Hindawi Aquaculture Nutrition Volume 2023, Article ID 6415693, 19 pages https://doi.org/10.1155/2023/6415693 Research Article Effect of Feeding Acid Oils on European Seabass Fillet Lipid Composition, Oxidative Stability, Color, and Sensory Acceptance Paula Albendea ,1,2 Alba Tres ,1,2 Magdalena Rafecas ,1,3 Stefania Vichi ,1,2 Roser Sala ,4 and Francesc Guardiola 1,2 1 Nutrition, Food Science and Gastronomy Department-XIA, Campus de l’Alimentació Torribera, Facultat de Farmàcia i Ciències de l’Alimentació, Universitat de Barcelona, Av Prat de la Riba, 171., 08921 Santa Coloma de Gramenet, Spain 2 Institut de Recerca en Nutrició i Seguretat Alimentària (INSA-UB), Universitat de Barcelona, Av Prat de la Riba, 171., 08921 Santa Coloma de Gramenet, Spain 3 Nutrition, Food Science and Gastronomy Department-XIA, Facultat de Farmàcia i Ciències de l’Alimentació, Universitat de Barcelona, Av Joan XXIII, 27-31., 08028 Barcelona, Spain 4 Animal Nutrition and Welfare Service (SNiBA), Animal and Food Science Department, Facultat de Veterinària, Universitat Autònoma de Barcelona, Travessera dels Turons., 08193 Bellaterra, Spain Correspondence should be addressed to Alba Tres; atres@ub.edu Received 25 October 2022; Revised 28 November 2022; Accepted 3 December 2022; Published 6 January 2023 Academic Editor: Erchao Li Copyright © 2023 Paula Albendea et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Acid oils (AO) are fat by-products of edible oil refining with a high energetic value, being an interesting option for a more sustainable aquaculture nutrition. This study was conducted to evaluate the effects of the partial replacement of fish oil (FO) in diets by two AO instead of crude vegetable oils on the lipid composition, lipid oxidation and quality of fresh European seabass fillets, and after their commercial refrigerated storage for 6 days. Fish were fed with five different diets, the added fat being FO (100%) or a blend of FO (25%) and another fat (75%): crude soybean oil (SO), soybean-sunflower acid oil (SAO), crude olive pomace oil (OPO), or olive pomace acid oil (OPAO). Fresh and refrigerated fillets were assessed for fatty acid profile, tocopherol (T) and tocotrienol (T3) composition, lipid oxidative stability, 2-thiobarbituric acid (TBA) value, volatile compound content, color, and sensory acceptance. Refrigerated storage did not affect T + T3 total content but increased secondary oxidation products (TBA values and volatile compound contents) in fillets from all diets. The FO substitution decreased EPA and DHA and increased T and T3 in fish fillets, but the recommended human daily intake of EPA plus DHA could still be covered with 100 g of fish fillets. Both a higher oxidative stability and a lower TBA value were found in SO, SAO, OPO, or OPAO fillets, obtaining the greatest oxidative stability in OPO and OPAO fillets. Sensory acceptance was not affected by the diet or the refrigerated storage, while the differences found in color parameters would not be perceived by the human eye. According to the oxidative stability and acceptability of flesh, SAO and OPAO are adequate replacements of FO as energy source in European seabass diets, which implies that these by-products can be upcycled, improving the environmental and economical sustainability of aquaculture production. 1. Introduction human consumption while the rest had other uses, as fish oil (FO) and fish meal production [1]. The commonest lipid As the years have gone, the importance of aquaculture pro- source used in fish feed has been FO due to its high percent- duction to cover food fish demand has significantly raised age of n-3 long-chain polyunsaturated fatty acids (n-3 LC- [1]. This increasing trend is expected to continue in the PUFA), such as eicosapentaenoic acid (C20:5 n-3, EPA) future, reaching to cover 62% of food fish production in and docosahexaenoic acid (C22:6 n-3, DHA), which are 2030 [2]. In 2018, 88% of fish production was intended for considered essential fatty acids (FA) for fish [3, 4]. 2 Aquaculture Nutrition The increase of aquaculture production has implied a sory quality of the fish products even after their storage raise in the demand of FO for fish feeding, while the produc- under commercial conditions. tion of FO has remained stable. As this could jeopardize The aim of this work was to evaluate the effect of the aquaculture sustainability, it has been very important to find partial replacement of FO in fish diets by AO instead of alternative fat sources to partially replace FO in fish diets. crude oils from two different vegetable sources (soybean This substitution should guarantee enough EPA and DHA and olive pomace) on lipid composition, oxidative stability, for an adequate growth and development of fish, especially color, and sensory acceptance of fresh and refrigerated for marine species such as European seabass (Dicentrarchus European seabass fillets. This study complements the infor- labrax) [5, 6]. Moreover, EPA and DHA content in fish diet mation published by Verge-Mèrida et al. [13] on growth influences their content in fish flesh, and, since these FA and digestibility parameters of the same specimens used in have shown health benefits, it also affects the nutritional the present work. value of fish flesh [7]. On the other hand, EPA and DHA have a great tendency to suffer oxidation reactions due to 2. Materials and Methods their high number of double bonds, and, consequently, oxi- dative rancidity is one of the main nonmicrobiological 2.1. Experimental Fats. Five different experimental fats were reasons of fish and fish food spoilage [8]. used: fish oil (FO), two crude vegetable oils with different The options that have drawn more attention to replace FA profile, crude soybean oil (SO) and crude olive pomace FO as energy source are vegetable fat sources, the most com- oil (OPO), and two AO available in the Spanish market mon being soybean, linseed, rapeseed, sunflower, palm, and with similar FA composition to theirs, soybean-sunflower olive oil [6]. These fats have low EPA and DHA percentages (55 : 45, w/w) acid oil (SAO) and olive pomace acid oil but high percentages of C18 polyunsaturated fatty acids (OPAO) (Table 1). Bunge Ibérica S.A.U. (Sant Just Desvern, (PUFA) or monounsaturated fatty acids (MUFA) [9]. Also, Spain) supplied SO and SAO, General d’Olis i Derivats S.L. some vegetable fats are usually characterized for being rich (Les Borges Blanques, Spain) provided OPO, RIOSA (Refina- in tocopherols (T), tocotrienols (T3), and other bioactive ción Industrial Oleícola S.A., Ibros, Spain) supplied OPAO, compounds, such as polyphenols [9]. Thus, introducing and AFAMSA (Agrupación de Fabricantes de Aceites Mari- vegetable fat sources to partially replace FO in fish feeding nos S.A., Mos, Spain) supplied the commercial degummed might cause both a FA profile modification of fish flesh, FO, which mostly came from tuna. leading to a reduction of its EPA and DHA contents, and The contents of moisture and volatile matter (M), I, U, an increment in antioxidant compounds in fish flesh, which T and T3, and FA and lipid class compositions, and acid could prevent the development of lipid oxidation reactions and peroxide values of these five fat sources (Table 1) and enhance its preservation [10]. were determined by triplicate as described by Varona Although crude oils are the most widely used vegetable et al. [15]. sources to replace FO, it is also possible to replace it with some fat by-products from the food industry such as acid 2.2. Animals and Diets. The animal housing and husbandry oils (AO). AO are fat by-products coming from edible oil was conducted at the Aquaculture Center facilities of the refining, characterized for having a similar FA profile to Institute of Agrifood Research and Technology (IRTA, Sant the crude oil and a high content of free fatty acids (FFA), Carles de la Ràpita, Spain). All the procedures were as they are obtained from the steps in which FFA are approved by the Animal Protocol Review Committee of removed from the crude oil [11]. Using AO as feed fats the Universitat Autònoma de Barcelona (UAB) and were alternatively to other lipid sources is a way to upcycle them, in accordance with the guidelines set by the EU Directive which would contribute to improving aquaculture sustain- 2010/63/EU [16]. A total of 300 European seabass (Dicen- ability. However, there are only a few trials performed with trarchus labrax) specimens of approximately 100 g were these by-products in the literature, focusing mainly on their randomly distributed into 15 tanks of 500 L of capacity (20 effects on lipid digestibility and productive parameters and fish per tank) with a sea water semirecirculation system showing controversial results [12, 13]. In fact, as described (IRTAmar®; IRTA, Sant Carles de la Ràpita, Spain). This by Varona et al. [11, 14], one disadvantage of AO is their system allowed water recirculation of between 1 and 1.5 tank high variability in various compositional parameters such volumes per hour (15 m3/h) and was equipped with an aer- as FFA content, insoluble impurities (I), or unsaponifiable obic biofilter for the removal/transformation of ammonia matter (U). This affects their nutritional value [11], and to nitrite and nitrite to nitrate. The fish were reared indoors consequently, this could impair the animal performance, and subjected to natural photoperiod and breed under decreasing the confidence of producers in these feeding fats. controlled conditions of water temperature (22:6 ± 0:8°C), Also, the fact that the content of certain antioxidants, such dissolved oxygen levels (7:3 ± 0:7 mg/L), pH (7.7-8.1), and as T and T3, is highly variable in these refining by- salinity (35:5 ± 0:5%). products could affect the oxidative stability and the quality The experimental diets were prepared in form of of fish fillets during their commercial shelf life [11, 14]. extruded pellets by the Service of Experimental Diets of the However, the effects of using AO in fish diets on the compo- University of Almería (Almería, Spain) using standard aqua- sition and quality of fish flesh and fish food products are feed procedures. The ingredients and the proximate compo- barely studied, which would be essential to upcycle them sition of the diets are presented in Table 2. All the diets assuring the nutritional value, oxidative stability, and sen- contained a total of 15.4% (w/w) added fat. In the control Aquaculture Nutrition 3 Table 1: Composition of the experimental fats used in this study. FO SO SAO OPO OPAO MIU (g/100 g)1 Moisture2 0:24 ± 0:01 0:05 ± 0:01 0:40 ± 0:01 0:36 ± 0:01 0:31 ± 0:01 Impurities 0:30 ± 0:05 0:21 ± 0:03 0:89 ± 0:15 0:44 ± 0:12 1:94 ± 0:21 Unsaponifiable 2:01 ± 0:12 0:53 ± 0:25 2:35 ± 0:34 1:64 ± 0:35 3:90 ± 0:35 Total 2:55 ± 0:13 0:80 ± 0:25 3:64 ± 0:37 2:44 ± 0:37 6:15 ± 0:41 FA composition (%)1; 3 C14:0 3:7 ± 0:13 ND ND ND ND C16:0 21:8 ± 0:50 10:7 ± 0:04 11:2 ± 0:05 13:3 ± 2:18 11:5 ± 0:05 C18:0 6:4 ± 0:09 3:4 ± 0:11 3:5 ± 0:01 2:7 ± 0:06 3:1 ± 0:01 SFA 34:9 ± 0:53 15:0 ± 0:24 16:3 ± 0:06 16:8 ± 2:18 16:0 ± 0:05 C16:1 n-7 4:7 ± 0:35 0:1 ± 0:01 0:1 ± 0:01 0:7 ± 0:02 1:1 ± 0:02 C18:1 n-9 16:0 ± 0:20 23:5 ± 0:02 30:5 ± 0:08 68:7 ± 1:76 63:0 ± 0:19 C18:1 n-7 2:8 ± 0:03 1:7 ± 0:02 1:4 ± 0:01 2:0 ± 0:12 2:1 ± 0:06 MUFA 28:4 ± 0:41 25:6 ± 0:06 32:2 ± 0:08 71:7 ± 1:77 66:5 ± 0:20 C18:2 n-6 1:9 ± 0:02 53:1 ± 0:35 48:3 ± 0:03 10:4 ± 0:15 16:5 ± 0:11 n-6 PUFA 2:8 ± 0:04 53:2 ± 0:35 48:3 ± 0:03 10:5 ± 0:15 16:6 ± 0:12 C18:3 n-3 0:9 ± 0:01 6:1 ± 0:01 3:0 ± 0:01 0:9 ± 0:01 0:7 ± 0:01 C20:5 n-3 6:3 ± 0:92 ND ND ND ND C22:6 n-3 26:0 ± 0:15 ND ND ND ND n-3 PUFA 33:6 ± 1:04 6:1 ± 0:01 3:0 ± 0:01 0:9 ± 0:01 0:7 ± 0:01 Total PUFA 36:4 ± 1:04 59:2 ± 0:35 51:3 ± 0:04 11:3 ± 0:15 17:3 ± 0:12 Trans C18:1 0:3 ± 0:01 0:2 ± 0:08 0:1 ± 0:02 0:1 ± 0:04 0:3 ± 0:2 T and T3 content (mg/kg)1 α-T 41:18 ± 1:51 170:93 ± 4:71 40:83 ± 0:40 66:44 ± 0:57 65:61 ± 2:24 β-T 0:32 ± 0:03 20:04 ± 0:03 2:90 ± 0:03 0:47 ± 0:15 1:51 ± 0:10 γ-T 0:21 ± 0:05 535:93 ± 10:96 36:44 ± 0:61 2:22 ± 0:01 1:91 ± 0:09 δ-T 0:65 ± 0:32 53:63 ± 1:63 20:95 ± 0:84 ND 0:55 ± 0:01 β-T3 0:41 ± 0:17 4:35 ± 0:62 0:36 ± 0:03 3:11 ± 0:05 1:29 ± 0:09 γ-T3 ND 9:23 ± 0:28 ND ND 0:99 ± 0:05 Total (T + T3) 42:77 ± 1:92 794:11 ± 13:72 101:48 ± 1:92 72:64 ± 0:44 71:86 ± 2:46 1; 4 Lipid class composition (%) TAG 85:7 ± 0:01 93:9 ± 0:08 29:3 ± 0:30 77:5 ± 0:15 36:3 ± 0:10 DAG 6:9 ± 0:03 4:2 ± 0:08 16:1 ± 0:24 8:4 ± 0:15 17:4 ± 0:01 MAG 4:4 ± 0:03 0:5 ± 0:01 1:3 ± 0:28 0:9 ± 0:17 1:4 ± 0:10 FFA 3:1 ± 0:06 1:5 ± 0:01 53:3 ± 0:34 13:2 ± 0:12 45:0 ± 0:01 Acid value (mg KOH/g)1 14:3 ± 0:19 2:8 ± 0:11 116:6 ± 1:8 23:3 ± 0:11 96:8 ± 0:45 Peroxide value (meq O2/kg)1 2:7 ± 0:44 1:5 ± 0:12 1:7 ± 0:08 6:0 ± 0:72 1:7 ± 0:46 Abbreviations: FO: fish oil; SO: crude soybean oil; SAO: soybean-sunflower acid oil; OPO: crude olive pomace oil; OPAO: olive pomace acid oil; MIU: moisture and volatile matter + insoluble impurities + unsaponifiable matter; FA: fatty acid; SFA: saturated fatty acids (sum of C14:0, C15:0, C16:0, C17:0, C18:0, C20:0, C22:0, and C24:0); MUFA: monounsaturated fatty acids (sum of C16:1 n-9, C16:1 n-7, C17:1 n-7, C18:1 n-9, C18:1 n-7, C20:1 n-9, C22:1 n-9, and C24:1 n-9); PUFA: polyunsaturated fatty acids (n-6 PUFA: sum of C18:2 n-6, C18:3 n-6, C20:2 n-6, C20:3 n-6, and C20:4 n-6; n-3 PUFA: sum of C18:3 n-3, C20:3 n-3, C20:5 n-3, and C22:6 n-3; total PUFA: sum of n-6 PUFA and n-3 PUFA); T: tocopherol; T3: tocotrienol; TAG: triacylglycerols; DAG: diacylglycerols; MAG: monoacylglycerols; FFA: free fatty acids; ND: not detected. 