Control of Phlebotomine Sand Flies With Vertical Fine

VECTOR CONTROL, PEST MANAGEMENT, RESISTANCE, REPELLENTS

Control of Phlebotomine Sand Flies With Vertical Fine-Mesh Nets ˜ O, R. FAIMAN, R. CUN

AND

A. WARBURG1

Department of Parasitology, The Kuvin Centre for the Study of Infectious and Tropical Diseases, The Institute for Medical Research Israel Canada, The Hebrew University, Hadassah Medical School, The Hebrew University of Jerusalem, Jerusalem 91120, Israel

J. Med. Entomol. 46(4): 820Ð831 (2009)

ABSTRACT Insecticide-treated vertical net barriers were used to intercept foraging sand ßies. Two different nets were draped on fenced enclosures (10 by 10 m; 2 m high) in the central Jordan Valley. One enclosure was draped with a deltamethrin-impregnated net (PermaNet, 225 holes/in2). The holes of this net are sufÞciently large to allow sand ßies to pass through but not without coming in close contact with the mesh. The other enclosure was covered with SpiderNet⫹ (1,240 holes/in2) and sprayed with beta-cyßuthrin. Sand ßies were captured inside and outside the enclosures before and after draping with the nets using CO2-baited CDC traps or CDC light traps. Both barrier types exhibited ⬎90% efÞcacy in blocking sand ßies from entering the enclosures (P ⬍ 0.01). The SpiderNet⫹ exhibited high efÞciency even before being sprayed with insecticide because the small mesh size physically prevented ßies from passing through. In MaÕale Adumim, a 60-m-long, 2-m-high PermaNet barrier was erected to intercept sand ßies approaching houses from their natural habitats. Sand ßies were monitored on all sides of the barrier using CO2-baited CDC traps or CDC light traps. Results showed a 60% reduction in the mean number of sand ßies trapped behind the net compared with the untreated areas adjacent to it (P ⬍ 0.05). Integrated vector control campaigns for reducing the burden of sand ßy bites should consider vertical Þne-mesh nets to reduce the numbers of sand ßies arriving at inhabited areas. KEY WORDS Phlebotomine sand ßies, insecticide-impregnated nets, vertical barrier, leishmaniasis

The leishmaniases are vector-borne diseases caused by Leishmania parasites transmitted by phlebotomine sand ßies (Diptera: Psychodidae) in tropical, subtropical, and temperate regions of some 88 countries (Desjeux 2004). Two Leishmania (Kinetoplastida: Trypanosomatidae) species causing cutaneous leishmaniasis (CL) in humans are endemic to Israel and the Palestinian Authority: Leishmania major, Yakimoff and Schokhor, 1914, and L. tropica, Wright, 1903. Until recently, these parasites were thought to be restricted to geographically deÞned areas, with L. major occurring primarily in the Jordan Valley, Negev, and Arava deserts causing almost all of the CL cases (Greenblatt et al. 1985). However, recent studies have suggested that the areas affected by both parasite species are expanding to the north of Israel and into the Palestinian West Bank (Baneth et al. 1998, Abdeen et al. 2002, Jacobson et al. 2003, Jaffe et al. 2004, Svobodova et al. 2006). Recently, certain neighborhoods of the town of MaÕale Adumim, located in the Judean Desert only 7 km from Jerusalem, have become hyperendemic for CL caused by L. tropica. In 2004, ⬎300 CL cases were reported to the Israeli Ministry of Health (MOH). The parasite is transmitted by Phlebotomus (Paraphleboto1

Corresponding author, e-mail: [email protected].

mus) sergenti Parrot, 1917, and the probable reservoir hosts are rock hyraxes (Procavia capensis Pallas,1766) (Schnur et al. 2004, Svobodova et al. 2006). Emergency measures implemented by the municipal department of sanitation included the destruction of rock-hyrax breeding habitats near homes and spraying of insecticide around the affected neighborhoods. Although the following year, a reduction in CL cases was apparent, the hyrax populations have evidently recuperated (some relocating closer to residential areas), and human CL cases are on the rise again (R. F., unpublished observation). The breeding sites of most sand ßy species are unknown. Based on observations of adult resting sites in nature and the behavior of laboratory-reared sand ßies, investigators generally assume that eggs are laid individually in small batches in protected moist microhabitats such as rock crevices, animal burrows, termite mounds, cracks in the soil, domestic animal shelters, holes in trees, and organic debris on the forest ßoor (Bettini 1989; Killick-Kendrick 1990, 1999; Alexander 2000; Feliciangeli 2004). Because sand ßies do not require water for larval development, they may be found in large numbers even in arid regions and deserts. Indeed, phlebotomine sand ßies are among the most prevalent hematophagous insects affecting humans in deserts and semiarid regions of North Af-

0022-2585/09/0820Ð0831$04.00/0 䉷 2009 Entomological Society of America

July 2009

FAIMAN ET AL.: CONTROL OF PHLEBOTOMINE SAND FLIES

rica, the Middle East, and parts of Asia. A recent report documents the unusually heavy burden of sand ßy bites endured by U.S. and coalition soldiers stationed in Iraq (Coleman et al. 2007, Dalton 2008). Most control efforts against sand ßies are aimed at interrupting contact between female sand ßies and humans. Depending on the application techniques, timing, and target species, sand ßies are known to be highly susceptible to insecticides (Alexander et al. 1995a, Alexander and Maroli 2003, Wilamowski and Pener 2003, Orshan et al. 2006). Residual formulations of DDT and pyrethroids have been used to control sand ßies both in the old world and the Neotropics (Hertig and Fisher 1945, Hertig and Fairchild 1948, Hertig 1949, Le Pont et al. 1989, Marcondes and Nascimento 1993). Residual insecticide house-spraying has been used with signiÞcant effectiveness against endophilic species, mostly in Latin America (Alexander et al. 1995a, Vieira and Coelho 1998). Other studies have tested the efÞcacy of insecticide-impregnated textiles such as curtains, bed nets, or bed covers with varying success (Basimike and Mutinga 1995, Elnaiem et al. 1999, Kroeger et al. 2002, Courtenay et al. 2007). Environmental modiÞcation involving the total eradication of rodents, destruction of burrow systems, and spraying herbicides to kill their food plants, has been demonstrably effective in controlling CL caused by L. major in foci in the Asian republics of the former USSR and in Tunisia (Vioukov 1987, Committee 1990). In recent years, several studies have shown that use of insecticide-impregnated collars on domestic dogs can reduce the numbers of biting sand ßies and lower the incidence of infantile visceral leishmaniasis (Maroli et al. 2001, Gavgani et al. 2002, Maroli and Khoury 2004). Insecticide barrier zones were Þrst applied in French Guiana by clearing forest around a village and fogging with insecticide to a radius of 400 m (Esterre et al. 1986). The trial proved effective in reducing the sand ßy density and the incidence of leishmaniasis, but was extremely harmful to the environment and labor intensive to an extent rendering it impractical (Alexander and Maroli 2003). Pyrethroids were used to reduce sand ßy populations in Guatemala by spraying ground vegetation and tree trunks in a 100-m radius (Perich et al. 1995). A mesh barrier sprayed with DDT was tested around the desert village of Kfar Adumim not far from the current study sites (see Materials and Methods). In this trial, a 2-m-wide strip was laid horizontally on a slope surrounding the village. Sand ßy activity was monitored above and below the strip. No signiÞcant difference in numbers of sand ßies was found, and marked sand ßies were shown to cross the barrier presumably by ßying directly over it to reach the peripheral houses unaffected (Orshan et al. 2006). Adult sand ßies are delicate insects and refrain from ßight activity even in light winds (Roberts and Kumar 1994, Roberts 1994). They advance toward their hosts in short ßights close to the ground. When they encounter a vertical obstacle, they proceed upward in short hopping ßights and pass over it. Despite these

