Diversity of performance patterns in dairy goats: multi-scale analysis of the lactation curves of milk yield, body condition score and body weight

In the dairy goat sector, reduced longevity is a key issue leading to higher replacement rates in the herd and a poor dilution of doe-rearing costs. There is a need to better understand the determinants of lifetime performance. Thus, the general objective of this work was to analyze the phenotypic variability of lifetime trajectories (milk yield (MY), body weight (BW) and body condition score (BCS)) through a 3-step approach: (1) characterize individual phenotypic lactation curves, (2) explore the associations between MY, BW and BCS curves at the lactation scale and (3) assess the diversity of phenotypic curves over successive lactations. Routine data from two experimental farms: Le Pradel (Dataset 1, Ardeche department, France) and MoSAR experimental farm (Dataset 2, Yvelines department, France) were used. Dataset 1 included 793 Alpine goats from 1996 to 2020. Dataset 2 included 339 Alpine and 310 Saanen goats from 2006 to 2022. Weekly MY records (Dataset 1) and daily MY records (Dataset 2) were fitted using a lactation model with explicit representation of perturbations. Monthly BW records (Dataset 1) and BCS records (Dataset 1&2) were fitted using the Grossman multiphasic model. Daily BW records (Dataset 2) were fitted using a weight model. Each individual lactation curve modeled for MY, BW and BCS was thus summarized by synthetic indicators of level and dynamics. Principal component analysis was performed on the MY, BW and BCS indicators separately, and clusters of phenotypic curves identified. At the lactation scale, associations between MY, BW and BCS clusters were evaluated by contingency tables with a chi-square test. Lifetime-scale bar plots were used to display cluster changes throughout parities. For MY curves, 4 and 3 clusters were found for primiparous and multiparous goats respectively. For BW, lumbar and sternal BCS curves, 3 clusters were found for all parities. At the lactation scale, no major association was found among phenotypic curves suggesting a diversity of energy partitioning strategies between life functions. At the lifetime scale, change among clusters occurred primarily between first and second lactation, whereas a pattern of stable cluster membership appeared for multiparous goats. Further analyses are needed to include reproductive performance in analyzing lifetime performance clusters, to better identify clusters or combinations of clusters at risk for culling.