1Mean ± standard deviation of the three determinations. 2 It included moisture and other compounds that volatilize under the determination conditions. 3 The percentage of each FA was obtained by peak area normalization. 4 The percentage of each lipid class was obtained by peak area normalization. 4 Aquaculture Nutrition Table 2: Ingredients, proximate composition, and gross energy of the five diets used in this study. FO SO SAO OPO OPAO Ingredient composition (g/kg) Wheat meal 110.3 110.3 110.3 110.3 110.3 Wheat gluten 155.9 155.9 155.9 155.9 155.9 Soya protein concentrate 266.0 266.0 266.0 266.0 266.0 Fish meal 202.4 202.4 202.4 202.4 202.4 Hydrolyzed fish protein 25.3 25.3 25.3 25.3 25.3 Krill meal 25.5 25.5 25.5 25.5 25.5 Soybean lecithin 9.6 9.6 9.6 9.6 9.6 Fish oil 153.9 38.5 38.5 38.5 38.5 Experimental fat 0.0 115.4 115.4 115.4 115.4 L-lysine 2.9 2.9 2.9 2.9 2.9 DL-methionine 1.0 1.0 1.0 1.0 1.0 Choline chloride 4.8 4.8 4.8 4.8 4.8 Betaine 1.0 1.0 1.0 1.0 1.0 Vitamin and mineral premix1 19.2 19.2 19.2 19.2 19.2 Vitamin C 1.0 1.0 1.0 1.0 1.0 Guar gum 19.2 19.2 19.2 19.2 19.2 Proximate composition (g/kg) Dry matter 926.1 925.7 932.3 926.2 930.4 Crude protein 418.3 405.3 396.2 413.1 414.3 Ether extract 190.5 190.4 182.8 186.9 180.0 Ash 72.2 72.4 73.2 72.4 73.4 Gross energy (MJ/kg) 21.7 21.8 21.8 21.7 22.0 Abbreviations: FO: fish oil diet; SO: crude soybean oil diet; SAO: soybean-sunflower acid oil diet; OPO: crude olive pomace oil diet; OPAO: olive pomace acid oil diet. 1 Provides per kg of feed: vitamin A (38,460 IU), vitamin D3 (3,8460 UI), vitamin E (192.23 mg of α-tocopheryl acetate), vitamin K3 (48.08 mg), vitamin B1 (57.69 mg), vitamin B2 (57.69 mg), calcium pantothenate (192.23 mg), nicotinic acid (384.60 mg), vitamin B6 (38.46 mg), vitamin B9 (28.84 mg), vitamin B12 (0.19 mg), biotin (5.77 mg), inositol (961.50 mg), betaine (961.50 mg), cobalt carbonate (1.25 mg), cupric sulphate (17.31 mg), ferrous sulphate (11.54 mg), potassium iodide (0.96 mg), manganese oxide (18.46 mg), sodium selenite (0.02 mg), zinc sulphate (14.42 mg), calcium carbonate (3,577 mg), potassium chloride (463.44 mg), and sodium chloride (769.20 mg). diet, the added fat source was 100% FO. In the other diets, a cate was composed by six different fish, obtained from 3 75% of the added fat was one of the vegetable fat sources tanks (2 fish from each tank) fed with the same diet. From (SO, SAO, OPO, or OPAO), and the remaining 25% was each fish, the two entire fillets without skin were taken as FO to guarantee enough quantity of essential FA for Euro- samples, so a total of twelve fillets (3 tanks per diet × 2 fish pean seabass (at least 1% of EPA + DHA on dry weight per tank × 2 entire fillets per fish) were used for one repli- [17]). Feed samples were taken, and the FA composition, T cate. All the samples were kept in zipper bags on ice until and T3 content, and lipid class composition were deter- the next morning. mined in duplicate in all diets (Table 3). Therefore, consid- In less than 24 h after slaughter, the twelve fillets per rep- ering the M and the FA profile of each diet (Tables 2 and licate were divided into two groups of six fillets (one fillet of 3), the EPA + DHA content was expressed on dry weight, each fish per group), using one group to constitute the fresh being 5.9% in FO diet and 1.8% in the rest of diets. Thus, sample and the other group to constitute the refrigerated in all cases, EPA + DHA content was higher than the 1% of sample. For fresh samples, the six fillets of each replicate recommended for European seabass [17]. were pooled and ground. After measuring the color, various The five diets were randomly assigned to different tanks aliquots of 25 g of the ground sample were immediately (three tanks per experimental diet), and the experimental vacuum packed in high-barrier multilayer bags (Cryovac feeding period lasted for 100 days. The effects of these diets BB3255; permeability to O2, 17 cm3/m2 per day per bar at on animal performance parameters have been published 23°C and 0% relative humidity, ASTMD-3985; Cryovac elsewhere [13]. Europe, Sealed Air S.L., Sant Boi de Llobregat, Spain) and were kept at -20°C until the chemical analyses were per- 2.3. Sampling of European Seabass Fillets. At the end of the formed. The other part of the ground sample was pooled experimental period, fish were fasted for 24 h and slaugh- with the ground sample remains from other replicates of tered by immersion in ice-cold water (hypothermia). Five the same diet and used to make hamburgers, which were replicates per dietary treatment were prepared. Each repli- cooked to perform the sensory acceptance test. Aquaculture Nutrition 5 Table 3: Fatty acid, tocopherol and tocotrienol, and lipid class compositions of the experimental diets1. FO2 SO2 SAO2 OPO2 OPAO2 3 FA composition (%) C14:0 3:74 ± 0:05 1:40 ± 0:01 1:52 ± 0:06 1:44 ± 0:07 1:51 ± 0:05 C16:0 23:43 ± 1:73 15:13 ± 0:15 15:79 ± 0:59 15:77 ± 0:75 15:29 ± 0:06 C18:0 6:77 ± 0:56 4:45 ± 0:10 3:88 ± 0:85 3:50 ± 0:58 4:16 ± 0:04 SFA 35:16 ± 1:82 21:69 ± 0:18 22:18 ± 1:04 21:27 ± 0:95 21:69 ± 0:10 C16:1 n-7 4:39 ± 0:18 1:54 ± 0:01 1:63 ± 0:06 1:98 ± 0:09 2:20 ± 0:01 C18:1 n-9 16:28 ± 0:61 22:20 ± 0:58 27:20 ± 1:31 50:45 ± 1:56 46:12 ± 0:58 C18:1 n-7 2:71 ± 0:07 1:87 ± 0:04 1:55 ± 0:10 1:58 ± 0:13 2:10 ± 0:45 C20:1 n-9 1:52 ± 0:04 0:67 ± 0:01 0:66 ± 0:02 0:76 ± 0:03 0:71 ± 0:01 MUFA 25:09 ± 0:64 26:38 ± 0:58 31:15 ± 1:32 54:98 ± 1:57 51:32 ± 0:73 C18:2 n-6 6:16 ± 0:18 37:03 ± 0:78 33:13 ± 1:46 11:73 ± 0:57 15:03 ± 0:01 C20:4 n-6 2:10 ± 0:09 0:64 ± 0:01 0:67 ± 0:02 0:68 ± 0:03 0:68 ± 0:01 n-6 PUFA 9:04 ± 0:20 38:21 ± 0:78 34:50 ± 1:47 12:99 ± 0:57 16:31 ± 0:04 C18:3 n-3 1:21 ± 0:02 4:72 ± 0:01 2:60 ± 0:08 1:21 ± 0:06 1:05 ± 0:06 C20:5 n-3 6:72 ± 0:31 2:20 ± 0:01 2:44 ± 0:07 2:36 ± 0:12 2:45 ± 0:01 C22:6 n-3 21:98 ± 0:89 6:51 ± 0:01 6:84 ± 0:22 6:88 ± 0:41 6:92 ± 0:07 n-3 PUFA 30:70 ± 0:94 13:72 ± 0:02 12:17 ± 0:24 10:76 ± 0:43 10:69 ± 0:10 Total PUFA 39:74 ± 0:97 51:93 ± 0:78 46:67 ± 1:48 23:75 ± 0:72 26:99 ± 0:11 T and T3 profile (mg/kg) α-T 168:99 ± 5:97 230:03 ± 3:49 186:32 ± 3:90 235:93 ± 4:29 217:13 ± 5:57 β-T 0:95 ± 0:06 3:61 ± 0:22 3:43 ± 0:10 1:93 ± 0:15 3:19 ± 0:18 γ-T 1:40 ± 0:02 69:95 ± 17:29 27:37 ± 0:66 4:11 ± 0:18 3:67 ± 0:38 δ-T ND 21:30 ± 4:50 16:68 ± 0:28 ND ND Total (T+T3) 177:07 ± 6:23 329:20 ± 29:37 240:30 ± 4:27 253:03 ± 4:79 233:95 ± 5:15 4 Lipid class composition (%) TAG 81:56 ± 0:77 86:74 ± 0:03 47:35 ± 0:10 76:06 ± 0:07 50:90 ± 0:26 DAG 8:76 ± 0:72 6:35 ± 0:33 14:17 ± 0:05 10:04 ± 0:01 15:53 ± 1:03 MAG 0:16 ± 0:02 0:47 ± 0:08 2:39 ± 0:01 0:93 ± 0:07 2:61 ± 0:21 FFA 9:53 ± 0:03 6:43 ± 0:22 36:09 ± 0:16 12:97 ± 0:14 30:97 ± 1:50 Abbreviations: For diet abbreviations (FO, SO, SAO, OPO, and OPAO), see Table 2; FA: fatty acid; SFA: saturated fatty acids (sum of C14:0, C15:0, C:16:0; C18:0, and C22:0); MUFA: monounsaturated fatty acids (sum of C16:1 n-9, C16:1 n-7, C18:1 n-9, C18:1 n-7, and C20:1 n-9); PUFA: polyunsaturated fatty acids (n-6 PUFA: sum of C18:2 n-6, C18:3 n-6, C20:2 n-6, and C20:4 n-6; n-3 PUFA: sum of C18:3 n-3, C20:3 n-3, C20:5 n-3, and C22:6 n-3; total PUFA: sum of n-6 PUFA and n-3 PUFA); T: tocopherol; T3: tocotrienol; TAG: triacylglycerols; DAG: diacylglycerols; MAG: monoacylglycerols; FFA: free fatty acids; ND: not detected. 1 The analytical methods applied are described in supplementary material. 2 Mean ± standard deviation of the two determinations. 3 The percentage of each FA was obtained by peak area normalization. 4 The percentage of each lipid class was obtained by peak area normalization. For refrigerated samples, the six fillets of each replicate (2 : 1, v/v) mixture. First, 20 mL of this mixture was added to were stored for 6 days under commercial conditions (at the sample and homogenized using a high-speed homoge- 2°C, packed in a modified atmosphere of CO2/N2/O2, 40/ nizer (PT 3100 Polytron, Kinematica, Lucerne, Switzerland) 30/30). After the refrigerated storage, fillets were pooled at 19,000 rpm for 20 s and filtered through Whatman No. 1 and ground. The color determination, the sampling for filter paper. The sample residue retained in the filter was chemical analysis, and the sensory acceptance test were done reextracted twice with 10 mL of the same solvent mixture as detailed for fresh samples. at 19,000 rpm for 20 s. Next steps of lipid extraction were carried out as described by Bou et al. [18]. Then, FA methyl 2.4. Determination of Fatty Acid Composition. The FA com- esters were obtained from the extracted lipid fraction by a position of fresh fish fillets (1.5-1.6 g) was analyzed by gas double methylation procedure and determined by GC-FID chromatography with flame ionization detector (GC-FID) [15]. Each compound was identified by comparing its reten- after extracting the lipid fraction with chloroform/methanol tion time with that of standards (Supelco 37 component 6 Aquaculture Nutrition FAME Mix, Supelco®, Merck KGaA, Darmstadt, Germany). with 1,1,3,3-tetraethoxypropane (Sigma-Aldrich®, Merck The percentage of each FA was obtained by peak area KGaA, Darmstadt, Germany) as MDA precursor. normalization. 