821

limitations, sand ßies are able to disperse over relatively long distances, reaching several hundred meters per night, particularly when attracted by a relatively distant blood source (Yuval and Schlein 1986, Schlein et al. 1989, Alexander et al. 1992, Morrison et al. 1993, Killick-Kendrick 1999). This report describes experiments designed to test the efÞcacy of vertical Þnemesh barriers treated with insecticides for intercepting blood-seeking sand ßies. Materials and Methods Sand Fly Monitoring. Sand ßies were trapped using miniature CDC light traps (model 512; John W. Hock, Gainesville, FL) powered by a 6-V rechargeable battery (model 3FM12; Amit Industries, Ashdod, Israel). Traps were suspended either upside-down in the updraft position, with trap entrance ⬇10 cm above the ground or in the down-draft position, with trap entrance ⬇40 cm above the ground. Traps deployed in the up-draft position were found to capture signiÞcantly more sand ßies than traps deployed in the standard, down-draft orientation with their opening ⬇40 cm above ground (R. F. et al., unpublished data). The traps were baited with 1.5 kg dry ice placed in a tightly closed insulated container, a rubber tube was afÞxed to the spout, and its distal end was placed near the opening of the trap. The traps were deployed between 1700 and 0600 hours the following morning. For certain experiments, CDC traps were operated with an incandescent light bulb and without CO2. Sand Fly Exclusion: Enclosures. A deserted exotic tree plantation 1 km to the east of Kibbutz Gilgal (32⬚00⬘ N, 35⬚27⬘ E; altitude, ⫺240 m) was selected after baseline trapping in several locations in the Jordan Valley. The Gilgal area in the central Jordan Valley is characterized by a warm and dry climate, with an average annual rainfall of 100 Ð150 mm and summer temperatures exceeding 45⬚C. Agricultural plots cover most of the area, with some open pristine habitats to the east. The phlebotomine sand ßy fauna of the region is well documented, comprising almost exclusively P. papatasi (Schlein et al. 1982). Two enclosures (10 by 10 m) were built using 12 iron poles surrounded by a 2-m-high chain-link fence (hole size, 5 cm) made of galvanized iron. Polyester net (1,200 holes/in2, BioNet; Meteor Inc., PetachTikva, Israel) was used to cover the ground inside the enclosures, to minimize the possibility of sand ßies emerging from the ground within them. A 1-m-wide gate was set in the center of one side to allow entrance. The enclosures were built ⬇150 m apart to rule out the possibility of one affecting the other. Four CO2-baited CDC traps (withough lights) were deployed inside each enclosure at the corners. Four control traps were positioned outside each enclosure, ⬇20 m away and surrounding each from four directions. All traps were deployed in the up-draft orientation. Baseline trapping was performed for eight nights after the enclosures were erected but before they were draped with the nets to verify that sand ßies were not deterred by the chain-link fence and the polyester net covering

822

JOURNAL OF MEDICAL ENTOMOLOGY

the ground inside the enclosures. After the eight baseline nights (and 24 h before the ninth trapping night), the enclosures were draped with nets. The west enclosure with SpiderNet⫹ (1,240 holes/in2; Meteor Inc.; Fig. 1A and B), and the east enclosure with the deltamethrine-impregnated PermaNet (225 holes/in2; Vestergaard Frandsen, Lausanne, Switzerland; Fig. 1C and D). Trapping continued for seven additional nights (15 nights in total). Sand Fly Exclusion: Linear Barrier. The town of MaÕale Adumim (population, 32,000) is situated ⬇7 km east of Jerusalem in the Judean Desert (31⬚47⬘ N, 35⬚18⬘ E; altitude, 400 m). Annual rainfall averages 200 Ð 400 mm, summers are hot, and winters are mild (Goldreich 1998). The perimeter fence below Sheizaf street in the southern neighborhood of Tsemach HaÕsade was selected for the study after baseline trapping. A 60-m-long segment was draped with the deltamethrine-impregnated barrier (PermaNet) to a height of 2 m. The lower fringe of the net was covered with soil to anchor it to the ground. The top section of the barrier was fastened at a 45⬚ angle toward the slope to form an external overhang (Fig. 1E). We assumed the slanted overhang would trap ascending sand ßies, prevent them from ßying over the barrier, and thereby increase their exposure to insecticide and improve the efÞcacy of the barrier. A 1-m-wide iron gate was set in the fence to allow access for trapping on the slope. The gate was meticulously covered with net. Sand ßies were captured on four sides of the mesh barrier, i.e., the garden wall west of the barrier (treatment), the wall extending on both sides of the barrier (north and south control), and the slope below the barrier. All traps were deployed with their opening ⬇10 cm above the ground in the up-draft orientation. Marking Sand Flies With Dyed-Sugar Baits. A solution of 10% sucrose and 5 g/liter food dye (Blue No. 1; Indigotine C.I. Stern, Natanya, Israel) was sprayed using a manual hand sprayer (Star, model m5; TeraFlex Pal-Yam, Tnuvot, Israel). The dyed solution was sprayed on vegetation and on rock surfaces along a 2to 3-m-wide strip. Trapped ßies were brought to the laboratory, anesthetized, immersed in phosphatebuffered saline (PBS) with a few drops of detergent, and examined under a stereoscopic microscope. Fed ßies were distinguished by blue color in their abdomen (Schlein 1987). Comparing the proportion of dye-fed sand ßies trapped in different locations enabled us to assess the direction from which the majority of sand ßies were arriving. Sand Fly Identification. Trapped sand ßies were transported on dry ice to the laboratory where they were counted, sexed, and preserved in 70% ethanol or

