Meta-analysis to predict the effects of extended lactations on milk production and reproductive performance of dairy cows.

The effects of extended lactation (EXTL) on milk production and reproductive performance of dairy cows were quantified using a meta-analysis approach. A systematic literature review was conducted to identify studies that compared milk yield and reproductive performance between cows managed within EXTL systems and those managed under conventional lactation length, resulting in 21 studies eligible for inclusion in the meta-analysis. Data collected from those studies included calving interval (CInt), voluntary waiting period (VWP), parity, breed, farm type (experimental vs. commercial), and whether cows were selectively assigned to EXTL based on production and cow characteristics or randomly allocated. Corresponding treatment means were extracted for energy-corrected milk yield per calving interval day (ECM/CInt), first service conception rate, and number of inseminations per conception. Mixed-effects models were developed for each outcome, with study as a random effect. The models included CInt as the main predictor for ECM/CInt and VWP for first service conception rate and number of inseminations per conception, along with other covariates that were retained if significant. A separate subgroup analysis was conducted to compare studies using selective versus random assignment to EXTL. The relationship between CInt length and ECM/CInt differed by parity, where each 10-d increase in CInt was associated with a 0.08 kg/d increase in ECM/CInt in primiparous cows, while no effect of CInt length on ECM/CInt was observed in multiparous cows. Selective assignment of multiparous cows to EXTL, based on individual cow characteristics and production traits, tended to be associated with 1.9 kg/d higher ECM/CInt compared with randomized allocation, although production of cows selectively assigned to EXTL was similar to that of conventionally managed cows. In addition, criteria used to identify cows suitable for EXTL were variable between studies, and no consistent or clearly effective selection method could be determined. Reproductive performance improved with extended VWP across parities; each 10-d increase in VWP was associated with a 0.02 unit decrease in inseminations per conception and higher odds of first service conception (1.1 × in multiparous cows and 1.04 × in primiparous cows). Overall, the results of this meta-analysis indicate that EXTL improves measures of reproductive performance in both parity groups and increases ECM/CInt in primiparous cows, whereas the response in milk yield of multiparous cows depends on whether the EXTL strategy was selectively implemented. These results indicate that EXTL, if implemented, should be applied as an individualized rather than a herd-level strategy.