2.8. Determination of Volatile Compound Content. The 2.5. Determination of Tocopherol and Tocotrienol Content. volatile compound content was determined in fresh and For the determination of the T and T3 content in fresh refrigerated fillets by headspace solid-phase microextraction and refrigerated fish fillets, 2 g of ground sample was homog- coupled with gas chromatography and mass spectrometry enized with a mixture of antioxidants in ethanol using a PT (HS-SPME-GC-MS). To perform the analysis, 1 g of sample 3100 Polytron for 30 s at 20,000 rpm and saponified with was weighed into a 10 mL screw-capped vial, and 0.5 mL of a methanolic KOH as described by Bou et al. [18]. The nonsa- 4 mg/L aqueous solution of 4-methyl-2-pentanol (97%, ponifiable fraction was extracted with petroleum ether. The Sigma-Aldrich®, Merck KGaA, Darmstadt, Germany) was solvent was evaporated under a nitrogen stream at 30°C, added as internal standard. Subsequently, 0.5 mL of an and the residue was dissolved in 99% n-hexane and injected aqueous antioxidant solution with 4% of EDTA and 0.4% into the chromatographic system. High performance liquid of propyl gallate (both from Sigma-Aldrich®, Merck KGaA, chromatography separation was carried out as described by Darmstadt, Germany), 2 mL of double deionized water, Aleman et al. [19], and T and T3 were detected using a and three glass balls were added. The vial was immediately 1260 Infinity II Fluorescence Detector (Agilent Technolo- closed and kept in ice until all sample set was prepared. gies, Santa Clara, CA, USA), setting the excitation and Then, the mixture was homogenized using an ultrasound emission wavelengths at 290 and 320 nm, respectively. Each bath at 4°C for 10 min. Samples were kept in ice at the dark sample was analyzed twice, and calibration curves were until the HS-SPME-GC-MS determination was carried out. prepared using α-, β-, γ-, and δ-T standards (Calbiochem®, The instrument consisted of an Agilent 6890N Network Merck KGaA, Darmstadt, Germany). The content of α-, β-, GC system with an Agilent 5975C Inert MSD quadrupole γ-, and δ-T3 was calculated by applying the calibration curve mass spectrometer (both from Agilent Technologies Santa obtained for the corresponding T analogue. Clara, CA, USA) and a PAL autosampler (CTC Analytics, Zwingen, Switzerland) configured to perform SPME. After 2.6. Ferrous Oxidation-Xylenol Orange (FOX) Method. The 10 min of sample conditioning at the extraction temperature FOX method was applied in each sample per triplicate to (45°C), the fiber of divinylbenzene/carboxen/polydimethyl- evaluate the primary oxidation and the oxidative stability siloxane (2 cm length, 50/30 thickness) from Supelco® of fresh and refrigerated fillets, as detailed by Grau et al. (Merck KGaA, Darmstadt, Germany) was exposed to the [20]. Briefly, 15 mL of cold methanol was added to 2 g of head space for 30 min and desorbed in the injector at sample, and a PT 3100 Polytron was used for sample 260°C for 10 min. To perform the separation of the different homogenization at 12,000 rpm for 30 s. Then, the mixture volatile compounds, a Supelcowax-10 capillary column was centrifuged (1,400 g, 3 min), and the supernatant (30 m × 0:25 mm i.d., 0.25 μm film thickness) from Supelco® methanol extract was collected. The reaction mixture was (Merck KGaA, Darmstadt, Germany) was used. The oven prepared in glass cuvettes, capped with Teflon caps. The temperature program began at 40°C (held 10 min, during reagents were added as described by Tres et al. [21], using fiber desorption time), 3°C/min up to 150°C, and 15°C/min 1030 μL of methanol and 70 μL of sample extract. The absor- up to 250°C (held for 5 min). Helium was used as gas carrier bance at 560 nm was measured using a UV-3600 spectro- with a constant flow of 1 mL/min. The temperatures of the photometer (Shimadzu, Kyoto, Japan) after incubation for ion source and the transfer line were 230 and 280°C, respec- 30 min (as a measure of the lipid hydroperoxides (LHP) tively, and the ionization energy was 70 eV. Data were present in the samples from the beginning, named LHP con- acquired in full scan mode in selected representative samples tent) and for 96 h (as a measure of the amount of LHP for the identification of compounds, which was carried out formed during this time which is considered a measure of by comparison of their mass spectra and retention times sample’s oxidative stability, named final LHP value). LHP with those of standard compounds or with those available concentration was expressed as mmol of cumene hydroper- in mass spectrum library Wiley 6 and in the literature. Then, oxide (CHP) equivalents/kg of sample, with reference to a the quantitative assessment of all samples was carried by calibration curve prepared using CHP as standard (technical selected ion mode, considering m/z 44, 45, 55, 56, 57, 58, grade, 80%, Sigma-Aldrich®, Merck KGaA, Darmstadt, 70, 81, and 98, which were representative for the compounds Germany). of interest. Data were then analyzed by an Agilent MSD ChemStation. Relative amounts of volatile compounds were 2.7. Determination of 2-Thiobarbituric Acid (TBA) Value. calculated by the internal standard method, expressing the The 2-thiobarbituric acid (TBA) value was determined twice results as μg of 4-methyl-2-pentanol equivalents/kg of per sample to evaluate the secondary oxidation of fresh and sample. refrigerated fillets, applying the method described by Grau et al. [22] on 1.5 g of sample. The TBA value was measured 2.9. Color Determination. Immediately after grinding the through third derivative spectrophotometry after an acid fillet samples, color was measured by CR-410 Konica aqueous extraction. The results were expressed as malon- Minolta colorimeter (Tokyo, Japan) based on CIE L∗ a∗ b∗ dialdehyde (MDA) concentration (μg/kg), using a calibra- color space. The color parameters determined were lightness tion curve prepared as described by Botsoglou et al. [23] (L∗ ), from dark (0) to light (100); redness (a∗ ), from green Aquaculture Nutrition 7 (–a∗ ) to red (+a∗ ); and yellowness, (b∗ ) from blue (–b∗ ) to eters evaluated in fish fillets (n = 50, 2 refrigeration times × 5 yellow (+b∗ ), as recommended by the International Com- dietary treatments × 5 replicates each; for sensory acceptance mission on Illumination in 1976 [24]. The instrument was n = 300, 2 refrigeration times × 5 dietary treatments × 30 rep- set for D-65 illuminant at a 2° observer angle and cali- licates each), except for FA profile, which was determined brated prior to the determinations with a standard white only in fresh samples. One-way ANOVA (SPSS GLM proce- plate. Five measures were taken for each replicate in ran- dure) was used to assess if there was a significant influence of dom different locations of the ground samples, and the the dietary treatment (FO, SO, SAO, OPO, and OPAO) on average L∗ , a∗ , and b∗ was calculated. all the parameters studied in fresh fillets (n = 25, 5 dietary The dimensionless parameter ΔE [25] was used to eval- treatments × 5 replicates each ; for sensory acceptance n = uate if the differences in color parameters between different 150, 5 dietary treatments × 30 replicates each) and in refrig- sample groups could be perceptible by the human eye. erated fillets (n = 25, 5 dietary treatments × 5 replicates each; pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi for sensory acceptance n = 150, 5 dietary treatments × 30 ΔE = ΔL∗2 +Δa∗2 +Δb∗2 : ð1Þ replicates each). Significant differences among dietary treat- ments found by one-way ANOVA in fresh or refrigerated In this study, two different ΔE parameters were obtained: fillets were evaluated by the Scheffé’s post hoc test (SPSS GLM procedure). Pearson’s correlation test was performed (i) ΔER was calculated for each dietary treatment to to study the correlations of the oxidative parameters deter- evaluate the differences between the color parame- mined in fillets with the FA profile (n = 25) and between ters of fresh and refrigerated fillets. In this case, all the parameters analyzed in fresh and refrigerated fillets ΔL∗ , Δa∗ , and Δb∗ in Equation (1) corresponded (n = 50). In all cases, differences were considered significant to the difference between the L∗ , a∗ , or b∗ means when p < 0:05. of the fresh samples (n = 5) and of the refrigerated samples (n = 5) of the same dietary treatment 3. Results and Discussion (ii) ΔED was used to study the differences in color As previously reported, AO present different compositional parameters of fresh or refrigerated fillets between traits than crude oils and a high variability in their composi- the control diet (FO) and the other four dietary tion [11, 14], and since this can affect their nutritional value, treatments. Thus, two different ΔED were obtained it is important to properly characterize them to evaluate for each dietary treatment: one for fresh fillets and their quality as feed fats. The two AO assayed in this study another for refrigerated fillets. In this case, the ΔL∗ , (SAO and OPAO) presented a higher FFA content than Δa∗ , and Δb∗ in Equation (1) corresponded to the SO and OPO (Table 1), as they are refining by-products difference between L∗ , a∗ , or b∗ means of fresh coming from neutralization, a refining step performed to (n = 5) or refrigerated samples (n = 5) from FO diet remove FFA from crude fats. These FFA percentages were and those from each one of the other four dietary in agreement with the ones reported by Varona et al. [11] treatments for various AO. Another characteristic of AO is that they also tend to accumulate substances present in the I and U A value of ΔE > 5 [26] was considered as the cut-off fractions, increasing their total MIU value [11]. The total value above which the instrumental color differences would MIU value is a parameter commonly used to evaluate the be perceived by the human eye. quality of feed fats as it represents the content of substances 2.10. Sensory Acceptance Test. Two nine-scale hedonic tests that can dilute their energy value. Accordingly, the total were performed to evaluate if there were significant differ- MIU values of the two AO used in this study were higher ences in overall acceptance between the dietary treatments than those of the rest of the fat sources, with the percentages in fresh or refrigerated fish fillets. In both cases, hamburgers of M, I, and U, and total MIU being similar or lower than the with 22.5 g ground fish fillet per unit were prepared and usual levels found for these type of AO in Spanish market cooked for 3 min using five different machines (model SS- [11]. For both SAO and OPAO, U was the fraction with 5515 750 W, Jocca, Zaragoza, Spain), one for each dietary the greatest contribution to the total MIU value (Table 1). treatment. A total of 30 regular fish consumers participated Remarkably, the recommendations published by FEDNA in each test. In each test, each participant evaluated five [27] for the use of AO in animal diets establish a total hamburgers (one per dietary treatment) and, for each of MIU value lower than 5%, which was fulfilled by the SAO them, indicated the degree of acceptability on a 9-point scale and slightly exceeded by OPAO. In fact, in the performance (1: “dislike extremely”; 5: “neither like nor dislike”; 9: “like parameters resulting from the experimental diets used in this extremely”). work and that were reported by Verge-Mèrida et al. [13], no effect was observed when SAO was used, but a significantly 2.11. Statistical Analysis. Statistical tests were carried out lower performance was found when fish were fed with with SPSS (version 20.0, IBM Statistics Inc.). Multifactor OPAO diet, showing the lowest weight (226.22 g vs. the ANOVA (SPSS GLM procedure) was used to study the range of 244.45–250.20 g), specific growth rate (0.80%/day influence of the refrigeration time (0 and 6 days) as one vs. the range of 0.88–0.90%/day), and average daily gain main factor and its interaction with the main factor dietary (1.25 g vs. the range of 1.43–1.49 g). Regarding the apparent treatment (FO, SO, SAO, OPO, or OPAO) on all the param- digestibility of dry matter, crude protein, or total FA, no 8 Aquaculture Nutrition Table 4: Fatty acid profile of fresh fillets coming from fish fed with the five experimental diets. FA (%) FO1 SO1 SAO1 OPO1 OPAO1 SEM2 p3 C14:0 2.5 a 1.5c 1.7b 1.6 bc 1.6 bc 0.026 <0.001 C16:0 19.2a 16.4bc 16.4bc 16.6b 16.0c 0.130 <0.001 C18:0 4.7a 4.6a 4.3b 4.0c 3.9c 0.053 <0.001 SFA 27.5a 23.2b 23.2b 22.9bc 22.3c 0.187 <0.001 C16:1 n-7 4.0a 2.3c 2.5bc 2.6b 2.7b 0.053 <0.001 C18:1 n-9 25.8d 27.8d 30.6c 43.3a 40.6b 0.538 <0.001 C18:1 n-7 1.8a 1.6ab 1.4ab 1.2b 1.6ab 0.118 0.022 C20:1 n-9 1.9a 1.4d 1.5cd 1.6b 1.6bc 0.027 <0.001 MUFA 34.0bc 33.4c 36.5b 49.2a 47.0a 0.560 <0.001 C18:2 n-6 11.4e 26.2a 23.7b 12.9d 15.1c 0.310 <0.001 C20:4 n-6 1.7a 0.7b 0.8b 0.8b 0.8b 0.024 <0.001 n-6 PUFA 13.9d 28.3a 25.6b 14.4d 16.6c 0.314 <0.001 C18:3 n-3 2.0c 3.8a 2.6b 1.9d 1.9cd 0.028 <0.001 C20:5 n-3 5.3a 2.7c 3.0b 2.9bc 3.0b 0.055 <0.001 C22:6 n-3 16.1a 7.7b 8.2b 7.9b 8.1b 0.293 <0.001 n-3 PUFA 24.3a 14.8b 14.4b 13.2b 13.7b 0.331 <0.001 Total PUFA 38.2b 43.0a 40.0b 27.6d 30.3c 0.483 <0.001 Abbreviations: For diet abbreviations (FO, SO, SAO, OPO, and OPAO), see Table 2; FA: fatty acid; SFA: saturated fatty acids (sum of C14:0, C16:0, C17:0, C18:0, C20:0, and C22:0); MUFA: monounsaturated fatty acids (sum of C16:1 n-9, C16:1 n-7, C18:1 n-9, C18:1 n-7, and C20:1 n-9); PUFA: polyunsaturated fatty acids (n-6 PUFA: sum of C18:2 n-6, C18:3 n-6, C20:2 n-6, and C20:4 n-6; n-3 PUFA: sum of C18:3 n-3, C20:3 n-3, C20:5 n-3, and C22:6 n-3; total PUFA: sum of n-3 PUFA and n-6 PUFA). See Table S1 in supplementary material for the complete FA composition including minor FA results. 