Comparison among methods of nutritional evaluation of dietary ingredients for rabbits

Abstract

A collaborative study has been carried out in four laboratories (UPV, UPM, CLO, INRA) with the aim of comparing different evaluation methods of feedstuffs for rabbits. Fifteen diets were designed to assay six ingredients using at least two of the following four evaluation methods: direct, substitution of a basal diet (BD), substitution of a reference feedstuff (alfalfa meal (AM)) by difference and extrapolation, and multiple regression. The experimental diets were supplied to a total of 202 New Zealand White or New Zealand×Californian growing rabbits (10–13 rabbits per diet) for assaying the digestibility according to the reference European method. The dietary digestible energy (DE) and digestible crude protein (DCP) determined were used to estimate the nutritional value of the six ingredients. Despite the differences between laboratories in the gross energy digestibility ( P=0.056) of the basal diet 1, which was used for estimating grape pulp (GP) by substitution of a basal diet or of alfalfa meal, no differences were found ( P>0.05) between the DE of grape pulp estimated by both methods (5.55 MJ/kg DM versus 5.27 MJ/kg DM). However, its energy value obtained by the direct method was significantly higher (7.41 MJ/kg DM, P<0.05) than by substitution, and the DE estimated by multiple regression (6.42 MJ/kg DM) presented an intermediate value and not significantly different ( P>0.05) from them. The DCP of grape pulp estimated by substitution and directly was not significatively ( P>0.05) different from zero, however, an estimation of 63 g/kg was obtained by multiple regression. There were no significant differences ( P>0.05) between the DE and DCP values estimated by the substitution or by the multiple regression (MR) methods for wheat bran (WB), wheat, full-fat soya and sunflower meal. Similarly, alfalfa meal had similar values when estimated directly and by multiple regression method (7.29 MJ DE/kg DM versus 7.43 MJ DE/kg DM, 93.1 g/kg DCP versus 91.8 g/kg DCP, respectively). The additivity of the DE and DCP values estimated by the different methods were tested by means of a residual study (difference between determined and estimated dietary values). The values estimated by substitution showed good additivity (from 0.95 to 1.62% of error for DE and from 4.45 to 4.60% for DCP). When values determined directly for grape pulp and alfalfa meal were evaluated, the error increased to 2.15% for DE, however, no change was observed for DCP. Based on the residual differences it seems that the DE of grape pulp and alfalfa meal was over and underestimated, respectively, by the direct method. The estimations from the multiple regression method under evaluate the dietary DE and DCP values, showing higher residues and as a consequence higher errors (3.9% for DE and 9.6% for DCP) than substitution, the reason of these high errors could be the correlation between some ingredients in the independent data sets used for multiple regression.