OBJECTIVE METHODS FOR ESTIMATION OF THE L EAN MEAT CONTENT OF PORK BELLY 1ST COMMUNICATION - DERIVATION OF REGRESSION EQUATIONS BY USING DIFFERENT CARCASS TRAITS

In the present study the suitability of regression equations derived f rom different kinds of breeds and samples to estimate the belly lean c ontent is verified. The comparison was based on regression equations d erived from the data sets from Grub and Bonn and on the relationship b etween two different kinds of methods to assess the quality of the bel ly: regression equations, using different carcass traits and video ima ging of the belly cut between 13th and 14th rib.The results were as fo llows: Because of a larger sample size, lower MSE, higher slaughter we ight and castrates used in the derivation of the dam breed functions, the new regression equations developed from the Grub data are probably more accurate in comparison to the so far applied functions derived f rom the Bonn data. Although the accuracy of the Pi regression equation s with a coefficient of determination of 58% was relative low, the ide ntification of Pi slaughter pigs having a poor belly quality should be possible. As shown by the example of the data recorded in Bonn, the a pplication of the regression equations derived from the data from Grub to different breeds should be possible. Depending on test station and breed the coefficients of correlation between the currently used meth ods to assess the quality of the belly - video imaging of the belly cu t and regression equation, using different carcass traits - are betwee n 0.61 and 0.81, hence the comparability of both methods is limited.