1 Data were expressed as the mean of the five replicates from each dietary treatment (n = 5). 2 Standard error of the mean. 3p values obtained by ANOVA (n = 25). Values in bold were significant (p < 0:05). Differences between dietary treatments found with Scheffé’s post hoc test were noted in the same row as a > b > c > d > e. effect was observed when AO were included in the diets EPA and DHA content in the whole fish, organs, and flesh [13]. Beyond this, to promote the upcycling of this type [10]. However, considering that the general recommenda- of refining by-products as vegetable feed fats to replace tion for the daily EPA plus DHA intake is 250 mg [29] and FO for aquaculture nutrition, it is important to compre- taking into account the amount of fat extracted from fish fil- hensively study their impact not only on fish performance lets in this study (data not shown), 100 g of FO fillets would parameters but also on the composition, oxidative stability, cover 302% of the recommended EPA plus DHA daily and sensory acceptability of the flesh, so that the interest intake, whereas fillets from the other four diets would cover of the various stakeholders (feed producers, farmers, and between 168% and 192%. Therefore, even though FO was consumers) is addressed. partially replaced with the AO or crude oils, the content of these FA in fish fillets was enough to cover the recom- 3.1. Fatty Acid Profile of Fish Fillets. The results of the FA mended EPA + DHA daily intake. Nevertheless, it is impor- profile of fresh fillets coming from each diet are presented tant to remind that these LC-PUFA are highly prone to in Table 4 (the complete FA composition including minor lipid oxidation, and, as it will be commented below, they FA can be found in Table S1 in supplementary material). might influence the oxidative stability of fish fillets. The differences in the FA profile of fish fillets between The decrease in saturated fatty acid (SFA) percentage dietary treatments were significant for all the determined (from 27% to 22%–23%) in fish fillets observed with the FA and for the sums of each FA type. Most of the trials partial substitution of FO by SO, SAO, OPO, or OPAO found in the literature showed that the FA profile of fish fil- was related to the FA profile of the feed fats and of the diets, lets mirrored the FA composition of the diet [6, 28]. In our as FO diet showed the highest content in SFA (35%). This study, fillets coming from FO diet had the highest content was also in concordance with previous trials. For example, in EPA (5%) and DHA (16%), whereas the other four diets Mourente and Bell [5] also found a reduction of SFA levels led to ≈3% of EPA and to similar levels of DHA (≈8%) in in European seabass fillets when a 60% of FO was replaced fillets (Table 4). This was in agreement with the EPA and by different blends of vegetable fat sources in fish diets with DHA reductions observed in the diets when FO was partially 16% added fat. However, the decrease in the SFA percentage substituted by vegetable sources (from 7% of EPA and 22% (from 24% to ≈21%) was lower than the one observed in this of DHA in the FO diet, to approximately 2% of EPA and study, as their vegetable blends included palm oil. Another 7% of DHA in the rest of diets) (Table 3). Similarly, previous work representative of this fact is the one carried out by studies reported that the main effect of the partial or total Izquierdo et al. [30] who, despite of working with another substitution of FO with vegetable oils is a reduction of species (gilthead seabream) and of replacing FO by 60% or Aquaculture Nutrition 9 80% of different vegetable oils (linseed, rapeseed, or S) in ied and its proportion between polar lipids, with structural diets with 17.6% added fat, observed a SFA reduction in functions, and neutral lipids [36, 37]. fillets (from 28% to the range of 24%–21%) similar to the one found in this study. 3.2. Tocopherol and Tocotrienol Composition of Fish Fillets. Other distinctive aspects of the FA profile of the fat Vitamin E has an important role during fish rearing, as it sources and diets were also reflected in the FA composition helps preventing some diseases and contributes to a correct of fillets (Table 4). For example, fillets from fish fed with development of fish [38]. Moreover, the T and T3 deposited SO diet showed the highest levels of linoleic (26%) and lino- in fish fillets might have an important role in lipid oxidation, lenic (4%) acids. In general terms, the FA profiles of SO and and thus, in their shelf life. The T and T3 contents in fresh SAO fillets were similar, although differences were found for and refrigerated European seabass fillets are presented in oleic (C18:1 n-9) and linoleic (C18:2 n-6) acids. Again, these Table 5. differences reflected the ones observed between the corre- The contents of total T + T3 and α-T, which was the sponding diets (Table 3) and between the two fats (SO and main tocol present in fish fillets, were not significantly SAO) (Table 1). In fact, the lower linolenic acid content in affected by refrigeration. The refrigerated storage signifi- SAO compared with SO evidenced that this fat by-product cantly reduced the β-T, γ-T, and β-T3 concentration in fil- contained AO from the refining of soybean and sunflower lets coming from all diets. Literature results on the T and oils [31]. Previous works performed using diets formulated T3 stability during storage are controversial, as they might with SO as a partial replacement of FO also revealed incre- depend on the storage conditions and on the many factors ments of linoleic and linolenic acids in fish muscle. For that can affect T and T3 content in fish, such as the biology example, Izquierdo et al. [32] observed that linoleic acid of the species, the source of vitamin E used in the diet, the content in seabass muscle raised from 5% to 17% and linole- presence of interacting compounds in the diet (such as sele- nic from 1% to 2% when using a diet with 19.7% of added fat nium, vitamin C, or astaxanthin), or the PUFA content [39]. and a 60% replacement of FO by SO. The lower linoleic acid For example, mackerel-minced muscle kept into plastic bags percentage and the lower increase found by them in compar- significantly reduced its α-T content from 3.6 mg/kg to ison to our results can be explained by the presence of a 1.0 mg/kg after 3 days of refrigeration at 4°C, and it contin- lower content of this FA in their diets (7% in FO and 26% ued dropping to 0.6 mg/kg after 7 and 11 days [40]. Another in SO). The same work showed an influence of the fish spe- trial also showed that when rainbow trout specimens coming cies on the tendency to mirror the FA profile of the diet, with from a diet with 100 mg/kg of added α-tocopheryl acetate the increase of the linoleic acid percentage observed in gilt- were packed into black nylon bags and refrigerated for 9 head sea bream flesh (from 6% to 24%) being greater than days at 1°C, the α-T concentrations in fillets decreased in seabass muscle [32]. Similarly, when Menoyo et al. [33] from 30.1 mg/kg to 19.1 mg/kg [41]. However, when sea- replaced 60 or 80% of FO by SO in diets with 17.6% added bass specimens fed diets with 4 different levels of α-toco- fat, the increases in the linoleic acid content of gilthead sea pheryl acetate were kept in boxes, covered with flake-ice bream (from 6% to 24% with the diet with 60% of SO and and black nylon bags, and stored in a refrigerated room to 28% with the diet with 80% of SO) were higher than the at 1°C for 12 days, no significant variations in the vitamin one observed in our study for European seabass. Although E content (α-T + γ-T) of fillets were observed [42]. this fact might be due to the different species studied, it The total T + T3 content and the tocol profile of fresh can be also related to the greater differences in the linoleic and refrigerated fillets depended on the diet (Table 5), with acid content between their diets (5% in FO diet, 30% in the differences being in concordance with those observed the diet with 60% of SO, and 39% in the one with 80% of between diets (Table 3) and between fat sources (Table 1). SO) than among the diets used in this study (Table 3). As animals are unable to synthetize T and T3, it is manda- Regarding MUFA, the highest MUFA values in the pres- tory to have enough levels in diets to, at least, fulfill the ent study were found in fillets from fish on OPO and OPAO essential requirements [43]. Since it might be difficult to diets (≈48%), being higher than those found in FO and SO reach them only with the T and T3 supplied by the added fillets. This agreed with the higher MUFA content in OPO fat (Table 1) and by the rest of the feed ingredients, the addi- and OPAO diets (Table 3) resulting from the addition of tion of vitamin E in fish diets is a common practice in aqua- these fats of olive origin (Table 1). Likewise, the study car- culture. In this study, the vitamin-mineral premix supplied ried out by Mourente et al. [34] showed an increment of 192.23 mg of α-tocopheryl acetate per kg of feed (Table 2), oleic acid in fish flesh (from 16% to 28%) when European making α-T the main tocol in all diets (Table 3) and, conse- seabass was fed with a diet with 17.7% of added fat and a quently, in fish fillets (Table 5). This also attenuated the dif- 60% of FO replacement by olive oil. However, a great num- ferences in the tocol profile and in the total T + T3 content ber of factors beyond the fat source used in the diet can observed between the diets (Table 3). However, the influence determine the FA profile of fish, such as the fish species used, of the type of fat used in each diet was still noticeable in the the feeding period, the added fat percentage, or the propor- diets (Table 3) and thus reflected in the total T + T3 content tion of FO replacement. For example, Nasopoulou et al. [35] and tocol profile of fresh and refrigerated fillets (Table 5). did not obtain significant differences in oleic acid content for For instance, the partial substitution of FO by vegetable oils gilthead sea bream muscle probably because in diets, FO was in fish diets increased T + T3 levels and modified the tocol substituted by only 8% of OPO. Another variable that affects profile of fish fillets (Table 5) because, among the fat sources, the magnitude of FA modification is the type of muscle stud- FO had the lowest total T + T3 concentration (Table 1). 10 Aquaculture Nutrition Table 5: Tocopherol and tocotrienol content of fresh and refrigerated fish fillets coming from fish fed with the five experimental diets. α-T β-T (mg/kg) γ-T (mg/kg) δ-T (mg/kg) β-T3 (mg/kg) T + T3 (mg/kg) (mg/kg) Effect of the diet on fresh fillets FO1 6.98c 0.09c 0.43c 0.05c 0.12b 7.68c 1 ab SO 9.12 0.29a 5.79a 1.08a 0.16b 16.43a 1 bc SAO 8.02 0.22b 1.96b 0.80b 0.16b 11.16b OPO1 10.44a 0.19b 0.71c 0.04c 0.22a 11.60b 1 ab OPAO 9.17 0.26a 0.72c 0.06c 0.23a 10.44b 2 SEM 0.351 0.008 0.124 0.020 0.009 0.457 pdiet3 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 Effect of the diet on refrigerated fillets FO1 6.48c 0.08d 0.39c 0.04c 0.08b 7.06c 1 ab SO 8.75 0.26a 5.20a 0.97a 0.12ab 15.28a 1 b SAO 8.15 0.21b 1.84b 0.73b 0.13ab 11.05b OPO1 10.03a 0.17c 0.64c 0.04c 0.19a 11.07b 1 ab OPAO 9.38 0.23ab 0.63c 0.04c 0.18a 10.45b 2 SEM 0.287 0.008 0.127 0.016 0.018 0.395 pdiet3 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 Effect of the refrigeration Fresh Fillets4 8.74 0.21 1.92 0.41 0.18 11.46 Refrigerated Fillets4 8.56 0.19 1.74 0.35 0.14 10.98 SEM2 0.144 0.004 0.056 0.008 0.006 0.191 prefrigeration5 0.363 < 0.001 0.028 < 0.001 < 0.001 0.089 Effect of the interaction between refrigeration and diet prefrigeration x diet5 0.724 0.617 0.173 0.011 0.891 0.686 1 Abbreviations: For diet abbreviations (FO, SO, SAO, OPO and OPAO) see Table 2; T, tocopherol; T3, tocotrienol. Mean of the different replicates for each dietary treatment (n = 5). 