Retrospective analysis of calving interval data from seasonal-calving pasture-based New Zealand dairy herds

This retrospective cross-sectional study described calving intervals (CIV) for spring-calving New Zealand dairy farms, evaluated the effects of cow-related factors, and investigated potential associations between CIV and future reproductive performance. Individual cow data were extracted from a veterinary dairy herd database for farms with at least 2 yr of data between June 2019 and May 2023. For the CIV analysis, observations were excluded from male animals, with >1 recorded calving event for a cow in a season, without a CIV, if the cow was born after June 1, 2020, or if CIV was <300 d. A descriptive analysis of CIV data was carried out, then herds with alternative calving patterns, cows that calved outside spring, and cows that had 2 consecutive calvings that skipped one or more seasons (carry-over cows) were excluded from the CIV analysis. A mixed multivariable linear regression model was constructed to evaluate associations between CIV and potential explanatory variables (age, breed, dairy season, and milk yield). A separate dataset was constructed to evaluate effect of CIV on odds of future reproductive success (a recorded calving event in the subsequent spring calving season). The same inclusion criteria were applied as for the analysis of CIV, except cows with only 2 dairy seasons of data or born after June 1, 2019, were excluded but carry-over cows were included. A mixed multivariable generalized linear regression model was constructed to evaluate associations between odds of reproductive success and the potential explanatory variables of previous CIV, age, breed, season, and mean milk solids yield. Mean (SD) CIV was 368.8 (23.96) d for spring-calving herds, compared with 381.9 (53.6) d for herds with alternative calving patterns. Calving intervals were longer for herds that carried some cows over between seasons (mean [SD] 376.3 [58.3] d) compared with herds that did not (mean [SD] 366 [23] d). The regression model confirmed that mean CIV varied between seasons, was shorter for cows with higher milk production, and the interval between first and second calvings was 7 d longer than the interval between subsequent calvings. Most variation in CIV occurred within the herd at the cow-season level. Probability of reproductive success (adjusted for age, season, and milk production) declined with CIV. When averaged across farms, age groups, seasons, and milk production, the predicted probability of a successful mating declined from 0.69 (95% CI = 0.66-0.72) at a CIV of 357 d, to 0.05 (95% CI = 0.03-0.07) at 538 d. Calving intervals could potentially be used to identify animals at higher risk of reproductive failure within seasonal-calving dairy herds. However, causal inferences are limited by the retrospective study design and limited information available. In seasonal-calving systems, CIV is closely linked to calving date, so prospective studies explicitly adjusting for calving date are warranted and we interpret CIV primarily as a timing-influenced, operational indicator that complements calving date for within-herd screening rather than a standalone causal determinant.