CONTROL OF TSETSE AND TRYPANOSOMOSIS IN THE SOUTHERN RIFT VALLEY OF ETHIOPIA: EVALUATION OF DELTAMETHRIN APPLICATIONS

LA LUTTE CONTRE LES TSETSE ET LA TRYPANOSOMOSE DANS LA PARTIE AUSTRALE DE LA VALLEE DU RIFT DE L’ETHIOPIE: EVALUATION DES APPLICATIONS DE DELTAMETHRINE

Jemere Bekele1 & Getachew Abebe2

1alemaya University, Faculty of Veterinary Medicine P.O. Box 38, Dire Dawa,
Ethiopia jemereb@yahoo.com
2Addis Ababa University, Faculty of Veterinary Medicine, P.O. Box 34, Debre Zeit,
Ethiopia gkibret@yahoo.com 

            On a procédé à un essai pour évaluer l’efficacité de la Deltamétrine (écrans à appât olfactif imprégnés à 0,4% et formulation en pour-on à 1% appliquée aux animaux) et  comparer la rentabilité correspondante des deux stratégies dans deux sites choisis (Universal Transverse Mercator Grids) de dimensions 10 x 10 km dans la zone du Projet austral d’éradication des tsétsé (STEP), au sud de la Vallée du Rift en Ethiopie. Les quadrillages sélectionnés étaient H3 et G5, désignés respectivement comme sites d’étude I et II. L’essai a duré de fin  septembre 2003 à début avril 2004, et a consisté en une enquête de référence, une intervention à l’insecticide (Deltaméthrine à 0,4 %), l’installation d’écrans imprégnés à appât olfactif sur le site I (Quadrillage H3), l’application topique de Deltaméthrine à 1 % sur les bovins sur le site II (Quadrillage G5) et en une surveillance mensuelle de l’incidence de la maladie et de la densité apparente des tsétsé. Après le déploiement de 460 écrans (imprégnés de Deltaméthrine à 0,4 % et à appât olfactif) à une densité de 4 écrans/km² sur le site d’essai I, l’abondance relative de la population de tsétsé (G. pallidipes) a baissé de son niveau pré-intervention de 1,35 mouches/piège/jour à 0,05 mouche/piège/jour pendant la dernière période de piégeage en avril 2004, soit un taux global de réduction de 88,88 %. Le coût des interventions de routine pour la formulation en pour-on offrait un ratio progressif de rentabilité allant de 34 762,54 $ EU à 102 927,94 $ EU d’efficacité par unité, et l’utilisation des écrans imprégnés à appât olfactif offrait un niveau de rentabilité allant de 43 184,93 $ EU à 942 652,00 $ EU d’efficacité par unité pour la lutte contre les tsétsé et la réduction de la maladie  (la trypanosomose) qui s’ensuit.

            A trial to evaluate the efficacy of Deltamethrin and comparing the corresponding cost-effectiveness of both strategies was conducted in two selected 10x10km Universal Transverse Mercator Grids of the Southern Tsetse Eradication Project (STEP) area in the Southern Rift Valley of Ethiopia. The Grids selected were H3 & G5 designated as study Sites I & II respectively. The trial was underway from late September 2003 to early April 2004. The accomplishment of the trial included baseline survey, intervention with insecticide (Deltamethrin 0.4%) impregnated odour-baited targets at Site I (Grid H3) and Deltamethrin 1% ‘pour-on’ application to cattle at Site II (Grid G5) and monthly monitoring of the incidence of disease and apparent density of tsetse fly. Following the deployment of 460 targets (0.4% Deltamethrin impregnated and odour-baited) at a density of 4 targets/km2 at the trial Site I, the relative abundance of tsetse fly (G. pallidipes) population declined from a pre-intervention catch of 1.35 flies/trap/day to 0.05 flies/trap/day during the final trapping time in April, 2004 with about 88.88% overall reduction achieved. The routine intervention cost for Pour-on formulation offered   Incremental Cost-Effectiveness Ratio (ICER) ranging from 34,762.54 – 102,927.94 USD per unit effectiveness and the use of impregnated odour-baited targets ranging from 43,184.93 – 942,652.00 USD per unit effectiveness for tsetse control and the associated disease (trypanosomosis) reduction.

Introduction

            Tsetse-transmitted trypanosomosis is the main constraint to livestock production on the continent of Africa, preventing full utilization of land to feed the rapidly increasing human population (Murray et al., 1991). It affects 37 sub-Saharan countries; extending over 10 million km2 of land (Erkelens et al., 2000).  In Ethiopia, the potential area of tsetse infestation has been estimated at 135,000 – 220,000 km2, based on maximum dispersals up to 1700 m above sea level and 2000 m (Slingenbergh, 1992 a & b). Five species of tsetse and four tsetse-borne trypanosome species are found in Ethiopia confined to Southern and South-western regions of the country (Langridge, 1976).