2 Standard error of the mean. 3p values obtained from ANOVA (n = 25) of fresh or refrigerated fillets. Values in bold were significant (p < 0.05). The differences between dietary treatments found in fresh or refrigerated fillets with Scheffé’s post hoc test (n = 25) were noted in the same column as a > b > c > d. 4 Pooled means of fresh or refrigerated fillets coming from the five dietary treatments (n = 25). 5p values obtained for the refrigeration (prefrigeration) and the interaction between the refrigeration time and the diet (prefrigeration x diet) from multifactor ANOVA (n = 50). Values in bold were significant (p < 0.05). Regarding the vegetable fats, the total T + T3 levels were sim- fillets with similar α-T content (Table 5). Some modifica- ilar for OPO and OPAO, and the highest amount was found tions in the tocol profile of fish flesh when vegetable oils for SO, and even if SAO presented the second highest T + T3 are added to fish diets have been reported before. For amount, it was 7.8 times lower than that of S (Table 1). Con- instance, Regost et al. [44] demonstrated that the α-T con- sequently, SO fish fillets showed the highest T + T3 content tent was lower in Atlantic salmon fed with FO and SO and FO fish fillets the lowest ones (Table 5). Concerning diets than with low-erucic acid rapeseed oil, and that the the tocol profile of fish fillets, the highest α-T content was highest γ-T level was observed for SO diet. Ng et al. found for fillets from OPO diet, and similar α-T levels were [45] showed an increase in total T + T3 levels (from 7.5 found for OPAO and SO fillets, whereas fillets from SO diet to 19.3 mg/kg) and T3 content (from 1.0 mg/kg to more had the highest levels of γ-T and δ-T, followed by the ones than 5.0 mg/kg) of catfish muscle with the replacement from SAO diet. This is also in concordance with the tocol of FO by palm fatty acid distillate in the diet (from 0% profile of diets (Table 3) and fat sources (Table 1); as in fat to 100%). Also, Trullàs et al. [46] reported a greater con- sources, the highest γ-T levels were found for SO followed tent of β − T + γ − T and total T + T3 in rainbow trout by SOA, with γ-T being the main compound in SO and when FO was replaced by 75% of different types of exper- α-T in SOA, which evidenced the presence of AO of sun- imental rapeseed oils in diets with 20.1% added fat. flower origin [31]. The use of OPAO diet led to a similar T + T3 content and profile in fish fillets as OPO diet, 3.3. Lipid Hydroperoxide Content and Oxidative Stability of except for a higher β-T content, which is in concordance Fish Fillets. Primary oxidation of fresh and refrigerated fillets with the similar tocol profile between these two fat sources was evaluated by the LHP content obtained by the FOX (Table 1). Comparing SO with SAO, SO had a much method after a 30 min incubation. The LHP content in fresh higher α-T content than SAO (Table 1), but these differ- fillets was too low to be quantifiable. However, the refriger- ences were attenuated in the diets (Table 3) and led to fish ated storage increased the LHP content, with the levels being Aquaculture Nutrition 11 higher in refrigerated FO fillets (0.10 mmol/kg) than in OPO 0.15 fillets (0.04 mmol/kg) (Figure 1). This might be related to the A different FA unsaturation of fillets from different dietary (mmol of CHP eq/kg) treatments (Table 4), as FO fillets showed the highest con- 0.10 LHP content AB AB tent of unsaturated FA, where as OPO fillets had the lowest. AB The oxidative stability of the different samples (the B higher the formation of LHP, the lower the oxidative sta- 0.05 bility) was evaluated by measuring the LHP by the FOX assay after an incubation period of 96 h (Final LHP value, Figure 2). 0.00 The effect of the interaction between the refrigeration FO SO SAO OPO OPAO time and the diet on the final LHP value was not significant. Refrigerated fillets Also, the refrigerated storage for 6 days at the conditions assayed in this study did not affect the oxidative stability of Figure 1: Primary oxidation of refrigerated fillets: lipid the fillets. However, there were significant differences for this hydroperoxide content (obtained after an incubation time of parameter between the diets (Figure 2) that were in concor- 30 min) from the five dietary treatments of interest. The results dance with the FA profile and the T and T3 content of fish were represented as mean ± standard deviation (n = 5) of CHP fillets. The lowest oxidative stability was found for FO fillets, concentrations (mmol/kg). The differences between dietary and it could be associated with its higher EPA and DHA treatments found in refrigerated fillets with Scheffé’s post hoc contents, which are FA with a great tendency to suffer lipid (n = 25) were noted as A>B. Abbreviations: For diet abbreviations oxidation [47]. This fact was supported by the significant (FO, SO, SAO, OPO, and OPAO), see Table 2; LHP: lipid Pearson’s correlations between the final LHP value and hydroperoxide; CHP: cumene hydroperoxide. EPA (r = 0:836; p < 0:001) and DHA percentages (r = 0:850; p < 0:001). As Figure 2 shows, SO and SAO diets led to fish of fish fillets packed in modified atmosphere. Generally, fillets with an intermediate oxidative stability. Fillets coming the results found in the literature are in concordance with from both diets were rich in linoleic and linolenic acid con- the development of secondary oxidation reactions in fish tents (Table 4), which are less susceptible to oxidize than during storage observed in our study. For instance, Kyrana EPA and DHA but more than MUFA. In addition, SO and and Lougovois [48] also found increases in the TBA values SAO fillets had a significantly higher T + T3 content than (from 370 to 650 μg MDA/kg) in skinned fillets from FO fillets (Table 5), which might help preventing oxidation ungutted European seabass stored in ice inside a refrigerator reactions. In fact, the final LHP value was negatively corre- (0°C–4°C) for 22 days. Likewise, TBA values gradually lated with the content of the main tocol (α-T) in fish fillets increased in fillets from rainbow trout packed into black (r = −0:807; p < 0:001). Despite SO fillets presented the high- nylon bags and stored at 1°C for 9 days and revealed that a est T + T3 levels (Table 5), FA unsaturation was similar in SO higher α-tocopheryl acetate content in the feed led to a and SAO fillets (Table 4), which led to similar behaviors in higher α-T in trout fillets and a lower TBA values [41]. terms of oxidative stability. The highest oxidative stability Concerning the effect of the diet on fillet TBA values, sig- was found for OPO and OPAO fillets, and this might also nificant differences were observed between FO and the other be related with their FA profile (Table 4), in which MUFA dietary treatments (Figure 3), with the highest TBA values were the main type of FA, and PUFA had less relevance than being found in FO fillets. This agreed with FO fillets showing in the other fillets. This is in agreement with the negative the lowest oxidative stability (Figure 2), a fact that might be correlation found between final LHP value and oleic acid associated with their highest n-3 LC-PUFA content and the percentage (r = −0:820; p < 0:001). Therefore, lipid oxidative lowest T + T3 levels. Accordingly, TBA values had a strong stability was clearly influenced by the botanical origin of the positive correlation with various LC-PUFA, such as arachi- fat source used, but there was not an observable impact of the donic acid (C20:4 n-6) (r = 0:943; p < 0:001), C20:3 n-3 type (crude or acid) of vegetable oil. In this sense, the upcy- (r = 0:879; p < 0:001), EPA (r = 0:936; p < 0:001), and DHA cling of SAO or OPAO as a replacement of FO in fish diets (r = 0:942; p < 0:001). Our observations about the impact could improve the lipid oxidative stability in a similar way of the diet on TBA values are in agreement with those than their corresponding crude oils. obtained by Regost et al. [44] in Atlantic salmon, who also observed higher TBA levels when FO was added (at 29.8%) 3.4. TBA Values of Fish Fillets. The presence of MDA as sec- to the diet instead of SO. Contrarily, Trullàs et al. [46] found ondary oxidation product in fish fillets from the different no significant differences in fillets of rainbow trout fed with a diets was evaluated by the TBA values (Figure 3). FO diet (20%) or those with vegetable fat sources (75% of FO The TBA values of fish fillets were significantly affected replacement), but in that case, FO fillets showed the highest by the refrigeration time (p < 0:001) and the diet used (for content in α-T. fresh fillets p < 0:001 and for refrigerated fillets p = 0:006), although there was no significant effect of the interaction 3.5. Content of Volatile Compounds of Fish Fillets. The between the two factors. The TBA values increased from formation of volatile compounds can lead to a loss of senso- the range of 112 μg MDA/kg–412 μg MDA/kg to 488 μg rial attributes, because they are contributors to the aromatic MDA/kg–1051 μg MDA/kg during the refrigerated storage profile of the fish. Nevertheless, the influence of volatile 12 Aquaculture Nutrition A A 2.0 2.0 (mmol of CHP eq/kg) (mmol of CHP eq/kg) Final LHP value Final LHP value 1.6 1.6 B 1.2 1.2 B B B 0.8 0.8 0.4 C C 0.4 C C 0.0 0.0 FO SO SAO OPO OPAO FO SO SAO OPO OPAO Fresh fillets Refrigerated fillets (a) (b) Figure 2: Oxidative stability: final lipid hydroperoxide value obtained after an incubation time of 96 h for fresh fillets (a) and refrigerated fillets (b) from the five diets of interest. The results were represented as mean ± standard deviation (n = 5) of CHP concentrations (mmol/kg). The differences between dietary treatments found in fresh or refrigerated fillets with Scheffé’s post hoc test (n = 25) were noted as A>B>C. Abbreviations: For diet abbreviations (FO, SO, SAO, OPO, and OPAO), see Table 2; LHP: lipid hydroperoxide; CHP: cumene hydroperoxide. A 1400 1400 1200 1200 (µg of MDA/kg) (µg of MDA/kg) 1000 1000 TBA values TBA values B 800 800 B B B 600 A 600 400 400 200 B B B B 200 0 0 FO SO SAO OPO OPAO FO SO SAO OPO OPAO Fresh fillets Refrigerated fillets (a) (b) Figure 3: Secondary oxidation: TBA values of fresh fillets (a) and refrigerated fillets (b) coming from the different dietary treatments. The results were represented as mean ± standard deviation (n = 5) of TBA values (μg of MDA/kg). The differences between dietary treatments found in fresh or refrigerated fillets with Scheffé’s post hoc test (n = 25) were noted as A>B. Abbreviations: For diet abbreviations (FO, SO, SAO, OPO, and OPAO), see Table 2; MDA: malondialdehyde. compounds on the odor depends not only on their concen- ent diets (Table 6). Heptanal might be associated with the tration but also on their odor threshold. In the present study, oxidation of n-9 MUFA (such as oleic acid) or n-6 PUFA a total of twelve compounds were identified in fresh and (such as linoleic or arachidonic acids) [49, 50]. For the alde- refrigerated European seabass fillets, including seven alde- hydes in refrigerated fillets, significant differences between hydes (propanal, pentanal, hexanal, heptanal, (Z)-4-hepte- dietary treatments were only found for propanal (Table 6), nal, octanal, and nonanal), three alcohols (1-penten-3-ol, with the levels in FO fillets (6 μg/kg) being higher than in (Z)-2-penten-1-ol, and 1-octen-3-ol), one ketone (2-octa- OPAO fillets (2 μg/kg). Propanal has been linked to the oxi- none), and one furane (2-pentylfuran) (Table 6). Generally, dation of n-3 PUFA [51], and, in this case, its relation with in meat and fish products, aldehydes are considered crucial the development of lipid oxidation could be confirmed by compounds, as they are characterized for possessing a lower its strong correlation with the TBA values (r = 0:751; p < odor threshold than alcohols and ketones [49]. 