Vol. 46, no. 4

frozen at ⫺70⬚C. A representative sample was dissected and mounted in HoyerÕs medium on microscope slides for taxonomic identiÞcation. Species were identiÞed based on the morphology of the pharynx, the external genitalia of males, and the spermathecae of females, using several keys (PerÞlÕev 1968, Artemiev 1978, Lewis 1982). Insecticide Decay Assay. Beta-cyßuthrin (1% in water) was sprayed on the SpiderNet (1 liter/10 m2) used in the west enclosure at Gilgal. PermaNet was received preimpregnated with delta-methrine and used in the eastern enclosure in Gilgal and in MaÕale Adumim. To evaluate the rate of decay of insecticides on the nets, two 30 by 30-cm samples were removed on days 1, 30, 60, 90, and 120 after spraying (or after draping of the PermaNet). Net samples were kept in airtight bags at 4⬚C. Assays were performed by independent laboratories using gas chromatography and mass spectrometry (FDA 1995). Decay analysis was performed by plotting the quantity of insecticide found in identical-sized samples at different times after the beginning of the experiment. Data Analysis. Sand ßy numbers were log-transformed [log(n ⫹ 1)] to normalize the distribution and control the variance in the case of non-normal distribution, caused chießy by aggregated distribution. This permitted the derivation of the geometric “Williams means” (Mw), which presented a more accurate and sensitive mean than the arithmetic mean and enabled the application of parametric tests (Williams 1937, Bidlingmayer 1985, Alexander 2000). To determine whether parametric tests were appropriate, the distribution and variance of the sand ßy catch data were tested using tests of normality; 1 ⫺ sample Kolmogorov-Smirnov Z test (K-S) for raw trapping data, and Shapiro-WilkÕs test (S-W) for transformed data and group means. Pearson product-moment and SpearmanÕs ␳ correlation tests were applied to the data gathered in MaÕale Adumim to validate the signiÞcance of the proportion between total sand ßy catch per trap and the dyedsugar-fed sand ßy catch per trap. Establishing a signiÞcant positive correlation enabled us to select the optimal location for the barrier. These tests were also used to conÞrm that ßuctuations in catch of control traps correlated with those observed in experimental traps before construction of the barrier nets. Mann-Whitney rank tests, two-sample t-tests, and paired sample t-tests were applied to compare the mean trap yields before and after the erection of the barrier nets. In Gilgal, comparisons were performed between the two enclosures and between each enclosure and its surrounding control traps. The same

Fig. 1. (Opposite). (A) The west enclosure in Gilgal draped with 2-m-high SpiderNet⫹. (B) Close-up of the SpiderNet⫹ depicting the very Þne mesh (1,240 holes/in2) with a sand ßy above the mesh shown at the same scale. (C) The east enclosure in Gilgal draped with 2-m-high PermaNet. (D) Close-up of the PermaNet depicting the relatively coarse mesh (225 holes/in2) with a sand ßy above the mesh shown at the same scale for reference. Sand ßies are able to pass through these holes but must come in contact with the insecticide impregnated material as they do so. (E) The linear PermaNet barrier erected below Sheizaf Street in MaÕale Adumim. The barrier was constructed of two 150-cm-wide strips, hence the visible overlap. Note the slanted overhang facing the slope. Monitoring traps were set along the stone support wall visible behind the net. (Online Þgure in color.)

July 2009


823

824


Table 1. Correlation tests between the mean catch of control traps set in the periphery and the mean (Mw) catch of experimental traps set inside the enclosures before draping the enclosures with the nets (Spearmans’ ␳ test) Correlated groups Control E Control W Control E Enclosure W Control E Control W

Enclosure E Enclosure W Control W Enclosure E Enclosure W Enclosure E

R

P

0.92 0.87 0.94 0.89 0.66 0.90

0.01 0.00 0.00 0.00 0.01 0.00

P values are signiÞcant at an ␣ of 5%. R, coefÞcient of correlation.

tests were used to test the signiÞcance of differences between the trap groups in MaÕale Adumim. Two-way analysis of variance (ANOVA) was used to show the effect of the trap locations and treatment condition on trap group means. All statistical analyses were performed using SPSS statistical software for Windows, version 13, and Microsoft OfÞce Excel 2003/7. Results Net Enclosures (Gilgal) Sand Fly Species. Preliminary sand ßy trapping was conducted to determine the most suitable locations for the study. In these trapping sessions, 3,367 sand ßies were caught. A randomly selected sample of 300 ßies was identiÞed (PerÞlÕev 1968, Artemiev 1978, Lewis 1982). The population comprised mostly P. papatasi (⬎99%) and a few P. sergenti (⬍1%). Sergentomyia spp., abundant throughout the region, were captured as well but were not included in this study. The average female-to-male ratio was 2.4. Mean sand ßy densities per trap, at the two locations, were 82.1 and 39.0 for the east and west, respectively. Enclosure Experiment. The sand ßy distribution within the Gilgal habitat, as assessed by trap yields was patchy. Therefore, the locations of control traps were selected arbitrarily and maintained throughout the experiments. To verify that the fenced enclosures did not affect the natural sand ßy distribution, correlation tests were conducted between the traps inside each enclosure before draping it with net and its respective control traps. A signiÞcant positive correlation conÞrmed that the control traps provided an accurate representation of the sand ßy population traversing the fenced enclosures, before draping them with nets. Similar correlations were found between the east and west controls and between the east and west enclosures (Table 1). Mean comparison tests showed no signiÞcant differences between enclosure and control groups (P ⬎ 0.1). The mean yields of the different traps were not normally distributed and were therefore log-transformed to derive the geometric means. Although transformation helped reduce the variation and normalize the distribution to a certain extent, the transformed data were still non-normally distributed, re-