            About 25,000 km2 areas that is agriculturally suitable and potential land in the Southern Rift Valley of Ethiopia are infested by tsetse flies. In this part of the country, almost all domestic animals in and adjacent areas of the valley are at risk of acquiring the disease at any time (Vreysen et al., 1999). Thus, priority for tsetse control has been given by Ethiopian Government and the ‘Southern Tsetse Eradication Project’ STEP) was launched in 1997. The programme is jointly supported by the Ethiopian Federal Government and Southern Nations, Nationalities and People Regional State and the International Atomic Energy Agency and is coordinated by the Ethiopian Science and Technology Commission. A phased dynamic approach with the Sterile Insect Technique (SIT) as the final eradication component is in use, as this area is regarded sufficiently isolated from other tsetse belts (Muturi, et al., 2000). An area-wide integrated tsetse eradication programme with the eventual aim to eliminate by sequential releases of male tsetse flies sterilized by ionizing radiation, the project is in operation engaged in suppression of tsetse fly in areas of high fly population densities by the use of conventional tsetse control methods. Emphasis must be put on acceptable levels of effective suppression of disease and tsetse with methods that are appropriate, efficacious and cost-effective. Therefore, this work was planned to conduct a comparative evaluation of the efficacies of Deltamethrin application as ‘pour-on’ or insecticide impregnated targets as currently employed available technique for the tsetse and trypanosomosis control and then determine the best method for further use. And also, to assess and determine the cost-effectiveness of both strategies which might aid in suppressing the wild fly in the field in the shortest possible time before the release of sterile male flies.

Material and methods

The Study Area

            The first block of the STEP area is situated in the Southern Rift Valley and most of the area lies within the administrative boundary of the Southern Nation Nationalities and Peoples Region. The trial was carried out in an approximate area of about 10x10km (100km2) grids each, selected at two sites located in two districts (Woredas) of Wolaita Zone of the Southern Nations Nationalities and Peoples’ State within the Southern Tsetse Eradication Project (STEP) area. The peasant associations (‘Pa’s) contained within the specified grids are Tora-Sedebo & part of Adecha considered as Site I is located in Damot Woyde district at about 36 kilometers eastward from Soddo close to Bilate river. The other, Abela-Mareka ‘Pa’ of Humbo Woreda (Site II) some 40 kilometers South of Wolaita Soddo town on the way from Soddo to Arba Minch in the area near the rift valley great lake (Lake Abaya). The coordinates of the area for Site I (Tora-Sedebo& Adecha) is 6047’50”N-6052’30”N and 37055’00”E-38000’00’’E or Grid H3 whereas that of Abela-Mareka is 6032’00”-6037’00”N and 37049’44”-37055’00”E or Grid G5. According to the information, supplied by the respective agricultural offices, cattle population for Site I was estimated at 1754 and that of Site II was nearly 1500. Livestock husbandry is traditional type with mixed farming. All cattle encountered were East-African zebu type.

Study Methodology

Sampling Method

            Random sampling was used to examine animals for baseline parasitological survey. Abundance of tsetse flies, frequency of trypanosomosis cases, altitude category, vegetation type, cattle grazing land and watering point were taken as criteria to determine the ideal habitat of tsetse flies for study site selection.

Pre-Intervention baseline data collection

Entomological data

            Tsetse flies were sampled by deploying about 71 NG2U traps (35 in Site I & 36 in Site II) baited with three-week-old bovine urine and acetone in two different dispensing bottles. Traps were set at approximate intervals of 200-250 meters. All trap positions were geo-referenced (by use of hand-held GPS, Garmin 48), altitude and vegetation type recorded. It was attempted to include different vegetation types like bushland (BUL), wooded grassland (WGL), and cultivated land (CUL) for trapping. Collection of trapped flies took place 72 hours after deployment.

Parasitological and haematological data

            A total of 323 cattle above one year old were randomly sampled and examined. Blood samples were collected from marginal ear veins using micro-haematocrit capillary tube and sealed on one side with cristaseal (Hawskely Ltd.). The capillary tube was then transferred to a haematocrit centrifuge and spun for 5 minutes at 1200 revolution per minute. The centrifuged capillary tube was measured for packed cell volume (PCV) values on the haematocrit reader. It was then cut 1mm below the Buffy coat and the contents of the tube expressed on to a slide, mixed and covered with a 22x22 mm cover slip. This slide was then examined under x40 objective using phase contrast or dark field microscopy to examine for the presence /absence of motile trypanosomes (Murray et al. 1977). Animals found positive for trypanosomes as well as showing the clinical signs were treated with curative dose of Berenil® (Diminazene aceturate) at dose rate of 3.5 mg/kg body weight intramuscularly. During this time, sentinel herd of cattle (95 in Site I and 75 in Site II) were selected and ear-tagged in each site.

