Large-scale snail control trial with trifenmorph in the Gezira irrigation scheme, Sudan

A large-scale snail control trial with trifenmorph in the Gezira irrigation scheme, Sudan M. A. AMIN,1 A. FENWICK,2 J. M. OSGERBY,3 A. P. WARLEY,3 & A. N. WRIGHT 3 A large-scale field trial was carried out during December 1973 to assess the effect of trifenmorph on Bulinus truncatus and Biomphalaria pfeifferi in 379 000 feddans (,J159 000 ha) of the Gezira irrigation system in the Sudan. The commercial formu- lation used (Frescon) is an emulsifiable concentrate containing 16.5 % trifenmorph. Five dispensers were used to add the commercialproduct to the water continuously for 7.5 days; 18 121 litres were used to treat 28.4 million m3 of water. In addition, each minor canal was hand-sprayed from the tail to 300 m upstream of the last open field outlet pipe; 360 litres of the commercial formulation were used for this operation. A minimum concentration of 0.035 mg trifenmorph per litre of water was produced at the head of each minor canal. The use of caged snails showed that a concentration as low as 0.015 mg/litre was sufficient to produce 100 % mortality in B. truncatus in 7.5 days; this is equivalent to a concentration x time product of 0.12 mg/litre days. The Gezira and Managil irrigation systems supply water to more than 800 000 ha (approximately two million feddans) of the Sudan. Since December 1970, a pilot area north of kilometre (K) 183 on the Gezira main canal (serving 46 200 ha (110 000 feddans)) has been treated with trifenmorph for the control of schistosomiasis (1). The commercial formulation used (Frescon) is an emulsifiable concentrate based on perchloroethylene containing 16.5 % trifen- morph. Satisfactory control of Biomphalaria pfeifferi and Bulinus truncatus, the host snails of Schistosoma mansoni and S. haematobium, respectively, can now be achieved in both the major and minor canals in the pilot area by applying trifenmorph from a drip-feed dispenser at K 183 followed by hand-spray- ing of the ends of the minor canals, where low rates of water flow would otherwise preclude adequate penetration of the molluscicide. One possible annual treatment regimen consists of two drip-feed appli- 1 Senior Lecturer, Department of Social and Preventive Medicine, University of Khartoum, P.O. Box 102, Khartoum, Sudan. Project Leader, London Khartoum Bilharzia Project, P.O. Box 2371, Khartoum, Sudan. ' Wellcome Trust Fellow, Department of Helminthology, London School of Hygiene and Tropical Medicine, on secondment to the London Khartoum Bilharzia Project. ' Shell Research Ltd, Woodstock Laboratory, Sitting- bourne Research Centre, Sittingbourne, Kent, ME9 8AG, England. cations (in September and December) followed by aerial spraying of all minor canals in March and May, when water flow is lowest. It is recognized that the treatment of the whole of the Gezira and Managil irrigation systems will be necessary for optimum snail control. It was there- fore decided to increase the treated area and, in the interests of economy, to use improved application techniques. The first phase of this increase was started in December 1973, when canals irrigating about 160 000 ha (379 000 feddans) were treated with trifenmorph from five application points (Fig. 1). This paper describes the details of this treatment and the results. The irrigation system The Gezira irrigation system contains three main types of distributive canal (main, major, and minor) and the field channels known as abu eshreens and abu sittas (Fig. 2). The minor canals are thought to be the most important foci in the transmission of schistosomiasis since snail populations there are large and the man/water contact is high. The main Gezira canal runs approximately due north from the Sennar dam for 204 km and feeds water to approximately 70 major canals along its length. The function of both the main and the major canals is to transport water rather than to irrigate land. 3554 - 573 _ BULL. WORLD HEALTH ORGAN., Vol. 54, 1976 M. A. AMIN ET AL. Fig. 1. Gezira canalization north of K 150 showing selected minor canals. The minor canals are positioned at right angles to the major canals at intervals of 1.42 km. They are, however, fed from the major canals in pairs so that the offtakes from the major canals are 2.84 km apart. The function of the minor canals is to pro- vide a sufficient head of water to irrigate the adjacent land. The water is taken from the minor canals through field outlet pipes into abu eshreens and abu sittas and thence to the crops. Trifenmorph