Article — Artikel The prevalence of serum antibodies to Ehrlichia ruminantium infection in ranch cattle in Tanzania: a cross-sectional study E S Swaia*, P F Mtuia A K Chang’ab and G E Machangea across farming systems in the country. ABSTRACT This investigation was designed to provide Serum samples collected in a cross-sectional survey of grazing cattle on Manyara Ranch, information on the risk of heartwater and Monduli district, Tanzania, were tested by indirect major antigenic protein 1 fragment B to assist in the development of appropri- (MAP 1-B) ELISA to determine the seroprevalence of Ehrlichia ruminantium and to assess ate control strategies. The prevalence of ranch-level risk factors for heartwater. Heartwater-exposed cattle were widespread on the E. ruminantium was assessed, and factors ranch and overall seroprevalence was 50.3 % (95 % CI, 44.9 –55.6), enough to indicate an associated with infection in extensively endemically unstable situation. Multivariate logistic regression modelling was used to raised indigenous cattle on Manyara identify risk factors associated with seropositivity. Two factors appeared to increase the Ranch, Tanzania, were explored. herd’s risk for contracting heartwater. Seroprevalence increased significantly with age (β = 0.19 per year of age, P < 0.001) and animals carrying ticks of any species were associated MATERIALS AND METHODS with an increased risk of infection with E. ruminantium (Odds ratio, OR = 3.3, P < 0.001). The force of infection based on the age seroprevalence profile was estimated at 18 per 100 Study site cattle year-risk. The current tick control measures on the ranch were associated with a decreased risk of infection with E.ruminantium (OR = 0.25 for no dipping and OR = 0.31 for Manyara Ranch (03°34’04’’S, 36°05’ low dipping, P < 0.001). Six tick species were identified; in order of frequency these were: 01”E) occupies 17 871 hectares (~45 000 Ambylomma variegatum 59.9 %, Rhipicephalus evertsi evertsi 13.9 %, Rhipicephalus pulchellus acres) of land and is located 92 km 12.5 %, Hyalomma truncatum 7.03 % and Rhipicephalus appendiculatus 6.07 %. The least southwest of Arusha city in the Maasai encountered tick was Rhipicephalus simus, which accounted for 0.38 %. The cattle seemed Steppe Heartland26. The ranch is lo- well adapted to their environment and capable of resisting the tick burden under this cated between 2 major national parks: extensive wildlife/livestock grazing and interaction system. Tarangire to the south and Lake Manyara Key words: E. ruminantium, indigenous cattle, risk factors, seroprevalence, Tanzania. to the north. Wildlife and livestock inter- act freely when grazing. The ranch is Swai E S, Mtui P F, Chang’a A K, Machange G E The prevalence of serum antibodies to Ehrlichia ruminantium infection in ranch cattle in Tanzania:a cross-sectional study. Jour- semi-arid with annual average rainfall of nal of the South African Veterinary Association (2008) 79(2): 71–75 (En.). Veterinary Investiga- 600–700 mm and is at an elevation of tion Centre, PO Box 1068, Arusha, Tanzania. 1200 m above sea level. The vegetation is mainly Acacia–Commiphora bushland, wooded savanna, with short grass. The rains are usually concentrated in 2 INTRODUCTION results in considerable economic losses seasons: end of March to May, and end of Heartwater (cowdriosis), caused by the due to loss of production, treatment costs October to December.The mean tempera- rickettsial organism Ehrlichia ruminantium and reduced initiatives for the upgrading ture ranges from 15 to 30 °C. The study (formely Cowdria ruminantium), is an in- of local breeds of livestock with more sus- was carried out in October and Novem- fectious disease of ruminants. The organ- ceptible exotic breeds 16. ber 2006. ism is transmitted by 3 host ticks of the Heartwater can be unequivocally diag- genus Amblyomma33. Heartwater is one of nosed by the identification of E. rumi- Cattle husbandry system the most important tick-borne diseases nantium by demonstration of rickettsial The cattle are maintained in a pastoral (TBD) affecting both wild and domesti- inclusion bodies in the endothelial cells of open grazing system from approximately cated grazing ruminants in Tanzania18, brain crush smears24. This method is, 07:30 to 16:30 daily. No supplementary and it has been estimated that the direct however, only applicable to clinical cases. feeding is provided. The animals are economic loss due to the disease is about Serological tests specifically targeting housed at night in bomas or kraals con- US$20.5–24.5 million10. The mortality rate bovines have been applied but are of structed from thorny tree branches to due to heartwater is around 15 % with a limited use under field conditions owing protect them from theft and predators. morbidity rate of 5 %, although the sever- to cross-reactions with other Ehrlichia spe- The breeding system used on the ranch ity of the