Non genetic factors influencing fertility problems in Murrah buffaloes

unknown. The data on 2449 reproduction records of Murrah buffaloes were analysed using SAS to study the effect of various non genetic factors on the incidence of fertility problems (repeat breeding, true anoestrus and anoestrus due to persistent corpus luteum). The association between non-genetic factors and fertility problems were analysed using Chi squares method. To study the effect of these independent variables simultaneously, multiple logistic regression method was used. Overall fertility problems and true anoestrus were significantly (p<0.01) higher in buffaloes calved during winter and significantly (p<0.01) lower in those calved during rainy season of the year. However, the association between anoestrus PCL and season of calving was not significant. Significantly (p<0.05) higher incidence of repeat breeding was reported in autumn season and third period of study. The incidence of anoestrus was significantly (p<0.01) higher in heifers and first calvers in comparison to older buffaloes. Parity showed non-significant association with incidence of repeat breeding in Murrah buffaloes. Highly significant association (p<0.01) was observed between uterine health problems and incidence of all the fertility problems, which increased with the severity of post partum uterine health complications. Logistic regression results also revealed higher ODDS ratio for incidence of overall fertility problems in those suffered from uterine health problems only (3.35) and both calving abnormalities and uterine health problems (3.16) in comparison to normal (1.00) animals. Similarly, significantly (p<0.01) higher ODDS ratio were observed for incidence of anoestrus (3.12 and 2.92) and repeat breeding (6.43 and 5.05). From the present investigation it has been concluded that multiple logistic regression can be used effectively to study the effect of various non-genetic factors on discrete dependent variables like fertility problems in animals.