application By applying trifenmorph from a drip-feed dis- penser to the main canal, all water courses below the treatment point are treated in a convenient way, provided that the water is flowing. At the ends of minor canals this is frequently not the case and these portions have to be treated separately with knapsack sprayers. Trifenmorph is hydrolysed in water to triphenyl- methanol, which is not molluscicidal. It is therefore necessary to know the rate of hydrolysis and the time taken for water to travel to a particular minor canal in order to calculate the rate of application. In previous work, the half-life of trifenmorph has been measured in major canals and shown to be 25-30 h (1). This is considerably less than that meas- ured in laboratory tests (c-70 h) with water of the same pH (2) and this difference is thought to be due to an interaction of trifenmorph with suspended silt. It was considered possible that the half-life in the main canals was different from that in the major canals and therefore an investigation of this possi- bility was incorporated into the present work. Flow rates in the main canals and all of the major canals were calculated from the design data of the irrigation scheme and from crop rotations and crop water use. A further objective was to compare these calculated values with values measured during mol- luscicide application. Calculations made with these data have shown that, for treatment areas greater than 42 000 ha (100 000 feddans), much greater econ- omy results from the use of more than one appli- Y dispenser El Sayr 1N |Ahmed Wad Yussef Waghara Major El Sari \>% 0 4 8 12 16 km I 574 SNAIL CONTROL IN THE SUDAN o _R Minor Canal Main Canal - Fig. 2. Schematic diagram of Gezira canal system. cation point. For the area of about 160 000 ha to be treated in the first phase, it was decided that one main and four supplementary dispensers should be used. The main Gezira canal has 13 regulators and the offtakes of all the major canals are situated immedi- ately upstream of these regulators. In previous appli- cations to the pilot area it was found convenient to site the molluscicide dispenser on the regulator at K 183, or close to it. At this position the water is very turbulent and good mixing should result. However, it is clearly more economical to apply trifenmorph just upstream of the main regulators since this reduces the time taken for the chemical to reach the major canals and therefore loss by hydrolysis. Minor canals The majority of the snails live in the minor canals, where water flow rates are low and there is an abundance of aquatic weeds. These minor canals are between 2 and 20 km long and they contain much more water than is required to irrigate the crops over a 24-h period. It follows that any trifen- morph entering from a major canal is initially diluted in the minor canal. The minor canals are fairly uniform in cross-sec- tional area (6-8 m2) throughout their length. Because of this, the water velocity must decrease along the length of the canal almost linearly. The velocity of water (v) at a point L metres from the end of a minor canal is given by: discharge at the point cross sectional area m sec The discharge at the point depends upon the amount of water required to irrigate land served by the minor canal below that point. Water is taken from the minor canal by field outlet pipes spaced approximately 300 m apart that, on average, irrigate 37.8 ha (90 feddans) each. The gross area served by a length L of minor canal is therefore: 90 L gross area = 292 feddans c O-n.0 , _. I I 575 M. A. AMIN ET AL. The amount of water required to irrigate this area (WD) depends on the cropping pattern and it is known for each month of the year, thus: discharge = 90 WD m3 month-'292 The water velocity may therefore be expressed as: 90 L WD 292 x CSA where CSA is the cross-sectional area of the minor canal and a is a constant for any month. Equation (1) may be expressed as: dL = -aLdt which on integration gives: t= - log (2) a Lo-b where Lo is the total length of the minor canal, b is the distance from the major canal, and t is the time taken to reach the point b. Let b = fLo (O<f<1) then t =- log (3) Thus, to a first approximation, the time taken for trifenmorph to flow down a given portion of a minor canal will be independent of the length of the minor canal. In principle, it is therefore necessary to ensure that the concentration at the head of all minor canals is equal to or greater than that which will give rise to a lethal concentration at the end of the