranean, the Levant, Turkey, the Arab peninsula, and stretches east as far as the Indian subcontinent and Western China (Lewis and Buttiker 1980, 1982, Lewis 1982, Lane 1986). In eastern Africa, it reaches as far south as Sudan and Ethiopia (Abonnenc 1972). This species is often abundant, largely domestic, and is an aggressive man biter (Lewis and Ward 1987). Apart from being a major nuisance species, P. papatasi is also the vector of Leishmania major, a zoonotic pathogen causing cutaneous leishmaniasis (Killick-Kendrick 1999). In Israel, this disease is highly endemic in the Jordan Valley, in the Dead Sea region and parts of the Negev (Schlein et al. 1984, Greenblatt et al. 1985, Wasserberg et al. 2003). The animal reservoirs of L. major are rodents; namely, the fat sand rat (Psammomys obesus Cretzschmar) and the Sundevall's Jird (Meriones crassus Sundevall). So far, Wagner's Gerbil (Gerbillus dasyurus Wagner) was the only other species with signs of an infection (Schlein et al. 1984, Wasserberg et al. 2002). During the last few decades, cutaneous leishmaniasis is re-emerging in the Middle East and researchers suspect this is connected to human activities that cause habitat disturbances. Oumeish (1999) and Wasserberg and colleagues (2003a) showed that in the last decade, the significant increase of leishmaniasis cases in the Negev is connected to the influx of susceptible human populations into zoonotic areas, and the resulting anthropogenic environmental changes, favorable to both the reservoir and the vector (Wasserberg et al. 2003b).
In some recent studies in Neot Hakikar, the sugar feeding behavior and infection rates of P. papatasi with Leishmania parasites were studied (Müller and Schlein 2004, Schlein and Jacobson 2002). In the western part of the date plantation, high population densities of P. papatasi were observed over several years. Catches with CDC-UV traps, without any kind of lures, yielded catches of up to several hundred sand flies per trap. About 20 years ago at this site, sand rats were common, but in the last 15 years, no sand rats were observed in this part of the oasis. The few gerbils living at this site should not be able to support huge sand fly populations (unpublished data of the authors). The goal of our study was to examine this atypical sand fly population, and to explore their resting sites in a largely natural oasis site.
MATERIALS AND METHODS
The experimental area is located in the Arabian Zone, a true desert, which centers on the Arabian Peninsula. Annual winter rainfall, from a few millimeters up to a maximum of 200 mm, is followed by a short period of blooming, and afterwards, the vegetation dries up rapidly (Orni and Efrat 1980). The vegetation, mainly thorny trees, bushes and shrubs, is very sparse and over large areas, averages one plant per one to ten square meters.
The study was conducted during two consecutive days, in August 2009, in Neot Hakikar, southern Israel, which lies in the Dead Sea Depression (390 m below sea level). Here, a date plantation was chosen as the study area, and in the eastern, more agricultural part of this oasis, there is a small settlement with gardens, vast fields and greenhouses. Weather conditions in the oasis remained sunny and dry over the study period.
The western part of the oasis is a nature reserve, with a mixture of salt marshes, wet and dry Salinas, and fresh water springs, surrounded by riparian vegetation. The plantation is surrounded by groves and thickets of trees and bushes like Tamarix nilotica Ehrenb. Bge. and T. passerinoides Del. Ex Desv. (Tamaricaceae), Prosopis farcta Macbride (Mimosaceae), Nitraria retusa Forssk. Asch. (Nitrariaceae) and chenopod bushes like Atriplex halimus L, A. leucoclada Boiss., Suaeda asphaltica Boiss., and S. fruticosa Forsk. (Chenopodiaceae). The area is known for its rich mosquito fauna (Margalit et al. 1973), but the only Phlebotomus species is Phlebotomus papatasi (Müller and Schlein 2004).
Description of the defined microhabitats
In the western, more natural part of the oasis, a total of 16 habitats were chosen for the study (Figure 1). In five of these, gerbil burrows were also investigated. In the center of the oasis, close to the springs, three types of habitats were defined:
Center 1 was below poplar trees with little or no undergrowth, but the ground was covered loosely in a 10 cm thick layer of a mixture of poplar leaves and other decaying organic matter. Here in the damp soil, a few gerbil burrows were found.
Center 2 was in the midst of a 2 to 3 m high reed thicket with a 0.5 to 1 m high layer of collapsed and dry reed on the slightly moist ground.
Center 3 was a heap of cut reeds, uprooted sedges, shrubs and top soil dumped about two years ago in the middle of a reed field.