Intervention

Insecticide impregnated odour-baited targets (Site I)

            A total of 460 impregnated (0.4% Dltamethrin) targets were deployed by the community participation assisted by technical staff. The targets were deployed at a density of 4 targets /km2 where they were placed at approximately 250 m apart. Acetone was the odour used in this case. Replenishment of odours for the targets was done three months after initial deployment.

Pour-on application (Site II)

            In this case, nearly one-third of the cattle in the area excluding calf below the age of 1 year were mass treated with Deltamethrin 1% W/V pour-on ready for use formulation (Appropriate Applications Ltd (USA)) at a dose rate of 10ml/100kg body weight. Pour-on application was subsequently repeated monthly throughout the whole monitoring period.

Intervention monitoring

            Entomological monitoring was conducted every month by deploying traps (NG2G) in previous sites of catch using odours (acetone and bovine urine). Parasitological and haematological examination was subsequently conducted every month.

Cost-Effectiveness of the trial

            Cost-effectiveness was assessed based on the information obtained during the course of the trial particularly with the intervention cost.

Statistical analysis

            All sorts of entomological and parasitological raw data were stored into a computer program Microsoft Excel Spreadsheet, edited and then imported into the statistical software called Intercooled Stata version 7.0 for Windows 98/95/2000/NT (Stata Corporation, Texas, USA) for different ways of analysis. The pre-intervention entomological data (trap catches) were compared by Student’s t-test while parasitological data by chi-square test. Intervention results in the relative abundance of fly population were assessed for each study site and vegetation type using descriptive statistics as well as Student’s t-test. The data were finally presented by calculating the de-transformed means (Geometric means). Calculating the incidence rates of each monitoring month and calculating the rate of reduction in the disease status by descriptive statistics assessed the situation of the disease during intervention and then McNemar’s test was justified.  In addition to this, the mean PCV value of sentinel animals before and after intervention was evaluated by the use of Student’s t-test for related samples. Only new cases were considered when calculating the incidence rate.

            Economic assessment based on incremental cost-effectiveness ratio (ICER) comparison between the two strategies was carried out for the interventions. It was determined by calculating the ratio of difference in the cost of two options to the difference obtained in effectiveness (change in disease magnitude) between two options. This could either be the difference before and after intervention (i.e. ‘intervention’ compared to ‘doing nothing’) or the difference between two intervention strategies being compared which help to discuss a shift from one strategy to another (the most efficient given CER ranking). 

                  ∆C = Ct5 – Ct0, where ∆C is the change in cost expense (extra cost) brought about by intervention;

Ct5 is the total cost expenditure due to intervention and

Ct0 is the cost expense before intervention (doing nothing).

On the other hand, change in effectiveness criteria was obtained mathematically by:

                  ∆ E = Et5 – Et0, where

                  E stands for effectiveness represented by prevalence of disease and t for time (before or after)

                  Et0 is the prevalence of disease before intervention (doing nothing),

                  Et5 is the prevalence of disease after intervention and

                  ∆E is the change in the prevalence of disease brought about by intervention.

  Finally, ICER = ∆C/∆E, where, ICER is incremental cost-effectiveness ratio.           

Results

Pre-intervention

Entomology

            Only G. pallidipes was the species of tsetse detected during the study. The magnitude of the catches did not show statistically significant difference (P > 0.05) between the two Sites of study. Of the total 240 flies caught during the pre-intervention survey, 88 (36.8%) were males and the rest 152 (63.2%) were females showing significantly higher picture than a 50:50 distribution (P< 0.01) (Table 1).

tAble 1. entomological baseline survey result recorded in both study sites.

               R.A: relative abundance of tsetse fly /trap/day; CI: 95% confidence interval

Parasitology

            Out of 171 cattle sampled and examined in study Site I, 39 were positive for three species of trypanosomes accounting for a prevalence rate of 23% whereas 32 animals out of 152 examined were positive accounting for a prevalence rate of 21% in study Site II. Regarding to the relative abundance of the species of trypanosomes detected Trypanosoma congolence had a higher frequency of 49 accounting for 70% of all infections detected keeping the highest proportion in both sites and followed by T. vivax (20%), T. brucei (8%) and mixed infections (2%). The overall parasitaemic status of animals examined from both study sites was evaluated by the use of chi-square test and the difference in the proportion of infection between them was found to be statistically insignificant (P>0.05).

Haematology

            In this study, a PCV measurement of 25% was regarded as a threshold value. Chi-square test was applied to evaluate the presence of association between disease and mean PCV values. For study Site I, a highly significant association was observed between mean PCV values and occurrence of parasitaemia (P<0.01). In the same manner, a highly significant association of PCV result to the presence of infection was noticed in the similar treatment conducted for study Site II (P < 0.01). The individual animal level PCV values recorded were further subjected for statistical treatments using one way Analysis of Variance (ANOVA) and as result aparasitaemic animals had significantly (P<0.01) higher mean PCV values (25.65%) than parasitaemic animals (18.8%). 

