LWT - Food Science and Technology 184 (2023) 115044 Available online 28 June 20230023-6438/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Reutilization of berry press residues in minced Baltic herring (Clupea harengus membras) – Effect on lipid oxidation and sensory characteristics during cold storage Tanja Kakko *, Annelie Damerau , Claudia Mejia Rios , Oskar Laaksonen , Baoru Yang Food Sciences, Department of Life Technologies, University of Turku, 20014, Turun yliopisto, Turku, Finland A R T I C L E I N F O Keywords: Fish Volatiles Antioxidants Generic descriptive analysis Gas chromatography-olfactometry A B S T R A C T Lipid oxidation is one of the main causes of quality deterioration in dark-muscled fish. This study investigated the effect of berry side streams of juice pressing or oil extraction at different concentrations on lipid oxidation and sensory quality of minced Baltic herring (Clupea harengus membras) stored at 1 C. Lipid oxidation was assessed by determining the peroxide value and volatile secondary oxidation products. Generic descriptive analysis and gas chromatography-olfactometry were used to investigate the odor, flavor, and odor-active compounds of minces. Based on the peroxide value and formation of hexanal, 1-penten-3-ol, and 3,5-octadien-2-one, a con- centration of 1.5 g/100 g of lingonberry-bilberry juice press residue and sea buckthorn juice press and oil extraction residue was sufficient to retard and reduce the extent of oxidation. Furthermore, the addition of these residues inhibited the formation of fishy odors during 3 days of storage. The control mince stored for 3 days showed the highest number and intensities of oxidation products such as (Z)-4-heptenal, 2,3-pentanedione, and (E,Z)-2,6-nonadienal, while their formation was retarded in the mince containing press residue from sea buck- thorn. Berry press residues were efficient in controlling lipid oxidation of minced Baltic herring during storage and have the potential to mask potentially unpleasant fishy odor and flavor. 1. Introduction Lipid oxidation is a main cause of quality deterioration in fish due to the high content of pro-oxidants, such as heme proteins and poly- unsaturated fatty acids (PUFAs), especially in dark-muscled fish. Although high PUFA content is desirable in terms of nutritional quality, the oxidation of PUFAs may give rise to potentially harmful compounds (Rundblad et al., 2017; Vieira et al., 2017) and produce aroma com- pounds that negatively impact the sensory quality (Fu et al., 2009; Hammer & Schieberle, 2013; Jiarpinijnun et al., 2022). Most processing, such as mincing, further accelerates oxidation in fish. Mincing increases the oxygen to surface ratio and exposes the lipids to pro-oxidants and enzymes, such as lipoxygenases. While low storage temperature delays the development of lipid oxidation, even freezing at temperatures close to 20 C is not sufficient to fully control it (Aydin & Gokoglu, 2014; Baron et al., 2007; Damerau et al., 2020). Baltic herring (Clupea harengus membras), a subspecies of Atlantic herring (Clupea harengus), is a small dark-muscled fish. PUFAs account for approximately one-third of the total fatty acids in Baltic herring (Aitta et al., 2021; Aro et al., 2000; Damerau et al., 2020). Baltic herring is commercially the most important catch species in Finland (based on quantity and value), but most of the catch is directed to low-value uses, such as feed (Sarkki & Pihlajamaki, 2019). One of the factors limiting the use of Baltic herring is its high susceptibility to oxidation and development of its distinctive odor and flavor that many consumers find unappealing (Pihlajamaki et al., 2019). Previously, several secondary lipid oxidation compounds were observed contributing to the odor of minced Baltic herring (Kakko, Aitta, et al., 2022). Retarding lipid oxidation of Baltic herring could therefore control the formation of potentially unpleasant odor and flavor. While there are several synthetic antioxidants available for control- ling lipid oxidation, rising consumer demand for natural products is driving the search for alternatives. For instance, synthetic phenolic an- tioxidants, such as butyl hydroxyanisole and dibutyl hydroxytoluene, have raised concerns due to their toxicity and spread in the environment (Wang et al., 2021). The antioxidative potential of extracts, fractions, and residues of different botanical origins has been widely studied in fish and meat (Bellucci et al., 2022; Cropotova et al., 2019; Mattje et al., * Corresponding author. E-mail address: tatese@utu.fi (T. Kakko). Contents lists available at ScienceDirect LWT journal homepage: www.elsevier.com/locate/lwt https://doi.org/10.1016/j.lwt.2023.115044 Received 15 March 2023; Received in revised form 22 June 2023; Accepted 28 June 2023 LWT 184 (2023) 115044 2 2019; Silveira Alexandre et al., 2022). The berry-processing industry produces high amounts of berry biomass with bioactive properties, such as antioxidative (Puganen et al., 2018), antimicrobial (Puupponen-Pimia et al., 2021), and antiviral (Granato et al., 2022) effects in vitro, but currently, the biomasses remain underutilized. Berries and berry byproducts have also been investigated as potential natural antioxidants in meat and fish (Dai et al., 2021; Damerau et al., 2020; Püssa et al., 2008; Sampels et al., 2010). For example, sea