canal, after making allowances for losses due to hydrolysis. Attempts have been made to calculate this minimum concentration but this depends on estimates not only of the flow time in the canal but also of the minimum concentration that is lethal to snails for a given period of exposure. METHOD AND MATERIALS Sampling procedures Water samples were collected in 800-ml glass bottles, which were usually filled by immersing them at the side of the canal. Where possible, the water was taken immediately downstream of a sluice gate. Analytical procedures Water samples were analysed for trifenmorph by means of two colorimetric procedures. The field pro- cedure (3) was used for the majority of the samples since it is quicker and it was possible to make up to 50 determinations in a working day. The labora- tory procedure (4) was used for a few samples to provide a cross-check and in those cases where analy- sis for triphenylmethanol was required. Extraction of most water samples (800 ml) was carried out in the field with cyclohexane and the extract, after it had been dried with anhydrous sodium carbonate, was transferred to the laboratory. Some water samples were transferred without ex- traction but with the addition of sodium carbonate to pH 9 to stabilize the trifenmorph. These pro- cedures have previously been shown to give good recoveries of trifenmorph (1). Separation of the trifenmorph in the extracts from triphenylmethanol was carried out on Alumina H with 1.9%. or 2.0%4 water either in the trifenmorph purification tubes (field procedure, the eluent being 25 parts of cyclohexane to 8 parts of diisopropyl ether by volume) or in 300 x 10-mm columns (labora- tory procedure, the eluent being 95 parts of cyclo- hexane to 5 parts of diethyl ether by volume). No problems were found in obtaining good recoveries (90-100%) of trifenmorph and usually good separ- ation (>95%/O) of the triphenylmethanol was achieved. However, some samples of diisopropyl ether gave premature elution of the triphenylmethanol in the trifenmorph fraction. This was possibly due to the presence of isopropanol in the solvent. Determination of the trifenmorph and triphenyl- methanol present was carried out on a Unicam SP 600 spectrophotometer. This instrument gave a straight-line relation of optical density to concen- tration with 10- or 40-mm path length cells as appropriate. The analytical limit was normally set by the absorption by untreated water. This ranged from 0.002 to 0.007 mg/litre for the field procedure and was rather lower and more consistent (<0.003 mg/ litre) for the laboratory procedure. Appropriate values were used for each set of samples. At 435 nm, a concentration of trifenmorph in the water of 0.01 mg/litre gave an optical density of 0.05 (10-mm cell) and 0.23 (40-mm cell). Silt content A semi-quantitative procedure was used to meas- ure the silt content of water in various parts of the 576 SNAIL CONTROL IN THE SUDAN irrigation system. Samples of water (800 ml) were allowed to settle for 16 h and the sediment in the bottom 25 ml was run into a second container. The latter consisted of a B 19 tube fused to a 2.5-mm internal diameter capillary tube graduated in 0.02-ml units to 1.0 ml. The slurry was left for 24 h to settle in this tube and the volume of the pellet was meas- ured. Although this procedure is not quantitative, it could be calibrated for comparison in a particular irrigation system. Duplicate samples gave consistent readings. Application of trifenmorph Trifenmorph was applied to the flowing water from simple dispensers (5). The emission rates used are shown in Table 1, together with rates of water dis- Table 1. Dispensing schedule for December 1973 Concentra- tion of Mean water Emision Quantity ofDispensing commercial discharge rate commercial point formulation (millions formulation added M3/day) (m/i)used (litres)(mg/litre) K 150 0.078 3.74 1225 13 280 K 183 0.050 1.00 200 2250 Waghara Branch 0.034 0.70 100 1207 Awamra Major 0.035 0.31 46 913 Abu Qutu Major 0.061 0.42 110 471 total 18 121 Table 2. Data on knapsack spraying of the tail ends of minor canals Length sprayed No. of canals Quantity of commercialLengthmsprayed No. of canals formulation used(in) (litres) 0 2 0 300 150 600 600 45 360 900 21 252 1200 8 128 1500 1 20 total 227 1360 charge and total volumes of the commercial formu- lation used. Table 2 shows the amount of the commercial formulation used for spraying the minor canals. Biological procedures Eight minor canals were selected for studies on caged snails. The canals were chosen to represent