A wide range of species are produced in aquaculture , which implies a great variability among aquatic species in terms of nutritional requirements compared to what is seen in livestock production. This poses a great challenge for fish nutritionists and aquafeed producers. Objective criteria, based on the use of body composition and nutrient/energy budget modelling, can be used to explore similarity relationships between species and thus contribute to mitigate this challenge.
To predict the body composition in fish, usually isometric and/or allometric models are used. The isometric model assumes fish composition is proportional to their weight, while the allometric model accounts for a non-proportional relationship between the two variables . In nutrition studies, assumptions of isometric behaviour for protein and/or ash in fish may not hold for dynamic body composition evaluations.
Nutrient-based models, consider energy and protein intake to predict growth and body composition. These models assume that maintenance metabolic expenditure can be defined by exponents, which determine the change in metabolic rate as a function of body weight . However, there is a disagreement among authors regarding the definition of these exponents for the metabolic weight of fish, which makes it unclear whether the same exponents should be applied to different species.
In this work, commercial fish species such as, Atlantic salmon ( Salmo salar) , gilthead seabream ( Sparus aurata) , European seabass (Dicentrarchus labrax) , Nile tilapia (Oreochromis niloticus) , rainbow trout (Oncorhynchus mykiss) , turbot (Scophthalmus maximus) and Senegalese sole (Solea senegalensis) were compared based on their body composition and growth analysis . The goal was to assess certain assumptions and hypotheses present in the literature, such as the use of isometric or allometric models to describe body composition, the applicatiom of "universal" or species-specific metabolic body weight exponents, the use of constant or non-constant protein/energy efficiency ratios, and the assumption of similarity between species with similar morphological and physiological traits (e.g., flatfish species).
Materials and methods
Data was collected from literature on the body composition and growth for the aforementioned species . For analysis of growth parameters, datasets on in vivo g rowth trials were used to estimate fasting maintenance and retention efficiency parameters, through linear regression techniques , for all species , except for sole due to limited data.
The study evaluated the similarity relationships between species in terms of body composition and protein/energy budgets using principal components analysis (PCA ). The respective parameter estimates of the species were plotted along the first two principal components and scatter plots of the individual parameters were generated. Additionally, the predictions of the calibrated models (species-specific) were compared between different species.
PCA analysis shows that seabream and seabass share similarities in their body composition parameters, as do salmon and trout. However, the flatfish species, turbot and sole, seem to have different body composition parameters. In turn, sole appears to be more similar to salmonids , whereas turbot displays certain similarities with tilapia.
In terms of general growth parameters, it seems that tilapia, seabream, salmon, and rainbow trout have some similarities , while seabass and turbot seem more different. Salmon and rainbow trout are closer to each other, while tilapia is closer to seabream. Turbot has the most distant estimations for fasting maintenance costs. Salmonids have higher retention efficiency for protein and energy, while Mediterranean species have lower retention efficiency for protein. Seabream has higher energy retention efficiency than seabass, but both have low protein retention efficiency.
Discussion and Conclusion
Results suggest that the similarity within salmonids and Mediterranean species are strong and consistent across parameters and thus clearer in the PCA projections. D espite having some differences in growth, Mediterranean species show similarities, especially in early life. However, flat species do not group as clearly. The variability observed between species in terms of model parameters and predictions may be related to taxonomy, physiological stages, ecological features, fish activity, body mass, as stated in previous studies. Results also suggest the metabolic body weight exponents for energy and protein are likely to be species-specific and differ between species , ranging from 0.6 to 0.9, which agrees with previous studies that challenge the theory of universal metabolic allometry.
This study enables the evaluat ion of certain assumptions in literature. Moreover, it shows that species belonging to the same family, geographical area, or with similar diets or body shapes may or may not exhibit similar body composition and growth patterns . Therefore, conducting the analysis of body composition and growth parameters between species is useful , as they complement each other, providing a full view of their developmental patterns, and allows to identify both similarities and differences between the species.
A. Raposo acknowledges financial support by Grant PD/BDE/150525/2019 (SANFEED Doctoral program, with support by FCT and SPAROS Lda , Portugal). This work was also funded by project 47175_FICA, supported by Portugal and the European Union through FEDER/ERDF, COMPETE 2020 and CRESC Algarve 2020, in the framework of Portugal 2020.