disease varies widely between cies.6,9. The MAP1-B ELISA based on the is natural mating, with service bulls run- cattle types, agro-ecological zones, truncated major antigenic protein 1 of ning freely with females year-round. socio-economic conditions and cattle pro- E. ruminantium has been shown to be of Calves are weaned at 8 months of age. duction systems10. In tropical and sub- value under field and experimental Young calves are isolated and penned tropical areas, the disease is endemic and conditions, especially in the regions until the return of their dams from graz- a where the distribution of cowdriosis is ing, and then allowed to suckle. Cattle are Veterinary Investigation Centre, PO Box 1068, Arusha, 14,15,17,33 Tanzania. unknown . vaccinated yearly against anthrax, haem- bHerd Health Division, Manyara Ranch, PO Box 2658, Although the disease is known to be orrhagic septicaemia, lumpy skin disease, Arusha, Tanzania. endemic in various parts of Tanzania, and female calves <10 months of age with *Author for correspondence. E-mail: [email protected]. there is very little information on the Brucella Strain 19. Control of ectoparasites, Received: September 2007. Accepted: April 2008. epidemiology of the disease between and mainly ticks, is carried out by using 0038-2809 Jl S.Afr.vet.Ass. (2008) 79(2): 71–75 71 Decatix® (2.5 % m/v, Deltamethrin, Statistical analysis Tick infestation Coopers, Zimbabwe). The frequency of Descriptive statistics for the animal and A total of 2600 ticks (mature and imma- dipping is coded as follows: none: no ranch-level explanatory variables (inclusive ture) were counted during sampling. Of dipping, low: once every 4 weeks; moder- of tick infestation) examined in the study the 360 cattle examined and sampled, 320 ate: twice every 4 weeks. Anthelmintic were developed using Epi-Info version were found to carry ticks, giving an infes- treatment using Albendazole (10 %, Kela 6.04d7. Cross-correlations7 were performed tation rate of 88.9 % and the overall mean N.V., Phenix, Belgium) is generally lim- on all explanatory variables to investigate (mean± SE) tick density of 8.12 ± 0.44 ited to calves. the potential for confounding amongst tick/cattle. Tick infestation rate was signif- χ2 them (defined as those with a coefficient icantly ( = 31.7, df = 4, P = 0.001) higher in low, intensive and non-dipped and Study animals and sampling of 0.7 or greater). The relationship between least in moderately dipped cattle. The A systematic sampling procedure was explanatory variables and outcome re- mean tick-specific densities per animal adopted and every 4th animal was selected sponse (seroconversion to E. ruminan- are detailed in Table 2. as the cattle passed through a crush2.A tium) were investigated in 2 steps by logis- 4 Six tick species were identified, with total of 360 cattle (all ages and sexes), out tic regression (using Egret for Windows A. variegatum being the most abundant of the total ranch herd size of 1481, were version 2.0). Explanatory variables inves- (59.9 %), followed by R. evertsi evertsi used in the study. The cattle types reared tigated were ‘presence of a tick of each (13.9 %), R. pulchellus (12.5 %), H. trun- on the ranch are Boran (75 %) and Tanza- species’ investigated as a binary variable catum (7.03 %) and R. appendiculatus nia short horn zebu (TSHZ) (20 %), and (i.e. yes or no); body score (coded as fat, (6.07 %). The least encountered tick was their crosses (5 %). Each selected animal medium, thin); dipping (yes or no), R. simus at 0.38 %. was subjected to a thorough, whole-body acaricide application frequency (coded as inspection for the presence of ticks. Other none, low, moderate); sex (female or Age–antibody prevalence profiles information recorded included: sex, age male) and age. An increasing seropositivity was re- (estimated by ear notch, examination of In the 1st step, the relationships between corded with respect to increasing age. The teeth and written records). The age of each explanatory and outcome variable estimated force of infection was at 0.18 each animal was transformed age centred were individually investigated. In the animals per cattle year-risk. The graphical to normalise the data. Tick infestation, 2nd step, any variables that were signifi- age seropositivity relationship and forces with both mature and immature stages, cantly associated at P <0.25 were in- of infection are shown in Fig. 1. was coded as low: ≤20; moderate: >20 cluded in multivariate logistic regression4 and ≤40; high: >40. Tick infestation was models, producing, by forwards and Factor influencing antibody response moderate, therefore whole body count- backwards substitution and elimination, to E. ruminantium ing was adopted8. Ticks on the study ani- the most parsimonious models in which The factors that were significantly asso- mals were identified to genera and all explanatory variables remained signif- ciated with E.