buck- thorn press residue was observed to retard lipid oxidation in mechani- cally separated chicken and turkey meat (Püssa et al., 2008). In a previous study conducted by our research group, both lingonberry-bilberry and sea buckthorn press residues were found to be as or even more effective than conventional antioxidants in retarding lipid oxidation of Baltic herring mince stored at 20 C for 10 months (Damerau et al., 2020). However, employing antioxidants of natural origin may present challenges concerning sensory quality and consumer acceptability. In the study by Damerau et al. (2020), berry additions had a negative impact on consumer perception, which emphasizes the need to optimize the dosage to minimize the effect on sensory properties. In another study, supercritical CO2 ginger extract and ginger essential oil reduced the pleasantness of fish burgers due to a strong ginger flavor (Mattje et al., 2019). While assessing the potential of antioxidants, investigation of sensory quality is important since the antioxidative effects based on instrumentally measured chemical oxidation indicators do not always correlate with sensory quality. For example, a combination of δ-tocopherol, ascorbyl palmitate, and lecithin was seen to inhibit hy- droperoxide formation in refined fish oil, but despite extremely low peroxide values, development of off-flavors was not inhibited (Hamilton et al., 1998). Conversely, Sveinsdottir et al. (2020) reported that dipping Atlantic mackerel (Scomber scombrus) fillets into a sodium erythorbate solution significantly prolonged the shelf life of frozen fillets based on sensory evaluation, although no clear inhibition of hydroperoxide or thiobarbituric acid reactive substance (TBARS) formation was observed. The aim of this study was to investigate whether the addition of lingonberry-bilberry juice press residue, sea buckthorn juice press res- idue, and sea buckthorn juice and oil extraction residue added at different concentrations (1, 1.5, and 3 g/100 g) retarded the formation of primary and secondary lipid oxidation products in minced Baltic herring stored at 1 C for 0–21 days. A further aim was to study the effect of the additions on the sensory quality of the mince and to examine connections between lipid oxidation, flavor, odor, and odor-active compounds of the minces during the first three days of storage. 2. Materials and methods 2.1. Materials For the 21-day storage test, fresh Baltic herring fillets (with skin) (Martin Kala Oy, Turku, Finland) were delivered to the laboratory, packed in ice, immediately portioned in sealable bags, and frozen at 80 C until the beginning of the storage test. For the sensory evaluation and investigation of odor-active compounds, Baltic herring fillets were kindly provided by Kalaset Oy (Uusikaupunki, Finland). The fillets used for sensory evaluation and gas chromatography–olfactometry (GC-O) (3-day storage test) were brought to the laboratory packed in ice within 24 h of being caught. Upon arrival, they were immediately processed and subjected to sensory analysis. The fillets provided by the two com- panies (Martin Kala Oy and Kalaset Oy) were caught in the same area and processed similarly and hence were comparable. Lingonberry-bilberry press residue after juice pressing (“LR”) was provided by Kiantama Oy (Suomussalmi, Finland). The press residue is mostly composed of lingonberry, but due to minor “contamination” with bilberry during commercial harvesting of lingonberry, as indicated by the anthocyanin profile of the press residue in our previous study (Damerau et al., 2020), the press residue also contains small amounts of bilberry and is referred to as lingonberry-bilberry juice press residue. Two sea buckthorn press residues, obtained after juice pressing (“SR”) and after juice pressing and supercritical CO2 oil extraction (“SRO”), were provided by Aromtech Oy Ltd. (Tornio, Finland). SR and SRO were from the same berry batch. All berry press residues were brought to the laboratory as dried and stored at 20 C prior to use. L-Ascorbic acid and α-tocopherol were purchased from Sigma (Sigma‒Aldrich Co, St. Louis, MO, USA). 2.2. Storage test 2.2.1. Sample preparation and storage A storage test of processed Baltic herring mince was conducted at 1 C for up to 21 days. The fish (Martin Kala Oy) were defrosted over- night at 4 C and processed into mince. Fish were minced using a Ken- wood food processor (Chef Titanium, Kenwood Limited, Havant, United Kingdom) with a meat grinder attachment (with holes 4 mm in diam- eter), and the antioxidant additions were mixed into the mince imme- diately. Mixing was conducted manually, the distribution of press residues was confirmed visually, and the control mince was mixed for an equivalent duration. Berry press residues were added at concentrations of 1 g/100 g, 1.5 g/100 g, and 3 g/100 g. These concentrations were chosen based on our earlier observations on 3 g/100 g LR and SR being effective (Damerau et al., 2020), but to reduce costs and effect on sen- sory quality, the aim was to determine whether a lower concentration (1 or 1.5 g/100 g) would suffice. In addition to LR, SR, and SRO, a com- bination of α-tocopherol and ascorbic acid (αT ‡ AA, 0.01 g/100 g and 0.2 g/100 g, respectively) was studied for comparison. This combination was chosen due to the synergistic effect of α-tocopherol and ascorbic