Evaluating the efficiency of a progesterone radioimmunoassay to predict pregnancy and fetal numbers in ewes.

Accuracy of pregnancy diagnosis in late gestation is essential for adjusting nutritional demands and management protocols in sheep production. The objective of this study was to evaluate the efficiency of progesterone (P4) concentrations measured via radioimmunoassay for determining pregnancy status and estimating the number of lambs expected. Blood samples were collected via jugular venipuncture at day 100 of gestation (term approximately 150 d) from 423 ewes over an 8-year period. Serum P4 concentration thresholds for predicting lamb numbers were established: non-pregnant (0 to 3.8 ng/mL), single lamb (3.9 to 12.5 ng/mL), and multiple lambs (12.6 ng/mL and above). Thresholds were compared with actual lambing outcomes to validate the effectiveness of the assay in detecting number of lambs. We hypothesized that using these P4 thresholds would provide reasonable accuracy in predicting the number of lambs, enabling more precise resource allocation and improved management decisions. The P4 radioimmunoassay demonstrated 99% accuracy for pregnancy detection and 73% overall accuracy in predicting number of lambs. Multiparous ewes demonstrated greater accuracy (84%) than nulliparous ewes (73%) in prediction of lamb number. Maternal serum P4 was positively associated with the number of lambs born (P < 0.001). The P4 × lamb birth weight interaction (P = 0.02) indicated that lamb birth weight modified the association between maternal P4 concentration and the number of lambs born, providing a biological explanation for some of the overlap in P4 concentrations between singleton- and twin-bearing ewes. Accurate prediction of pregnancy status and fetal number can improve flock management by identifying non-pregnant ewes for culling and enabling nutritional and management strategies tailored to ewes carrying single or multiple fetuses, ultimately improving animal welfare, productivity, and profitability. The reliability of the model across parities and years supports its efficacy in mixed-age flocks.