ReviewAre fish what they eat? A fatty acid’s perspective
Introduction
With aquaculture’s rapid expansion and stagnating landings from wild fisheries, the relative contribution of farmed fish to the total fish consumption is increasing, with cultured products now being the largest contributor of seafood intake in humans [1]. Notably, fish are the main source of long-chain polyunsaturated fatty acids (LC-PUFA, >C18) of the omega-3 (n-3) series for human consumption, characterized in particular by docosahexaenoic acid (DHA, 22:6n-3) and eicosapentaenoic acid (EPA, 20:5n-3). It has been widely observed that the fatty acid (FA) profile in the edible tissues of farmed fish generally reflect those of the diets [2]. This means that fish diets (commonly referred to aquafeeds), via the modulation of fish FA profiles, will exert an increasingly significant effect on the intake of nutritionally beneficial n-3 LC-PUFA in humans.
In modern compounded aquafeeds, the inclusion of n-3 LC-PUFA-rich fish oil has decreased in unison with increased utilization of terrestrially sourced oils such as vegetable and rendered animal oils and fats [3]. This marked switch from marine to terrestrial oil sources has occurred in response to the limited availability, increasing cost, and perceived environmental concerns surrounding the utilization of traditionally used fish oil [4]. Novel alternative oils containing n-3 LC-PUFA are currently being developed and are at different stages of commercial implementation. Generally, these consist of genetically modified oilseed crops and single celled oils [5]. However, commercial scale-up and product pricing are currently viewed as major roadblocks for widespread commercial adoption. Considering the lack of n-3 LC-PUFA in current commercially available oils, an evident decrease in the n-3 LC-PUFA concentration of fish fillets is unavoidable [2,3]. Consequently, the characteristic FA of the alternative oils currently utilised, typically C18 PUFA (mainly 18:2n-6 and 18:3n-3), monounsaturated FA (MUFA), and saturated FA (SFA), have increased in commercial aquafeed formulations and, as a consequence, in farmed fish. By affecting the FA composition of farmed fish, the composition of aquafeeds clearly influences dietary FA intake in humans. Therefore, a greater understanding of the ‘diet-fish’ fatty acid relationship is central, and of paramount importance for the judicious future use of raw materials for inclusion in aquafeed, whilst simultaneously mitigating potential downstream risks to human health related to changes to aquafeed formulations.
Multiple factors influence the transfer of dietary FA into farmed fish. It is known that species-specific metabolic differences and FA-specific metabolic pathways, such as selective incorporation, β-oxidation, and de novo synthesis/lipogenic activity, influence the incorporation of dietary FA into fish tissues [4,6]. For example, a primary metabolic factor influencing LC-PUFA deposition relates to the LC-PUFA biosynthetic capacity of the species (Fig. 1). In general, it has been established that freshwater and euryhaline fish species can synthesize LC-PUFA, from their C18 substrates, i.e., 18:3n-3 and 18:2n-6, whereas most marine carnivorous fish cannot. Specifically, it is hypothesised that the abundance of LC-PUFA in marine ecosystems, which originate in a variety of lower trophic microbial, alage and invertebrate organisms, has rendered the LC-PUFA bioconversion pathways largely redundant [[6], [7], [8], [9]]. Conversely, freshwater and anadromous species have adapted to a relative paucity of dietary LC-PUFA and demonstrate a higher capacity for LC-PUFA biosynthesis. Based on a parallel hypothesis, it has been recently demonstrated that the de novo LC-PUFA synthesis capability of a species is correlated to, and possibly predicted by, the trophic level of fish, whereby fish occupying lower trophic levels (i.e., trophic level < 3) are capable of de novo LC-PUFA synthesis, whereas those occupying higher trophic levels (i.e., trophic level > 4) are unable, or exhibit a limited capacity to synthesize LC-PUFA from C18 precursors [10]. Notably, fish are an incredibly diverse taxonomic grouping, with roughly 460 species currently cultured globally, including 7 hybrids [11]. As a result, a considerable diversity in species-specific FA metabolism exists.
