AN INVESTIGATION ON RELATIONSHIP BETWEEN LACTATION MILK YIELD, SOMATIC CELL COUNT AND UDDER TRAITS IN FIRST LACTATION TURKISH SAANEN GOAT USING DIFFERENT STATISTICAL TECHNIQUES
E. Eyduran, I. Yilmaz, A. Kaygisiz* and Z. M. Aktas**
Igdir University, Faculty of Agriculture, Department of Animal Science,76000, Iğdır, Turkey.
*Kahramanmaras Sütcü Imam University, Faculty of Agriculture, Department of Animal Science ; **Graduate School of Natural and Applied Sciences, Kahramanmaras, Turkey.
Corresponding author mail: firstname.lastname@example.org
This study aimed to determine the relationship between lactation milk yield, (LMY), somatic cell count (SCC), and udder traits using some statistical techniques, viz. multiple linear regression, stepwise regression, use of factor analysis in multiple regression, and regression tree. Forty six Turkish Saanen Goat were used in the present investigations . These goats were milked with milking machine twice a day. The traits recorded were: daily milk yield (DMY), udder upper height (UUH), udder width (UW), udder length (UL), udder diameter (UD), teat angle (TA), udder volume (UV), somatic cell count (SCC), lactation length (LL), and lactation milk yield (LMY), Respectively. In the multiple linear regression analysis, significant effects of and LT (P<0.01) and TA (P<0.01) on LMY trait were observed. According to result of Stepwise Regression Analysis model (SRAM); TA, LL and UD had positively significant effect on the LMY trait without multicollinearity problem. Kaiser-Meyer-Olkin (KMO=0.6) and, Bartlett’s test of spherity (P<0.01) also confirmed that factor analysis was applicable for the data. Present results from using factor analysis scores with multiple regression analysis (FASMLRA) reflected that five factors, explained 87.7 % of variation and only UV, TA, UL, UD and LL could significantly increase LMY (P<0.01) for factors 2-4. FASMLRA with the lowest VIF values gave more advantageous results. Regression Tree (RT) Method illustrated that, the goats with TA > 500 among all the goats (Node 2) had the highest LMY (297.357 lt) or Node 4 as a group of the goats with LL > 191 days among the goats with TA < 500 provided higher LMY average except for Node 2. If there is a verification at genetic level, the relationship between lactation milk yield (LMY), somatic cell count (SCC), and udder traits might contribute to further goat breeding to increase milk yield in the future.
Key words: Factor Analysis Scores, Milk yield, Udder trait, Stepwise Regression, Regression Tree, Saanen Goat.