In the transition zone, between the natural vegetation with springs and the date plantation, another three habitats were defined:
Transition 4 consisted of patchy, not- too-dense and not-too-high vegetation, on dry soil, dominated by reeds and sedges, with a layer of mulch up to 5 cm high.
Transition 5 was a tamarisk thicket without any undergrowth, but the dry ground was covered with a 5 to 10 cm high layer of tamarisk leaves.
Transition 6 was a dense thicket of Atriplex and Suaeda bushes with no undergrowth, but the dry ground was covered with a 2 to 5 cm high layer of partially decaying leaves. Numerous gerbil burrows were found here.
A further four habitats were characterized in the date plantation surrounding the springs:
Plantation 7 consisted of areas with dusty, bare ground, from which the undergrowth is removed by mechanical means on an annual basis.
Plantation 8 consisted of patches of dry, dusty ground with a dense cover of P. farcta, and a layer of mulch up to 5 cm high. Several gerbil burrows were found here.
Plantation 9 consisted of a heap of cleared P. farcta shrubs, date palm leaves, and root-stocks, dumped several years ago in the middle of the date plantation at a level of 1 to 3 m high and partially overgrown by P. farcta.
Palm Trunks 10 includes the lower part of palm trunks, from the ground up to 1 m, surrounded by rough 10 to 20 cm long stumps of the cut palm leaves. This formed deep cavities filled with mulch and rotting organic matter.
In the outer periphery of the date plantation, towards the desert, two habitats were defined:
Periphery 11 was a thicket of Suaeda bushes without any undergrowth. The dry ground was covered with a 5 to 10 cm high layer of leaves. Most of the gerbil burrows were found here.
Periphery 12 was between scattered Atriplex and Tamarix bushes growing on dry, coarse desert alluvium.
In the semi-desert surrounding the oasis, two habitats were characterized:
Desert 13 was between groups of Suaeda and Tamarix bushes on dry, dusty silt.
Desert 14 was between scattered Nitraria bushes, growing on bare, coarse, desert alluvium. A few gerbil burrows were also found here.
In the eastern, more agricultural part of the oasis, only two habitats were selected with sand rat burrows in both of them:
Natural 15 was in the midst of a Suaeda and P. farcta grove on dry silt. Several sand rat burrows were found here.
Plowed 16 was in direct proximity to Natural 15 and was originally covered by the same type of vegetation. However, about two to three years ago, it was cleared and plowed in preparation for agricultural fields. Here, along intersecting dirt roads, numerous sand rat burrows were found. The plowed silt soil was about 40 cm deep and contained plenty of cut roots and other coarse plant material.
Collecting resting sand flies
Sand flies were caught above the ground with six emergence traps in each of the defined micro-habitats for two consecutive days. Traps were operated for 24 h on both days; they were set at 16:00 and emptied the next day at 16:00, at which time the traps were set again. These emergence traps consisted of a cubic wooden frame 80 × 80 × 80 cm. The bottom of this cube was open, and the four sides were wrapped with a one meter wide, durable, off-white cotton cloth. The roof was covered with a white, rigid plastic net that has 0.3 cm square holes and 0.1 cm wide netting. The net was secured to the frame every 10 cm with thin wire. The cloth, 20 cm longer than the sides of the frame, formed a flap used to seal the bottom of the cage to the ground, using iron bars to weigh it down. (Figure 2).
To evaluate resting in the trunks of palm trees, two rectangular frames, one on ground level, the other 1 m above the ground, were constructed in a way that the supported netting was kept at least 10 cm from the trunk. The upper opening of the cage was closed with cloth, while the lower end sat directly on the ground. The ground / trunk facing side of the nettings were painted with castor oil with a large paint brush (Figure 2).
At seven of the defined habitats, rodent burrows were sampled with exit traps for resting and hatching sand flies. In the western part of the oasis, in five habitats (center 1, transition 6, plantation 8, periphery 11, desert 14), burrows of gerbils and jirds were explored, while in the eastern part, in two habitats (natural 15 and plowed 16), sand flies were collected from burrow exits of sand rats. In each habitat category, six burrows were sampled for two consecutive days with exit traps made from 1.5 liter plastic bottles. The bottles were cut on the upper third, the bottom was removed and covered with fine gauze, the upper part was turned round and inserted into the lower part to form a fish-basket-like conical trap. The internal side of the bottom part was painted with castor oil (Figure 2).
Sand flies were picked off traps with forceps from the ground-trunk facing sides of the netting, as well as from the interior of the oil coated bottles, on a daily basis and were stored in vials with 70% ethanol. At the same time, the oil was freshly repainted to remove dirt and non target insects.
The catches of the first day were a combination of resting flies and flies that hatched overnight. The catches from the following day were only hatching sand flies. The average daily number of resting P. papatasi per habitat sampled was calculated by using the following equation: [(average catch day 1) – (average catch day 2)] and results are presented in Tables 1 and 2. The data obtained with soil emergence traps are presented as number of sand flies/m2/day; the data from tree trunks are presented as number of sand flies/trunk/day; and data obtained with exit traps are presented as number of sand flies exiting single rodent burrows/day.
Table 1. Mean number of P. papatasi (± SD), caught by emergence traps, in 16 habitats on the first and second day. Daily resting rates per square meter are given.
Females and (Males)
Average first day
Average second day
Daily resting/ m2
Western part of the oasis
1.50 ± 0.56
(1.83 ± 0.54)
4.50 ± 1.34
(3.50 ± 0.99)
9.67 ± 1.75
4.83 ± 1.14
(7.33 ± 1.73)
(3.17 ± 0.48)
2.33 ± 1.05
(1.83 ± 0.98)
3.00 ± 1.03
0.50 ± 0.22
(3.50 ± 1.34)
(0.33 ± 0.21)
(0.17 ± 0.17)
1.17 ± 0.47
(0.83 ± 0.31)
0.17 ± 0.17
9.83 ± 3.90
2.67 ± 0.67
(8.83 ± 2.57)
(2.17 ± 0.48)
Palm trunks 10
11.83 ± 5.24
1.83 ± 0.65
0.67 ± 0.33
(1.67 ± 0.56)
(0.83 ± 0.31)
0.17 ± 0.17
Eastern part of the oasis
0.83 ± 0.167
0.33 ± 0.21
(0.67 ± 0.42)
(0.17 ± 0.17)
4.67 ± 1.09
1.50 ± 0.43
(4.50 ± 1.29)
(1.00 ± 0.37)
Table 2. Mean number of P. papatasi (± SD) caught emerging from rodent burrows in seven selected habitats within the first day, the second day. Calculated daily resting rates are given.
Females and (Males)
Average first day
Average second day
Gerbil and jird burrows in the western part of the oasis
1.50 ± 0.67
(0.67 ± 0.33)
6.50 ± 2.38
1.17 ± 0.48
(5.50 ± 1.48)
(0.83 ± 0.40)
3.00 ± 1.21
1.00 ± 0.52
(2.83 ± 0.70)
(0.67 ± 0.21)
6.00 ± 2.60
2.50 ± 1.06
(4.67 ± 2.22)
(2.00 ± 0.58)
1.00 ± 0.45
0.50 ± 0.22
(0.67 ± 0.33)
(0.33 ± 0.21)
Fat sand rat burrows in the eastern part of the oasis
5.33 ± 1.31
1.83 ± 0.60
(4.17 ± 1.11)
(1.50 ± 0.50)
23.33 ± 4.39
5.33 ± 1.50
(27.67 ± 7.50)
(4.00 ± 1.21)
Altogether, 1,249 sand flies (659 females and 590 males) were collected. The 96 emergence traps on the ground and surrounding the date trunks caught 241 females and 211 males on the first day. On the following day, 64 females and 46 males were caught. At the same time, from all of the rodent burrows, 280 females and 277 males were caught with exit traps. On the second day, the traps caught 74 females and 56 males. On the first day, about four times as many flies were collected as on the following day, indicating that the bulk of the trapped sand flies were resting in the investigated habitats.
Close to the overgrown springs there were two major resting sites: one in the heap of uprooted and rotting weeds (center 3), with an average of 7.55 females/m2/day and 6.49 males/m2/day, and the other in an overgrown reed thicket (center 2) with an average of 7.02 and 5.46 females/m2/day and males/m2/day, respectively (Table 1). At the third site, near the water in the leaf litter below a group of poplar trees (center 1), only about one-third of the flies were resting (2.34 females /m2/day and 2.86 males/m2/day).
In the slightly drier transition zone (transition 6), between the spring and the date plantation, most flies were resting in the mulch layer of Atriplex and Suaeda brush (3.90 females/m2/day and 4.95 males/m2/day), and in the dry vegetation layer of a low but dense thicket (transition 4) of reeds and sedges (3.64 females/m2/day and 2.86 males/m2/day). In the leaf litter of the nearby tamarisk grove (transition 5), no flies were caught.
