The prevalence of subclinical mastitis and associated risk factors at the herd, cow and quarter level were studied using multilevel modelling in smallholder dairy farms in Jimma, Ethiopia. Forty-two herds, out of the 55 dairy farms located in Jimma (76%), were visited, a questionnaire was performed, and 635 quarters belonging to 176 lactating cows were screened to detect the presence of subclinical mastitis using the California Mastitis Test (CMT). Sixty-two percent of the cows and 51% of the quarters were subclinically infected. Quarters from cows in later stage of lactation (>180 DIM) [opposed to early lactation (<90 DIM)] were more likely to be subclinically infected (OR=2.40, 95% CI=1.44-3.99). Overall, quarters from cows milked by squeezing (as opposed to stripping) were less likely to be subclinically infected (OR=0.45, 95% CI=0.29-0.71), although quarters from cows with tick-infested udders were more likely to be subclinically infected when milked by squeezing (as opposed to stripping). The milking technique did not influence the likelihood of infection in cows without ticks on the udder. This study stresses the high prevalence of subclinical mastitis in smallholder dairy farms in Jimma and a lack of awareness of the existence of the disease among dairy farmers. Implementation of a mastitis prevention program adapted to the local needs, including a focus on milking technique, application of appropriate tick control measurements as well as fertility management, allowing cows to be dried-off at a more appropriate moment, are needed. To conclude, milking by squeezing instead of stripping, but not of cows with tick-infested udders, as well as fertility management could reduce the subclinical mastitis prevalence (and incidence) on the short term.
Keywords: California Mastitis Test; Jimma town, Ethiopia; Multilevel modelling; Risk factors; Smallholder dairy farms; Subclinical mastitis.
Copyright © 2013 Elsevier B.V. All rights reserved.