Intervention

Entomological results

            Student’s t-test was applied to evaluate the significance of decline in the catch of tsetse flies in both sites during last monitoring as compared to the pre-intervention catch. In both sites a statistically significant (i.e. Site I: P < 0.01 and Site II: P< 0.01) reduction was revealed (Fig.1).  However, the difference between the results recorded at final monitoring of both cases was insignificant (P>0.05).    

Fig.1.  Changes in tsetse fly catches record during Pre-intervention and after intervention.

Parasitological results

            The monthly monitoring results (cumulative incidence rate of trypanosome infection) in cattle of both sites are presented in Table 2 (A&B) below. Result assessment of the status of parasitaemia during the intervention phase indicated that there was a slow reduction pattern in the incidence rate of trypanosomosis among the sentinel animals in both study places (Site I & II). Evaluation of the magnitude of decline in the risk of infection during intervention revealed that 83.25% reduction was achieved in the incidence rate of trypanosome infection in Site I and 90.5 % reduction in Site II. The corresponding prevalence rates estimated for the first and last monitoring at Site I were 11% & 9% while the same results for Site II were 10% & 4.75% respectively.

            Therefore, this indicates that decline of prevalence is significantly associated with the progress of intervention (P< 0.01) and the final monitoring result P< 0.01) in Site I (Fig. 2A).  The same type of assessment of the results recorded at Site II revealed that there is a highly significant relationship between the decline of the status of parasitaemia and progress of intervention i.e. pre-intervention to first monitoring P<0.01) and pre-intervention to the last monitoring P<0.01) (Fig.2B).

            From the result, the pre-intervention   trypanosomosis prevalence of 21% had dropped to 4.75% at post intervention with a 77.4% overall reduction attained in study Site II. Moreover, this reduction   was evident as compared to first monitoring visit result (10%) which had almost declined by half. While, the same treatment of the results from study Site I demonstrated an apparent reduction in the proportion of animals becoming parasitaemic i.e. 23% prevalence rate of the pre-intervention study declined to 9% at post intervention monitoring (60% overall reduction).

Table 2. sequential parasitological monitoring result of sentinel cattle in sites i (a) and  ii (b)

(A)

(B)

(Fig. 2A) Pattern of reduction in the incidence of disease during intervention at Site I (A)

(Fig. 2B)   Pattern of reduction in the incidence of disease during intervention at Site II (B)

Haematological results

            Application of paired t-test to assess the magnitude of change occurred in the mean PCV value of animals examined in site I showed an obvious difference (P<0.01) with a statistical significance, and Site II had also proved a highly significant increment in the monthly mean PCV value (P<0.01). The increment in the mean PCV value in Site I (Fig.3A) followed nearly continuous pattern, while the mean PCV values recorded in Site II (Fig.3B) showed continuous increment only until the third month of monitoring visit and thereafter maintained a slightly stable situation during the rest of  the visits as compared to Site I.

(Fig. 3A)

(Fig, 3B)

Fig.3.Comparison of mean PCV values during the intervention period in Site I (A) &II (B)

Table 3. Costs breakdown for strategy in Site I (Government’s perspective)

(†): Scale of Ethiopian Government ,  (±): Wage rate Birr 5.00 at Study Sites

Cost-effectiveness analysis for the intervention

Effectiveness

Site I: The prevalence of trypanosomosis recorded before the intervention was 23% (0.23).  The corresponding prevalence estimated from incidence rate during the final monitoring at this site was 9% (0.09). Then the difference is calculated with the confidence interval as follows.

                     ∆E1 = Et5-Et0

                           = 0.09-0.23

                           = 0.14; 95% CI (0.005-0.27)

(Cost breakdown : Table 3)

Site II: The prevalence of trypanosomosis recorded before the intervention is 21% (0.21).  The corresponding prevalence estimated from incidence rate during the final monitoring in this site is 4.8% (0.048). Then the difference is calculated with the confidence interval as follows.

                     ∆E2 = Et5-Et0,

                            = 0.048-0.21

                            = 0.162; 95%CI (0.034-0.29)

(Cost breakdown : Table 4)

Incremental Cost-Effectiveness Ratio (ICER) determination

            This applies to both strategies for comparison based on Sensitivity analysis, which depends on the minimum and maximum value calculated. Results signify in the uncertainty interval that, the ICER lies between 43,184.93 – 942,652.00 USD for Site I (impregnated targets) and 34,762.54-102,927.94 USD for Site II (pour-on) considering the cost of intervention as an overall effect. The average ICER for Site II was 24,332.1USD per unit effectiveness while that of Site I was 49,478.07 USD per unit effectiveness. The uncertainty interval obtained shows an overlapping result and thus the difference between the two strategies is insignificant.

