Analyses of udder health in Valle del Belice dairy sheep using Somatic Cell Count (SCC)

Intramammary infections (IMI) are a complex of inflammatory diseases of the mammary gland. Mastitis is one of the most frequent IMI affecting small dairy ruminants. Direct selection against clinical mastitis is difficult because mastitis is not widely recorded; therefore, somatic cell count (SCC) is promoted as selection criterion for mastitis resistance. A dataset containing 2,475 first-lactation Valle del Belice ewes from 14 flocks recorded from 1998 to 2003 was analysed. In total 116 sires with at least four daughters with a record were included in the pedigree file. Test-day IMI events were coded as a binary trait using a SCC cut-off value of 750,000 cells/ml. Analyses of risk factors for the binary IMI trait were based on a logistic regression model. The herd effect was the major risk factor for IMI. The risk of culling due to IMI increased for late season of lambing. A single trait threshold sire model was applied. The heritability calculated using a logit function was 0.16±0.072, very close to the expected value of 0.14 calculated with Dempster and Lerner’s formula treating IMI as a continuous variable by using somatic cell scores. INTRODUCTION – Intramammary infections (IMI) are a complex of inflammatory diseases which are defined as an inflammation of the mammary gland resulting from the introduction and multiplication of pathogenic micro-organisms. Mastitis is one of the most frequent IMI affecting small dairy ruminants. IMI are mainly of economical, hygienic (consumption of dairy products) and legal importance in Europe (EU directives 46/92 and 71/94 defining the bacteriological quality of milk). Although management is the most effective way to prevent IMI, selection for IMI resistance is an alternative to be considered, at least to prevent any detrimental effect of milk yield on udder health. Direct selection against clinical mastitis is difficult because clinical mastitis is not widely recorded; on the contrary, somatic cell count (SCC) is promoted as selection criterion for mastitis resistance. However, the relationship between SCC and mastitis is far from clear. Some authors (Coffey et al., 1986; Kehrli and Shuster, 1994) were concerned by the recommendation of continuously decreasing SCC by selection and argued that this trend could impair the cow’s capacity for leukocyte recruitment and therefore its ability to respond to IMI. Animals with very low SCC would be more susceptible to mastitis. A punctual approach with a single threshold is a simple methodology which proposes the punctual or instantaneous discrimination between ‘healthy’ and ‘infected’ udders (Bergonier et al., 2003). In ewes, with the fluoro-optoelectronic method, single thresholds were proposed, surprisingly ranging from 200,000 to 1.5x10 cells/ml (Ftenakis 1996; Bergonier and Berthelot, 2003). The objective of this study was the phenotypic and genetic analysis of IMI estimates from SCC with a single trait threshold model. Corresponding author: Valentina Riggio [email protected] Session S6.8 2 MATERIAL AND METHODS – The dataset contained 2,475 first-lactation Valle del Belice ewes from 14 flocks recorded from 1998 to 2003. In total 116 sires with at least four daughters with a record were included in the pedigree file. All first-lactation test-day records were required to have SCC information. The average number of SCC test-day records per ewe was 7.62. For the threshold model, the cut-off value was fixed at 750,000 cells/ml (Di Marco et al., 1997). Mastitis or IMI events were declared with a binary trait L set to one, when the SCC was higher than the cut-off value in one test-day record within lactation, while if SCC within lactation was still lower than the cut-off value the ewes were considered as ‘doubtful or healthy’ and the binary trait L was set to zero. Analyses of the risk factor for the binary IMI trait were based on a logistic regression model using the SAS LOGISTIC procedure (SAS®, 2000). The overall significance of main effects in the model was assessed by a Wald chi-square test. This statistic takes the form of a squared ratio of an estimate to its standard error and asymptotically follows an approximate chi-square distribution with one degree of freedom. The odds ratio (OR) and OR 95% confidence interval were computed according to Hosmer and Lemeshow (1989). The OR measures how much more (or less) likely the outcome is among observations with a given level of a risk factor, compared with those with a reference level of the risk factor. A single trait threshold sire model was applied to the data using the ASReml package (Gilmour et al., 2002). The underlying liability of IMI in first-lactation was modeled as: ( ) ijklm l jk i ijklm e s y sea h L + + × + + = μ , where L is the IMI event with mean μ for ewe m, daughter of sire l (l = 1,...,116) with random effect sl and lambing i in herd hi at season j (j = 1,..,3) and year k (k = 1998,....,2003). The binary trait is not normally distributed, so it is necessary to perform analyses that account for this distribution. Binary analyses were carried out, fitting a Generalized Linear Model, assuming a binomial distribution and using a logit function: