In vivo in vitro digestibility estimations of poutlry meal in Atlantic salmon (salmo salar)

Rendered animal-by products (i.e. poultry meal) have a high protein content, no anti-nutritional factors and a favourable amino acid composition compared to plant protein sources. Nevertheless, an impediment limiting their efficient utilisation emanates from high levels of variability in meal digestibility on a batch to batch basis. As a consequence of well documented logistical and economic reasons, alternative methods for the estimation of ingredient digestibility are required, and these have to be cheap, reliable, time-effective and efficient. In this study 22 different poultry meal samples were sourced (11 commercial samples and 11 specifically designed meals of varying raw material freshness and cooking conditions) and their digestibility was evaluated using three contrasting techniques. The digestibility methodologies tested consisted of an in vivo feeding trial using Atlantic salmon reared at two different temperatures (10C° and 20C°) and two in vitro methodologies: pepsin digestion chambers and a bioreactor with a semi-permeable membrane. An in depth chemical characterisation of samples, including amino acid racemization rate, was implemented. In vivo apparent protein digestibility coefficients of the 22 samples ranged from 5 to 108%, and a larger variation of individual poultry meal apparent protein digestibility values was observed at 20°C compared to 10°C. Cooking temperature in meals prepared with fresh raw material had no significant effects on digestibility. However increasing cooking temperatures were found to significantly decrease the digestibility of meals prepared with aged raw materials. All data originating from this trial will be summarised and discussed towards developing improved poultry meal production practices and favouring uptake of this product by the aquafeed sector.