0:001). In our study, the main aldehyde observed in fish There was not a significant effect of the interaction fillets was hexanal, which has been associated with n-6 between the two main factors (refrigeration time and diet) PUFA oxidation, and it is commonly used as an indicator on any volatile compound. The content of all the identified of lipid oxidation [49, 50]. However, in this work, hexanal compounds suffered a noticeable increment after the refrig- was not significantly affected by the diet, contrary to what erated storage (Table 6). was observed for propanal. This is in agreement with some Regarding the differences observed between dietary authors that have proposed propanal and heptanal as better treatments, in fresh fillets, heptanal was the only aldehyde indicators of rancidity in fish, whereas hexanal would significantly affected by the diet (p = 0:024), but Scheffé’s usually be better correlated with meat flavor deterioration post hoc test was not able to separate the means of the differ- [52, 53]. The other aldehydes found in fish fillets could also Table 6: Volatile compounds (μg/kg) of fresh and refrigerated European seabass fillets. Propanal Pentanal Hexanal Heptanal (Z)-4-Heptenal Octanal Nonanal 1-Penten-3-ol (Z)-2-Penten-1-ol 1-Octen-3-ol 2-Octanone 2-Pentylfuran Aquaculture Nutrition Effect of the diet on fresh fillets FO1 2.7 7.0 46.6 9.6 1.2 3.1 7.0 34.8 5.7 49.8a 4.5 9.9b 1 SO 1.7 5.4 42.8 5.4 0.7 1.9 6.4 23.2 4.6 24.3b 3.0 13.0ab SAO1 1.7 5.1 36.9 4.8 1.2 2.2 6.2 15.9 2.2 21.8b 3.5 22.5a OPO1 2.2 6.7 32.3 4.6 0.8 2.1 5.8 26.8 5.9 23.9b 3.8 9.5b OPAO1 1.9 4.1 25.6 4.4 0.8 1.3 4.7 22.6 5.0 21.2b 4.0 11.0b 2 SEM 0.458 1.616 9.024 1.166 0.319 0.500 1.157 4.447 1.141 4.065 0.465 2.296 pdiet3 0.516 0.708 0.511 0.024 0.765 0.192 0.722 0.083 0.202 < 0.001 0.200 0.004 Effect of the diet on refrigerated fillets FO1 6.0a 14.3 100.6 15.1 2.1 4.6 11.8 84.9a 15.4a 126.5a 4.9 14.1 SO1 3.8ab 17.5 100.3 15.6 2.5 3.6 11.0 40.5b 7.9ab 64.6b 7.0 33.1 1 ab SAO 3.6 9.0 91.2 8.2 2.3 3.7 9.0 42.6b 6.0b 66.9b 5.0 33.9 OPO1 3.3ab 14.8 62.0 9.8 1.7 3.2 10.1 49.9ab 10.7ab 60.4b 5.3 12.2 OPAO1 2.5b 8.9 57.2 13.9 2.5 4.6 10.0 46.4ab 8.8ab 65.1b 5.4 18.5 2 SEM 0.560 2.824 12.710 3.301 0.910 0.909 3.010 7.829 1.586 11.956 0.519 6.236 pdiet3 0.005 0.192 0.060 0.469 0.971 0.774 0.977 0.005 0.008 0.004 0.067 0.068 Effect of the refrigeration Fresh Fillets4 2.1 5.6 36.8 5.8 0.9 2.1 6.0 24.6 4.7 28.2 3.8 13.2 Refrigerated Fillets4 3.8 12.9 82.2 12.5 2.2 3.9 10.4 52.9 9.8 76.7 5.5 22.4 2 SEM 0.228 1.022 4.908 1.096 0.302 0.326 1.010 2.829 0.615 3.953 0.220 2.081 prefrigeration5 < 0.001 < 0.001 < 0.001 < 0.001 0.006 < 0.001 0.005 < 0.001 < 0.001 < 0.001 < 0.001 0.004 Effect of the interaction between refrigeration and diet prefrigeration x diet5 0.108 0.437 0.559 0.595 0.941 0.622 0.989 0.112 0.141 0.179 0.050 0.364 1 Abbreviations: For diet abbreviations (FO, SO, SAO, OPO and OPAO) see Table 2. Mean of the different replicates for each dietary treatment (n = 5) expressed as 4-metil-2-pentanol equivalents. 2 Standard error of the mean. 3p values obtained from ANOVA (n = 25) of fresh or refrigerated fillets. Values in bold were significant (p < 0.05). The differences between dietary treatments found in fresh or refrigerated fillets with Scheffé’s post hoc test were noted in the same column as a > b. 4 Pooled means of fresh or refrigerated fillets coming from the five dietary treatments (n = 25). 5p values obtained for the refrigeration (prefrigeration) and the interaction between the refrigeration time and the diet (prefrigeration x diet) from multifactor ANOVA (n = 50). Values in bold were significant (p < 0.05). 13 14 Aquaculture Nutrition be linked to oxidation reactions of some specific FA, octanal pounds and the TBA values but also for the total aldehyde and nonanal might be associated to n-9 MUFA oxidation, content (r = 0:715; p < 0:001). Also, the differences in vola- pentanal might come from the oxidative degradation of n- tile compounds observed between FO fillets and fillets from 6 PUFA, and (Z)-4-heptenal might be formed during n-3 the rest of the diets agreed with their higher TBA values. PUFA oxidation [49], but results did not reveal any signifi- cant effect of the dietary treatment on any of them. 3.6. Color of Fish Fillets. The behavior of the color parame- Similar to aldehydes, the presence of some alcohols, as ters (L∗ , a∗ , and b∗ ) in the different diets was the same in the three identified in fish fillets in this study, might be fresh than in refrigerated fillets (Table 7). The refrigerated related with lipid oxidation. For example, 1-octen-3-ol was storage had a significant impact on L∗ and a∗ , whereas no the most important alcohol found in fish fillets, which is effect was found for b∗ . The lightness of the fillets increased well-known because of its low odor threshold [49] and might after the refrigerated period and the a∗ values decreased, be generated from n-6 PUFA oxidation [54]. In this study, implying that refrigerated fillets lost redness and gained the highest 1-octen-3-ol content was found in FO fillets greenness compared with fresh fillets. (50 μg/kg in fresh fillets and 126 μg/kg in refrigerated fillets). The ΔER parameter that compares instrumental color This is in agreement with the highest arachidonic acid parameters between fresh and refrigerated fillets reflected amount obtained for FO fillets even though they did not that the greatest color stability was found for SO fillets have the highest total n-6 PUFA content (Table 4). In addi- (0.6), whereas FO fillets were the ones which experienced tion, 1-octen-3-ol was the volatile compound that most cor- the highest color change after the refrigerated storage (2.4). related with arachidonic acid content (r = 0:785, p < 0:001). Since ΔER values were in all cases lower than 5 [26], these In refrigerated fish fillets, the diet also had a significant differences in color parameters might not be perceived by impact on the content of the other two alcohols identified: the human eye. 1-penten-3-ol and 2-pentenol. Refrigerated FO fillets Color differences among fresh fillets from the studied showed higher levels of 1-penten-3-ol than refrigerated SO dietary treatments were significant for L∗ and a∗ . The high- and SAO fillets and higher content of (Z)-2-penten-1-ol est L∗ value was observed for SO fresh fillets, where as FO than SAO fillets (Table 6). This alcohol has been associated and OPAO fresh fillets had the lowest values. However, FO with the oxidation of n-3 PUFA, such as linolenic acid, fresh fillets were the ones with highest a∗ . Accordingly, com- EPA, and DHA [50, 55]. The isomer (E)-2-pentenol has been pared to FO fresh fillets, the highest similarities in color were identified by other authors in sardines, turbot, mussels, and found for OPAO fresh fillets (ΔED = 0:8) and the greatest clams [56], whereas (Z)-2-penten-1-ol has been proposed differences for SO fresh fillets (ΔED = 2:4). In the case of as a potential marker for salmon freshness [57] and for oxi- refrigerated fillets, significant differences were found only dation of krill oil during storage [58]. In fact, some authors in a∗ between FO fillets and SO and OPO fillets. However, have related the formation of this last alcohol with the oxida- considering all three parameters, OPO fillets (ΔED = 0:8) tion of n-3 PUFA, concretely of EPA and DHA [50, 59]. had the greatest color similarities to FO refrigerated fillets, The only ketone identified in this study was 2-octanone, whereas OPAO fillets (ΔED = 0:9) were the least similar to which was a minor volatile compound in fresh (≈4 μg/kg) FO fillets. As ΔED in all cases was lower than 5 [26], the and refrigerated fish fillets (≈6 μg/kg). Some studies have color changes of fresh and refrigerated fillets due to the par- linked the formation of this ketone to the oxidation of unsat- tial substitution of FO for SO, SAO, OPO, or OPAO might urated FA, but any of them have specified the type of FA that not be perceptible by the human eye. can be involved [60–62]. However, Parlapani et al. [63] The facts that are behind the color changes in fish and reported a bacterial origin for 2-octanone in a model sub- fish products are not as clear as in the case of meat [64] strate of seabream (made of sterile flesh fish juice agar) only and are less studied in white fish muscle than in flesh rich when it was inoculated with different bacteria (Pseudomonas in pigments, such as salmon. Two different trials showed and Carnobacterium/Lactobacillus) and stored in modified that 60% or 80% FO replacement by vegetable oils (SO or atmosphere (CO2: 60%, O2: 10%, and N2: 30%) for 18 days linseed oil) reduced a∗ and b∗ in seabream fillets [30, 33]. at 0°C or for 2 days at 15°C. In the present study, yellowness (b∗ ) was the most stable The last compound found in fish fillets was 2- parameter as it was not influenced by the refrigeration, and pentylfuran which has been associated with n-6 PUFA oxi- there was no strong correlation between the lightness, dation [49]. The diet had a significant effect on the content redness, or yellowness and lipid oxidation parameters. of this compound in fresh fillets but not in refrigerated fil- Contrarily, another study related the increase of b∗ in dark lets. Fresh SAO led to higher values (22 μg/kg) than FO, muscle of yellowtail after 2 days of storage in ice to the OPO, or OPAO fillets (10 μg/kg–11 μg/kg). development of lipid oxidation [65]. These differences Despite some volatile compounds can be formed by between our results and the ones found in the literature other reactions not related to FA, such as amino acid degra- can be due to the different conditions (type of muscles, fish dation or microbial action [49], the relation between the vol- species, or processing and storage conditions) used in each atile compound formation and lipid oxidation in this study case. was clear. The significant increase of all volatile compounds in fish fillets after the refrigeration agreed with the develop- 3.7. Sensory Acceptance of Fish Fillets. Refrigeration had no ment of oxidation reactions (Figure 3). Moreover, strong significant effect on the overall acceptance of European sea- correlations were found not only for different volatile com- bass filets (Table 7). Despite ANOVA showed significant Aquaculture Nutrition 15 Table 7: Color parametres (L∗ , a∗ , b∗ , and ΔE values) and consumer overall acceptance of fresh and refrigerated fish fillets. Color parameters Overall acceptance L∗ a∗ b∗ ΔER1 (Fresh vs refrigerated) ΔED1 (F vs other dietary treatments) Sensory scores Effect of the diet on fresh fillets FO2 73.33b 3.81a 9.11 2.35 6.57 2 a SO 75.35 2.57b 8.99 0.64 2.38 5.97 SAO2 74.67ab 2.92b 8.87 1.39 1.63 5.67 2 ab b OPO 74.84 3.00 9.35 1.18 1.73 5.30 OPAO2 73.24b 3.01b 8.94 1.65 0.82 5.03 SEM3 0.412 0.142 0.102 0.350 pdiet4 0.004 < 0.001 0.032 0.023 Effect of the diet on refrigerated fillets FO2 75.47 2.85a 9.32 6.13 SO2 75.61 2.04b 9.25 0.83 6.57 SAO2 75.70 2.28ab 8.88 0.89 5.77 2 OPO 75.57 2.10b 9.51 0.78 5.40 OPAO2 74.73 2.33ab 9.09 0.93 5.70 SEM3 0.463 0.142 0.180 0.326 pdiet4 0.609 0.006 0.125 0.115 Effect of the refrigeration Fresh Fillets5 74.29 3.06 9.05 5.71 Refrigerated Fillets5 75.42 2.32 9.20 5.91 SEM3 0.196 0.064 0.065 0.151 prefrigeration6 < 0.001 < 0.001 0.145 0.335 Effect of the interaction between refrigeration and diet prefrigeration x diet6 0.266 0.508 0.806 0.481 Abbreviations: For diet abbreviations (FO, SO, SAO, OPO and OPAO) see Table 2; L∗ , Lightness; a∗ , redness; b∗ yellowness. 1ΔE values were calculated by Equation (1) (ΔE > 5 would mark differences in instrumental color perceptible by the human eye [26]). 2 Mean of the different replicates for each dietary treatment (n = 5 for color parameters and n = 30 for sensory scores). 