long, medium-length, and short canals throughout the whole area under treatment. Caged snails were positioned at the tail end of each canal and at inter- vals of 600 m from the tail, at points just upstream of the field outlet pipes for the abu eshreens. The snail cages measured 20 x 20 x 20 cm and consisted of a mild steel rod frame covered with poly(ethylene terephthalate) curtain netting. Each cage was submerged 1-2 m from the canal bank at a depth of 0.5-1.0 m and held in place by a cord and peg. At each sampling station, 1-3 cages were positioned before the arrival of the trifenmorph and removed after different periods of exposure (2-8 days). Adult Bulinus truncatus and Biomphalaria pfeifferi were collected from the ends of minor canals outside the treated area immediately prior to the bioassay. The snail species of major interest for this study was B. truncatus, since it is by far the less susceptible of the two to trifenmorph; it was therefore included in all of the cages. B. pfeifferi is, however, the most important species from a public health point of view and was therefore included in most of the cages. Each cage contained 30-50 snails of each species. After the removal of a cage from the water, all of the snails were transferred to the flat gauze of a dip net and those snails that were obviously dead were counted and discarded. Those obviously or possibly alive were placed in plastic cups with damp cotton wool, returned to the field laboratory, and placed in fresh water. Mortality counts were carried out after allowing a recovery period of 12 h. A water sample was taken from each sampling station at the beginning of the exposure period and at the removal of each cage. This was immediately extracted with cyclohexane for subsequent analysis in the laboratory. The treatment was also assessed by the routine sampling of the natural snail population. Thirteen minor canals were sampled before and after treat- ment by means of 40 dip-net scoops at the head of the minor canal and 20 scoops at 300-m intervals down its entire length. 577 M. A. AMIN ET AL. mg/litre Fig. 3. Trifenmorph concentrations during 6-hour application from K 150. RESULTS AND DISCUSSION Application of trifenmorph above a regulator Trifenmorph was applied to the water for 1 h at the western side of the main canal, 500 m upstream of the regulator at K 169, to determine whether ade- quate mixing took place before the canal divided. Water samples were taken at 10-min intervals from below the regulator and from the heads of the three main offtakes-the NW Branch Canal, Debeiba Major, and Abu Usher Major. Analysis of the sam- ples showed that inadequate mixing had occurred and that most of the trifenmorph entered the NW Branch Canal and Debeiba Major (both on the western side of the main canal). It was therefore decided that all subsequent applications would be from the regulator at K 150. Study of the half-life oftrifenmorph in main and major canals Trifenmorph was applied for 6 h from the eastern side of the main canal immediately below the regu- lator at K 150. The water discharge was 3.89 mil- lion m3/day and the delivery rate of the commercial formulation was 1640 ml/min, to give a calculated trifenmorph concentration of 0.1 mg/litre. Water samples were collected from 11 sites throughout the treatment area and these were analysed for trifen- morph to determine the time of arrival at each site and the concentration-time profile. These profiles are shown in Fig. 3. The half-life of trifenmorph was determined by plotting log concentration against time (Fig. 4) and was found to be 31.5 h. This is in good agreement with values reported by Amin & Fenwick (1). Samples taken 500 m below the dispenser at K 150 indicated that most of the trifenmorph was on the same side of the canal as the dispenser. Therefore, even though the water was extremely turbulent at the point of addition, there was inadequate mixing across the main canal. The dispenser was sub- sequently moved so that the formulation could be added to the centre of the main canal; this position resulted in acceptably uniform mixing. Water velocity and discharge Water velocities and discharges were calculated from the design data of the irrigation system, and from data for the cropping pattern and crop water requirements reported by Farbrother (6). This has 578 SNAIL CONTROL IN THE SUDAN Uo1U 0.09 _ 0.08 0.07 0.06 0.054 e 0.04 !t 0~~~~~~~~~~ 0.02 0010 0 10 20 30 40 Time (h) 50 60 70 Fig. 4. Variation of trifenmorph concentration with time in 6-hour application from K 