acid, particularly the ability of ascorbic acid to regenerate α-tocopherol (Packer et al., 1979). Their concentrations were the same as in our previous study (Damerau et al., 2020) and within the same range as previous reports by others (EFSA ANS Panel, 2015; Hamilton et al., 1998; Wang et al., 2017; Yanishlieva & Marinova, 2001). Mince without any additions was used as a control (C). The fish minces subjected to lipid extraction and PV measurement were stored in plastic containers (30 g in 350 mL container). Minces used for volatile analyses were stored in SPME vials (3 g in a 20 mL vial) to capture the volatiles formed during storage. Both containers and vials were stored at a temperature of 1 C. At each designated time point during storage (0, 3, 7, 10, and 14 d for volatiles, 0, 3, 7, 10, 14 d, and 21 d for PV measurement), the vials and containers were transferred to 80 C until analysis. 2.2.2. Lipid extraction, PV measurement, and volatile analysis Lipids were extracted using 2:1 (mL/mL) chloroform-methanol with 0.5 g/L BHT (2,6-di-tert-butyl-4-methylphenol) according to the method by Cavonius and Undeland (2017) with minor modifications. A 2-g aliquot of mince was first extracted by homogenizing the sample for 2 min in 15 mL chloroform-methanol, followed by another extraction with 6 mL chloroform-methanol. A 3 mL volume of 20 g/L aqueous NaCl was added to the pooled extracts to promote phase separation, followed by vortexing for 30 s and centrifugation at 1000g for 2 min. The organic phase was collected and used for PV analyses. Samples were kept on ice during the extraction, and extraction was performed in duplicate for each sample. Peroxide values were measured in duplicate from each lipid extract according to (Lehtonen et al., 2011). Volatile compounds were extracted by headspace solid phase microextraction (HS-SPME) and analyzed by gas chromatography‒mass spectrometry (GC-MS) as previously reported by our group (Kakko, Damerau, et al., 2022). Briefly, HS-SPME extraction of volatiles was achieved by 20 min incubation and 30 min extraction at 40 C using a 1 cm divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS) fiber (50/30 μm film thickness; Supelco, St. Louis, MO, USA). Volatiles were desorbed in a 240 C injector for 6 min and separated on a Supelco SPB-624 column (60 m  0.25 mm i.d., 1.4 μm film thickness; Supelco, T. Kakko et al. LWT 184 (2023) 115044 3 U.S.A.). MS was operated in electron ionization mode. 1-Penten-3-ol, hexanal, and 3,5-octadien-2-one (E,Z or E,E) were identified using commercial reference compounds and/or the NIST library (version 2.3, National Institute of Standards and Technology, Gaithersburg, Mary- land, USA) and used as indicators of secondary lipid oxidation. 2.3. Sensory analysis of minces with and without LR and SRO 2.3.1. Panelists and training Nine panelists, consisting of staff and students of the Food Sciences unit at the University of Turku, Finland, were recruited for the generic descriptive analysis (GDA). All panelists had prior experience in sensory analysis. National ethical guidelines (the Finnish National Board on Research Integrity) involving humans were adhered to, but no formal documentation process is available. The panelists were informed about the study protocols and their right to withdraw from the study at any time, and their informed consent to participate and allow storage of personal data during the study was obtained. Microbial analyses of minces were conducted at a certified laboratory (Eurofins Scientific Finland Oy, Raisio, Finland) to ensure that there were no risks for con- sumption. The panelists were trained to evaluate the odor and flavor characteristics of Baltic herring minces during four sessions. The first three sessions included building the vocabulary, selecting reference samples (Supplemental Table S1), determining their intensities, and placing samples on the scale (0–10, 0 ˆ not at all, 10 ˆ very intense). In the final training session, the panelists evaluated six samples, after which they received feedback on their performance. 2.3.2. Sample preparation and sensory evaluation Sensory analysis was conducted on minces stored for 0 or 3 days at 1 C, and minces were not frozen at any point. To limit the number of samples for sensory analyses, two antioxidant additions, LR (1.5 g/100 g) and SRO (1.5 g/100 g), were selected for odor and flavor profiling, while SR was excluded due to having less pronounced antioxidative effects in the 21-day storage test. Samples were evaluated both raw and after sous vide cooking, but raw samples were assessed only for their odor. To prepare the sous vide cooked samples, raw minces (stored for 0 or 3 days, with or without LR and SRO) were mixed with NaCl (0.65 g/100 g), vacuum-packed in vacuum seal bags in portions of 25 g and cooked in a 70 C water bath for 20 min. Cooked samples weighing approximately 12 g were served in 50 mL glass bowls with lids and were tempered on a 65 C hot plate for 15–20 min prior to serving to the panelists. Raw samples were served in brown glass bottles (6 g in 30 mL bottle). Odor and flavor/taste attributes were evaluated on a scale of 0–10 (0 ˆ not at all, 10 ˆ very intense), and data were collected using Compu- sense Cloud 19 (Compusense Inc., Guelph, Ontario, Canada). Evalua- tions were held in 5 sessions (cooked samples in duplicate, raw samples in triplicate in separate sessions), and 6 samples coded with three-digit codes were served monadically in a randomized order on each session. Cooked samples were evaluated separately from raw samples. Due to the availability of the fish, in most sessions, the panelists evaluated only fresh or only stored samples. Assessors were provided with a low-salt cracker and active charcoal filtered water to cleanse their palate be- tween samples. 