In addition to metabolic factors and dietary FA composition, environmental and other dietary factors are known to affect the incorporation of dietary FA into fish tissues by exerting influence on FA metabolism. Temperature adaption, sometimes termed homeoviscous adaptation, differs among fish species and influences the FA requirements and biosynthesis of that species owing to the integral role of FA in regulating cellular membrane fluidity [12]. Differences in FA deposition are also apparent within the same species, as changes in water temperature have been shown to modulate FA metabolism [13]. Further, salinity has been shown to affect LC-PUFA biosynthesis, especially in euryhaline species [[14], [15], [16]]. Dietary factors also affect general FA metabolism, including uptake, deposition and β-oxidation. For example, feeding duration, dietary lipid/energy levels, dietary protein/lipid ratio, and some micronutrients (such as carnitine, some amino acids, and some vitamins and minerals that act as co-enzymes/co-factors) reportedly affect the metabolism and deposition of FA into fish tissue [[17], [18], [19], [20], [21], [22], [23], [24], [25]]. Interactions among different dietary FA (for example, competition between EPA and arachidonic acid (ARA, 20:4n-6), and omega-3 sparing effects of MUFA and SFA) have also been reported to modulate FA metabolism and deposition [[26], [27], [28], [29], [30], [31]]. In addition, the digestive process and the subsequent absorption of FA is strongly influenced by both the chemical composition and the physical properties of the diet. Specifically, strong relationships exist between the digestibility of FA and dietary lipid source in many species, and this may be in complex interaction with environmental conditions, including water temperature [[32], [33], [34], [35], [36], [37]]. When examined as a whole, a multitude of aspects influencing nutrient digestibility, such as rearing condition, feeding strategy, total dietary fiber and ash content, and anti-nutritional factors, amongst others, present a myriad of complex and often inter-related factors that influence the incorporation of dietary FA into fish tissues [[38], [39], [40]].
A variety of models have been developed in an attempt to describe, and, in some circumstances, predict the ‘diet-fish’ fatty acid relationship. A dilution model was firstly introduced by Robin et al. [41] to describe the change in FA profile of brown trout (Salmo trutta) and turbot (Scophthalmus maximus) following a change in dietary lipid source. This simple, yet elegant, dilution model considers that dietary FA are incorporated into tissues without any mobilization or turnover of pre-existing ones. Essentially, this method considers fish as “inert containers”, whose FA composition is the direct sum of the initial FA composition of the fish itself, plus the FA composition of the oils “poured” into it, via the feed administered. This model appears applicable (and quite reliable) for some FA, yet less reliable for others, such as DHA, 18:3n-3, 20:1n-9, 18:1-isomers and 20:1n-11, which commonly undergo more extensive metabolism and/or selective retention [41]. Moreover, the dilution model proved more reliable when applied to neutral lipid (NL), than to polar lipid (PL) [41]. Further studies by Jobling [42,43] used the dilution model to verify its efficacy of predicting modifications of C18 FA in Atlantic salmon (Salmo salar) undergoing a dietary change from vegetable to fish oil. These studies showed that the proportions of the three tested FA (18:1-isomers, 18:2n-6 and 18:3n-3) in the total lipids of salmon tissues conformed closely to predictions made on the basis of the dilution model [44]. However, the applicability of this model across fish species was not uniform, and reductions in the accuracy of predictions were found to decrease when applied to Atlantic cod (Gadus morhua). Accordingly, the dilution model is now considered a useful model to describe FA changes in fatty fish species and fatty tissues (such as salmon and liver, respectively), but not reliable for leaner species and leaner tissues (such as cod and muscle, respectively) [45]. To date, this model has been implemented in limited fish species, including salmonids, turbot, Atlantic cod, gilthead sea bream (Sparus aurata), tilapia (Oreochromis niloticus) and Murray cod (Maccullochella peelii peelii) [[46], [47], [48]], and on a limited subset of FA, mainly C18 FA, EPA, and DHA. One additional limitation of this model is, in some instances, a lack of predictive accuracy. As such, it may be more appropriately defined as a descriptive model. Addtionally, analysis of the final lipid content of the fish is required for the subsequent calculations. Further development, validation and possible application of this model to other fish species, and for a suite of other FA, is therefore desirable. Nevertheless, based on currently available results, there may be implications regarding its widespread applicability.