Discussion 

            In the present study, the apparent densities of tsetse flies caught were 1.35flies/trap/day in Site I and 0.91flies/trap/day in Site II. These findings correspond well to the results obtained during a study by Muturi et al. (2000) where they had caught about 1.4 flies/trap/day in the Southern rift valley of Ethiopia and also with the findings of Leak et al. (1993) who for over 50 months observation period reported 1.42 flies/trap/day. Muturi et al. (2000) reported that G. pallidipes was caught in all lowland areas under 1600 m above sea level. They also stated that this species was widespread being detected in all types of vegetation, the highest relative density being detected in the bush land vegetation.  According to Leak (1999) vegetation is vital for providing shade and maintains a suitable microclimate for tsetse as well as a habitat for their hosts. Higher catches of flies during this work at bush land class of vegetation could be related to the wide dispersion of flies to this area as a result of the transition from rainy season to early spring.  Catching of female flies showed a greater degree of deviation from the expected 50:50 ratio with a significant relationship.  Females accounted for 63.2% catch during this study. The result is similar to the study reports of Muturi et al.(2000) indicating  71% and Vreysen et al. (1999) reported the catch of female G. pallidipes to account for more than 62% in the same area under concern.  Leak (1999) described this as in unbiased sample female would comprise between 70-80% of an average population. Phelps and Lovemore (1994) associated such higher catches of female G. pallidipes to be attributable to their longer life span (average of 8 weeks) than males living of about 4 weeks.

Table 4. Costs breakdown for strategy in Site II (Government’s perspective)

            The present result agrees with the previously estimated 20-40% prevalence range for the project area and is similar to the prevalence (25.9%) reported by Muturi et al. (2000) in the low altitude strata (<1600m above sea level) as the current study sites were in a similarly lower altitude category.   The detection of a high proportion of T. congolence in both sites of this study (70%) appear to be consistent with previous report  by Abebe and Jobre (1996) in South-west Ethiopia where they reported T. congolence (59%) and T. vivax (31%) infection, Muturi et al. (2000) reported 66.86% T. congolence and 20.57% T. vivax infection in the same project area in the Southern rift valley and also Van den Bossche et al. (2004) indicated that most of the trypanosome infections (90.9%) are due to T. congolence and the remainder to T . vivax infection.

            Measurements of PCV value for each animal sampled in the pre-intervention survey in both sites was analysed. Marked differences noticed depending on the infection status of animals. Pparasitaemic animals had generally lower mean PCV value than the corresponding aparasitaemic ones. About 94.87% of the parasitaemic animals in Site I and 81.25% in Site II had their PCV below 25%. Again parasitaemic animals of Site I had significantly lower mean PCV than the same class of animals in Site II. This is most likely attributable to either the difference in constant challenge of animals by tsetse, management factors associated with widespread use of trypanocidal drugs, as no marked difference in the overall mean PCV of animals (i.e. both parasitaemic and aparasitaemic) between the two sites; however majority of the animals had relatively low PCV value in both cases. The appearance of parasitologically negative animals with PCV values of less than the threshold value set (25%) may be due to the inadequacy of the detection method used (Murray et al., 1977) or delayed recovery of the anaemic situation after current treatment with trypanocidal drugs. And also the occurrence of positive animals with PCV of greater than 25% might be are recent infections. Trypanosome infection and mean PCV values obtained in the present study in the parasitaemic animals were found to be highly associated. It was generally accepted that mean PCV value is affected by many factors other than trypanosomosis. However, these factors are likely to affect both trypanosomosis positive and negative animals (Van den Bossche and Rowlands, 2001).

            During the intervention study, the population of tsetse flies decreased over the trial period. The apparent density of tsetse flies caught during last monitoring time was particularly low as compared to the pre-intervention catch result. The overall reduction in the tsetse population achieved in Site I is 88.8% attributable to the use of effective odours i.e. bovine urine and acetone with frequent follow up. Consequently, subsequent replacements of lost and damaged targets and replenishment of odours was accomplished promptly.

            On top of these, the relative abundance of tsetse flies caught at every month visit in Site II dropped to nil starting at 4th month of trapping. The efficacy determined in this case was about 94.88 % reduction in the overall fly population attributable to the pour-on intervention. This substantiates the relative superiority of Deltamethrin 1% pour-on when used to control tsetse flies obtaining a good result with in a short period of time as compared to 0.4% Deltamethrin impregnated odour-baited targets. The efficacy of insecticide-treated cattle in controlling trypanosomosis hinge on the importance of cattle in the diet of the tsetse, the proportion of the total cattle population treated at regular intervals and the invasion pressure from tsetse. According to Van den Bossche et al. (2004), the importance of cattle in the diet of tsetse varies spatially but is usually high in areas where people and cattle have encroached into a tsetse-infested game area and where, because of cultivation, the density of large game animals has subsequently declined.  This phenomenon is definitely true in the current study area.