150. given good agreement with the values found in the trial (Tables 3 and 4), particularly for flow times in the main and branch canals. From the design data, it is possible to calculate water discharge and vel- ocity only at the head of each major canal. Therefore, in order to estimate the flow time in these canals it is necessary to make some assumption regarding the decrease of velocity with distance travelled. If it is assumed that this decrease is exponential in nature, then the following will be true: v=vo e-PL (4) where v is the velocity at distance L from the head of the major canal, vo is the velocity at the head of the canal, and p is the rate constant. Table 3. Calculated and K 150, in hours observed flow times from Observed Calculated K 169 8.0 8.0 K 183 14.7 14.7 K 197 23.0 23.8 Waghara Branch K 11 26.75 26.6 Waghara Major K 10 39.7 39.7 Waghara Major K 18 48.0 50.5 a end of Awamra Major 44.75 54.3 a end of Kassember Major 36.0 31.0 a end of Ugud Major 64.5 64.2 a a Calculated with the equation: time = 106 (e0.06L_1). Table 4. Calculated and observed rates of water discharge in millions m3/day Observed Calculated mean discharge discharge K 150 3.74 3.89 K 183 1.00 1.07 Waghara Branch 0.70 0.89 Awamra Major 0.31 0.27 Abu Quta Major 0.42 0.48 Equation (4) may be written: dL = voeVLdt which on integration gives: 1t=-eP-I PVO (5) From the observations made in the present work the best fit is obtained when p=0.06 km-'. Distribution and operation of dispensers Previous calculations had shown that five dis- pensers would be required for the treatment. The main dispenser was positioned in the centre of the 579 M. A. AMIN ET AL. main canal at K 150 and four booster dispensers were placed at the side of the regulators at K 183, at the head of the Waghara Branch canal, at the head of Awamra Major, and at the head of Abu Quta Major (see Fig. 1). The main dispenser was switched on at 13 h 00 on 8 December and the delivery rate was adjusted to give a concentration of 0.078 mg of the active ingredient per litre of water. The booster dispensers were switched on when the trifenmorph front arrived and water samples above and below the boosters were analysed for trifen- morph at intervals to check that the required con- centrations were being produced. There was inevi- tably a considerable lag between the collection of samples and their analysis and the results were of little practical value to the dispenser operators. The delivery rate from each dispenser was therefore adjusted daily in accordance with variations in dis- charge and the calculated concentration required at each point. Trifenmorph concentrations The trifenmorph concentrations below each dis- penser were calculated to produce a concentration of 0.035 mg/litre at the head of the minor canal farthest from the dispenser. This value was chosen to give the best degree of snail control without undue overdosing. The actual concentrations measured at the heads of a number of minor canals are com- pared with the expected values in Table 5. There was very good agreement between the two sets of figures. Penetration of minor canals The rate of flow of water in minor canals is much less than that in major canals and it was not practi- cal to measure this by timing the passage of the trifenmorph front. Most of the water samples col- lected from these canals were used primarily to estimate the concentrations to which caged snails were exposed, but an attempt was made to use the analytical data to give an estimate of the half-life of trifenmorph in these canals. Since minor canals are fairly uniform in cross- section throughout their length, it is to be expected that the water velocity will decrease almost linearly along the length of the minor canal (see equation 1). Consequently, the time taken for trifenmorph to flow down a given portion of a minor canal should be independent of the length of the canal (see equation 3). Table 5. Comparison of trifenmorph concentrations (mg/litre) calculated and found at the heads of some minor canals Canal Found Calculated Difference Abu Gabal 0.033 0.035 +0.002 Ibraheim Amri 0.044 0.047 +0.003 Magerein 0.033 0.035 +0.002 Malier 0.045 0.045 0 Ugud Tail 0.035 0.035 0 Nuweila 0.047 0.054 +0.007 Ahmed Wad Yussef 0.041 0.050 +0.009 Tekeina 0.061 0.055 -0.006 Hassabala 0.042 0.052 +0.010 Sari 0.055 0.035 -0.020 Um Sineit 0.038 0.050 +0.012 If it is assumed that the trifenmorph entering the minor canal from the major canal is not diluted in the process (and this is probably true after the first day or so of application), the