Vol. 46, no. 4

quiring the use of both parametric and nonparametric statistic tests. The transformed Mw was plotted on a graph depicting the variation of sand ßy abundance in the trap groups during the 15 trapping nights of the experiment (Fig. 2A). The sand ßy population remained relatively stable and low during the Þrst six nights in all the groups (0 Ð35 ßies per trap). An increase in numbers occurred during the seventh night (1 June) when sand ßy numbers reached 20 Ð90 per trap, and gradually increased in the control traps until the Þnal night. The yields of traps placed inside the enclosures did not differ signiÞcantly from the controls until the nets were deployed after the eighth night (Mann-Whitney; East: Z ⫽ ⫺0.21, P ⬎ 0.1; West: Z ⫽ ⫺0.53, P ⬎ 0.1). Draping with nets resulted in a sharp decrease in the numbers of ßies captured within the enclosures, which remained consistently and signiÞcantly lower than the numbers collected in the control traps for the duration of the experiment (Fig. 2A). Plotting the geometric means (Mw) of the different trap groups on a bar graph showed that the differences between the enclosure groups and their respective controls before draping with the nets were relatively small and insigniÞcant. The addition of the nets caused a substantial reduction in the numbers of sand ßies captured inside both enclosures (Fig. 2B; East: Z ⫽ ⫺3.24, P ⬍ 0.00; West: Z ⫽ ⫺3.24, P ⬍ 0.00). To rule out the possibility that the net barriers were impeding the ßow of CO2, thereby reducing the attractiveness of traps placed inside the enclosures, we used light traps deployed in identical formations (in the down-draft position). A highly signiÞcant difference in the number of ßies caught in the control traps compared with enclosure traps was observed (Fig. 2C; East: Z ⫽ ⫺2.56, P ⬍ 0.05; West: Z ⫽ ⫺2.50, P ⬍ 0.05), essentially supporting the previous Þndings. Linear Net Barrier (Ma’ale Adumim) Selection of the Study Area. Baseline trapping conducted along the southern peripheral fence of MaÕale Adumim during JuneÐAugust 2007 (13 nights) identiÞed several suitable sites in which sand ßy abundance was high. Among these, Sheizaf Street provided the most suitable experimental siteÑallowing vehicle access to the fence and access by foot to the slope below it. The fence was located below the peripheral row of houses, at a distance of 20 Ð30 m, in which there were no suitable habitats for sand ßy breeding (Figs. 1E and 3). Sand ßies were constantly abundant at the site throughout the season, and the local residents who were well acquainted with the nuisance facilitated our work. Sand Fly Species Identification. A total of 19,934 sand ßies were caught in MaÕale Adumim throughout the trapping season (513 trap-nights in all). Of these, 150 randomly selected ßies were identiÞed and shown to comprise four species: 87% P. sergenti, 5% P. papatasi, 4% P. syriacus, and 4% P. tobbi. The average female-to-male ratio was 0.9. Mean sand ßy density per trap at the Sheizaf location was 72.3.

July 2009


825

Fig. 2. Sand ßy catches in and around the vertical net enclosures. Traps were deployed in the updraft position without lights. (A) Total number of sand ßies captured by four CO2-baited CDC traps in the different locations. Bold black 䡺 ⫽ Control West; bold gray 䉫 ⫽ Control East; light black ‚ ⫽ West enclosure with SpiderNet⫹; light gray E ⫽ East enclosure with PermaNet. Enclosures were draped before trapping night 9. The SpiderNet⫹ was treated with beta-cyßuthrin before trapping night 11. (B) Mean (Mw ⫾ SE) sand ßy catches per trap before (two bar pairs on the left) and after (two bar pairs on the right) draping of nets. In each pair, the black bar represents average trap yield inside the enclosure and the gray bar represents the trap yields for the respective control traps. E, east; W, west. The differences between enclosures and control traps were highly signiÞcant (East: Z ⫽ ⫺3.24, P ⬍ 0.00; West: Z ⫽ ⫺3.24, P ⬍ 0.00). (C) Mean sand ßy catches (Mw ⫾ SE) in the two enclosures (two bars on the left) and the control areas (two bars on the right) using CDC traps with lights but without CO2. The differences between enclosures and control traps were highly signiÞcant (East: Z ⫽ ⫺2.56, P ⬍ 0.05; West: Z ⫽ ⫺2.50, P ⬍ 0.05).

Sergentomyia spp., abundant throughout the region, were captured as well but were not included in this study because they do not pose any known health threat to humans. Marking Sand Flies With Dyed-Sugar Baits. To assess whether sand ßies trapped along the gardens support wall originated from the slope below the fence, we sprayed the vegetation immediately below the fence with a dyed sucrose solution to mark foraging sand ßies (Schlein 1987). A 2- to 3-m-wide,

120-m-long strip of vegetation (Fig. 1E, bottom left corner) adjacent to the fence was sprayed with blue colored sucrose solution. Sixty-four percent of the ßies captured along the bottom of the gardenÕs supporting wall on the night after the spraying were marked. A strong positive correlation was found between the total number of sand ßies per trap and the number of dyed sugarÐfed ßies (R ⫽ 0.93, P ⬍ 0.01), i.e., the larger the yield of sand ßies per trap, the higher the proportion of dyed sugar-fed ßies (data not shown).

826


N

Wall Fence

Sheizaf

2

Street

3

1

Barrier

5

4

50 m

Slope

Fig. 3. Schematic diagram of Sheizaf Street study site. The locations of traps are depicted in bold circles (groups 1, 2, 3, 4, and 5). The support wall is indicated in a bold black curved line and the fence in a gray line. The barrier net is marked in the central part of the fence.

A 60-m-long section of the fence was draped with the deltamethrin-impregnated PermaNet (Figs. 1E and 3). The raw trapping data (absolute count) were obtained from 26 traps in Þve locations during 12 trapping nights from a total of 312 trapping nights. The Þve trapping locations were: (1) Control North (six traps), along the garden wall, north of the barrier section of the fence; (2) Control South (six traps), along the garden wall, south of the barrier section of the fence; (3) Barrier (eight traps), along the garden wall, opposite the net; (4) North Slope (three traps), on the northern section of the slope outside the fence opposite the net; and (5) South Slope (three traps), below the fence and south of the net (Fig. 3). During the seven trapping sessions before the net was draped on the fence, sand ßy numbers ßuctuated between 60 and 100 ßies per trap, and their distribution was not uniform, necessitating the validation of the control groups. Pearson product-moment correlation tests were applied to the group means to verify that the control groups along the wall and those on the slope correlated with the means of traps placed along the wall opposite the fence where the barrier was to be constructed. Control N (Fig. 3, group 1) and Control S (Fig. 3, group 2) correlated signiÞcantly with the Barrier group and with each other, indicating that the control traps represented the barrier traps reliably (Table 2). The control group Slope N (Fig. 3, group 4) was excluded because sand ßy numbers at this site were consistently much higher than the rest of the groups (data not shown).