2.4. Gas chromatography-olfactometry (GC-O) of raw minces with and without SRO 2.4.1. Panelists and training The odors of raw C and raw SRO (1.5 g/100 g) minces after 0 d and 3 d of cold storage were further studied by GC-O. Six panelists (4 women, 2 men, age 25–33), 4 of whom had also taken part in the GDA, were recruited and trained in three sessions to describe and evaluate the in- tensity of odor compounds. The first session was group training that included dilutions of standard compounds pipetted on 1 cm  1 cm paper in a brown glass bottle (30 mL). The intensity of odors was evaluated on a scale of 0–4 (1 ˆ very mild, barely noticeable, 2 ˆ mild, 3 ˆ fairly strong, and 4 ˆ strong; 0 ˆ not detected). The second and third training sessions included evaluation of a standard compound mix and one of the four Baltic herring mince samples, respectively, on the GC-O. 2.4.2. HS-SPME-GC-O Volatile compounds for GC-O were extracted using HS-SPME. For the extraction, 20  0.1 g of mince was weighed in a 90 mL Erlenmeyer flask and incubated at 35 C for 20 min prior to 35 min of extraction at the same temperature. A 2 cm DVB/CAR/PDMS fiber was used to extract the volatiles, and desorption of volatiles occurred at 240 C for 5 min. GC-O evaluations were conducted using an HP 6890 Series GC with a flame ionization detector (FID) (Hewlett Packard, Palo Alto, CA, U.S.A.). All four samples were assessed twice by the 6 panelists: once on a DB-WAX column (60 m, 0.25 mm, 0.25 μm; Agilent Technologies, Santa Clara, CA, U.S.A.) and once on an SPB-624 column (30 m, 0.25 mm, 1.40 μm, Supelco Inc., Bellefonte, PA, U.S.A.). The oven temperatures for both columns were the same, 40 C held for 3 min, followed by a ramp of 10 C/min until 220 C was reached, and then held for 10 min. The sniffing time was approximately 20 min. Panelists evaluated the samples in a randomized order. Recorded audio files were processed using Au- dacity 3.0.2 (The Audacity Team). The average odor intensity for each compound was calculated by including only the intensities given by the assessors who detected the odor (i.e., null values were not included). Compounds detected by at least 2 out of 6 panelists (Nasal Impact Fre- quency, NIF 33%) were considered odor-contributing compounds. Odor-active compounds were identified using external standards or analysis using HS-SPME-GC-MS. External volatile standards included ethyl isovalerate, 3-methylbutanal, 2,3-butanedione, (Z)-2-penten-1-ol, 2-methylpropanal, 1-octen-3-one, 2,3-pentanedione, (E,E)-2,4-hex- adienal, 2,4-heptadienal, (Z)-4-heptenal, heptanal, and hexanal (all from Sigma‒Aldrich, St. Louis, MO, USA), 1-penten-3-ol (Fluka Chem- icals, Neu Ulm, Switzerland), 2-methylbutanal, propanal (Acros Or- ganics, Geel, Belgium), 1,5-octadien-3-one, and 1,5-octadien-3-ol (Eptes Sarl, Vevey, Switzerland). TRACE 1310 gas chromatograph coupled with an ISQ7000 single quadrupole mass spectrometer and TriPlus RSH autosampler (Thermo Fisher Scientific, Waltham, Massachusetts, U.S.A.) were used for the identification of volatiles by HS-SPME-GC-MS. The MS was operated in electron ionization mode. The transfer line and ion source temperatures were 220 C and 250 C, respectively, and mass spectra were collected in the range of m/z 40–300 in scan mode. GC-MS data were processed using Chromeleon 7.0 (Thermo Scientific™), and identification was performed by comparison to the NIST library (version 2.3, National Institute of Standards and Technology, Gaithersburg, Maryland, U.S.A.) 2.5. Statistical analyses The PVs, volatile peak areas, and intensities of sensory attributes of minces with different additions, within the same time point, were compared using one-way analysis of variance (ANOVA) and Tukey’s test in SPSS Statistics 27 (IBM, Armonk, New York, U.S.A.). The intensities of odor and flavor attributes after 0 or 3 days of storage within the same type of mince were compared using paired samples t tests. A 95% con- fidence level was used. Partial least squares (PLS) regression using Un- scrambler (version 10.4.1, Camo Process AS, Oslo, Norway) was used to model how odor-active compounds explain differences in observed odor attribute intensities in GDA (X variables were weighed intensities, i.e. intensity*NIF, of odor-active compounds; Y variables were observed odor intensities in GDA). PanelCheck software (version 1.4.2, Nofima, Tromsø, Norway) was used to evaluate the agreement, sensitivity, and reproducibility of the panel in GDA (Tomic et al., 2009). T. Kakko et al. LWT 184 (2023) 115044 4 3. Results and discussion 3.1. Lipid oxidation during storage of minced Baltic herring 3.1.1. Changes in peroxide value during 0–21 days of storage PVs (Fig. 1) after 0, 3, 7, 10, 14 and 21 d of storage were analyzed as markers of primary lipid oxidation. The p values of differences between samples within each time point are presented in Supplemental Table S2. The PV of the control mince reached its peak after 