concentration of trifen- morph at a pointfdown the minor canal (Cf) will be: Cf=Cm exp(-kt) (6) where Cm is the concentration of trifenmorph in the major canal at the head of the minor canal and k is the first-order rate constant for the hydrolysis of trifenmorph. Substitution for t, the time taken to travel down the minor canal to the point f, in (6) from (3) gives: Cf = Cm exp k log (1-f) (7) or: lo(Cf= klo -log(1-f)\Cm a (8) Thus, if an equilibrium distribution of trifenmorph has been achieved within the minor canal, it should be possible to calculate the constant k/a and hence k from the analytical data. These data are somewhat inadequate for this pur- pose since the majority of samples were collected from points more than half-way down the canals. They do indicate, however, that beyond the half-way mark log Cf/Cm decays linearly with distance tra- 580 SNAIL CONTROL IN THE SUDAN Table 6. Estimated half-life of trifenmorph in some minor canals Canal Length (km) Half-life (h) k/a ± Standard error Azrag 19 83.6 0.417 ± 0.0132 (3 %) Hassabala 11 117 0.298 ± 0.0234 (7.8 %) Tekeina 7 75.3 0.463 ± 0.0113 (2.44 %) Sari 4 62.8 0.555 ± 0.0702 (12.56%) Ahmed Wad Yussef 10 86.3 0.404 ± 0.0438 (10.8 %) Nuweila 6.2 55.6 0.627 ± 0.0339 (5.4 %) Ibraheim Amri 8 66.3 0.526 ± 0.0187 (3.5 %) average 78.1 0.47 veIled and this appears to be independent of canal length. Values of k/a, calculated by regression analy- sis, are shown for 7 minor canals in Table 6. Despite the variability of these values, they suggest that the rate of hydrolysis of trifenmorph in minor canals is much less than in other, faster flowing canals. The average half-life is 78 h, which is very similar to the half-life determined in laboratory experiments (2). It has been suggested that the much lower half-life values found in major canals (25-30 h) may be a 0.15 0.10 ._~~~~~~ 0 result of the interaction of trifenmorph with sus- pended silt. Qualitative measurements of the silt content of the water from a number of canals were therefore made. Fig. 5 shows the variation of silt content with water velocity in the main and major canals and clearly indicates that there is a consider- ably lower silt content in the minor canals. This is not surprising, since both the main and the major canals are designed to carry water at a velocity that is too high to allow significant deposition of silt but Relative silt content Fig. 5. Variation of silt content with water velocity. 581 M. A. AMIN ET AL. Table 7. Mortality of B. truncatus in cages in 8 minor canals during trifenmorph application Exposure trifenmorph concentration (mg/litre) (days) 0.001- 0.006- 0.011- 0.016- 0.021- 0.026- 0.031-(daYs) 0 0.005 0.010 0.015 0.020 0.025 0.030 0.035 2 0/36 5/46 0/12 1/86 98/211 67/155 161/267 21/23(11 %) (1 %) (46%) (43% ) (60%) (91 %) 3 - - - - 21/37 36/49 42/46 27/31 (55%) (73%) (91 %) (87%) 4 0/25 - - 33/104 188/228 101/108 205/214 51/52(31 %) (32 %) (93 %) (96 %) (98%) 5 - 4/30 - 39/68 47/63 45/47 96/103 22/22(13 %) (57 %) (75 %) (96%) (93 %) (1 00 %) 6 - 7/42 21/25 65/80 99/118 109/109 311/316 t60/60 - (17%) (84%) (81 %) (84%) (100%) (99%) (100%) 7 - - - - 106/125 55/55 - -(85%) (1 00 %) 8 - 21/38 29/30 - - 22/22 57/57 - (55 %) (97 %) (1 00 %)) (100 %) Table 8. Counts of natural snail populations in some minor canals: total numbers of snails recovered in routine dip net samples Post-treatment Pre-treatment Minor canal after 3 weeks after 7 weeks B. pfeifferi B. truncatus B. pfeifferi B. truncatus B. pfeifferi B. truncatus Treated Karsh el Fil 2592 1 453 0 3 0 9 Abu Wafi 5581 5209 5 23 4 51 Tekeina 1 969 3806 0 0 0 24 Beibesh 1 009 991 0 4 0 26 El Sari 2 014 2102 18 47 29 67 El Semeir 743 844 6 12 16 65 UmWizin 1 818 1 850 0 140 0 168 Nuweila 3 606 4 215 57 57 158 85 Wad Gamil 820 639 0 0 3 13 Dakeen 9185 9 862 37 42 50 88 Hassaballah 11 876 16 476 0 35 0 202 Aftas 2 849 2 954 48 94 124 90 Ibraheim Amri 3 299 2 459 289 299 303 793 Total 47361 52860 460 756 687 1 681 Untreated Toba NDa ND ND ND 5784 1 646 Wad el Nil ND ND 1 431 226 2450 485 a ND = not done. 582 SNAIL CONTROL IN THE SUDAN 100 90 80 - 70 - 60 % kill Bulinus 50 truncatus 40 30_ 26 10 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 c.t.