Evaluation of Different Test-Day Milk Recording Protocols by Wood’s Model Application for the Estimation of Dairy Goat Milk and Milk Constituent Yield

SIMPLE SUMMARY: Considering the seasonal trend of milk production due to application of estrus synchronization treatments in autumn or in spring season, as well as to meet meat production and market requirements for religious holidays, one may find it useful to simplify the milk and milk constituent yield quantification during the entire period of lactation. The present work aimed to verify the application of Wood’s model to a different number of test-day milk recording protocols in order to reduce the frequency of National Breeding Association test-day milking record protocol. Wood’s model fit well on goat milk production, showing high R(2) values. Moreover, the milk yield, fat-corrected milk yield, fat yield, and protein yield estimation applying Wood’s model to four recordings (first, second, fourth, and sixth month records) showed results superimposable to International Committee of Animal Recording (ICAR) application and Wood’s model application on test-day records every 15 days. ABSTRACT: Goats have important social and economic roles in many countries because of their ability to survive and be productive in marginal areas. The overarching aim of this study was to compare the application of Wood’s model to different test-day milk recording protocols for estimation of total milk, fat, and protein yield in dairy goats. A total of 465 goats were used (Garganica, 78; Girgentana, 81; Jonica, 76; Maltese, 77; Red Mediterranean, 76; Saanen, 77). Milk yield was recorded every 15 days throughout lactation of 210 days, for a total of 14 collection days, during both morning and afternoon milking sessions. Milk samples were collected and analyzed for protein and fat. The fat-corrected milk was standardized at 35g fat/kg of milk. Wood models showed high R(2) values, and thus good fitting, in all the considered breeds. Wood model applied to first, second, fourth, and sixth month recordings (C) and ICAR estimation showed total milk yield very close to Wood’s model applied to all 14 recordings (A) (p > 0.38). Differently, Wood’s model applied to the first, second, third, and fourth month recording (B) estimation showed great differences (p < 0.01). This could be applied for farms that had the necessity to synchronize flock groups for kidding in order to produce kid meat. In farms that apply the estrus induction and/or synchronization for kidding, it would be possible to perform only four test-day milk recordings and to apply the Wood’s model on them in order to obtain the estimation of total milk, fat, and protein yield during lactation for animals inscribed, or to be inscribed, to the genealogical book.