10 d, after which it declined over the remaining storage period. In the berry press residue- containing minces, however, the highest PV was observed after 14 d or later. In addition to delaying hydroperoxide formation, most berry press residues also decreased the total extent of the formation. SR1.5 and SR3, as well as αT ‡ AA, however, had a PV similar to that of C at the beginning of storage, but at 14 days, the PV was significantly (p < 0.05) higher. The higher PVs observed in SR and αT ‡ AA could be caused by their possible pro-oxidativity effects. Another explanation might be the stabilization of hydroperoxides, which would lead to their accumula- tion. In a previous study, chub mackerel (Scomber japonicus) mince with pomegranate seed extract showed higher levels of hydroperoxides dur- ing frozen storage compared to the control, but the TBARS levels were considerably lower (Ozalp Ozen et al., 2011), indicating that the higher PV observed in the mince with pomegranate seed extract was not due to pro-oxidativity of the extract. Compared to SR, SRO minces had lower PVs, especially when comparing SR3 and SRO3 (8.4 vs. 4.3, respectively, on Day 14). SR has been reported to contain a significant amount of lipids but also lipid soluble antioxidants, such as tocopherols (Damerau et al., 2020). The SRO was from the same berry batch as SR but had been further subjected to supercritical CO2 extraction of oil and therefore contained less lipids and lipid soluble antioxidants than SR. In LR-containing minces, the PV increase was slow during the storage period and did not reach its peak during the storage period of 21 days. LR, SR, and SRO affected the PV differently depending on the con- centration. In the case of SR, the difference between SR3 and SR1.5 was significant only on Days 3, 10, and 21 (Supplemental Table S2), and in most cases, the former had a similar or higher PV compared to C. In the case of SRO, SRO3 had a significantly lower PV compared to SRO1 and SRO1.5 at most time points. However, SRO1.5 had significantly lower PV than the control at Days 0, 3, and 10. The PVs of LR3 and LR1.5 did not differ statistically significantly at any point. However, due to the rapid degradation of hydroperoxides to form secondary oxidation products, PV by itself is not a sufficient measure of lipid oxidation and should be considered together with secondary oxidation indicators (Ross & Smith, 2006). 3.1.2. Formation of volatile secondary oxidation products during 0–14 days of storage 1-Penten-3-ol, hexanal, and 3,5-octadien-2-one were chosen as sec- ondary oxidation indicators (Fig. 2) since they have been previously found to be suitable indicators for lipid oxidation produced from different precursors and were among the most abundant volatiles in the samples. 1-Penten-3-ol is reported to be formed from n-3 fatty acids through scission of the C17 alkoxy radical, which is formed from 17- OOH through O–O cleavage (Hammer & Schieberle, 2013; Lee et al., 2003). In previous studies, 1-penten-3-ol was shown to be one of the most abundant volatiles formed during lipid oxidation of herring (Sampels et al., 2010) and Baltic herring (Damerau et al., 2020). Hexanal, which can be formed by scission of the fatty acid chain on either side of the radical and is mainly produced from oxidation re- actions of n-6 fatty acids (Gomez-Cortes et al., 2015), has been widely used as a lipid oxidation indicator and was identified as an odor-active compound in Baltic herring (Kakko, Aitta, et al., 2022). 3,5-Octadie- n-2-one can be produced during docosahexaenoic acid (DHA) autoxi- dation via 14-OOH-13,16-epidioxide (Noble & Nawar, 1975). At most time points, volatile formation was highest in the control sample, whereas berry press residue and αT ‡ AA addition had an inhibitory effect. At all concentrations, LR significantly inhibited the formation of 1-penten-3-ol and 3,5-octadien-2-one after 3–14 days of storage when compared to the control (p < 0.05) (Fig. 2 and Supple- mental Tables S3–S5). On the other hand, SR1 showed the least inhi- bition of the formation of these compounds among the berry press residue additions. Similar inhibition of volatile secondary oxidation product formation in fish by other natural antioxidants has also been reported in previous studies (Farvin et al., 2012; Joaquin et al., 2008; Sampels et al., 2010). For example, milk protein concentrate added to herring (Joaquin et al., 2008) and ethanol extracts of potato peel added to horse mackerel (Farvin et al., 2012) decreased the formation of 1-pen- ten-3-ol during storage. In general, based on the PVs and the content of the volatile secondary oxidation products, LR and SRO additions showed the most pronounced antioxidative effects in minced Baltic herring (Figs. 1 and 2). Fig. 1. Peroxide values (PV, meq/kg fish, n ˆ 4) of Baltic herring minces containing no additions (con- trol, C) (A–D), α-tocopherol and ascorbic acid (αT ‡ AA) (A), 1.0, 1.5, or 3.0 g/100 g lingonberry-bilberry juice press residue (LR) (B), 1.0, 1.5, or 3.0 g/100 g sea buckthorn juice press residue (SR) (C), or 1.0, 1.5, or 3.0 g/100 g sea buckthorn juice press and oil extraction residue (SRO) (D), after 0, 3, 7, 10, 14, and 21 days at 1 C. An asterisk (*) indicates a statistically significant (p < 0.05) difference between the samples at the given time point. T. Kakko et al. LWT 184 (2023) 115044 