(mg/Q days) 0.11 0.12 0.13 0.14 0.15 Fig. 6. Variation of percentage kill of Bulinus truncatus with concentration x time product in mg/litre days. is low enough to prevent erosion of the canal banks. That silt is deposited in the minor canals is shown by the extensive dredging operations carried out in these canals. Biological results The data on caged snail mortality and trifenmorph concentration at each sampling station are shown in Table 7. Natural snail counts, both pre- and post- treatment, are given in Table 8. Data analysis and interpretation The measurements showed that there was a con- siderable delay in reaching a plateau of trifenmorph concentration near the tail ends of the minor canals. The major difficulty, therefore, was to determine a mean value for trifenmorph concentration for each cage exposure time. The logarithms of observed con- centrations were plotted against time for each 24-h period of exposure per cage and a mean 24-h ex- posure value was obtained per cage by means of midpoint values. These midpoint values were sum- med and divided by the number of days of exposure to give a mean value for concentration per cage exposure time. A plot of percentage snail mortality against con- centration x time (Ct, mg/litre x days) (Fig. 6) shows the general activity function. The predicted Ct values for 100%4 mortality were obtained by extrapolation from the data that gave less than 1OO° and greater than zero mortality. The observed minimum values that gave 100%o mortality were also calculated for comparison and both sets of figures are given in Table 9. The predicted values for two canals (El Sari and El Sayr) were not calculated because the data points were very few and suggested a decrease in mortality with an increase in Ct value. In general, despite the approximate methods used for determining Ct, the predicted values compared favourably with the observed values. The results from one canal, the Ibraheim Amri, lay outside the common trend. The snail mortality in this canal was lower than expected from the measured concen- trations and the natural snail samples also showed A0 X 0* A~~~ x x 0 A 0 Minor Canals O Ahmed wad yussef x A x Tekeina 0 0 Sayr A Azrag 0 U* Nuweila O Sari 6 Hassabala 0 0.16 0.17 583 n M. A. AMIN ET AL. Table 9. Concentration x time mortality of B. truncatus (Ct) values for 100% Predicted Ct Mean observed Minor canal (mg/litre days) minimum Ct Azrag 0.141 0.159 Nuweila 0.116 0.088 Ahmed Wad Ussef 0.125 0.126 Tekeina 0.164 Hassabala 0.126 0.079 Ibraheim Amri 1.200 Sari - 0.104 Sayr - 0.092 mean value 0.110 this canal to have the lowest mortality. At present it is not possible to offer a satisfactory explanation for this anomaly but the result was so atypical that it was disregarded for the purpose of determining a representative value for Ct. Since the observed values for Ct showed quite good agreement with those predicted (Table 9), it seemed reasonable to take a mean value for Ct from the former. With log transformation a mean value of 0.12 mg/litre days was calculated. With this value, and assuming Ct to be constant with time, the following concentration regimes should give 100% mortality of B. truncatus: 0.020 mg/litre for 6 days; 0.015 mg/litre for 8 days; or 0.010 mg/litre for 12 days. The samples of natural snail populations (Table 8) indicated 99% snail control 4 weeks after treatment. The application was for 7.5 days and a trifenmorph concentration of about 0.015 mg/litre penetrated to within 600-1200 m of the tail ends of all the canals studied. The Ct value of 0.12 mg/litre days, there- fore, showed good agreement with the observed effect and confirmed the correct choice of applied dose. It was apparent that the extent of penetration of trifenmorph down the minor canals depended on the number of abu eshreens that were being used towards the tail ends. Observations of the use of these were made during application but attempts to relate these to the degree of trifenmorph penetration were unsuccessful. It seemed to be impossible to predict the pattern of water use in any given minor canal over a period as short as 7-8 days. Conse- quently, the length of canal that needed to be hand-sprayed could be determined only by inspec- tion in the field. The procedure adopted during the present study was to spray the water from a point 300 m above the last open abu eshreen to the tail end. This proved to be a very successful method of killing the snails in this region. CONCLUSIONS It was shown that, in order to get uniform mixing across the main canal below K 150, trifenmorph must be added to the highly turbulent water passing through the central regulator. In future applications it would be wise to apply the formulation from all of the dispensers to the centre rather than to the side of the canal. When the treatment area is increased and even greater volumes of water have to be treated, it may be necessary