Table 2. Pearson product-moment correlations between the means (Mw) of control traps and the means of the barrier traps before draping of the net Correlated groups Control N Control N Control S Slope S Slope S

Control S Barrier Barrier Barrier Slope N

R

P

0.68 0.81 0.67 0.68 0.70

0.01 0.00 0.02 0.01 0.01

Positive correlations support the validity of the control traps.

Vol. 46, no. 4

Sand ßy catches showed a strong non-normal distribution (K-S normality test, P ⬍ 0.01). Derivation of the geometric means (Mw) by log transformation reduced the variation and normalized the distribution of the means (S-W normality test, P ⬎ 0.05), enabling the use of parametric statistical tests. The transformed means of the different groups over 12 trapping nights of the experiment were depicted on a graph. The control curve represents the mean of the north (Fig. 3, group 1), south (Fig. 3, group 2), and the south slope (Fig. 3, group 5) control groups pooled together. The barrier curve represents the mean of the eight wall traps positioned opposite the 60-m section of the fence draped with the net (Fig. 3, group 3). The sand ßy catches ßuctuated with a decreasing trend over time (Fig. 4A). To differentiate between the seasonal decline and the effect of the barrier and to further reduce the variance, the mean sand ßy catches during the last 3 d before draping of the nets (nights 5Ð7) were compared with those obtained during the Þrst three nights after the draping (nights 8Ð10). The three nights before the intervention exhibited similar nightly averages, as did the three nights after it, markedly reducing the variance caused by the natural ßuctuations in sand ßy population densities (Table 3). Two-sample t-test results comparing the mean (Mw) sand ßy yields before and after treatment showed no signiÞcant difference between means of any group of traps before draping the net (F ⫽ 0.59, df ⫽ 3, P ⬎ 0.1). Hence, before intervention, the sand ßy distribution in all trapping groups was similar. After intervention, all comparisons involving the barrier group were significantly different from controls (F ⫽ 9.62, df ⫽ 3, P ⬍ 0.01), conÞrming that the net effectively reduced the number of sand ßies in the barrier group. To strengthen the above results, we tested the changes recorded in the different locations before and after the draping of the net (Fig. 4B). The average decrease in sand ßy densities between nights 5 and 10 in control traps was 34%. This change was attributed to end-of-season decline in sand ßy density. The decrease recorded over the same period for traps in the barrier group was 83%ÑsigniÞcantly higher than the rest of the groups (Fig. 4B; t ⫽ 5.9, df ⫽ 2, P ⬍ 0.05). To ameliorate the graphic illustration, sand ßy catches of individual traps were plotted as bar graphs along the x-axis (Fig. 5A). The left black bars represent catches of traps in the southern control group (Fig. 3, group 2). The gray bars in the center represent the eight barrier group traps (Fig. 3, group 3). The bars on the right represent the six traps in the north control group (Fig. 3, group 1). To preclude putative artifacts stemming from alterations in the ßow of CO2 through mesh compared with unhindered ßow in open spaces, the same trapping setup was repeated using light-traps instead of CO2-baited traps. Not surprisingly, the trap yields were somewhat smaller (25%). However, the relative sand ßy catches remained very similar and a signiÞcant difference between the control and barrier trap groups was recorded (Fig. 5B; F ⫽ 12.6, df ⫽ 2, P ⬍ 0.01).

July 2009


827

Fig. 4. Large-scale barrier experiment using the delta-methrine impregnated PermaNet in MaÕale Adumim. (A) Mean (Mw ⫾ SE) sand ßy catches of the control (gray line) and barrier (black line) groups during the 12 nights of the experiment. The net was draped before the eighth night (arrow). (B) Mean (Mw ⫾ SE) sand ßy catches for the Þve trapping groups before (gray bars) and after (black bars) deployment of the nets. **Statistically signiÞcant difference (t ⫽ 5.9, df ⫽ 2, P ⬍ 0.05).

Twelve months after the barrier was draped, it remained almost entirely intact, and the experiment was repeated for three nights in July 2008. Results were similar to those obtained the year before. SigniÞcant differences between the means (Mw) of traps in control groups (Control N and Control S) and those of the barrier group were noted (Fig. 5C; F ⫽ 6.1, df ⫽ 2, P ⬍ 0.05). Discussion It is generally agreed that sand ßies advance in short ßights along the ground, and when they encounter a Table 3.

vertical barrier, they proceed upward in short ßights. Although there is no published evidence documenting this behavior, several observations support its validity. First, in experiments conducted in Gilgal, sticky sheets were placed on the fence of a turkey coop to monitor sand ßies attracted to the birds inside. Eighty-four percent of the trapped sand ßies were concentrated in the bottom-most 30 cm of the sticky surfaces (A.W., unpublished data). Furthermore, in preliminary experiments conducted in MaÕale Adumim to monitor sand ßies approaching houses, unbaited CDC traps placed at ground level caught almost nine times more sand ßies than traps placed 1 m above the ground and

Comparison of trapping means before and after draping the net (t-test) Pretreatment (n ⫽ 3)

Groups Control N Control S Control mean Control N Slope S

Mw⫾ SE Barrier Barrier Barrier Control S Barrier

57.8 ⫾ 4.6 46.7 ⫾ 12.4 52.2 ⫾ 5.2 57.8 ⫾ 4.6 61.1 ⫾ 8.0

Posttreatment (n ⫽ 3) P

56.7 ⫾ 7.7 56.7 ⫾ 7.7 56.7 ⫾ 7.7 46.7 ⫾ 12.4 56.7 ⫾ 7.7

0.91 0.53 0.71 0.45 0.52

Mw⫾ SE 38.5 ⫾ 5.5 32.7 ⫾ 3.1 35.6 ⫾ 2.6 38.5 ⫾ 5.5 34.0 ⫾ 6.4

P 9.9 ⫾ 1.0 9.9 ⫾ 1.0 9.9 ⫾ 1.0 32.7 ⫾ 3.1 9.9 ⫾ 1.0

0.01 0.00 0.00 0.40 0.02

Differences between the various traps groups before draping the nets were all insigniÞcant, whereas differences in the mean catch between barrier traps and control traps after draping the net were highly signiÞcant. Numbers in bold indicate statistical signiÞcance at ␣ ⫽ 5%. N, north; S, south; n, nights.

