Effects of plant protein diets on health and welfare of turbot juveniles Psetta maxima

This study aimed to assess the health and welfare of turbot juveniles fed with experimental diets containing graded levels of a mixture of plant protein, by means of blood chemistry analyses including stress and metabolic indicators. Four extruded isonitrogenous (51%) and isolipidic diets (16%) were manufactured by Skretting ARC, Stavanger, Norway: a reference diet (diet D) contained 50% fish meal and three other diets in which fish meal was progressively reduced to 35% (diet C), 20% (diet B) and 5% (diet A) by proportionally increasing the level of wheat gluten meal, soybean meal and soy protein concentrate. Turbot juveniles (Initial weight 9.68 g ± 0.98 g) were reared in 500 l square recirculating tanks at 18°C (n= 55 fish per tank). The trial was performed in triplicate and lasted 9 weeks. Blood was withdrawn from the caudal vein, serum was collected after centrifugation and stored at -80°C for analyses of cortisol (COR), glucose (GLU), osmolality (OSM), non-esterefied fatty acids (NEFA), triglycerides (TAG), cholesterol (CHO), total protein (TP), urea (BUN), aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phophatase (ALP). Biometric measurements and autoptical analysis were also performed. One-way ANOVA and discriminant analysis were applied to data set to test the effects of dietary treatment. Experimental diets significantly affected growth and metabolic status of turbot juveniles (ANOVA p<0.05 for PT, LT, COR, TAG, CHO, NEFA, TP, BUN, ALP). At the end of experiment, a progressive reduction of size and serum concentration of TAG, CHO, NEFA, TP, BUN levels were observed in fish fed C, B and A diets, whereas a significant cortisol increase was found only in fish fed diets containing equal or less than 20% of fish meal. Although not significantly, the levels of GLU and ALT were higher in A and B diets compared to C and D. Only CHO, NEFA and BUN were different between C and D diets. Osmolality as well as hepatosomatic and splenosomatic indices did not change among diets. Discriminant analysis of all data set shows a separation of the groups fed the experimental diets compared to reference group along a physiological gradient (x-axis) according to the following variables COR, GLU, ALT, TP, BUN, CHO and TAG. Physiological status of group C was intermediate and more similar to group D than the other groups. The diets A and B induced physiological stress, metabolic depression and growth reduction. These effects were worst in fish fed with diet A than diet B. In addition the ALT leakage into the blood that occurred in these fish could indicate a possible liver dysfunction. These results could indicate a minimum threshold of 20% for fish meal in the diets to avoid negative effects on the whole body composition, nitrogen retention and transaminase enzyme activities. The final body weight of turbot fed diet C with 15% of fish meal replacement reduced (17%) compared to diet D as a consequence of reduction in the feed intake (data not shown). The clinical parameters gave no indication of fish stress, suggesting a small reduction of lipid metabolism and protein catabolism. Results on growth and metabolic status of turbot obtained with diet C are encouraging. Further investigations are need in order to evaluate the potential influence of this experimental diet on immune function also by extending the duration of the experiment, to gain a comprehensive evaluation of the health status in this species.