A MULTIVARIATE STATISTICAL ANALYSIS OF MILK YIELD AND QUALITY IN INTENSIVE DAIRY PRODUCTION SYSTEMS IN PARANÁ STATE, BRAZIL

Abstract

Background: Dairy farm manager must apply strategic planning to manage the complexity of dairy production systems and produce high-quality milk. Objective. To identify and group dairy farms on the basis of milk yield and composition, animal data, and farm characteristics. Methodology. Official milk records and production variables of 26 dairy farms and 45,343 cows in the region of Arapoti, Paraná, southern Brazil (24°49′36″S 49°49′36″W, 860 m elevation), were assessed. Milk yield and composition data were analyzed using multiple correspondence analysis (MCA) and ascending hierarchical classification (AHC) followed by simple linear regression analysis. Results. AHC afforded five clusters of dairy farms, grouped according to milk yield and composition. MCA showed that cow age influenced mammary gland health, whereas feed management and lactation stage influenced milk composition. Regression analysis indicated that the relationships between milk composition parameters and lactation number had the same behavior in all clusters; milk protein and lactose contents decreased with increasing lactation number. A relationship was found among cow age, lactation stage, and mammary gland health, as well as among milk yield, cow age, and mammary gland health. Somatic cell count increased with cow age. Low milk yield was associated with high somatic cell counts. Mammary gland health was affected by lactation stage, and milk composition was influenced by feeding management and lactation number (cow age). Implications. The comprehension of the relationship between milk production and composition on dairy cattle farms can suggest specific decision-making for group of milk farmers, regarding nutritional and milk quality according to the production, composition and somatic cell count. Conclusion. A relationship was found among cow age, lactation stage, and mammary gland health as well as among milk yield, cow age, and mammary gland health. Multivariate statistical analysis helped to understand the relationship between milk production and composition in these dairy cattle farms. Based on these grouping results, we can suggest specific decision-making for each group of producers, regarding nutritional and milk quality according to the production, composition and somatic cell count.

Keywords

• parity
• mammary gland
• milk quality
• multiple correspondence analysis.