828


Vol. 46, no. 4

Fig. 5. (A) Means (Mw ⫾ SE) of sand ßy catches using CO2-baited traps along the garden wall in the order of placement from south to northÑcontrol south, six traps (Fig. 2, group 2); barrier group, eight traps (Fig. 2, group 3); and control north, six traps (Fig. 2, group 1) during the Þve nights posttreatment (F ⫽ 42.2, df ⫽ 2, P ⬍ 0.01). (B) Mean (Mw ⫾ SE) sand ßy catches along the same stretch of fence using the same setup with light-traps (without CO2; F ⫽ 12.6, df ⫽ 2, P ⬍ 0.01). (C) Exact same experiment as in A performed 1 yr after the net was erected (F ⫽ 6.1, df ⫽ 2, P ⬍ 0.05).

40 times more ßies than traps placed 2m above the ground (P ⬍ 0.01). In contrast, on a vertical wall, ground level traps captured twice as many ßies as traps positioned at 1 m and almost four-fold more ßies than traps placed 2 m above ground (P ⬍ 0.01). In these experiments, traps were deployed without any attractant (unpublished data). These results showed that sand ßies normally keep close to the ground. When they encounter a vertical obstacle, they proceed upward to pass over it. We assumed the vertical nets would comprise obstacles and block foraging sand ßies. Consistent with this hypothesis, the numbers of sand ßies cap-

tured inside the protected enclosures using CO2baited CDC traps without lights were ⬎90% lower than in open areas close by (Figs. 2 and 3). The small scale of the enclosure experiment was believed to be a limitation preventing us from drawing sweeping conclusions. In Africa, several mosquito species only rarely descended into small enclosures, but when large enclosures were constructed, similar numbers of mosquitoes were captured inside and outside them (Gillies and Wilkes 1978). Although such a scenario is not likely to explain our Þndings, because sand ßies tend to stay close to the ground and are not likely to pass over a 2-m-high, 10-m-long

July 2009


enclosure, we decided to test net barriers on a larger scale. Despite several limitations, the relatively short linear net in MaÕale Adumim provided a 50 Ð 60% reduction in sand ßy numbers (Table 3; Figs. 4 and 5). Those ßies captured behind the net may have comprised populations returning from their foraging in and around houses to their natural breeding habitats. Others still may have passed around the barrier on either side or through its holes. The mesh size used was rather large, tentatively allowing passage of ßies through it. However, the Þnding that the net remained effective a year after its initial deployment (Fig. 5C), once insecticide levels have dropped considerably, indicates that physical blocking of the ßies may have been more important than contact with insecticides. One concern arising while interpreting the results was that utilization of CO2-baited traps deployed behind net barriers may not provide an accurate assessment of the numbers of sand ßies. Conceivably, changes in air movement caused by the netting could alter the CO2 plume and reduce trapping efÞciency (Cooperband and Carde 2006). To control for such possibilities we measured CO2 seepage through the two experimental nets using a CO2 detector. Results showed that CO2 seeping through the PermaNet was unaffected by the mesh, whereas seepage through the SpiderNet⫹ was considerably slower (unpublished data). Therefore, to make sure that lower sand ßy yields within enclosures were not an artifact caused by less CO2 making it through the mesh, CDC light traps (without CO2) were used in both study sites to complement the results obtained with CO2-baited traps. Results conÞrmed that much fewer ßies were captured in the enclosures than the surrounding habitat in Gilgal or behind the barrier in MaÕale Adumim regardless of the attractant used (Figs. 2C and 5C). Evaluating the efÞcacy of different insecticides for controlling sand ßies was not one of the aims of this work. Deltamethrin was used because the PermaNet came preimpregnated with it. Deltamethrin is one of the insecticides used for treating bednets for mosquitoborne disease control (Pates and Curtis 2005) and has also been shown efÞcacious in controlling several species of sand ßies (Alexander et al. 1995b, Alexander and Maroli 2003, Moosa-Kazemi et al. 2007, Ritmeijer et al. 2007). Beta-cyßuthrin was selected for treating the SpiderNet⫹ because data existed showing that, under desert conditions such as those prevailing in the Judean Desert, Beta-cyßuthrin was effective against phlebotomine sand ßies (Tetreault et al. 2001, Orshan et al. 2006). The results of the experiments reported here have established the feasibility of using vertical Þne-mesh barriers to block sand ßies approaching human dwellings from the periphery. It is anticipated that this type of intervention used on a large scale could affect a mass reduction of sand ßies both inside and outside houses in entire neighborhoods. The barriers can be deployed with relative ease and inßict minimal harm to the environment. Integrating vertical barriers with

829

other control strategies will be necessary to achieve acceptable levels of control under different settings. For example, use of diffusible prallethrin in bedrooms has proven effective against endophilic sand ßies, providing excellent relief from biting females (Sirak et al. 2008). Acknowledgments We thank G. Muller for suggesting the study with vertical nets and the residents of Sheizaf Street in MaÕale Adumim for their cooperation. PermaNet was provided gratis by Vestergaard Frandsen, Switzerland. This research was supported by a grant from the Deployed War-Fighter Protection (DWFP) Research Program, funded by the U.S. Department of Defense through the Armed Forces Pest Management Board (AFPMB) and The Israel Science Foundation (Grant 135/ 08). Additional funding provided by Grant SCHO 448/8 Ð1 from the Deutsche Forschungsgemeinschaft (DFG): “Emergence of cutaneous leishmaniasis in the Middle East: an investigation of Leishmania tropica in The Palestinian Authority and Israel.”