            Tsetse intervention activity and subsequent five month monitoring showed that the incidence rate of trypanosomosis in sentinel animals of Site I declined from 10.75% to 1.8% during 1st and 5th monitoring visits respectively. The result of this finding shows that there was a reduction of about 83.26% in the incidence of trypanosomosis and the trend of reduction was slow. Such slow rate of reduction in the incidence of disease could partly be due the time taken for the intervention being shorter than the average required, as it was seen comparatively with the experience in other places. Study conducted in Zambia by Lumamba et al. (1997) and Zimbabwe by Chamisa and Mweempwa (1997) proved that tsetse control using insecticide impregnated odour-baited targets could take not less than 9 months to show a marked drop in the disease incidence as well as achieve scanty fly catch result. Additionally, the sensitivity of the diagnostic method used under such a field condition, could also play some role. However, the final result obtained (i.e. drop of incidence rate to 1.8%) should not be held in low esteem as compared to the pre-intervention time point prevalence obtained in Site I (i.e. 23%). And therefore, a 60 % overall reduction in the parasitaemic status was achieved attributable to the intervention underwent. Livestock owners of the site were able to appreciate the change and expressed their observations of reduction in the disease recurrence in comparison to the previous worsening situations to their animals. And they were also able to observe the subsequent avert from the fly harassment to the animals while grazing. This was because they found that their animals were able to graze the whole day after intervention that was previously regarded as impossible phenomenon.

            No matter how the status of parasitaemia progress, gradual improvement was observed in the monthly mean PCV value of the sentinel animals in Site I. The PCV improvement showed a strongly positive correlation with intervention time while negatively correlated with mean tsetse catch and monthly incidence rate of trypanosomosis.   The mean PCV values  followed a constant increment throughout the intervention time. The intervention parasitological status in study Site II also resulted in a subsequent decline of trypanosome incidence falling to 0.95% at final monitoring visit. An overall reduction in the incidence rate was 90.5%. Together with the drop of parasitaemia, a marked increase with relatively stable pattern of mean PCV was noticed. This result agrees with the report of Van den Bossche et al. (2004) where the monthly incidence of trypanosomosis was negatively correlated with the time elapsed since the start of Cyfluthrin applications in the control of G. m. morsitans in two districts of the Eastern province of Zambia. The overall reduction in the corresponding prevalence was estimated by computing (Thrusfield, 1995) from the incidence rate result and compared with the pre-intervention prevalence. This has showed a 77.4% reduction. This indicates a better success as compared to that reported by Mulatu et al. (1997) from tsetse control campaign using an insecticidal Cypermethrin ‘pour-on’ application to village zebu cattle in Ghibe in South west Ethiopia where trypanosomosis in adult cattle reduced from 41% to 16% (a reduction of 61%) and the number of curative trypanocidal treatments per animals  reduced by 50% despite very high levels of drug resistance detected in the area. Again this overall prevalence reduction result achieved is far higher than the result obtained in Site I i.e. 60% and signifies the relatively better effectiveness attained. 

            The significant increment noticed in the mean PCV value of sentinel cattle could be attributed to the reduction in the incidence of trypanosomosis due the intervention and the subsequent improvement in the health status of the sentinel animals.  This result agrees with the work of Leak et al. (1995) in which the mean PCV of cattle rose from a mean of 23.8% of pre-control period to 26.8% following the start of the tsetse control trial with Cypermethrin and monthly mean PCV values were negatively correlated with monthly trypanosome incidence. Similar result was obtained by Van den Bossche et al. (2004) in which the increase in the average PCV of the herd is also best explained by a decline in the incidence of trypanosomosis although consistently high average PCV were only observed after the incidence of trypanosomosis was reduced to zero in their work. A similar result was reached from third   monitoring onwards in the current study. Though an immediate effect on the incidence of trypanosome infections were not noticed, the use of Deltamethrin pour-on in the current study seem to have resulted in an immediate improvement in the monthly mean PCV of the sentinel cattle in the study Site II.  Such increases in the herd mean PCV was also observed by Van den Bossche et al. (2004).

            In general, it was accepted that seasonal fluctuations might occur. Regarding this phenomenon, the study was conducted in the period during which trypanosomosis is normally anticipated to occur, associated with poor pasture in the surrounding and thus animals being forced to move in to bushes and riverine forests where they  get in contact with tsetse flies and acquire infection.  This was a usual incident noticed in both of the study areas. Furthermore, seasonal reduction in the tsetse fly population does not assume a rapid fall phenomenon, but rather shows gradual tendency to reduce. In spite of this fact, a sharp drop in the tsetse fly population density to nil (Site II) and a comparable decline (Site I) together with the resultant falling in the disease status was observed from both interventions.  To this effect, a markedly significant improvement in the PCV of sentinel animals in both study sites was noticed and this would be most likely attributed to the interventions they underwent.  