5 Interestingly, SRO was more effective than SR, despite the removal of lipids and lipid soluble antioxidants by supercritical CO2 extraction. This may be attributed to the removal of lipids leading to the concentration of water-soluble antioxidants, such as flavonol glycosides, which are not removed during supercritical CO2 extraction (Linderborg et al., 2012). Püssa et al. (2008) suggested that flavonol glycosides were significant contributors to the antioxidative effect of SR in mechanically deboned chicken and turkey. In addition, the high content of oil in the SR may interfere with the dispersion of phenolic compounds in the fish mince and reduce their access to reactive radical species. Lingonberry press residue was previously shown to be a potent antioxidant during pH-shift processing of herring and salmon based on reduced levels of malon- dialdehyde and 4-hydroxy-(E)-2-hexenal (Zhang et al., 2022). Regarding different concentrations of berry press residues in the present study, based on their inhibitory effect on the formation of sec- ondary volatile compounds, the addition of 3 g/100 g and 1.5 g/100 g LR, SR, and SRO performed similarly. Especially during the first seven days of storage at 1 C, and in the case of LR and SRO, there were only a few statistically significant (p < 0.05) differences between the 1.5 g/100 g and 3 g/100 g concentrations (Supplemental Tables S3–S5). Reducing the concentration of added press residues is desirable for minimizing the changes in sensory properties and for limiting the costs of the addition. 3.2. Flavor and odor profile of raw and cooked Baltic herring minces 3.2.1. Panel performance The performance of the GDA panel was evaluated using PanelCheck software. According to 3-way ANOVA, Tucker-1 PCA, and p-MSE plots (data not shown), out of the attributes in which significant differences were observed, most disagreement was found in fish oil odor. The disagreement was, nevertheless, less significant than the sample effect. Marine odor and berry/fruit-like odor were most agreed upon. In the case of cooked minces, the berry-/fruit-like odor was most difficult for the panel, whereas fishy flavor and marine odor and flavor were among the best understood attributes due to low and nonsignificant pan- elist*sample interactions. Although some panelists had difficulties in repeatability or sensitivity of individual attributes, no one performed systematically weakly, and therefore, the results of all panelists were Fig. 2. Changes in selected secondary volatile oxidation compounds, 1-penten-3-ol (A–D), hexanal (E–H), and 3,5-octadien-2-one (I–L) as peak areas (n ˆ 3) in Baltic herring minces without any additions (control, C) (A–L) and with α-tocopherol and ascorbic acid (αT ‡ AA) (A, E, I), 1–3 g/100 g lingonberry-bilberry juice press residue (LR) (B, F, J), 1–3 g/100 g sea buckthorn juice press residue (SR) (C, G, K), or 1–3 g/100 g sea buckthorn juice press and oil extraction residue (SRO) (D, H, L) during the 14-day storage period. An asterisk (*) indicates a statistically significant (p < 0.05) difference between the samples at the given time point. T. Kakko et al. LWT 184 (2023) 115044 6 included for other statistical analyses. 3.2.2. Effect of berry press residue addition on the odor and flavor of Baltic herring mince during storage The odor profiles of raw and cooked minces and flavor/taste profiles of cooked minces (Fig. 3, Supplemental Table S6) were assessed to investigate the effect of LR (1.5 g/100 g) and SRO (1.5 g/100 g) addi- tions on the sensory quality of the minces during storage. In the case of raw samples, the C, LR1.5, and SRO1.5 at Day 0 differed statistically significantly (p < 0.05) only in terms of berry-/fruit-like odor. During the 3-day storage, however, the raw control mince exhibited significant increases in total intensity of odor, fishy odor, marine odor, and fish oil odor along with a decrease in berry-like/fruit-like odor intensity. LR1.5 showed a slight increase in fishy (3.3 vs. 4.0) and marine odor (2.0 vs. 3.5), and SRO1.5 did not show any significant (p < 0.05) differences on Day 0 vs. Day 3. A similar improvement in sensory quality by antioxi- dant treatment was reported by Sveinsdottir et al. (2020). In their study, dipping Atlantic mackerel fillets into sodium erythorbate was seen to increase the shelf life of fillets during frozen storage due to inhibition of rancid flavor development. In another study, tannic acid was seen to efficiently reduce the formation of TBARS, hydroperoxides, and fishy odor during storage of fish emulsion sausages (Maqsood et al., 2012). A previous study by Fu et al. (2009) reported that hemoglobin-induced oxidation caused an oxidized oil odor in silver carp (Hypophthalmichthys molitrix) mince. In the present study, fish oil odor increased in raw minces during the 3-day storage, but after cooking, there was no difference in fish oil odor between C-0d and C-3d. This could be due to the masking effect of other odors formed during sous vide cooking. Previously, sous vide cooking was seen to significantly alter the volatile profile of European sea bass to a higher extent than steaming and boiling (Nieva-Echevarría et al., 2017). Unlike the raw minces at Day 0, the cooked minces at Day 0 differed significantly in several odor attributes, indicating that cooking influ- enced the odor of the minces differently. For