References Cited Abdeen, Z. A., S. S. Sawalha, C. L. Eisenberger, H. M. Khanfar, C. L. Greenblatt, O. Yousef, L. F. Schnur, K. Azmi, A. Warburg, K. A. Bader, C. L. Jaffe, and G. Baneth. 2002. Epidemiology of visceral leishmaniasis in the Jenin District, West Bank: 1989 Ð1998. Am. J. Trop. Med. Hyg. 66: 329 Ð333. Alexander, B. 2000. Sampling methods for phlebotomine sandßies. Med. Vet. Entomol. 14: 109 Ð122. Alexander, B., and M. Maroli. 2003. Control of phlebotomine sandßies. Med. Vet. Entomol. 17: 1Ð18. Alexander, B., C. Ferro, D. G. Young, A. Morales, and R. B. Tesh. 1992. Ecology of phlebotomine sand ßies (Diptera: Psychodidae) in a focus of Leishmania (Viannia) braziliensis in northeastern Colombia. Mem. Inst. Oswaldo Cruz 87: 387Ð395. Alexander, B., C. Jaramillo, M. C. Usma, B. L. Quesada, H. Cadena, W. Roa, and B. L. Travi. 1995a. An attempt to control Phlebotomine sand ßies (Diptera: Psychodidae) by residual spraying with deltamethrin in a Colombian village. Mem. Inst. Oswaldo Cruz. 90: 421Ð 424. Alexander, B., A. L. Morales, J. Becerra, and C. A. Rojas. 1995b. Entomological aspects of a Leishmaniasis control project based on comunity participation in riverine communities of SW Colombia. Bol. Direccion Malariol. Saneamiento Ambient 35: 29 Ð 40. Artemiev, M. 1980. Revision of the sandßies of the subgenus Adlerius (Diptera Phlebotominae, Phlebotomus). Zool. Zh. 59: 1177Ð1193. Baneth, G., G. Dank, E. Keren-Kornblatt, E. Sekeles, I. Adini, C. L. Eisenberger, L. F. Schnur, R. King, and C. L. Jaffe. 1998. Emergence of visceral leishmaniasis in central Israel. Am. J. Trop. Med. Hyg. 59: 722Ð725. Basimike, M., and M. J. Mutinga. 1995. Effects of permethrin-treated screens on phlebotomine sand ßies, with reference to Phlebotomus martini (Diptera: Psychodidae). J. Med. Entomol. 32: 428 Ð 432. Bettini, S. 1989. Sandßy breeding-sites. Life Sci. 163: 179 Ð 188. Bidlingmayer, W. L. 1985. The measurement of adult mosquito population changesÑsome considerations. J. Am. Mosq Control Assoc. 1: 328 Ð348.

830


Coleman, R. E., D. A. Burkett, V. Sherwood, J. Caci, S. Spradling, B. T. Jennings, E. Rowton, W. Gilmore, K. Blount, C. E. White, and J. L. Putnam. 2007. Impact of phlebotomine sand ßies on U.S. Military operations at Tallil Air Base, Iraq: 2. Temporal and geographic distribution of sand ßies. J. Med. Entomol. 44: 29 Ð 41. Committee, W. E. 1990. Control of the leishmaniases. Report of a WHO Expert Committee. World Health Organ Tech. Rep. Ser. 793: 1Ð158. Cooperband, M. F., and R. T. Carde. 2006. Comparison of plume structures of carbon dioxide emitted from different mosquito traps. Med. Vet. Entomol. 20: 1Ð10. Courtenay, O., K. Gillingwater, P. A. Gomes, L. M. Garcez, and C. R. Davies. 2007. Deltamethrin-impregnated bednets reduce human landing rates of sandßy vector Lutzomyia longipalpis in Amazon households. Med. Vet. Entomol. 21: 168 Ð176. Dalton, R. 2008. Entomology: battleÞeld insectica. Nature (Lond.) 454: 18 Ð19. Desjeux, P. 2004. Leishmaniasis: current situation and new perspectives. Comp. Immunol. Microbiol. Infect. Dis. 27: 305Ð318. Elnaiem, D. A., A. M. Elnahas, and M. A. Aboud. 1999. Protective efÞcacy of lambdacyhalothrin-impregnated bednets against Phlebotomus orientalis, the vector of visceral leishmaniasis in Sudan. Med. Vet. Entomol. 13: 310 Ð 314. Esterre, P., J. P. Chippaux, J. F. Lefait, and J. P. Dedet. 1986. [Evaluation of a cutaneous leishmaniasis control program in a forest village of French Guyana]. Bull. W.H.O. 64: 559 Ð565. [FDA] Food and Drug Administration. 1995. Pesticide analytical manual (PAM). Food and Drug Administration, Washington, DC. Feliciangeli, M. D. 2004. Natural breeding places of phlebotomine sandßies. Med. Vet. Entomol. 18: 71Ð 80. Gavgani, A. S., M. H. Hodjati, H. Mohite, and C. R. Davies. 2002. Effect of insecticide-impregnated dog collars on incidence of zoonotic visceral leishmaniasis in Iranian children: a matched-cluster randomised trial. Lancet 360: 374 Ð379. Gillies, M. T., and T. J. Wilkes. 1978. The effect of high fences on the dispersal of some West African mosquitoes (Diptera: Culicidae). Bull. Entomol. Res. 68: 401Ð 408. Goldreich, Y. 1998. The climate of Israel: observations research and applications. Bar Ilan Univ. and Magnes Publishers, Ramat Gan, Israel. Greenblatt, C. L., Y. Schlein, and L. F. Schnur. 1985. Leishmaniasis in Israel and vicinity, pp. 415Ð 426. In K.-P. Chang and R. S. Bray (eds.), Leishmaniasis. Elsevier, Amsterdam, The Netherlands. Hertig, M. 1949. Phlebotomus and residual DDT in Greece and Italy. Am. J. Trop. Med. Hyg. 1: 773. Hertig, M., and L. R. Fisher. 1945. Control of sandßies with DDT. Bull. U.S. Army Med. Dept. 88: 97. Hertig, M., and G. B. Fairchild. 1948. The control of Phlebotomus in Peru with DDT. Am. J. Trop. Med. Hyg. 1: 207. Jacobson, R. L., C. L. Eisenberger, M. Svobodova, G. Baneth, J. Sztern, J. Carvalho, A. Nasereddin, M. E. Fari, U. Shalom, P. Volf, et al. 2003. Outbreak of cutaneous leishmaniasis in northern Israel. J. Infect. Dis. 188: 1065Ð1073. Jaffe, C. L., G. Baneth, Z. A. Abdeen, Y. Schlein, and A. Warburg. 2004. Leishmaniasis in Israel and the Palestinian Authority. Trends Parasitol. 20: 328 Ð332. Killick-Kendrick, R. 1990. Phlebotomine vectors of the leishmaniases: a review. Med. Vet. Entomol. 4: 1Ð24. Killick-Kendrick, R. 1999. The biology and control of phlebotomine sand ßies. Clin. Dermatol. 17: 279 Ð289.