            The Incremental Cost-Effectiveness Ratio (ICER) obtained (24,332.1USD per unit effectiveness) in Site II was lower than the corresponding ICER calculated for Site I (49,478.07 USD per unit effectiveness) on average market cost basis disbursed for intervention. However, there appeared an overlapping in the uncertainty intervals of the two scenarios analysed, which implies lack of justifiable difference between both strategies based on the present intervention conducted.

            The cost-effectiveness results obtained in the present trial offered a good evidence that tsetse and trypanosomosis suppression can be achieved with a relatively lower cost than the average costs estimated by different workers in other African countries. Sustainability of such work results conducted within a portion of the project area (STEP) depends on the continuity of the activity of the project itself. Of course, the project is now carrying out activities of massive suppression of tsetse fly by applying pour-on and impregnated targets with the involvement of community for labour- intensive activities like deploying of targets. Therefore, the risk of re-invasion of tsetse-suppressed sites is likely to be minimal. 

            Based on this present study, the application of both insecticide impregnated (0.4% Deltamethrin) odour-baited targets at a density of 4 targets /km2 and pour-on formulation (Deltamethrin 1% at a rate of 1ml/ 10kg body weight to cattle) reduced the relative abundance of tsetse fly population and trypanosomosis incidence with the resultant improvement in the overall mean PCV value of sentinel animals. The application of pour-on  resulted in the decline of tsetse fly catch (relative density) and in the incidence of trypanosomosis faster than the use of insecticide impregnated odour-baited targets and therefore proved more efficacious. The incremental cost-effectiveness ratio (24,332.1 USD per unit effectiveness) of Site II appeared lower than the corresponding ICER (49,478.07 USD per unit effectiveness) of Site I based on average calculation. However, the uncertainty interval rules out the overall effect of this result given significance consideration.

            The trial confirmed that the use of the pour-on formulation of Deltamethrin was more efficacious but did not prove more cost-effective than odour baited target. Furthermore, it was found easy to apply, required less labour and highly appreciated by user community. However, this advantage would be maintained if and only if a sufficient number of   cattle were allowed within the environment to be cleared of tsetse by applying pour-on. Otherwise, it would be difficult to have the anticipated result. On the other hand, the use of targets impregnated with insecticide (0.4% Deltamethrin) is generally known to be labour-intensive, offers a slow reduction rate of trypanosomosis and is associated with accessibility risks (such as irregular topography and wild animals). In spite of these facts, the technique is still regarded as an alternative wherever tsetse fly pocket areas exist in the absence of cattle.  And also, a substantial result could be obtained by applying targets if community participation is initiated in the control scheme with the associated training of how to deploy and maintain targets. Based on the conditions and the fact that they have not shown considerable variation in the cost-effectiveness assessment conducted, in the current activity of Southern Tsetse Eradication Project (STEP), the integrated use of both methods in the suppression of tsetse fly population for the subsequent application of sterile male flies (G. pallidipes) is recommended as none of them could achieve its goal independently. Moreover, further study is recommended to assess the situation with regards to the cost and benefits acquired by the society from the suppression programme carried out  in the period of the project life.

            The authors would like to thank the Faculty of Veterinary Medicine, Addis Ababa University and the Southern Tsetse Eradication Project for the financial and logistic supports during the study period.

References

Abebe, G. and Jobre, Y. (1996). Trypanosomiasis: A Threat to Cattle Production in Ethiopia. Revue. Med. Vet. 147, 897-902.

Chamsa, A. and Mweempwa, C. (1997). Control of Tsetse flies G. pallidipes Austen and G. morsitans morsitans Westwood in the Kariba hills, an area common to Zambia and Zimbabwe.   In: Proceedings of the International Scientific Council for Trypanosomiasis Research and Control (ISCTRC), 24th Meeting held in Maputo,Tanzania. OAU/STRC Publication No. 119.

Erkelens, A.M., Dwinger, R.H., Bedane, B., Slingenbergh, J. H.W., and Wint, W.( 2000). Selection of Priority Areas for Tsetse Control in Africa; A Decision Tool Using GIS in Didessa Valley, Ethiopia, as a Pilot Study. In, Animal Trypanosomosis: Diagnosis and Epidemiology. FAO/IAEA Coordinated Research Programme on the Use of Immunoassay Methods for Improved Diagnosis of Trypanosomosis and Monitoring Tsetse and Trypanosomosis Control Programmes. International Atomic Energy Agency, Vienna, Austria.

Langridge, W.P. (1976). A Tsetse and Trypanosomiasis Survey in Ethiopia. Ministry of Agriculture, Ethiopia and Ministry of Overseas Development, United Kingdom.

Leak, S. G. A., Mulatu, W., Rowlands, G. J. and d’Ieteren, G. D. M. (1995). A trial of a Cypermethrin ‘Pour-on’ insecticide to control G. pallidipes, G. fuscipes fuscipes and G. morsitans submorsitans (Diptera: Glossinidae) in South-West Ethiopia. Bull.  Entomol.  Res.  85, 241-251.

