Influence of dietary inclusion of a wet processed faba bean protein isolate on post-smolt Atlantic salmon (Salmo salar).

Legumes such as soybean, peas and lupin have attracted considerable interest as potential sources of protein to replace finite and limiting supplies of marine fishmeal (FM) as major ingredients for aquafeeds. In this respect, faba beans ( Vicia faba ) represent a widespread and relatively unexploited legume crop in Europe with potentially favourable characteristics. However, for carnivorous species such as Atlantic salmon ( Salmo salar ), protein levels in legumes are generally too low and require to be concentrated to be direct replacements for FM. Previously we showed that a faba bean protein concentrate, produced by air classification, containing 55% protein could partially replace FM and/or soy protein concentrate (SPC) in feeds for parr and post-smolt salmon. In the present study, a faba bean protein isolate (BPI), produced by a wet process, with almost ~ 80% crude protein was investigated in feeds for Atlantic salmon in seawater. Four dietary treatments were tested including one with high inclusion of FM (400 g kg − 1 ) and three with low FM (216 g kg − 1 ) and increasing inclusions of BPI (0, 70 and 140 g kg − 1 ) substituting for SPC (236, 125 and 45 g kg − 1 ). Growth performance in fish was unaffected with the lower level of dietary BPI, but was reduced in fish fed the higher level, mainly due to feed intake being reduced initially. Histological analysis of the distal intestine showed inflammation in fish fed both diets containing BPI, but especially at 140 g kg − 1 . The high dietary level of BPI affected the transcriptome of pyloric caeca with almost 2000 differentially expressed genes (DEG) compared to fish fed FM, whereas fish fed SPC or the lower level of BPI showed no DEG compared to fish fed FM. In contrast, the liver transcriptome was generally affected similarly by both BPI and SPC. The combined data suggested that the BPI utilised contained a factor that was detrimental above a certain threshold and, although this factor could be an artefact of the protein isolation method, an effect of a known ANF could not be excluded with saponin the most likely candidate. Overall, however, the results of the present study confirmed that protein concentrates or isolates derived from faba beans can replace FM and/or SPC up to a certain level in feeds for Atlantic salmon. Statement of relevance A study testing novel raw materials for fish feed. [ABSTRACT FROM AUTHOR]