7th World Congress on Genetics Applied to Livestock Production, August 19-23, 2002, Montpellier, France CHANGES AND VARIATION IN BODY CONDITION MEASURES OF A LACTATING COW I.L. Mao, K. Sloniewski, P. Madsen, J. Jensen and N.C. Friggens Danish Institute of Agricultural Sciences (Foulum) DK-8830, Tjele, Denmark INTRODUCTION The fundamentals of efficiency for production and metabolic stress in lactating dairy cows involve the interplay of energy expenditure and energy sources during lactation. Both milk yield and dietary intake, their amounts and compositions, can be recorded readily, at least on experimental farms, but body reserves of a living lactating cow cannot be measured. Body reserves is also a function of amount, or weight, and composition, or condition. Weight can be measured, but condition can be only approximated. Body condition score (BCS) is used to approximate the body condition of a dairy cow, with its crudeness unknown because of its subjectivity. It reflects largely adipose tissue stores and not protein reserves, but this is only speculation. Ultrasound measure of muscle (UTM) and backfat (UTBF) may be more objective measures, but indicators of exact what aspects of body reserves remain unknown. We in Denmark are in the process of attempting to profile body energy state of a lactating cow using repeated measures of several body condition related variables such as BCS, UTM and UTBF, and laboratory results of biochemical parameters from liver biopsy and blood samples in a data set from an experimental farm. This paper presents the changes in BCS, UTM and UTBF and their phenotypic and genetic variations during lactation. MATERIAL AND METHODS Data. Included in analyses was a total of 3247 BCS records on 241 cows from up to 5 lactations from July of 1997 to December of 2000. Scores taken beyond 411 days postpartum were excluded. The 3 rd to 5 th lactations were grouped as parity 3+. There were 1074 records on Danish Red cows, 1414 on Holsteins and 759 on Jerseys. A total of 1618 records with 202 recording dates were in parity 1, 847 with 181 in parity 2, and 782 with 123 in parity 3+. The ultrasound measures, from August of 1991 to October of 2000, had a total of 8707 both UTM and UTBF on 158 Danish Red, 190 Holstein and 100 Jersey cows in the three lactation groups. Models, methods and algorithms. Below analyses were done for BCS, UTM and UTBF individually. Preliminary analyses using SAS were done on a variety of models to test all the fixed effects in the experimental design such as breed, genetic lines, feeding level, parity, days- in-milk (DIM), and scoring date. They were done also to screen fixed interaction effects. Random effects in these fits were cows and cow by parity interactions, the results of which were to provide priors for subsequent analyses. The fixed effects were not of primary interest, and the results are not presented in this paper. Within each of the three parities by each of the three breeds, data was analyzed by univariate linear mixed random regression models. The fixed effects in the model were two genetic lines, Session 01. Breeding ruminants for milk production Communication N° 01-31