3 Standard error of the mean. 4p values obtained from ANOVA (n = 25 for color parameters and n = 150 for sensory scores) of fresh or refrigerated fillets. Values in bold were significant (p < 0.05). The differences between dietary treatments found in fresh or refrigerated fillets with Scheffé’s post hoc test were noted in the same column as a > b. 5 Pooled means of fresh or refrigerated fillets coming from the five dietary treatments (n = 25 for color parameters and n = 150 for sensory scores). 6p values obtained for the refrigeration (prefrigeration) and the interaction between the refrigeration time and the diet (prefrigeration x diet) obtained from multifactor ANOVA (n = 50 for color parameters and n = 300 for sensory scores). Values in bold were significant (p < 0.05). differences between dietary treatments in fresh fillets, In the literature, it has been reported that the use of Scheffé’s post hoc test could not separate the means of the vegetable oils in fish feeds can alter the sensory characteris- different diets. After the refrigerated storage, the diet used tics of fresh fish flesh [6]. However, in agreement with our did not significantly affect the overall acceptance of fish observations, the results obtained by Montero et al. [69] fillets. revealed that the organoleptic properties of fresh European In our study, the significant increase of TBA values and seabass fillets were not modified due to the inclusion of veg- volatile compound content observed during the refrigerated etable oils (60% of SO, rapeseed, or linseed oils) in diets. storage did not result in a reduction of the overall acceptabil- Also in other studies, the acceptability of fresh seabream fil- ity. A great number of trials associate the loss of the sensory lets [30, 32] and of seabass fillets [32] was not affected by the acceptance with the secondary oxidation products measured dietary supplementation with vegetable oils (SO, rapeseed, by TBA values [66–68]. In seabass fillets, sensory rejection or linseed oils at 60% or 80%). was found by trained panelists at TBA values between 700–2,400 μg MDA/kg [67]. The TBA values found for fresh 4. Conclusions and refrigerated European seabass fillets in our study were not higher than these levels. This indicates that lipid oxida- The partial substitution of FO with SO, SAO, OPO, or tion during the processing and storage of fish fillets may OPAO in fish diets reduced EPA and DHA contents and not have developed enough to affect the sensorial properties, increased T + T3 in fish fillets. Even though, the recom- which agreed with the results obtained in the overall accep- mended EPA plus DHA amount for human daily intake tance test. (250 mg) could be covered with 100 g of fish fillets coming 16 Aquaculture Nutrition from all experimental diets. These changes in fillet composi- OPAO: Olive pomace acid oil tion led to a decrease of lipid oxidation reactions in fish PUFA: Polyunsaturated fatty acid fillets coming from SO, SAO, OPO, and OPAO diets, SAO: Soybean-sunflower (55 : 45, w/w) revealed by a lower final LHP value, a lower TBA value, acid oil and a lower formation of some volatile compounds (e.g., 1- SFA: Saturated fatty acid octen-3-ol) in fresh and refrigerated fillets. Moreover, the SO: Crude soybean oil partial substitution of FO for SO, SAO, OPO, or OPAO T: Tocopherol did not affect the overall acceptability and may not have T3: Tocotrienol influenced consumer’s perception of the color of the fish fil- TAG: Triacylglycerol lets. According to the results obtained in this study, in gen- TBA: Thiobarbituric acid eral, the partial replacement of FO by AO instead of crude U: Unsaponifiable matter oils of similar FA composition similarly affected the oxida- tive parameters, the color, or the overall acceptance of fish Data Availability fillets. Overall, the findings show that the final FA composi- tion, T content, and oxidative stability of fish fillets might be All data will be available upon request. more dependent on the botanical origin of acid oils (and thus, on their FA composition and T content), than on the Conflicts of Interest fact that they are acid or crude oils, at least when their qual- The authors report no conflicts of interest. ity (MIU value) is close to the recommended values [11, 27]. A clear development of lipid oxidation during the refrig- erated storage of European seabass fillets was observed with Authors’ Contributions the increment of TBA values and of the volatile compound P. Albendea was responsible of the conceptualization, vali- content of fish fillets, which occurred in all experimental dation, formal analysis, investigation, data curation, writing diets. However, the storage conditions (CO2/N2/O2; 40/30/ the original draft, and visualization. A. Tres managed the 30; 2°C; 6 days) were adequate to prevent a loss of overall conceptualization, methodology, validation, data curation, acceptance of fish fillets and might have prevented color writing the review, editing, visualization, and supervision. changes perceptible by the human eye. Overall, from the M. Rafecas was responsible of the conceptualization, meth- quality flesh point of view, it could be possible to upcycle odology, writing the review, editing, and funding acquisi- these refining by-products to partially substitute FO in fish tion. S. Vichi was in charge of the conceptualization, diets instead of using crude oils of similar botanical origin. methodology, validation, writing the review, and editing. R. Sala managed the conceptualization, resources, methodol- Abbreviations ogy, validation, writing the review, editing, and project administration. F. Guardiola was responsible of the concep- AO: Acid oil tualization, resources, methodology, validation, writing the CHP: Cumene hydroperoxide review, editing, visualization, supervision, project adminis- DAG: Diacylglycerol tration, and funding acquisition. DHA: Docosahexaenoic acid EPA: Eicosapentaenoic acid Acknowledgments FA: Fatty acid FFA: Free fatty acid The authors acknowledge Nutrion Internacional S.L.U., FO: Fish oil Refinación Industrial Oleícola S.A., and Bunge Ibérica FOX: Ferrous oxidation-xylenol orange S.A.U. for freely providing the vegetable fat sources used in GC-FID: Gas chromatography with flame ioni- this study. We also thank Francisco Javier Alarcón López zation detector from Ceimar-University of Almería (Service of Experimental HS-SPME-GC-MS: Headspace solid-phase microextraction Diets) for helping in the manufacturing process of the exper- coupled with gas chromatography and imental feeds used and the staff at IRTA aquaculture center mass spectrometry for their technical support. This research has been supported I: Insoluble impurities by MCIN/AEI/10.13039/501100011033 and by “ERDF A LHP: Lipid hydroperoxide way of making Europe” through the project (AGL2015- M: Moisture and volatile matter 64431-C2-2-R), by MCIN/AEI/10.13039/501100011033 MAG: Monoacylglycerol and by “ESF Investing in your future” through the grant MDA: Malondialdehyde (RYC-2017-23601), and by the Spanish Ministry of Univer- MIU: Moisture and volatile matter + insoluble sities through the predoctoral contract within the FPU pro- impurities + unsaponifiable matter gram (FPU18/01010). MUFA: Monounsaturated fatty acid n-3 LC-PUFA: n-3 long-chain polyunsaturated fatty Supplementary Materials acid ND: Not detected The supplementary material details the analytical methods OPO: Crude olive pomace oil applied to diets to determine fatty acid composition, Aquaculture Nutrition 17 tocopherol and tocotrienol content and lipid class composi- and fatty acid distillates used in animal feeding,” MethodsX, tion, and the complete FA profile of fish fillets, including vol. 8, article 101334, 2021. minor FA (Table S1). (Supplementary Materials) [16] EU, “Directive 2010/63/EU of the European Parliament and of the council of 22 September 2010 on the protection of animals References used for scientific purposes,” Official Journal of the European Union, vol. 276, pp. 33–79, 2010. [1] FAO, Fishery and Aquaculture Statistics 2018, FAO, Rome, [17] NRC, Nutrient Requirements of Fish and Shrimp, National Italy, 2020. Academies Press, Washington DC, USA, 2011. [2] World Bank, Fish to 2030: Prospects for fisheries and aquacul- [18] R. Bou, F. Guardiola, A. Tres, A. C. Barroeta, and R. Codony, ture. Report number 83177-GLB, Washington DC, USA, “Effect of dietary fish oil, α-tocopheryl acetate, and zinc sup- 2013, http://documents.worldbank.org/curated/en/2013/12/ plementation on the composition and consumer acceptability 18882045/fish-2030-prospects-fisheries-aquaculture. of chicken meat,” Poultry Science, vol. 83, no. 2, pp. 282–292, [3] FAO, Use of fishmeal and fish oil in aquafeeds - Further 2004. thoughts on the fishmeal trap, Rome, Italy, 1997, https://www [19] M. Aleman, C. D. Nuchi, R. Bou et al., “Effectiveness of antiox- .fao.org/3/y3781e/y3781e.pdf. idants in preventing oxidation of palm oil enriched with heme [4] J. R. Sargent and A. G. J. Tacon, “Development of farmed iron: a model for iron fortification in baked products,” Euro- fish: a nutritionally necessary alternative to meat,” The Pro- pean Journal of Lipid Science and Technology, vol. 112, no. 7, ceedings of the Nutrition Society, vol. 58, no. 2, pp. 377–383, pp. 761–769, 2010. 1999. [20] A. Grau, R. Codony, M. Rafecas, A. G. Barroeta, and [5] G. Mourente and J. G. Bell, “Partial replacement of dietary fish F. Guardiola, “Lipid hydroperoxide determination in dark oil with blends of vegetable oils (rapeseed, linseed and palm chicken meat through a ferrous oxidation-xylenol orange oils) in diets for European sea bass (Dicentrarchus labrax L.) method,” Journal of Agricultural and Food Chemistry, vol. 48, over a long term growth study: effects on muscle and liver fatty no. 9, pp. 4136–4143, 2000. acid composition and effectiveness of a fish oil finishing diet,” [21] A. Tres, C. D. Nuchi, R. Bou, R. Codony, and F. Guardiola, Comparative Biochemistry and Physiology Part B: Biochemistry “Assessing rabbit and chicken tissue susceptibility to oxidation and Molecular Biology, vol. 145, no. 3–4, pp. 389–399, 2006. through the ferrous oxidation-xylenol orange method,” Euro- [6] C. Nasopoulou and I. Zabetakis, “Benefits of fish oil replace- pean Journal of Lipid Science and Technology, vol. 111, no. 6, ment by plant originated oils in compounded fish feeds: a pp. 563–573, 2009. review,” LWT, vol. 47, no. 2, pp. 217–224, 2012. [22] A. Grau, F. Guardiola, J. Boatella, A. Barroeta, and R. Codony, [7] N. Smichi, B. E. Abdelmalek, N. Kharrat et al., “The effects of “Measurement of 2-thiobarbituric acid values in dark chicken storage on quality and nutritional aspects of farmed and wild meat through derivative spectrophotometry: influence of vari- sea bass (Dicentrachus labrax) muscle: in vitro oils digestibility ous parameters,” Journal of Agricultural and Food Chemistry, evaluation,” Fisheries Research, vol. 188, pp. 74–83, 2017. vol. 48, no. 4, pp. 1155–1159, 2000. [8] S. Maqsood and S. Benjakul, “Retardation of haemoglobin- [23] N. A. Botsoglou, D. J. Fletouris, G. E. Papageorgiou, V. N. Vas- mediated lipid oxidation of Asian sea bass muscle by tannic silopoulos, A. J. Mantis, and A. G. Trakatellis, “Rapid, sensi- acid during iced storage,” Food Chemistry, vol. 124, no. 3, tive, and specific thiobarbituric acid method for measuring pp. 1056–1062, 2011. lipid peroxidation in animal tissue, food, and feedstuff sam- [9] F. D. Gunstone, Vegetable Oils in Food Technology: Composi- ples,” Journal of Agricultural and Food Chemistry, vol. 42, tion, Properties and Uses, Blackwell Publishing Ltd., Oxford, no. 9, pp. 1931–1937, 1994. UK, 2nd edition, 2011. [24] A. R. Robertson, “The CIE 1976 color-difference formulae,” [10] G. M. Turchini, B. E. Torstensen, and W. K. Ng, “Fish oil Color Research and Application, vol. 2, no. 1, pp. 7–11, 1977. replacement in finfish nutrition,” Reviews in Aquaculture, [25] CIE, Colorimetry, CIE, International Commission on Illumi- vol. 1, no. 1, pp. 10–57, 