Variability in the composition of lupin (Lupinus angustifolius) meals influences their digestible nutrient and energy value when fed to rainbow trout (Oncorhynchus mykiss)

Abstract Over a series of seven separate experiments 76 different lupin ( Lupinus angustifolius ) meals were assessed for their digestible dry matter, protein, amino acid and energy characteristics when fed to rainbow trout. Two reference diets (a common basal diet and a reference lupin meal) were also included in each experiment. Minimal variance in the digestibility parameters of both reference diets was observed among the experiments ensuring that there was a high degree of robustness in the across-experiment evaluations. Using simple and multiple-regression techniques, principal diet and ingredient composition factors affecting diet and ingredient digestibilities and ingredient digestible values were explored within the dataset. Using simple linear regression, it was shown that nitrogen digestibility of the lupin meals was negatively influenced by ingredient lignin content, but positively affected ingredient protein ( N  × 6.25) content. The energy digestibility of the lupin meals was positively affected by a range of compositional features including protein, sum of amino acids and negatively affected by carbohydrate content. The digestible nutrient and energy content of the meals reflected the combined effects of both ingredient digestibilities and ingredient composition. The digestible nitrogen content of the lupin meals was positively affected by protein ( N  × 6.25), sum of amino acids and energy content, but was negatively affected by lignin and carbohydrate content. The digestible sum of amino acids was also positively affected by protein, sum of amino acids, but only negatively affected by carbohydrate content, not lignin content. The digestible energy content of the lupin meals was also positively affected by protein, sum of amino acids and their own energy density, but only negatively affected by carbohydrate content. Multiple linear regression modelling supported that together ingredient protein and lignin content were the strongest predictors of digestible protein value, explaining close to 60% of the variability in this parameter. Discrete reassessment of these relationships using non-linear analysis methods provided a stronger interpretation of the ingredient composition effects. In recognition of this, a functional model including terms for ingredient protein and lignin content on the digestible protein value was defined. This study demonstrates that within one raw material type that not only does significant variability in the digestible value of the raw materials exist, but that it is possible to identify compositional features of that raw material that are intrinsically influencing their own digestible value.