The prediction of red steppe cows dairy productivity based on correlation relationship and regression models

Purpose. The estimation of the value and probability level of correlations between hematological parameters, nitrogen, calcium and phosphorus balances, the digestibility of dry matter, crude protein, crude fiber and the dairy productivity of Red Steppe cows bred in Volgograd region, the regression equations of linear dependence were found for high correlations. Materials and Methods. The scientific and economic experiment was conducted in the conditions of the industrial complex for raising dairy cattle of the Red Steppe breed and producing milk "Lenin Breeding Plant-Collective Farm". According to the principle of analogues, 20 lactating first-calf cows with an average live weight of 490.0±1.54 kg were selected. The indicators were studied from the beginning of the lactation period during the first three months (90 days). The digestibility of nutrients in feed in lactating cows was assessed based on the results of the balance experiment in accordance with the recommendations of the All-Russian Institute of Animal Husbandry. Milk productivity was recorded individually for each cow every ten days based on control milkings, the fat and protein content in milk was determined on the automated measuring complex "Laktan" (OOO VPK "SibagroPRIBOR", Russia). Morphological blood analysis was performed on the automatic analyzer URIT-3020 Vet (Urit Medical Electronic Co., Ltd., China), biochemical analysis – on the semi-automatic analyzer URIT-800 Vet (Urit Medical Electronic Co., Ltd., China), ascorbic acid in blood serum was determined on the chromatograph Agilent 1200 (Agilent Technologies Inc., USA). Results. The highest positive correlation coefficient between the serum iron level and the hemoglobin content in the blood of lactating cows was established. Also the highest positive relationship between milk yield and red blood cells was found. The correlations between dietary calcium and phosphorus intake, its levels in serum, in milk and milk yield were very strong or complete and statistically very significant. The correlation coefficient between serum total protein and milk yield was very strong, meanwhile a weak negative correlation between milk protein content and milk yield was found. A strong positive relationship was established between milk fat content and dry matter digestibility, as well as between crude protein digestibility and milk yield. Linear regression equations based on the data obtained were modeling aimed to predict quantitative and qualitative indicators of milk productivity. Conclusion. The use of correlation coefficients and linear regression models, as well as variance and regression analysis, allows dairy breeders to assess the relationship between different traits and make informed decisions about ways to increase the productive performance of cows. Understanding correlative relationships and their use in breeding work makes it possible to carry out tandem breeding of a lot of parameters at the same time. Therefore, this research has not only scientific, but also practical significance.