Three methods for estimating in vivo body fat in sheep, by the assessment o
f body water using the deuterium oxide (D2O) dilution technique, adipose ce
ll size (ACS) measurement and body condition scoring (BCS), were studied in
mature ewes and compared to results obtained after slaughter and chemical
analyses. Twenty Lacaune dairy ewes were slaughtered at three stages of lac
tation (30 days [n = 8], 60 days [n = 8] and 110 days [n = 4]), with a mean
body weight (BW) of 72.7 +/- 7.6 kg, 16.0 +/- 4.7 kg of body lipids (LIP)
and 42.7 +/- 3.8 kg of total body water. Body fatness ranged from 10.3 to 2
8.8% of empty body weight. Although the ewes were relatively fat, the mean
adipose cell volume of subcutaneous pericaudal or sternal adipose tissue (2
09 or 287 pl, respectively) was in the lower range of previously reported v
alues, while cells from omental adipose tissue had a larger mean volume (80
4 pi) than has been previously reported. Due to differences in the frame of
ewes, prediction of body lipids through D2O, ACS and BCS was improved by i
ncluding body weight in the model. The best equations using ACS were obtain
ed with subcutaneous fat tissues (sternal tissue: residual coefficient of v
ariation (rCV) = 13.9% and pericaudal: rCV = 17.8%), followed by internal t
issue (omental: rCV = 19.5%). The rCVs of the prediction of body lipids wit
h these in vivo methods were in general agreement with published works on c
attle or sheep i.e. D2O: 8.8 %, pericaudal ACS: 13.9 % and BCS: 16.7%. The
interest and limitations of the methods studied are discussed in terms of d
ifferent experimental objectives and constraints. ((C) Elsevier / Inra).