Other models have been developed and tested with the aim of identifying predictive or descriptive relationships between dietary and fish FA composition. For example, linear correlations have been described for Atlantic salmon [[49], [50], [51]] and Atlantic cod [52], however, inconsistent results have hampered further development of a predictive model. More recently, a curvilinear regression model was introduced by Ballester-Lozano et al. [53] to describe the relationships between muscle FA composition and two independent variables; namely, dietary FA composition and muscle lipid level. For SFA (14:0, 16:0, 18:0) and MUFA (16:1n-7, 18:1n-7, 18:1n-9, 20:1n-9), the overall variance in muscle FA composition was primarily explained by dietary FA composition, and secondly by muscle lipid level. The latter independent variable also explained the majority of the variance observed for both ARA and DHA. However, a statistically significant contribution was not found for 18:2n-6, 18:3n-3, EPA and 22:5n-3 [53]. The authors further extended this model (termed “dummy regression model”) to turbot, Dover sole (Solea solea), and European sea bass (Dicentrarchus labrax) using data on gilthead sea bream as a reference subgroup dataset and data from turbot, Dover sole and European sea bass as the respective “dummy” variables [54,55]. This model appears robust in predicting FA composition in the tested fish species, although the relative contribution of each independent variable to the total variance was found to vary within and among FA and fish species. In particular, the contribution of muscle lipid content was reported to be highly dependent on fish species. Additionally, major interaction effects between muscle lipid content and the “dummy” variables were found for both SFA and MUFA [54]. Given species specific physiological differences inherent in lipid metabolism, this predictive model of muscle FA composition was considered applicable to marine farmed fish.
Recently, Mock et al. [56] completed a systematic review and analysis on the effect of diet on the concentration of n-3 LC-PUFA in the muscle tissue of post-smolt Atlantic salmon. Overall, the study highlighted an unexpected paucity of studies that met the selection criteria, primarily due to a limited feeding duration and a tendency to utilise juvenile fish in growth trials. Nevertheless, some significant regression models were generated, particularly for muscle DHA and n-3 LC-PUFA, suggesting the possibility to develop predictive models. However, given the conclusiveness of the findings was hindered by a limited dataset, future, more robust models, will likely be more mechanistic in nature and require a more comprehensive suite of input variables.
The authors of the present review are aware that a series of predictive models, particularity for estimating final EPA and DHA content in fish muscle of some cultured species (e.g. Atlantic salmon) have been conceived. However, these models are IP protected and have been developed by private companies. Therefore, details are not disclosed in the scientific literature, nor are they made publicly available.
Another quantitative method developed to explain the relationship between diet and fish FA composition was introduced and tested by Turchini et al. [57,58], with following modifications by Turchini et al. [59,60]. The “whole-body fatty acid balance method” aimed at quantifying the apparent in vivo FA metabolism, including deposition/retention, β-oxidation, de novo synthesis and bioconversion (elongation and desaturation). Although not intended as a predictive model, it has been proven to be an accurate descriptive model and has been applied to a variety of aquatic species; from freshwater [61,62] to marine [63], from tropical [64] to temperate [65], under different conditions and nutritional scenarios [22,29,37,66], as well as in terrestrial animals and in cell culture [[67], [68], [69], [70]]. Collectively, these studies suggest that amongst different species all capable of the LC-PUFA biosynthetic pathway, there are remarkable differences in the efficiency of elongation and desaturation of C18 PUFA into LC-PUFA.
Considering all the metabolic, environmental, and dietary factors known to affect the metabolic fate of dietary FA, it is overwhelmingly apparent that the ‘diet-fish’ fatty acid relationship are inevitably complex. This is confirmed by the aforementioned limitations of both predictive and quantitative models in describing this relationship. Despite the clear importance of this research topic within the scientific and industrial fish nutrition sectors [2,4,56,71], to date, the relationships among various fish species and different FA types have not yet been comprehensively reviewed. From the perspective of FA, the answer to the popular question: “are fish what they eat?”, has been partially accepted as “mostly yes”. However, given that the progression of the aquaculture sector relies on increasingly marginal improvements for sustained economic and environmental sustainability [72], the absence of a thorough and comprehensive analysis on this topic is a limitation for future improvements. Therefore, an attempt to respond to this important question may likely be surmised as an attempt to answer a subtly different, yet complimentary question: “To what extent, are fish what they eat?”
To this end, the present review used a quantitative analytical approach, based on data published in the last two decades (1998 - 2018), with the aim of providing a novel and comprehensive contribution to our understanding of the ‘diet-fish’ fatty acid relationship in cultured fishes by summarizing the extent to which fish FA profiles resemble that of their respective diets. The muscle (fillet) tissue was selected as the primary target tissue of this analysis owing to its position as the major body compartment in teleost fish species, as well being the primary edible component from a consumer perspective. However, where available, other tissues were considered in order to make comparisons between various tissue types. The likely effects of environmental and experimental factors on the ‘diet-fish’ fatty acid relationship were also considered and are discussed herein. Moreover, we emphasized the role of fish oil finishing (or wash-out) strategy, which has been used to restore the LC-PUFA content in response to the utilization of alternative oils lacking/void of these health promoting FA. It is envisaged that this review will facilitate a more comprehensive understanding of the complex relationships between FA profiles in diets and those in fish. Ultimately, this is expected to benefit the further optimization of dietary FA management and manipulation and the subsequent delivery of FA to farmed fish. Progress towards this goal will ensure FA quality in cultured fish and seafood for human consumption, within the constraints of global resource availability.
Section snippets
Systematic review and data collection
A systematic review of the available scientific literature published in the last two decades (1998 - 2018) was conducted. The selection of studies conformed to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA statement) [73]. The literature search was implemented by utilizing Web of Science Database (Web of Science Core Collection), and searching for: (topic) “fatty acid”, AND (topic) “diet”, AND (topic) “fish”, AND (topic) “muscle OR fillet OR flesh”, AND
General description
Several interesting results from the D value and R2overall values analysis were noted (Supplementary Table S2). There was a significant negative correlation between D values and R2overall values (R2 = -0.624, P < 0.001; n = 523 tissues). Conceptually, this is intuitive as both parameters are a description of the same phenomenon, albeit from opposite directions. A tissue having a low D value and a high R2overall value demonstrates a high similarity to the dietary FA profile. In general, all R2
Factors influencing the ‘diet-fish’ fatty acid relationship
Incorporation of FA into tissues is modulated by various metabolic factors including selective incorporation, β-oxidation, lipogenic activity and FA elongation and desaturation processes. Interrelated factors, such as environment, and the size, age and physiological condition of fish further modulate and affect these metabolic activities. Therefore, the final FA composition in a given fish tissue depends upon the initial FA content, cumulative net intake, digestibility (and associated
Re-storing long-chain polyunsaturated fatty acids in fish muscle
As discussed in the sections above, it is clear that the first and most obvious consequence of replacing fish oil with alternative lipid sources in aquaculture feed is the reduction of LC-PUFA in the fish muscle, consequently diminishing its nutritional quality. In response, a finishing or wash-out feeding strategy, whereby a fish oil-rich diet is fed during the final phase of production has been proposed. This method has been suggested as an efficient method to re-store the LC-PUFA content in
Conclusion
The deposition of FA into fish tissues has an inherently high level of plasticity. Like other dietary nutrients, FA in fish tissues are strongly affected by the diet, however, the content and composition of FA in fish tissues are more susceptible to a variety of other, often interrelated, factors. Therefore, although the FA profile of fish tissues is reflective of the dietary FA composition, this notion remains overly simplistic and, in fact, detrimental to future efforts to optimise the
Author contributions
Houguo Xu: Conceptualization, Data curation, Writing - original draft; Giovanni Turchini: Conceptualization, Writing – substantial review & editing; David Francis: Writing - substantial review & editing; Mengqing Liang: Methodology, Supervision, Funding acquisition; Thomas Mock: Writing – substantial review & editing, Data curation; Artur Rombenso: Writing – review & editing; Qinghui Ai: Conceptualization, Writing – original draft; Resources. All authors approved the submission and publication
Declaration of Competing Interest
All the authors declare that they do not have any conflict of interest.
Acknowledgement
This work was supported by National Key R&D Program of China (2018YFD0900400), National Natural Science Foundation of China (31525024), Central Public-interest Scientific Institution Basal Research Fund, CAFS (2020TD48, 2018HY-ZD0505), and China Agriculture Research System (CARS-47-G15).
Reference (322)
Omega-3 long-chain polyunsaturated fatty acids and aquaculture in perspective
Aquaculture
(2015)- et al.
Biosynthesis of long-chain polyunsaturated fatty acids in marine fish: Characterization of an Elovl4-like elongase from cobia Rachycentron canadum and activation of the pathway during early life stages
Aquaculture
(2011) - et al.
The effects of dietary fatty acids on liver fatty acid composition and Δ6-desaturase expression differ with ambient salinities in Siganus canaliculatus
Comp Biochem Phys B
(2008) - et al.
Influence of salinity and linoleic or α-linolenic acid based diets on ontogenetic development and metabolism of unsaturated fatty acids in pike perch larvae (Sander lucioperca)
Aquaculture
(2019) - et al.
Growth performance, feed utilisation and fatty acid deposition in Atlantic salmon, Salmo salar L., fed graded levels of high-lipid/high-EPA Yarrowia lipolytica biomass
Aquaculture
(2012) - et al.
Effects of dietary iron supplementation on growth performance, fatty acid composition and fatty acid metabolism in rainbow trout (Oncorhynchus mykiss) fed vegetable oil based diets
Aquaculture
(2012) - et al.
Targeted dietary micronutrient fortification modulates n-3 LC-PUFA pathway activity in rainbow trout (Oncorhynchus mykiss)
Aquaculture
(2013) - et al.
Dietary micronutrients and in vivo n-3 LC-PUFA biosynthesis in Atlantic salmon
Aquaculture
(2016) - et al.
Retro-engineering the protein sparing effect to preserve n-3 LC-PUFA from catabolism and optimise fish oil utilisation: A preliminary case study on juvenile Atlantic salmon
Aquaculture
(2017) - et al.
Effects of dietary arachidonic acid on growth performance, survival, immune response and tissue fatty acid composition of juvenile Japanese seabass, Lateolabrax japonicus
Aquaculture
(2010)
Dietary arachidonic acid differentially regulates the gonadal steroidogenesis in the marine teleost, tongue sole (Cynoglossus semilaevis), depending on fish gender and maturation stage
Aquaculture
Fish oil replacement with different vegetable oils in Murray cod: evidence of an “omega-3 sparing effect” by other dietary fatty acids
Aquaculture
Marginal efficiencies of long chain-polyunsaturated fatty acid use by barramundi (Lates calcarifer) when fed diets with varying blends of fish oil and poultry fat
Aquaculture
Effects of water temperature and diets containing palm oil on fatty acid desaturation and oxidation in hepatocytes and intestinal enterocytes of rainbow trout (Oncorhynchus mykiss)
Comp Biochem Phys B
Fish oil substitution by vegetable oils in commercial diets for gilthead sea bream (Sparus aurata L.); effects on growth performance, flesh quality and fillet fatty acid profile: recovery of fatty acid profiles by a fish oil finishing diet under fluctuating water temperatures
Aquaculture
Coping with sub-optimal water temperature: Modifications in fatty acid profile of barramundi as influenced by dietary lipid
Comp Biochem Phys A
Digestibility, growth and nutrient utilisation of Atlantic salmon parr (Salmo salar L.) in relation to temperature, feed fat content and oil source
Aquaculture
Dietary n-6/n-3 LC-PUFA ratio, temperature and time interactions on nutrients and fatty acids digestibility in Atlantic salmon
Aquaculture
Fatty acid profile of fish following a change in dietary fatty acid source: model of fatty acid composition with a dilution hypothesis
Aquaculture
Are modifications in tissue fatty acid profiles following a change in diet the result of dilution? Test of a simple dilution model
Aquaculture
Lipid and fatty acid dynamics in Atlantic cod, Gadus morhua, tissues: Influence of dietary lipid concentrations and feed oil sources
Aquaculture
The time course of fish oil wash-out follows a simple dilution model in gilthead sea bream (Sparus aurata L.) fed graded levels of vegetable oils
Aquaculture
Replacement of fish oil with rapeseed oil in diets of Atlantic salmon (Salmo salar) affects tissue lipid compositions and hepatocyte fatty acid metabolism
J Nutr
Substituting fish oil with crude palm oil in the diet of Atlantic salmon (Salmo salar) affects muscle fatty acid composition and hepatic fatty acid metabolism
J Nutr