The highest numbers of resting sand flies were trapped from the trunks of palm trees (11.83 females/trunk/day and 8.67 males/trunk/day) while at the same time, on the cleared ground between the palms (plantation 7), only a single male was caught in an area of 4.8m2. In the parts of the plantation that were overgrown with P. farcta shrubs (plantation 8), some resting was observed (1.56 females/trunk/day and 1.30 males/trunk/day), but the resting activity clearly concentrated on a large heap of rotting palm leaves (plantation 9) with 11.17 females/trunk/day and 10.39 males/trunk/day.
Towards the drier periphery of the oasis, between scattered Atriplex and Tamarix bushes, on bare ground (periphery 12), apart from a single female, no flies were collected, while in the leaf litter of a nearby Suaeda grove (periphery 11), an average of 1.81 females/day and 1.31 males/day were caught. At both habitats outside the oasis, in the semi desert, between groups of Suaeda and Tamarix bushes on bare ground (desert 13), and also between groups of Nitraria retusa bushes, on bare soil (desert 14), no resting sand flies were recorded. About 4 km to the east, near Neot Hakikar village, in natural Suaeda and Prosopis brush land (natural 15), few resting flies were seen (0.78 females and 0.78 males), while nearby, on a cleared and plowed plot (plowed 16), almost seven times as many flies were observed (4.95 females/day and 5.46 males/day).
In the western part of the oasis (center 1, transition 6, plantation 8, and periphery 11), only a few gerbil and jird burrows were found, while at the eastern part of the oasis most burrows were from sand rats. There was little variation in the numbers of resting sand flies in rodent burrows on the western side in most habitats (Table 1). Only in the desert surrounding the oasis (desert 14), were numbers low (0.50 females/burrow/day and 0.34 males/burrow/day). In the eastern part of the oasis, only sand rat burrows were investigated. Here, in a natural Suaeda and Prosopis grove (natural 15), a similar numbers of resting sand flies were caught from the burrows as in most habits in the western part of the oasis (3.50 females/burrow/day and 2.67 males/burrow/day). Right next to this grove, a plot of land was cleared of vegetation and plowed about 2 years ago (plowed 16). Almost seven times as many resting sand flies were collected from burrows found here (18.00 females/burrow/day and 23.67 males/burrow/day).
Knowledge about diurnal resting sites of sand flies is scanty and often anecdotal (Alexander and Maroli 2003, Alexander 2000). Only a few researchers have tried to identify resting sites per se, and the preferential tools were often different types of aspirators or sheets of sticky paper inserted in cracks and crevices (Feliciangeli 2004). It appears that it is difficult to aspirate sand flies when they are attached to a substrate, and only flying specimens are easily caught. Sticky papers, on the other hand, often yielded small or no catches (Hogsette et al. 2008). The six emergence traps used for each of the 16 habitats in this study covered together over 60 m2.
Several types of emergence traps are described in the literature, but they vary greatly in features such as size, shape, material and trapping technique (Casanova 2001, Bettini et al. 1986, Rutledge and Ellenwood 1975, Vieira et al. 1999). Most of the traps are adapted to very specific questions and tasks, and cannot necessarily be used for multiple types of projects, especially if these take place in different types of habitats. The new design of the emergence trap, described in the current study, proved to be efficient in catching resting sand flies. Moreover, the traps were easy to build and maintain, inexpensive, and allowed us to cover large areas of suspected resting / breeding sites in a short time. In previous attempts to identify resting and breeding habitats in Neot Hakikar oasis, we used different types of sticky papers and bulky emergence traps built from small collapsible tents. Not only did we have little success, but these approaches were also very time consuming (unpublished data).
For the present study, we selected 16 different microhabitat types in Neot Hakikar oasis to better understand the factors influencing resting preferences of P. papatasi, and the influence of human interference with the natural habitats. Indeed, P. papatasi collected by emergence traps showed clear preferences for resting sites (Tables 1 and 2), but without any significant differences between females and males. In habitats with bare soil and little shade (plantation 7, periphery 12, desert 13 and desert 14), no or very little resting sand flies were found outside rodent burrows. It appeared that for resting on the ground, vegetation cover, type of vegetation, and the presence of a mulch layer were the decisive factors.
Sand flies resting on the ground preferred habitats with a thick, dense layer of vegetation as found in the partially collapsed reed thicket close to the springs (center 2) and in the coadunate patches of reeds and sedges (transition 4). Other preferred natural resting habitats were in the mulch layers and leaf litter below thickets of bushes and trees (center 1, transition 6, plantation 8, periphery 11 and natural 15). The only exceptions were tamarisk groves, which had a fairly pronounced mulch layer, but sand flies were neither resting nor breeding here (in part unpublished data of the authors). The reason could be that tamarisk trees are depositing salty exudates and salt-rich leaf litter in their vicinity and this significantly changes the soil chemistry (Ladenburger et al. 2006, Lesica and DeLuca 2004).
The highest numbers of resting sand flies recorded in this study (11.83 females/day and 8.67 males/day) were from the lower meter of the trunks of palm trees. This can probably be explained with the loose structure of the stumps of the cut palm leaves surrounding the trunk. Between some of the leaf stumps, there were up to 20 cm deep cavities filled with mulch and rotting organic matter, providing some moisture and relatively low temperatures ranging from 25 to 29° C, while temperatures on the soil surface could exceed 50° C. Surprisingly, one meter above the ground, little or no resting sand flies were observed (data not shown), despite the fact the aforementioned stumps were quite similar. During several projects in South America (Ferro et al. 1997, Hanson 1961, Rutledge and Mosser 1972), and in Africa (Mutinga et al. 1989, Mutinga and Odhiambo 1986), sand flies were collected from tree-holes, trunks, and buttressed roots. To the best of our knowledge, this is the first time that sand flies in general, and P. papatasi in particular, were found to be associated with tree trunks in great numbers, in the arid habitats of the Middle East (Killick-Kendrick 1999).
Apart from the date trunks, most resting P. papatasi were found in disturbed habitats, especially in the two large piles of organic waste (center 3 and plantation 9). The first pile covered almost 450 m2, while the second was a little more than twice as big. On a daily basis, potentially close to 30,000 sand flies were resting here. In South America, sand flies were often found resting and even breeding in leaf litter and rotting organic matter on forest floors (Hanson 1961, Rutledge and Mosser 1972, Rutledge and Ellenwood 1975a, 1975b), but in arid places like Israel, P. papatasi are mainly found resting, or suspected to rest, apart from rodent burrows in animal sheds, caves, in cracks in the ground and in piles of stones (Lewis and Buttiker 1982, Yuval 1991, Sawalha et al. 2003, Chelbi et al. 2007, Hogsette et al. 2008). In the Jordan Valley, several studies found P. papatasi in large numbers on fallow fields (Schlein et al. 1989, Yuval et al. 1988, Schlein and Yuval 1987), and after using white silk parachutes on plowed fields as emergence traps, it was found that these are not only dispersal sites but are also important for breeding and resting (Schlein and Yuval unpublished data).
In the eastern part of the oasis, relatively large numbers of resting sand flies (Table 1) were caught in a cleared and plowed field (plowed 15), a remarkable observation considering its sheer size of about 20 hectares and its proximity to the human settlement. Potentially, up to 2,000,000 sand flies might have rested there on a daily basis and as the continuous catches of the second day as well as the sex ratio of females and males (0.91) suggests (Feliciangeli 2004) it was probably also a major breeding site.
In some preliminary experiments in the natural habitats of Neot Hakikar, sand flies continued hatching after the mulch layer and upper layers of soil were carefully removed (unpublished data of the authors). This indicates that breeding occurred in the deeper more compact layers where resting for adults was probably not possible. The upper layers of undisturbed soil were either friable and almost dust like, or too compact to form cracks and crevices. Only the plowed field, with its bulked soil and multiple cut roots and branches, probably offered sufficient possibilities for sand flies to reach deeper, suitable layers with sufficient humidity and tolerable temperatures for resting and breeding.
From the same microhabitat, catches of resting P. papatasi from single exits of rodent burrows were generally higher than catches from one square meter of soil (Table 1 and 2). In the western part of the oasis, the differences between catches of resting sand flies in burrows were relatively small (center 1, transition 6, plantation 8, periphery 11) apart from the desert site (desert 14) with its low catches (0.50 females/day and 0.34 males/day). Nevertheless, it is safe to assume that the few burrows in this vast oasis do not play an important role for breeding and resting P. papatasi.
In general, fewer sand flies were found resting and breeding in burrows in this study than at other locations outside oases like the Jordan Valley, the northern part of the Dead Sea and parts of the Negev (unpublished data of the authors). These findings can possibly be attributed to the fact that in these more arid areas, suitable resting and breeding areas are a limiting factor, restricting resting and breeding activities to the favorable microclimate of burrows, whereas in Neot Hakikar the entire population seems to be dispersed over large areas of suitable microhabitats. It also appears that disturbing the natural environment is further increasing already existing sand fly populations, which is a situation also observed by Wasserberg et al. (2002, 2003a, 2003b) in several studies in the Negev.
This study was funded in part by the United States Department of the Army, Space and Missile Defense Command, Deployed War Fighters Protection Project (DWFP) grant # W81R6091285002 to explore applications of Attractive Toxic Sugar baits.