S.W.P. Cloete1,2, A.J. Scholtz2 & M.A. Snyman4

1Department of Animal Sciences, University of Stellenbosch, Private Bag X1, Matieland, 7602, South Africa; Directorate Animal Sciences, Private Bag X1, Elsenburg, 7607, South Africa; 1Grootfontein Agricultural Development Institute, Private Bag X529, Middelburg (EC), 5900 

#Corresponding author: Schalk Cloete



Background: Genetic parameters for ovine milk production and composition  are scant in the literature. However, evidence suggests that milk production in the dam underlies variation in direct and maternal breeding values for lamb weaning weight. This study investigated the genetics of ovine milk production.


Aim: To estimate genetic parameters and trends in a divergent selection experiment that ran from 1986 to the present.


Methodologies: A total of 1679 ewe-year milk records of Merino ewes from a divergent selection experiment for maternal multiple rearing ability were used in this study. The ewes were milked by hand, using the oxytocin method, to determine their milk production over 3 hours, either 22 (SD=3) or 84 (SD=4) days into lactation. Milk composition was analysed at the Stellenbosch milk testing laboratory. Data recorded included total milk yield (TMY), butterfat % (BFP), protein % (PP), lactose (LP) and somatic cell count (SCC). These data were analysed to estimate genetic parameters for these traits. Animal solutions, reflecting direct breeding values, were averaged according to birth year to derive genetic trends for TMY.


Results: The line selected in the upward selection (H Line) had a higher mean (±SE) daily milk production than the line selected for a reduced lamb output (L Line) both 3 weeks and 12 weeks into lactation (respectively 1437±56 vs. 1220±81 ml and 913±38 vs. 707±54 ml). No line differences were found for the percentage traits or SCC. Heritability estimates were 0.14±0.04 for TMY, 0.20±0.05 for BFP, 0.39±0.03 for PP, 0.26±0.04 for LP and 0.14±0.03 for SCC at 3 weeks of lactation. Corresponding estimates for 12 weeks into lactation were 0.20±0.05, 0.31±0.04, 0.47±0.03, 0.41±0.03 and 0.11±0.03. Animal permanent environment affected TMY at 3 weeks and at 12 weeks (respectively 0.08±0.04 and 0.15±0.04) while BFP (0.19±0.05) and LP (0.09±0.03) were also affected at 3 weeks. Genetic correlations of TMY with BFP and PP were unfavourable, while the latter traits were favourably correlated. Genetic correlations between the same traits after 3 and 12 weeks were high, ranging from 0.58±0.13 for SCC to 0.99±0.09 for TMY. Regressions of annual breeding values on birth year indicated that TMY increased by respectively 1.29±0.20 (r=0.77) and 2.74±0.17 (r=0.95) ml per annum after 3 and 12 weeks of lactation in the H Line. Corresponding trends in the L line amounted to 3.50±0.35 (r=0.88) and 3.32±0.24 (r=0.93) ml per annum.


Discussion: These results conformed to expectations based on lactation records in dairy cattle, namely that all milk traits were heritable, that yield was unfavourably correlated to BFP and PP and that the latter traits were favourably correlated. Moreover, results pertaining to realised genetic gains suggested that selection for lamb output resulted in a favourable correlated response in TMY over nearly 30 years of selection.   


Conclusions: The oxytocin method proved to be suitable for deriving genetic (co)variance components for milk traits in ewes managed on pasture. Further research should focus on genetic correlations of milk traits with other traits of economic importance.



49th SASAS congress, Stellenbosch