Vol. 46, no. 4

Kroeger, A., E. V. Avila, and L. Morison. 2002. Insecticide impregnated curtains to control domestic transmission of cutaneous leishmaniasis in Venezuela: cluster randomised trial. BMJ 325: 810 Ð 813. Le Pont, F., J. M. Padilla, P. Desjeux, A. Richard, and J. Mouchet. 1989. Impact de pulvÕrisations de deltramÕthrine dans un foyer de leishmaniose de Bolivie. Ann. Soc. Belge Med. Trop. 69: 223Ð232. Lewis, D. 1982. A taxonomic review of the genus Phlebotomus (Diptera: Psychodidae). Bull. Br. Mus. Nat. Hist. (Ent.) 45: 121Ð209. Marcondes, C. B., and J. A. Nascimento. 1993. [Evaluation of the effectiveness of deltamethrin (K-othrine CE) in the control of Lutzomyia longipalpis (Diptera: Psychodidae), in the municipality of Santa Rita, Paraiba, Brazil]. Rev. Soc. Bras. Med. Trop. 26: 15Ð18. Maroli, M., and C. Khoury. 2004. [Prevention and control of leishmaniasis vectors: current approaches]. Parassitologia 46: 211Ð215. Maroli, M., V. Mizzon, C. Siragusa, A. D’Oorazi, and L. Gradoni. 2001. Evidence for an impact on the incidence of canine leishmaniasis by the mass use of deltamethrinimpregnated dog collars in southern Italy. Med. Vet. Entomol. 15: 358 Ð363. Moosa-Kazemi, S. H., M. R. Yaghoobi-Ershadi, A. A. Akhavan, H. Abdoli, A. R. Zahraei-Ramazani, R. Jafari, B. Houshmand, A. Nadim, and M. Hosseini. 2007. Deltamethrinimpregnated bed nets and curtains in an anthroponotic cutaneous leishmaniasis control program in northeastern Iran. Ann. Saudi Med. 27: 6 Ð12. Morrison, A. C., C. Ferro, A. Morales, R. B. Tesh, and M. L. Wilson. 1993. Dispersal of the sand ßy Lutzomyia longipalpis (Diptera: Psychodidae) at an endemic focus of visceral leishmaniasis in Colombia. J. Med. Entomol. 30: 427Ð 435. Orshan, L., D. Szekely, H. Schnur, A. Wilamowski, Y. Galer, S. Bitton, and Y. Schlein. 2006. Attempts to control sand ßies by insecticide-sprayed strips along the periphery of a village. J. Vector Ecol. 31: 113Ð117. Pates, H., and C. Curtis. 2005. Mosquito behavior and vector control. Annu. Rev. Entomol. 50: 53Ð70. Perfil’ev, P. P. 1968. Phlebotomidae (Sandßies). Israel Program for ScientiÞc Translations available from the Clearinghouse for Federal ScientiÞc and Technical Information, SpringÞeld, VA. Perich, M. J., A. L. Hoch, N. Rizzo, and E. D. Rowton. 1995. Insecticide barrier spraying for the control of sand ßy vectors of cutaneous leishmaniasis in rural Guatemala. Am. J. Trop. Med. Hyg. 52: 485Ð 488. Ritmeijer, K., C. Davies, R. van Zorge, S. J. Wang, J. Schorscher, S. I. Dongu’du, and R. N. Davidson. 2007. Evaluation of a mass distribution programme for Þnemesh impregnated bednets against visceral leishmaniasis in eastern Sudan. Trop. Med. Int. Health. 12: 404 Ð 414. Roberts, D., and S. Kumar. 1994. Using vehicle-mounted nets for studying activity of Arabian sand ßies (Diptera: Psychodidae). J. Med. Entomol. 31: 388 Ð393. Roberts, D. M. 1994. Arabian sandßies (Diptera: Psychodidae) prefer the hottest nights? Med. Vet. Entomol. 8: 194 Ð198. Schlein, Y. 1987. Marking of Phlebotomus papatasi (Diptera: Psychodidae) by feeding on sprayed, coloured sugar bait: a possible means for behavioural and control studies. Trans. R Soc. Trop. Med. Hyg. 81: 599. Schlein, Y., B. Yuval, and R. L. Jacobson. 1989. Leishmaniasis in the Jordan Valley: differential attraction of dispersing and breeding site populations of Phlebotomus

July 2009


papatasi (Diptera: Psychodidae) to manure and water. J. Med. Entomol. 26: 411Ð 413. Schlein, Y., A. Warburg, L. F. Schnur, and A. E. Gunders. 1982. Leishmaniasis in the Jordan Valley II. Sandßies and transmission in the central endemic area. Trans. R. Soc. Trop. Med. Hyg. 76: 582Ð586. Schnur, L. F., A. Nasereddin, C. L. Eisenberger, C. L. Jaffe, M. El Fari, K. Azmi, G. Anders, M. Killick-Kendrick, R. Killick-Kendrick, J. P. Dedet, F. Pratlong, et al. 2004. Multifarious characterization of Leishmania tropica from a Judean desert focus, exposing intraspeciÞc diversity and incriminating Phlebotomus sergenti as its vector. Am. J. Trop. Med. Hyg. 70: 364 Ð372. Sirak-Weizman, M., R. Cunio, R. Faiman, A. Al-jawabreh, and A. Warburg. 2008. Control of phlebotomine sandßies in conÞned spaces using diffusible repellents and insecticides Med. Vet. Entomol. 22: 405Ð 412. Svobodova, M., J. Votypka, J. Peckova, V. Dvorak, A. Nasereddin, G. Baneth, J. Sztern, V. Kravchenko, A. Orr, D. Meir, et al. 2006. Distinct transmission cycles of Leishmania tropica in 2 adjacent foci, northern Israel. Emerg. Infect. Dis. 12: 1860 Ð1868.

831

Tetreault, G. E., A. Zayed, H. A. Hanafi, G. M. Beavers, and B. C. Zeichner. 2001. Susceptibility of sand ßies to selected insecticides in north africa and the middle east. J. Am. Mosq. Control Assoc. 17: 23Ð27. Vieira, J. B., and G. E. Coelho. 1998. [Visceral leishmaniasis or kala-azar: the epidemiological and control aspects]. Rev. Soc. Bras. Med. Trop. 31(Suppl 2): 85Ð92. Vioukov, V. 1987. Control of transmission. The leishmaniases in biology and medicine, pp. 909 Ð928. In W. Peters and R. Killick-Kendrick (eds.), The leishmaniases. Academic, London, United Kingdom. Wilamowski, A., and H. Pener. 2003. EfÞcacy of microencapsulated insecticides against the sandßy, Phlebotomus papatasi Scopoli. J. Vector Ecol. 28: 229 Ð233. Williams, C. B. 1937. The use of logarithms in the interpretation of certain entomological problems. Ann. Appl. Biol. 24: 404 Ð 414. Yuval, B., and Y. Schlein. 1986. Leishmaniasis in the Jordan Valley. III. Nocturnal activity of Phlebotomus papatasi (Diptera: Psychodidae) in relation to nutrition and ovarian development. J. Med. Entomol. 23: 411Ð 415. Received 6 October 2008; accepted 30 March 2009.