Factors affecting ewe's milk fat and protein content and relationships between milk yield and milk components

Test-day records of purebred Tsigai, Improved Valachian and Lacaune ewes were used to analyse the effect of environmental factors on milk fat and protein content. There were 121 424 and 121 158 measurements of fat and protein content for Tsigai, 247 742 and 247 606 measurements of fat and protein content for Improved Valachian and 2 194 measurements of fat and protein content for Lacaune ewes lambing between 1995 and 2005. Overall means and standard deviations for fat and protein content were 7.77 ± 1.606% and 5.94 ± 0.690% for Tsigai, 7.48 ± 1.446% and 5.82 ± 0.620% for Improved Valachian, and 6.97 ± 1.514% and 5.62 ± 0.692% for Lacaune. For fat content, analyses showed a highly significant (P < 0.01) effect of flock-test day and a highly significant (P < 0.01) or significant (P < 0.05) effect of the month of lambing, with the only exception of the month of lambing in Lacaune. The effect of litter size was highly significant (P < 0.01) or significant (P < 0.05) in Improved Valachian and Lacaune. For protein content, analyses showed a highly significant (P < 0.01) effect of flock-test day and a highly significant (P < 0.01) or significant (P < 0.05) effect of the month of lambing. The effect of litter size was highly significant (P < 0.01) in Tsigai and Improved Valachian. Covariates of days in milk which modelled the shape of lactation curves were insignificant, except for Improved Valachian fat content (Ali-Schaeffer regression adopted for sheep). The model explained about 50% of fat and protein variation in the breeds, with coefficients of determination between 0.517 and 0.587 for fat content and between 0.495 and 0.527 for protein content. Fat and protein content were almost equally correlated with milk yield in the three breeds. Lactation curves were constructed on the basis of solutions of a statistical model employed in the analyses.