Leak, S.G.A. (1999). Tsetse Biology and Ecology: Their Role in the Epidemiology and Control of Trypanosomosis. CAB International. Wallingford (UK).

Leak, S. G. A., Mulatu, W., Authie, E., d’Ieteren, G., Peregrine, A., Rowlands, G. J., Trail, J. (1993). Epidemiology of Bovine trypanosomosis in the Ghibe Valley, Southwest Ethiopia, I.Tsetse challenge and its relationship to trypanosome prevalence in cattle; 2. Factors associated with variations in trypanosome prevalence, incidence of new infections and prevalence or recurrent infections. Acta Trop.  53, 121-134; 135-150.

Lumamba, D.N., Mweempwa, C., Mubanga, J. and Leroy, E  (1997). Eradication of G. morsitans morsitans  and G. pallidipes in the Kariba hills of Siavonga , Southern Zambia.  In: Proceedings of the International Scientific Council for Trypanosomiasis Research and Control (ISCTRC), 24th Meeting held in Maputo,Tanzania. OAU/STRC  Publication No. 119.

Mulatu, W., Swallow, B.M., Rowlands, G. J., Leak, S. G. A., D’Ieteren, G.D.M. and Nagda, S.M. (1997). Economic benefits to farmers of six years of application of an insecticidal ‘pour-on’ to control tsetse in Ghibe, Southwest Ethiopia. In: Proceedings of the International Scientific Council for Trypanosomiasis Research and Control (ISCTRC), 24th Meeting held in Maputo, Tanzania. OAU/STRC Publication No. 119.

Murray, M., Stear, M. J., Trail, J. C. M., d’Ieteren, G.D.M., Agyemang, A. and Dwinger, R.H. (1991). Trypanosomiasis in cattle: Prospects for control, breeding for disease resistance in farm animals (Owen, J. B., Axford, R.F.E., Eds), CAB International, Wallingford.

Murray, M., Murray, P.K. and McIntyre, W.I.M. (1977). An improved parasitological technique for the diagnosis of African trypanosomosis. Trans. R. Soc. Trop. Med. Hyg. 71, 325-326.

Muturi, K.S., Msangi, S., Munstermann, S., Clausen, P., Getachew, A., Getachew, T., Bergenie, B. and Assefa M. (2000). Trypanosomosis Risk Assessment in Selected Sites of the Southern Rift Valley of Ethiopia. I. Distribution, Density and Infection Rates of Tsetse Flies. II. Epidemiology of Bovine Trypanosomosis. In: International Scientific Council for Trypanosomiasis Research and Control (ISCTRC). Proceedings of the 25th meeting held in Mombassa, Kenya.  OAU/STRC Publication No. 120.

Phelps, R. J. and Lovemore, D.F. (1994). Vectors: Tsetse flies. In: Coetzer, J. A. W., Thomson, G. R. and Tustin, R.C (ed.).  Infectious Disease of Livestock. Oxford University Press. Cape Town.

Slingenbergh, J. H. W. (1992a). Consolidation of tsetse and trypanosomosis control in the upper Didessa valley. Consultancy report of the tsetse control specialist to the Fourth Livestock Development Project, FAO, and Rome.  pp. 255.

Slingenbergh, J. H. W. (1992b). Tsetse Control and agricultural development in Ethiopia, World Anim.Rev.70/71, 30-36.

Stata Corporation (2000). Intercooled Stata version 7.0 for Windows 95/98/NT. University Drive East College Station, Texas, USA.

Thrusfield, M.V. (1995). Veterinary Epidemiology. 2nd edition. Blackwell Science, Oxford, pp. 183.

Van den Bossche, P., and Rowlands, G.J. (2001). The relationship between the parasitological prevalence of trypanosomal infections in cattle and herd average packed cell volume. Acta Trop. 78: 163-170.

Van den Bossche, P., Munsimbwe L., Mubanga, J., Jooste, R., and Lumamba, D. (2004). A large-scale trial to evaluate the efficacy of a 1%Pour-on formulation of  Cyfluthrin (cylence,Bayer) in controlling bovine trypanosomosis in Eastern Zambia. Trop. Anim. Hlth. Prod., 36(1), 33-43.

Vreysen, M. J. B., Mebrate, A., Menjeta, M., Bancha, B., Woldeyes, G., Musie, K., Bekele, K. and Aboset, G.  (1999). The Distribution and Relative Abundance of Tsetse Flies in the Southern Rift Valley of Ethiopia: Preliminary Survey Results. In: Proceedings of the International Scientific Council for Trypanosomiasis Research and Control (ISCTRC), 25th Meeting held in Mombassa, Kenya. OAU/STRC Publication No. 120.