Variation and repeatability of a.m. and p.m. records for milk yield and composition in Massese ewes
2000
Acciaioli, A. | Franci, O. | Lucifero, M. (Florence Univ. (Italy). Dipartimento di Scienze Zootecniche) | Pugliese, C.
Two hundred and eighteen lactations of Massese ewes belonging to three flocks were studied. A.m. and p.m. milk yields were recorded monthly and milk samples were collected and analysed for content of fat, protein and lactose. A mixed model was used to estimate variance components and to calculate repeatability between milkings on the same test-day (r sub(td)) and among all-a.m. (r sub(am)) or all-p.m. milkings (r sub(pm)) across test-days. Repeatability within test-day (r sub(td)) was high for milk yield (0.88) and % protein (0.9) but was rather low for % fat (0.54) and fat yield (0.69). Repeatability across test-days was higher for milk yield (r sub(am) = 0.53; r sub(pm) = 0.52) than for milk composition (eg, % fat: r sub(am) = 0.31; r sub(pm) = 0.21). Generally, repeatability was higher with a.m. than p.m. recording but it was highest with pooled test-day recording. Using a complete test-day model, a significant interaction milking-time x lactation emerged for % fat and fat yield, indicating that the exclusive use of either an all-a.m. or an all-p.m. recording scheme can produce misleading estimates of lactation fat content. The rank correlation between estimates for lactations with different schemes indicated that, compared with the complete method (both a.m. + p.m. milkings once a month), the scheme alternating a.m. and p.m. monthly milkings was more accurate than the only-a.m. or only-p.m. scheme, even if loss of accuracy exceeded 15% in the case of % fat
显示更多 [+] 显示较少 [-]