EVALUATION OF THE NON INVASIVE TOBEC (TOTAL-BODY ELECTRICAL-CONDUCTIVITY) PROCEDURE FOR PREDICTION OF CHEMICAL-COMPONENTS OF MALE BROILERS WITH SPECIAL CONSIDERATION OF DIETARY-PROTEIN LEVEL

TOBEC (total body electrical conductivity) measurement as a non invasi ve procedure for the estimation of body chemical composition was used to calculate calibration curves for the prediction of crude protein ma ss (CPM), crude water mass (CWM), crude ash mass (CAM) and fat-free ma ss (FFM) of male broiler chickens. A growth experiment with 3 protein levels (130, 230 and 330 g CP/kg diet, isoenergetic with 13.3 MJ AME(N )/kg) was combined with TOBEC measurements and body chemical analysis in order to obtain the values necessary for calibration. A total of 19 6 TOBEC measurements and body chemical analysis were undertaken in tim e intervals of two days beginning with hatch until day 17 of age. Diff erent dietary protein levels resulted in marked differences in body we ights and body chemical compositions but in similar TOBEC-responses fo r a given mass of FFM, CPM, CAM or CWM. Values for birds fed a diet wi th 130 g CP/kg diet tended to be more variable. Linear broken relation ships were found between FFM, CPM, CAM and CWM, respectively, and TOBE C values (E#). A set of different regression equations is given and yi elded high proportions of variance accounted for the piece wise regres sion model (R(2) ranged from 0.83 to 0.99). In spite of these high det erminations the prediction of crude fat mass (CFM) by subtracting the FFM from the body weight resulted in most cases in weak determinations between observed and predicted CFM (R(2) ranged from 0.38 to 0.86). T he highest R(2) was observed when the E# was expressed per unit metabo lic body weight to the power of 0.67 and regressed on FFM expressed to the same power. In conclusion, FFM, CPM and CWM may be predicted reas onably well by TOBEC. However, these high determinations are not high enough to predict CFM accurately. In addition, the application of such regressions to an individual bird seems to be impossible. Assessment for groups of animals should be possible if errors of estimation, stan dard deviations and differences to be detected are taken into account in the calculation of the number of birds necessary for the TOBEC meas urements.