Identified factors influencing the occurrence of abnormal lactation curves in Iranian buffaloes

Abstract

Atypical lactation curves (ALCs), mostly characterized by the lack of any distinguished peak, are the subject of this study. Logistic regression was used to assess the influences of some non-genetic factors on the occurrence of ALCs in Iranian buffaloes. The data set consisted of 24,679 lactation records from 11,478 buffaloes that calved during the years 1996 to 2012. As a reasonable statistical criterion, odds and odds ratio were used to evaluate the probable differences between the various levels of the factors. Genetic analysis of the ALC trait was also carried out using Bayesian approach to estimate heritability and repeatability parameters. The results revealed that almost 44% of the lactation curves were atypical. All factors but ecotype factor had significant effects on the occurrence of ALCs. The frequency of ALCs increased until the third parity and decreased afterward, therefore the highest and the lowest values of the trait were observed in the third and 10th parities, respectively. Unlike the parity effect, ageing of the buffaloes caused the odds values to be continually increasing. The incidence of the ALCs for buffaloes calved out of season (July to December) was more likely (almost 15%) than those calved in season (January to June). The structure of the data also affected the trait significantly, such that the lactations with 4 recorded test-days were 1.75 times more likely to have atypical curves than those with 8 recorded test-days. Heritability and repeatability of the occurrence of ALCs were estimated very low and were 0.020 and 0.036, respectively.