Prediction of body composition in rabbit females using total body electrical conductivity (TOBEC)

The purpose of this experiment was to evaluate the accuracy of a total body electrical conductivity technique (TOBEC; EM-SCAN Inc., Springfield, IL, USA) to estimate in vivo body composition of steady state rabbit females. Eighty-four female rabbits (New Zealand X Californian, weight range: 3126-5465 g) at different physiological states were used in the experiment. After TOBEC assessment (five passages through a SA-3203 type chamber), animals were weighed, slaughtered and stored at - 20 degreesC until chemical analyses (dry matter, lipids, proteins, ash and energy). A multiple regression analysis was employed to develop appropriate prediction equations for chemical composition using TOBEC assessment (E-value) and weight of the animals as independent variables. The mean variability of the E-value for an animal (CV) was 1.21%. Very weak correlations were obtained with the prediction equations of proteins and ash (R-2 < 0.55). On the other hand, TOBEC appears to be accurate in predicting water and energy in reproducing rabbits with high R-2 (greater than or equal to 0.84) and low CV (less than or equal to 4 and less than or equal to 12%, respectively) for these parameters. The high correlation between TOBEC values and body lipids (R-2 greater than or equal to 0.82) validates the method in predicting this parameter but the correlation is associated with a high CV value (24%) which limits its reliability. The inclusion of the physiological state of the doe as a fixed effect in the statistical model was not advantageous except for prediction of body proteins. A group of 25 rabbit does was used to validate the predictive equations, showing that TOBEC is an accurate method to predict water, energy and lipid body content. (C) 2002 Elsevier Science B.V. All rights reserved.