Estimation of Probability of Pregnancy Based on Health Status and Estrus Intensity in Organic Dairy Cows

Our first objective was to quantify the associations between health-related events (HRE) before insemination, the relative increase in estrus intensity (REI) at insemination, and the probability of pregnancy per artificial insemination (P/AI) in organic dairy cows. Quanti-fying these associations may aid on-farm decision-making, such as setting the voluntary waiting period, choice of type of semen, do-not-breed and culling decisions. A second ob-jective was to develop predictive models to estimate P/AI based on readily available data, and present common goodness-of-fit results also used in the machine learning community. All data were collected from a certified organic dairy farm in the western USA from 2019 to 2021. Health-related and reproduction data were obtained through DRMS (Raleigh, NC, USA). Activity data were collected using pedometers (IceRobotics, Stirling, UK) mounted on the rear legs. The REI, defined as walking steps per hour before insemination divided by the cow’s baseline steps per hour, was available for 17,238 inseminations from 4,759 cows. The REI was categorized as ≤200% (6,999 inseminations), >200-400% (4,685), >400-600% (2,929), or >600% (2,625). The HRE were available for 65,684 inseminations from 13,365 cows. The HRE was categorized as mastitis (prior to 9,114 inseminations), metabolic (displaced abomasum, ketosis, milk fever; 1,941), reproductive disease (metritis, endometritis, pyometra, retained fetal membranes; 4,907), lameness (4,058), 2 different diseases (4,022), ≥3 different diseases (813), or as healthy (none of these diseases prior to insemination; 40,829). Combinations (COMBO) between REI categories and 0, 1, or ≥2 HRE were also created for 16,415 inseminations in 4,647 cows. Data were split into training and test sets. The training data were used to fit 3 logistic regression models that included either HRE, or REI, or COMBO. Each of the 3 models also included the covariates of prior 3-mo herd P/AI and days in milk (DIM), and the fixed effects of parity, insemination season, days after the previous insemination or days to 1st insemination. A random effect ac-counted for repeated inseminations within cow. Parameter estimates, odds ratios, and the least-square means of the estimated P/AI of the fixed effects were obtained from the logistic regression models. The models’ estimates were applied to the test datasets, and discrimi-nation and calibration statistics were calculated to judge goodness-of-fit. Unadjusted mean P/AI were 31%, 28% and 28% for the HRE, REI and COMBO training datasets. For the HRE model, estimated P/AI ranged from 20% (≥3 different HRE) to 30% (healthy). The estimated P/AI associated with the 4 REI categories were not different from 27% in the REI model. The estimated P/AI associated with the combinations of HRE and REI in the COMBO model varied from 18% after ≥2 HRE and >200-400% REI, to 30% when insemina-tions were in healthy cows with REI >600%. Inseminations in older cows, in the spring, and outside 18-24 d after the previous insemination were also associated with lower esti-mated P/AI. The area underneath the Receiver Operating Characteristic curve ranged from 0.57 (COMBO) to 0.60 (HRE) for the test data, indicating fair discrimination ability of the models. The Brier score ranged from 0.19 to 0.21, indicating moderate performance of the prediction models. Calibration plots showed that the prediction models produced unbi-ased estimated P/AI. In conclusion, the results showed no conclusive evidence of greater estimated P/AI related to greater REI as a measure of estrus activity. More health-related events were associated with lower estimated P/AI. Combinations of low REI and more HRE were associated with notably decreased estimated P/AI. The logistic regression mod-els produce unbiased estimated P/AI. These predictive models may inform insemination and culling decisions in organic dairy cows. A variety of goodness of fit statistics were calculated to allow comparisons of the current logistic regression analyses with future analyses made by other machines learning techniques.