2009. nation, Viena, Austria, 3rd edition, 2004. [11] E. Varona, A. Tres, M. Rafecas, S. Vichi, A. C. Barroeta, and [26] C. Costa, S. D’Andrea, R. Russo, F. Antonucci, F. Pallottino, F. Guardiola, “Composition and nutritional value of acid oils and P. Menesatti, “Application of non-invasive techniques to and fatty acid distillates used in animal feeding,” Animals, differentiate sea bass (Dicentrarchus labrax, L. 1758) quality vol. 11, no. 1, p. 196, 2021. cultured under different conditions,” Aquaculture Interna- [12] C. Trullàs, R. Fontanillas, A. Tres, and R. Sala, “Vegetable re- tional, vol. 19, no. 4, pp. 765–778, 2011. esterified oils in diets for rainbow trout: effects on fatty acid [27] FEDNA (Fundación Española para el Desarrollo de la Nutri- digestibility,” Aquaculture, vol. 444, pp. 28–35, 2015. ción Animal), “Normas FEDNA de control de calidad de [13] G. Verge-Mèrida, A. C. Barroeta, C. Ferrer et al., “Olive pom- ingredientes para piensos-Especificación técnica de materia ace and soybean-sunflower acid oils as alternative fat sources prima, grasas y aceites,” 2002, http://www.fundacionfedna in European seabass (Dicentrarchus labrax) diets: effects on .org/sites/default/files/8_Grasas_y_aceites.pdf. performance, digestibility and flesh fatty acid composition [28] K. Kwasek, A. L. Thorne-Lyman, and M. Phillips, “Can human and quality parameters,” Animals, vol. 12, no. 9, p. 1198, 2022. nutrition be improved through better fish feeding practices? A [14] E. Varona, A. Tres, M. Rafecas, S. Vichi, R. Sala, and review paper,” Critical Reviews in Food Science and Nutrition, F. Guardiola, “Oxidative quality of acid oils and fatty acid dis- vol. 60, no. 22, pp. 3822–3835, 2020. tillates used in animal feeding,” Animals, vol. 11, no. 9, p. 2559, [29] EFSA, Dietary reference values for nutrients. Summary report, 2021. EFSA Supporting Publication, 2017. [15] E. Varona, A. Tres, M. Rafecas, S. Vichi, A. C. Barroeta, and [30] M. S. Izquierdo, D. Montero, L. Robaina, M. J. Caballero, F. Guardiola, “Methods to determine the quality of acid oils G. Rosenlund, and R. Ginés, “Alterations in fillet fatty acid 18 Aquaculture Nutrition profile and flesh quality in gilthead seabream (Sparus aurata) levels of vitamin E,” Food Chemistry, vol. 68, no. 3, pp. 289– fed vegetable oils for a long term period. Recovery of fatty acid 293, 2000. profiles by fish oil feeding,” Aquaculture, vol. 250, no. 1–2, [43] G. Mourente, J. G. Bell, and D. R. Tocher, “Does dietary pp. 431–444, 2005. tocopherol level affect fatty acid metabolism in fish?,” Fish [31] Codex Alimentarius, “Standard for named vegetable oils (CXS Physiology and Biochemistry, vol. 33, no. 3, pp. 269–280, 2007. 210-1999),” in Codex Alimentarius, International food stan- [44] C. Regost, J. V. Jakobsen, and A. M. B. Rørå, “Flesh quality of dards, Rome, Italy, 1999Adopted in 1999; Revised in 2001, raw and smoked fillets of Atlantic salmon as influenced by die- 2003, 2009, 2017, 2019; Amended in 2005, 2011, 2013, 2015, tary oil sources and frozen storage,” Food Research Interna- 2019. tional, vol. 37, no. 3, pp. 259–271, 2004. [32] M. S. Izquierdo, A. Obach, L. Arantzamendi, D. Montero, [45] W. K. Ng, Y. Wang, and K. H. Yuen, “Palm vitamin E for aqua- L. Robaina, and G. Rosenlund, “Dietary lipid sources for seab- culture feeds,” Journal of Oil Palm Research, vol. 8, pp. 1–7, ream and seabass: growth performance, tissue composition 2008. and flesh quality,” Aquaculture Nutrition, vol. 9, no. 6, [46] C. Trullàs, A. Tres, J. Saldo, R. Fontanillas, and R. Sala, “Qual- pp. 397–407, 2003. ity characteristics of fillets of rainbow trout fed acid or re- [33] D. Menoyo, M. S. Izquierdo, L. Robaina, R. Ginés, C. J. Lopez- esterified rapeseed oils as dietary fat sources,” Aquaculture, Bote, and J. M. Bautista, “Adaptation of lipid metabolism, tis- vol. 480, pp. 22–31, 2017. sue composition and flesh quality in gilthead sea bream [47] M. B. Betancor, E. Atalah, M. J. Caballero et al., “α-Tocopherol (Sparus aurata) to the replacement of dietary fish oil by linseed in weaning diets for European sea bass (Dicentrarchus labrax) and soyabean oils,” The British Journal of Nutrition, vol. 92, improves survival and reduces tissue damage caused by excess no. 1, pp. 41–52, 2004. dietary DHA contents,” Aquaculture Nutrition, vol. 17, no. 2, [34] G. Mourente, J. E. Good, and J. G. Bell, “Partial substitution of pp. e112–e122, 2011. fish oil with rapeseed, linseed and olive oils in diets for Euro- [48] V. R. Kyrana and V. P. Lougovois, “Sensory, chemical and pean sea bass (Dicentrarchus labrax L.): effects on flesh fatty microbiological assessment of farm-raised European sea bass acid composition, plasma prostaglandins E2 and F2α, immune (Dicentrarchus labrax) stored in melting ice,” International function and effectiveness of a fish oil finishing diet,” Aquacul- Journal of Food Science and Technology, vol. 37, no. 3, ture Nutrition, vol. 11, no. 1, pp. 25–40, 2005. pp. 319–328, 2002. [35] C. Nasopoulou, G. Stamatakis, C. A. Demopoulos, and [49] R. Ma, X. Liu, H. Tian et al., “Odor-active volatile compounds I. Zabetakis, “Effects of olive pomace and olive pomace oil on profile of triploid rainbow trout with different marketable growth performance, fatty acid composition and cardio pro- sizes,” Aquaculture Reports, vol. 17, pp. 100–312, 2020. tective properties of gilthead sea bream (Sparus aurata) and sea bass (Dicentrarchus labrax),” Food Chemistry, vol. 129, [50] D. Jimenez-Alvarez, F. Giuffrida, P. A. Golay et al., “Profiles of no. 3, pp. 1108–1113, 2011. volatile compounds in milk containing fish oil analyzed by HS- [36] J. G. Bell, J. McEvoy, D. R. Tocher, F. McGhee, P. J. Campbell, SPME-GC/MS,” European Journal of Lipid Science and Tech- and J. R. Sargent, “Replacement of fish oil with rapeseed oil in nology, vol. 110, no. 3, pp. 277–283, 2008. diets of Atlantic salmon (Salmo salar) affects tissue lipid com- [51] F. Shahidi, “Headspace volatile aldehydes as indicators of lipid positions and hepatocyte fatty acid metabolism,” The Journal oxidation in foods,” in Headspace Analysis of Foods and Fla- of Nutrition, vol. 131, no. 5, pp. 1535–1543, 2001. vors, R. L. Rouseff and K. R. Cadwallader, Eds., vol. 488 of [37] B. E. Torstensen, L. Frøyland, and Ø. Lie, “Replacing dietary Advances in Experimental Medicine and Biology, pp. 113– fish oil with increasing levels of rapeseed oil and olive oil – 123, Springer, Boston, MA, 2001. effects on Atlantic salmon (Salmo salar L.) tissue and lipopro- [52] M. A. Augustin, L. Sanguansri, and O. Bode, “Maillard reac- tein lipid composition and lipogenic enzyme activities,” Aqua- tion products as encapsulants for fish oil powders,” Journal culture Nutrition, vol. 10, no. 3, pp. 175–192, 2004. of Food Science, vol. 71, no. 2, pp. E25–E32, 2006. [38] I. U. Udo and O. E. Afia, “Optimization of dietary vitamin [53] C. F. Ross and D. M. Smith, “Use of volatiles as indicators of E (tocopherols) in fish: a review,” Nigerian Journal of Agri- lipid oxidation in muscle foods,” Comprehensive Reviews in culture, Food and Environment, vol. 9, no. 4, pp. 99–107, Food Science and Food Safety, vol. 5, no. 1, pp. 18–25, 2006. 2013. [54] R. J. Hsieh and J. E. Kinsella, “Lipoxygenase generation of spe- [39] C. Afonso, N. M. Bandarra, L. Nunes, and C. Cardoso, cific volatile flavor carbonyl compounds in fish tissues,” Jour- “Tocopherols in seafood and aquaculture products,” Critical nal of Agricultural and Food Chemistry, vol. 37, no. 2, Reviews in Food Science and Nutrition, vol. 56, no. 1, pp. 279–286, 1989. pp. 128–140, 2016. [55] J. D. Wood and M. Enser, “Manipulating the fatty acid compo- [40] M. Pazos, L. Sánchez, and I. Medina, “α-Tocopherol oxidation sition of meat to improve nutritional value and meat quality,” in fish muscle during chilling and frozen storage,” Journal of in New Aspects of Meat Quality. From Genes to Ethics, P. P. Agricultural and Food Chemistry, vol. 53, no. 10, pp. 4000– Purslow, Ed., Elsevier Science, Sawston, UK, 2017. 4005, 2005. [56] C. Prost, A. Hallier, M. Cardinal, T. Serot, and P. Courcoux, [41] M. Yildiz, E. Şener, and H. Gün, “Effect of refrigerated storage “Effect of storage time on raw sardine (Sardina pilchardus) fla- on fillet lipid quality of rainbow trout (Oncorhynchus mykiss vor and aroma quality,” Journal of Food Science, vol. 69, no. 5, W.) fed a diet containing different levels of DL α-tocopherol pp. S198–S204, 2004. acetate,” Turkish Journal of Veterinary and Animal Sciences, [57] R. L. Wierda, G. Fletcher, L. Xu, and J. P. Dufour, “Analysis of vol. 30, no. 1, pp. 143–150, 2006. volatile compounds as spoilage indicators in fresh king salmon [42] M. Pirini, P. P. Gatta, S. Testi, G. Trigari, and P. G. Monetti, (Oncorhynchus tshawytscha) during storage using SPME-GC- “Effect of refrigerated storage on muscle lipid quality of sea MS,” Journal of Agricultural and Food Chemistry, vol. 54, bass (Dicentrarchus labrax) fed on diets containing different no. 22, pp. 8480–8490, 2006. Aquaculture Nutrition 19 [58] F. S. H. Lu, I. Bruheim, B. O. Haugsgjerd, and C. Jacobsen, “Effect of temperature towards lipid oxidation and non- enzymatic browning reactions in krill oil upon storage,” Food Chemistry, vol. 157, pp. 398–407, 2014. [59] H. Lee, S. A. Kizito, S. J. Weese et al., “Analysis of headspace volatile and oxidized volatile compounds in DHA-enriched fish oil on accelerated oxidative storage,” Journal of Food Sci- ence, vol. 68, no. 7, pp. 2169–2177, 2003. [60] J. S. Del Pulgar, M. Roldan, and J. Ruiz-Carrascal, “Volatile compounds profile of sous-vide cooked pork cheeks as affected by cooking conditions (vacuum packaging, temperature and time),” Molecules, vol. 18, no. 10, pp. 12538–12547, 2013. [61] S. Ji, S. Gu, X. Wang, and N. Wu, “Comparison of olfactome- trically detected compounds and aroma properties of four dif- ferent edible parts of Chinese mitten crab,” Fisheries Science, vol. 81, no. 6, pp. 1157–1167, 2015. [62] S. Takeungwongtrakul, S. Karnjanapratum, P. Kaewthong, and S. Nalinanon, “Change in fatty acid profile, volatile com- pounds and FTIR spectra of samrong seed oil during storage,” International Journal of Agricultural Technology, vol. 16, no. 2, pp. 475–484, 2020. [63] F. F. Parlapani, A. Mallouchos, S. A. Haroutounian, and I. S. Boziaris, “Volatile organic compounds of microbial and non-microbial origin produced on model fish substrate un-inoculated and inoculated with gilt-head sea bream spoilage bacteria,” LWT, vol. 78, pp. 54–62, 2017. [64] H. Rehbein and J. Oehlenschläger, Fishery Products: Quality, Safety and Authenticity, Wiley-Blackwell, John Wiley & Sons, Chichester, UK, 2009. [65] J. H. Sohn, Y. Taki, H. Ushio, T. Kohata, I. Shioya, and T. Ohshima, “Lipid oxidations in ordinary and dark muscles of fish: influences on rancid off-odor development and color darkening of yellowtail flesh during ice storage,” Journal of Food Science, vol. 70, no. 7, pp. s490–s496, 2005. [66] S. Cakli, B. Kilinc, A. Cadun, T. Dincer, and S. Tolasa, “Quality differences of whole ungutted sea bream (Sparus aurata) and sea bass (Dicentrarchus labrax) while stored in ice,” Food Con- trol, vol. 18, no. 5, pp. 391–397, 2007. [67] M. Kostaki, V. Giatrakou, I. N. Savvaidis, and M. G. Kontominas, “Combined effect of MAP and thyme essential oil on the microbiological, chemical and sensory attributes of organically aquacultured sea bass (Dicentrarchus labrax) fil- lets,” Food Microbiology, vol. 26, no. 5, pp. 475–482, 2009. [68] P. Masniyom, S. Benjakul, and W. Visessanguan, “Shelf-life extension of refrigerated seabass slices under modified atmo- sphere packaging,” Journal of the Science of Food and Agricul- ture, vol. 82, no. 8, pp. 873–880, 2002. [69] D. Montero, L. Robaina, M. J. Caballero, R. Ginés, and M. S. Izquierdo, “Growth, feed utilization and flesh quality of Euro- pean sea bass (Dicentrarchus labrax) fed diets containing veg- etable oils: a time-course study on the effect of a re-feeding period with a 100% fish oil diet,” Aquaculture, vol. 248, no. 1-4, pp. 121–134, 2005.