Data paper: Milk production, body weight and body condition score during lactation in dairy goats: time-series data and model-based curves from two French herds

Abstract

We collected data from routine measurements carried out on two experimental farms: the MoSAR experimental farm (Yvelines, France) and the Le Pradel experimental farm (Ardèche, France). At MoSAR, milk yield (MY) was recorded at each milking using an automated system. BW was measured with a walk-over scale installed in the milking parlour, providing BW readings at each milking, which were then averaged to obtain daily BW values. Body condition score (BCS) was scored individually at the lumbar and sternal regions on a scale of 0–5. At Le Pradel, MY was recorded individually according to the national performance control method, while BW was measured using a standard scale. BCS assessments at this farm followed the same methodology as MoSAR. The dataset includes MY, BW, lumbar BCS (BCS_L) and sternal BCS (BCS_S) data collected from 1996 to 2020. Dataset 1 included 1 442 goats (1 132 Alpine and 310 Saanen) with a total of 4 068 lactations. After filtering outliers and selecting goats based on criteria established in a previous study, we obtained datasets 2–5. Dataset 2 contained 400 520 MY records from 3 421 lactations, while Dataset 3 contained 216 269 BW records from 2 925 lactations. Dataset 4 contained 18 848 BCS_L records from 1 462 lactations. Dataset 5 contained 20 510 BCS_S records from 1 630 lactations. Time series data for each lactation were modelled using different approaches tailored to record frequency (daily, weekly, or monthly), following a methodology developed in a previous study to characterise phenotypic dynamics. These historical data sets are unique due to their longitudinal nature, covering a large goat population. In particular, the MoSAR experimental farm provides high-frequency MY and BW measurements, making these datasets particularly valuable for precision livestock research.