to apply trifen- morph from a number of outlet pipes spaced across the canal at the main dispensing position. Predicted flow times and water discharges agreed well with those measured in the preliminary 6-h treatment. Operating procedures can now be worked out in advance of any application so that the dis- penser operators need perform only very simple calculations. These would involve adjusting the deliv- ery rate of a particular dispenser as the water dis- charge varied in order to maintain the required con- centration of trifenmorph in the water. It was found that the half-life of trifenmorph in the main and major canals was 31.5 h. This agrees well with previous determinations (1). In the minor canals, the half-life was approximately 78 h, which is in agreement with that determined in laboratory experi- ments (2). The difference in half-life between the main, major, and minor canals appears to be cor- related with the silt content of these waters, although the detailed mechanism of the processes involved is not yet fully defined. Although it is not possible to derive from the results the minimum concentration of trifenmorph required to give 100% mortality of B. truncatus over 7.5 days, a concentration of 0.015 mg/litre was adequate for this. It is therefore possible that a lower concentration would be equally satisfactory. The results similarly showed that a concentration x time product of 0.12 mg/litre days is adequate for 100% control of this species but, of course, extra- polation from this to other combinations of concen- trations and time may not be valid. A very high level of control of the natural snail 584 SNAIL CONTROL IN THE SUDAN 585 population was achieved in this trial. It is possible that slightly lower concentrations of trifenmorph would have been satisfactory and it would clearly be worth investigating this point. This could be done in future by reducing the concentration produced at only one booster dispenser and carefully measur- ing the snail control in that area. This was the first time that multiple dispensers had been used for trifenmorph application and few problems were encountered. There is no reason to expect that a future increase in the size of the treat- ment area would reveal any unforeseen problems, although such an increase would have to be preceded by a careful examination of the distributive canals in the new areas. The overall scheme now available for treating the major and minor canals in this irri- gation system is easy to handle, but for complete control of all transmission sites it will be necessary to spray the tail ends of minor canals and some abu eshreens by hand. RESUME ESSAI DE LUTTE A GRANDE ECHELLE CONTRE LES MOLLUSQUES A L'AIDE DE TRIFENMORPHE DANS LE RESEAU D'IRRIGATION DE LA GEZIREH, SOUDAN Un essai pratique a grande echelle a ete effectue au cours de decembre 1973 afin d'evaluer l'effet du trifen- morphe sur Bulinus truncatus et Biomphalaria pfeifferi dans environ 159 000 hectares du reseau d'irrigation de la Gezireh au Soudan. La formulation commerciale uti- lisee (Frescon) est un concentre emulsifiable contenant 16,5% de trifenmorphe. Cinq distributeurs ont ete utilises pour introduire de maniere continue le produit commer- cial dans l'eau pendant 7,5 jours; on a ainsi utilise 18 121 litres pour traiter 28,4 millions m3 d'eau. En outre, chaque canal tertiaire a ete pulverise A la main A partir de l'extremite d'aval jusqu'A 300 metres en amont de la derniere buse de prise d'eau ouverte; 1360 litres de la formulation commerciale ont et utilises pour cette operation. Une concentration minimale de 0,035 mg de trifen- morphe par litre d'eau a e produite a la tete de chaque canal tertiaire. L'utilisation de mollusques enfermes dans des cages a montre qu'une concentration aussi faible que 0,015 mg/l suffisait pour produire 100% de mortalite chez B. truncatus en 7,5 jours; cela correspond a une valeur de 0,12 pour le produit (#concentration (mg/I) x temps d'exposition (jours) >. REFERENCES 1. AMIN, M. A. & FENWICK, A. Annuals of tropical medicine and parasitology, 71 (1977). 2. BEYNON, K. I. ET AL. Bulletin of the World Health Organization, 37: 53 (1967). 3. BEYNON, K. I. & THOMAS, G. R. Bulletin of the World Health Organization, 37: 47 (1967). 4. BEYNON, K. I. & WRIGHT, A. N. Pesticide science, 6: 515 (1975). 5. SHELL CHEmiCAL Co. Frescon-a molluscicide for the better control of schistosomiasis. London, Shell Print- ing, 1974, p. 12. 6. FARBROTHER, H. G. Annual report, Agronomy and Crop Physiology Section, Gezira Research Station, Sudan, 1971.