Logistic regression model reveals major factors associated with total bacteria and somatic cell counts in goat bulk milk

Abstract

The objective of this study was to identify on-farm risk factors associated with microbiological traits and somatic cell counts of goat bulk milk. A cross-sectional study was carried out on 160 goat dairy herds. Bulk tank milk samples were collected aseptically for qualitative (Salmonella enterica) and quantitative (aerobic mesophilic bacteria, total coliforms, E. coli, Staphylococcus spp. and Staphylococcus aureus) microbiological analyses. Additionally, information on various herd characteristics, general management practices, and milking procedures was collected by means of a questionnaire. A model for risk factors associated with the investigated traits was built by means of logistic regression after selection of putative variables by univariable analysis. Absence of pre-dipping, non-discharge of foremilk before milking and confirmed mastitis cases in the herds were identified as major risk factors for high aerobic plate counts. Low milk yield (< 25 kg / milk / day) was identified as a major risk factor for high somatic cell counts (SCC) in goat milk. Other variables such as long-term establishment of the herds (>7 years), the use of ponds as water source, long lactation periods (> 210 days), practice of milking in late pregnancy and non-discharge of foremilk before milking were also identified as risk factors for high SCC. The information generated by this study can support intervention measures targeting the improvement of goat milk quality, which is of paramount importance for the continuous development of the goat dairy industry.