Altered fatty acid compositions in Atlantic salmon (Salmo salar) fed diets containing linseed and rapeseed oils can be partially restored by a subsequent fish oil finishing diet
J Nutr
Prediction of fillet fatty acid composition of market-size gilthead sea bream (Sparus aurata) using a regression modelling approach
Aquaculture
Fatty acid metabolism in the freshwater fish Murray cod (Maccullochella peelii peelii) deduced by the whole-body fatty acid balance method
Comp Biochem Phys B
Fish oil replacement in rainbow trout diets and total dietary PUFA content: (II) Effects on fatty acid metabolism and in vivo fatty acid bioconversion
Aquaculture
Genetically improved farmed Nile tilapia and red hybrid tilapia showed differences in fatty acid metabolism when fed diets with added fish oil or a vegetable oil blend
Aquaculture
Full substitution of fish oil with camelina (Camelina sativa) oil, with partial substitution of fish meal with camelina meal, in diets for farmed Atlantic salmon (Salmo salar) and its effect on tissue lipids and sensory quality
Food Chem
Influence of high contents of dietary soybean oil on growth, feed utilization, tissue fatty acid composition, heart histology and standard oxygen consumption of Atlantic salmon (Salmo salar) raised at two temperatures
Aquaculture
Growth, feed utilization and flesh quality of European sea bass (Dicentrarchus labrax) fed diets containing vegetable oils: a time-course study on the effect of a re-feeding period with a 100% fish oil diet
Aquaculture
Effects of total replacement of fish oil by vegetable oils in the diets of sharpsnout seabream (Diplodus puntazzo)
Aquaculture
Comparison of effects of vegetable oils blended with southern hemisphere fish oil and decontaminated northern hemisphere fish oil on growth performance, composition and gene expression in Atlantic salmon (Salmo salar L.)
Aquaculture
Growth performance, feed utilisation and fatty acid deposition in Atlantic salmon, Salmo salar L., fed graded levels of high-lipid/high-EPA Yarrowia lipolytica biomass
Aquaculture
DHA is essential, EPA appears largely expendable, in meeting the n−3 long-chain polyunsaturated fatty acid requirements of juvenile cobia Rachycentron canadum
Aquaculture
Effect of replacing dietary menhaden oil with pollock or soybean oil on muscle fatty acid composition and growth performance of juvenile Pacific threadfin (Polydactylus sexfilis)
Aquaculture
Tallow in Atlantic salmon feed
Aquaculture
A revisit to fishmeal usage and associated consequences in Chinese aquaculture
Rev Aquacult
Fish oil replacement in finfish nutrition
Rev Aquacult
Omega-3 Long-Chain Polyunsaturated Fatty Acids, EPA and DHA: Bridging the Gap between Supply and Demand
Nutrients
The future of genetic engineering to provide essential dietary nutrients and improve growth performance in aquaculture: advantages and challenges
J World Aquacult Soc
Metabolism and functions of lipids and fatty acids in teleost fish
Rev Fish Sci
Highly unsaturated fatty acid synthesis in Atlantic salmon: Characterization of Elovl5-and Elovl2-like elongases
Mar Biotechnol
Genes for de novo biosynthesis of omega-3 polyunsaturated fatty acids are widespread in animals
Sci Adv
Trophic levels predict the nutritional essentiality of polyunsaturated fatty acids in fish—introduction to a special section and a brief synthesis
N Am J Aquacult
The state of world fisheries and aquaculture
Effect of temperature on the incorporation into phospholipid classes and metabolismvia desaturation and elongation of n-3 and n-6 polyunsaturated fatty acids in fish cells in culture
Lipids
Dietary lipids and temperature interact to influence tissue fatty acid compositions of Atlantic salmon, Salmo salar L., parr
Aquac Res
Influences of low salinity and dietary fatty acids on fatty acid composition and fatty acid desaturase and elongase expression in red sea bream Pagrus major
Fisheries Sci
Cited by (117)
Proximate and fatty acids composition in the muscle of wild and farmed sardine (Sardinella brasiliensis)
2024, Food Chemistry AdvancesFatty acids profiling of goose barnacle (Pollicipes pollicipes) tissues to evaluate nutritional quality and confirm harvesting location
2024, Journal of Food Composition and Analysis