Nine mathematical models were compared for their ability to predict daily milk yields (n = 294,986) in standard 305-d and extended lactations of dairy cows of Costa Rica. Lactations were classified by parity (first and later), lactation length (9 to 10, 11 to 12, 13 to 14, 15 to 16, and 16 to 17 mo), and calving to conception interval (1 to 2, 3 to 4, 5 to 6, 7 to 8, and 9 to 10 mo). Of the nine models, the diphasic model and lactation persistency model resulted in the best goodness of fit as measured by adjusted coefficient of determination, residual standard deviation, and Durbin-Watson coefficient. All other models showed less accuracy and positively correlated residuals. In extended lactations, models were also fitted using only test-day records before 305 d, which resulted in a different ranking. The diphasic model showed the best prediction of milk yield in standard and extended lactations. We concluded that the diphasic model provided accurate estimates of milk yield for standard and extended lactations. Interpretation of parameters deserves further attention because of the large variation observed. As expected, the calving to conception interval was found to have a negative effect on milk yield for cows with a standard lactation length. In extended lactations, these negative effects of pregnancy on milk yield were not observed.

The aim was to determine the relationship between kinematic parameters of boar spermatozoa and fertility rates of sow, as well as to assess the effect of sperm clusters on the fertility capacity of the ejaculate. Semen samples were collected from 11 sexually mature boars. Samples were analyzed by an ISAS^®v1 CASA-Mot system for eight kinematic parameters. Ejaculate clusters were characterized using multivariate procedures, such as principal factors (PFs) analysis and clustering methods (the k-means model). Four different ejaculate clusters were identified from two kinematic PFs which involved linear trajectory and velocity. There were differences (p < 0.05) between the sperm kinematic variables by sire line. There was no statistical difference (p > 0.05) between dam lines and ejaculate clusters in fertility variables. The discriminant ability of the different kinematics of sperm variables to predict litter size fertility was analyzed using receiver operating characteristics (ROC) curve analysis. Curvilinear velocity (VCL), average path velocity (VAP), amplitude of lateral head displacement (ALH), and beat-cross frequency (BCF) showed significant, albeit limited, predictive capacity for litter size fertility variables (range: 0.55–0.58 area under curve, AUC). The kinematic analysis of the ejaculates in clusters did not have a predictive capacity for litter size variables.

The aim of this study was to determine risk factors associated with milk fever (MF) occurrence in Costa Rican grazing dairy cattle. A total of 69,870 cows from 126 dairy herds were included in the study. Data were collected in the Veterinary Automated Management and Production Control Program software by the Population Medicine Research Program of the Veterinary Medicine School, National University of Costa Rica, from 1985 to 2014. To determine the risk factors for MF, 2 logistic regression mixed models were evaluated. The first model used breed, month of calving, ecological life zone, herd nested within ecological life zone, and parity as fixed effects. The second model excluded first-lactation animals and cows without production information, had the same fixed effects of the first model, and added previous MF case, previous lactation length, previous dry period length, previous corrected 305-d milk yield, and calving interval length as fixed effects. Both models used animal and year as random effects. Of the 235,971 recorded lactations, 4,312 (1.83%) reported MF event. The significantly associated risk factors for MF occurrence, ranked by their highest odds ratio (OR), were parity (OR = 52.59), previous dry period length (OR = 4.21), ecological life zone (OR = 3.20), breed (OR = 3.04), previous corrected 305-d milk yield (OR = 2.39), previous MF case (OR = 2.35), and month of calving (OR = 1.36). The findings of this study are the first data reported using an epidemiological approach to study risk factors for MF in Costa Rican dairy cattle. Some of these results might be used to improve preventive management practices at the farms to reduce the incidence of this metabolic disease in grazing dairy herds.