instance, cooked SRO1.5 at 0 d had a lower intensity of marine odor (2.5 vs. 3.8) and fish oil odor (1.6 vs. 2.5) compared to cooked C at 0 d. Cooked LR1.5 at 0 d had a lower fishy flavor (3.6 vs. 5.1) compared to C at 0 d. In the case of cooked minces at Day 3, significant differences in fishy odor were observed between C vs. both LR1.5 and SRO1.5. In a study by Cropotova et al. (2019), the addition of an antioxidant (commercial preparation containing rosemary extract) was seen to promote the oxidative stability of Atlantic mackerel during sous vide cooking, although sensory quality was not evaluated. The rancidity of samples was considered low (<2) in all samples (Fig. 3). The reference for rancid odor was fish oil oxidized by micro- wave heating, and hence, the rancidity was perhaps different from the rancid odors formed during cold storage of the fish minces. In general, the term “rancid” may refer to different sensory perceptions in different products since oxidation may produce different odor- and flavor-active compounds depending on the mechanism and precursor fatty acids (Fu et al., 2009; Jacobsen, 1999). In the raw control mince, a significant increase in fishy odor was observed after 3 days of storage at a low temperature (1 C), which demonstrates the impact of lipid oxidation on the sensory quality of Baltic herring and fish in general. The present study showed that in addition to delaying lipid oxidation, measured by PV and secondary oxidation-derived volatiles, the added berry press residues inhibited oxidative deterioration from a sensory perspective and thus helped to maintain the sensory quality during the storage period. 3.3. Odor-active compounds in raw minces with or without added sea buckthorn press residue As the most significant differences in odor, based on GDA, were found between raw control mince and raw mince with 1.5 g/100 g SRO, their odor-active compounds at 0 d and 3 d were investigated using GC- O (Table 1 and Table 2). Considerably more odor-contributing com- pounds (NIF 33%) were detected in the stored control mince (C-3d) than in SRO1.5-0d, SRO1.5-3d, and C-0d (22 vs. 12, 11, and 8 odor- active compounds, respectively, on the semipolar column). 3-Methylbu- tanal (described as musty, chocolate, solvent) and (Z)-1,5-octadien-3- one (green, raw, metal, pelargonium) were detected in all samples by all panelists and on both columns, but the NIFs and intensities of most odor- active compounds differed between minces. Most odor-active com- pounds, such as 2,3-pentanedione (butter, caramel, fatty), hexanal (grass, green, fresh), (Z)-4-heptenal (fish, rancid, green), (E,E)-2,4-hex- adienal (musty, potato, green, wax), and (E,Z)-3,5-octadien-2-one (fruity, musty), had the highest NIFs and intensities in C-3d. This is in Fig. 3. Odor profiles of raw (A) and cooked (B) and flavor/taste profile of cooked (C) Baltic herring minces based on the generic descriptive analysis (n ˆ 9*3 for raw samples, n ˆ 9*2 for cooked samples). Sample abbreviations: CˆControl, SRO1.5 ˆ mince with 1.5 g/100 g sea buckthorn juice press and oil extraction residue, LR1.5 ˆ mince with 1.5 g/100 g lingonberry juice press residue, 0d ˆ after 0 days of storage, 3d ˆ after 3 days of storage at 1 C. Attributes marked with an asterisk (*) differed statistically significantly (p < 0.05) between minces. T. Kakko et al. LWT184(2023)115044 7 Table 1 NIFs (Nasal Impact Frequencies), intensities, and weighed intensities (NIF*intensity)a of odor-contributing (NIF 33%) compounds detected on the semipolar column (SPB-624) in Baltic herring mince with (SRO1.5-0d, SRO1.5-3d) and without (C-0d, C-3d) sea buckthorn press residue, after 0 days (0d) or 3 days (3d) storage at 1 C. Code Compound Identificationb RI SPB- 624 Description NIF Intensity NIF*Intensity C_0d C_3d SRO1.5_0d SRO1.5_3d C_0d C_3d SRO1.5_0d SRO1.5_3d C_0d C_3d SRO1.5_0d SRO1.5_3d A1 unknown 505 stale, musty 17% 33% 33% 17% 1.0 1.0 1.0 2.0 0.2 0.3 0.3 0.3 A2 unknown 526 citrus, bilberry, flowery 0% 0% 17% 33% 0.0 0.0 1.0 1.8 0.0 0.0 0.2 0.6 A3 2-methylpropanal std, ms, O, RI 600 solvent, chocolate, musty, green 50% 67% 50% 50% 1.7 2.6 1.5 2.0 0.8 1.8 0.8 1.0 A4 2,3-butanedione std, ms, O, RI 640 butter, caramel, fatty 67% 50% 67% 67% 2.0 1.7 1.5 1.1 1.3 0.8 1.0 0.8 A5 unknown 683 sweet, fruity, creamy 17% 50% 33% 33% 1.0 2.2 1.3 1.0 0.2 1.1 0.4 0.3 A6 3-methylbutanal ‡ 2- methylbutanal std, ms, O, RI 701 musty, green, chocolate, solvent 100% 100% 100% 100% 2.6 3.1 2.1 2.7 2.6 3.1 2.1 2.7 A7 2,3-pentanedione std, ms, O, RI 741 caramel, butter, fatty 17% 100% 17% 0% 1.0 2.3 1.0 0.0 0.2 2.3 0.2 0.0 A8 unknown 765 fruity 0% 17% 33% 17% 0.0 1.0 1.0 1.0 0.0 0.2 0.3 0.2 A9 unknown 781 rancid, wax, soap 17% 83% 0% 0% 1.0 2.5 0.0 0.0 0.2 2.1 0.0 0.0 A10 (Z)-2-penten-1-ol std, ms, O, RI 834 no common descriptor 0% 33% 0% 0% 0.0 1.3 0.0 1.0 0.0 0.4 0.0 0.0 A11 hexanal std, ms, O, RI 849 grass, leaf, fresh 33% 50% 17% 33% 1.0 1.7 1.0 1.3 0.3 0.8 0.2 0.4 A12 ethyl-3-methylbutanoate std, ms, O, RI 882 sweet, fruity, candy, citrus 17% 33% 100% 83% 1.5 2.3 2.0 2.2 0.3 0.8 2.0 1.8 A13 Heptanal std, ms, O, RI 942 potato-like 17% 33% 17% 17% 1.5 1.0 1.0 2.0 0.3 0.3 0.2 0.3 A14 (Z)-4-heptenal std, ms, O, RI 954 fish, rancid, green, oil 33% 100% 33% 33% 1.0 3.1 1.0 2.0 0.3 3.1 0.3 0.7 A15 (E,E)-2,4-hexadienal std, O, RI 986 musty, potato, green, wax 0% 83% 33% 50% 0.0 1.8 1.0 1.3 0.0 1.5 0.3 0.7 A16 (Z)-1,5octadien-3-ol or (E)- 1,5-octadien-3-one std, O, RI 1031 stale/musty, mushroom, green 67% 50% 67% 33% 1.7 2.7 1.5 1.5 1.1 1.3 1.0 0.5 A17 (Z)-1,5-octadien-3-one std, O, RI 1042 green, waxy, soil, metal, pelargonium 100% 100% 100% 100% 2.9 3.3 2.3 2.7 2.9 3.3 2.3 2.7 A18 (E,Z)-2,4-heptadienal ms, O, RI 1070 green, cucumber 17% 33% 0% 0% 2.0 2.5 0.0 0.0 0.3 0.8 0.0 0.0 A19 unknown 1097 no common descriptor 0% 0% 33% 17% 0.0 0.0 1.0 1.0 0.0 0.0 0.3 0.2 A20 (E,Z)-3,5-octadien-2-one ms, RI 1136 fruity, musty 17% 67% 17% 0% 1.0 1.3 1.0 0.0 0.2 0.8 0.2 0.0 A21 E,E)-3,5-octadien-2-one ms, RI 1156 soap, cucumber, cooked rice 17% 50% 17% 0% 1.0 1.3 2.0 0.0 0.2 0.7 0.3 0.0 A22 unknown 1208 no common descriptor 0% 33% 0% 0% 0.0 1.0 0.0 0.0 0.0 0.3 0.0 0.0 A23 (E,Z)-2,6-nonadienal std, ms, O, RI 1228 cucumber, fatty 0% 33% 0% 17% 0.0 2.5 0.0 1.0 0.0 0.8 0.0 0.2 A24 unknown 1260 no common descriptor 0% 33% 0% 17% 0.0 1.0 0.0 1.0 0.0 0.3 0.0 0.2 A25 unknown 1361 musty, grain-like/ cereal-like 0% 50% 0% 0% 0.0 1.3 0.0 0.0 0.0 0.7 0.0 0.0 A26 unknown 1384 no common descriptor 33% 33% 0% 0% 1.5 2.0 0.0 0.0 0.5 0.7 0.0 0.0 Literature and databases for comparison included Ahonen et al. (2022); Aitta et al. (2021); An et al. (2020); Selli et al. (2006); Serot et al. (2002), NIST Chemistry WebBook, https://webbook.nist.gov, and Flavornet database, https://www.flavornet.org. a n ˆ 6. b Identification was based on comparison to a commercial reference compound (std), comparison of ms spectra match (recorded by HS-SPME-GC-MS analysis) to the NIST library (ms), similarity of odor description to previous literature (O), and/or observed LRI close to LRI reported by previous literature or NIST. T . K akko et al. LWT184(2023)115044 8 Table 2 NIFs (Nasal Impact Frequencies), intensities, and weighed intensities (NIF*intensity)a of odor-contributing (NIF 33%) compounds detected on the polar column (DB-WAX) in Baltic herring mince with (SRO1.5) and without (C) sea buckthorn press residue, after 0 days (0d) or 3 days (3d) storage at 1 C. Code Compound Identificationb RI DB- WAX Description NIF Intensity NIF*Intensity C-0d C-3d SRO1.5- 0d SRO1.5- 3d C- 0d C- 3d SRO1.5- 0d SRO1.5- 3d C- 0d C- 3d SRO1.5- 0d SRO1.5- 3d B1 propanal std, ms, O, RI 801 no common descriptor 0% 33% 0% 0% 0.00 1.00 0.00 0.00 0.0 0.3 0.0 0.0 B2 2-methylpropanal std, ms, O, RI 817 chocolate, solvent, coffee 50% 67% 50% 67% 2.50 2.00 2.00 2.13 1.3 1.3 1.0 1.4 B3 3-methylbutanal‡2-methylbutanal std, ms, O, RI 912 solvent, cacao, stale 100% 100% 100% 100% 2.17 3.00 2.50 2.58 2.2 3.0 2.5 2.6 B4 unknown 926 stale 0% 50% 0% 0% 0.00 2.00 0.00 0.00 0.0 1.0 0.0 0.0 B5 unknown 958 fresh, sweet 0% 0% 0% 50% 0.00 0.00 0.00 1.33 0.0 0.0 0.0 0.7 B6 2,3-butanedione std, ms, O, RI 974 caramel, sweet, butter 100% 83% 67% 0% 1.83 2.00 1.50 0.00 1.8 1.7 1.0 0.0 B7 unknown 1011 solvent, rancid oil, wax 0% 100% 17% 33% 0.00 2.33 1.50 2.00 0.0 2.3 0.3 0.7 B8 2,3-pentanedione ‡ unknown std, ms, O, RI 1051 caramel, butter, sweet, fresh, fruity 33% 100% 100% 100% 1.00 2.83 1.67 1.92 0.3 2.8 1.7 1.9 B9 ethyl-3-methylbutanoate std, ms, O, RI 1066 pineapple, fresh, fruit 33% 0% 33% 67% 1.00 0.00 2.50 2.13 0.3 0.0 0.8 1.4 B10 hexanal std, ms, O, RI 1078 green, fresh 0% 33% 0% 0% 0.00 2.25 0.00 0.00 0.0 0.8 0.0 0.0 B11 (Z)-3-hexenal O, RI 1131 green, grass 17% 33% 17% 33% 2.00 1.50 2.00 1.00 0.3 0.5 0.3 0.3 B12 heptanal std, ms, RI 1172 no common descriptor 17% 33% 0% 17% 1.00 1.25 0.00 2.00 0.2 0.4 0.0 0.3 B13 unknown 1219 pungent, musty/stuffy 50% 0% 17% 33% 2.00 0.00 2.00 2.25 1.0 0.0 0.3 0.8 B14 (Z)-4-heptenal std, ms, O, RI 1237 fish, stale, pungent, sea 50% 100% 17% 33% 2.33 3.33 2.00 2.50 1.2 3.3 0.3 0.8 B15 (Z)-2-penten-1-ol ‡ 1-octen-3-one or 2-octanone std, ms, O, RI 1300 mushroom, stale 50% 100% 100% 100% 2.67 2.33 1.67 2.50 1.3 2.3 1.7 2.5 B16 unknown 1349 bergamot, pelargonium 0% 67% 0% 0% 0.00 1.50 0.00 0.00 0.0 1.0 0.0 0.0 B17 (Z)-1,5-octadien-3-one std, O, RI 1377 green, raw, metal, pelargonium 100% 100% 100% 100% 2.83 3.08 2.83 3.33 2.8 3.1 2.8 3.3 B18 (E)-1,5-octadien-3-ol std, RI 1453 no common descriptor 0% 33% 0% 17% 0.00 2.00 0.00 1.00 0.0 0.7 0.0 0.2 B19 2,4-heptadienal (E,Z or E,E) ms, O, RI 1473 potato 0% 33% 33% 50% 0.00 2.00 1.50 1.83 0.0 0.7 0.5 0.9 B20 (Z)-1,5-octadien-3-ol std, O, RI 1485 green, mushroom 33% 83% 0% 0% 1.00 1.70 0.00 0.00 0.3 1.4 0.0 0.0 B21 (E)-2-octenol ms, O, RI 1621 green, grass, cucumber 0% 67% 0% 17% 0.00 2.13 0.00 1.00 0.0 1.4 0.0 0.2 B22 unknown 1643 no common descriptor 0% 33% 0% 0% 0.00 2.50 0.00 0.00 0.0 0.8 0.0 0.0 B23 unknown 1740 no common descriptor 0% 33% 0% 0% 0.00 2.00 0.00 0.00 0.0 0.7 0.0 0.0 B24 unknown 1770 no common descriptor 33% 17% 17% 50% 1.00 2.00 2.00 1.67 0.3 0.3 0.3 0.8 Literature and databases for comparison included Ahonen et al. (2022); Aitta et al. (2021); An et al. (2020); Selli et al. (2006); Serot et al. (2002), NIST Chemistry WebBook, https://webbook.nist.gov, and Flavornet database (Acree and Arn), https://www.flavornet.org. a n ˆ 6. b Identification was based on comparison to a commercial reference compound (std), comparison of ms spectra match (recorded by HS-SPME-GC-MS analysis) to the NIST library (ms), similarity of odor description to previous literature (O), and/or observed LRI close to LRI reported by previous literature or NIST. T . K akko et al. LWT 184 (2023) 115044 9 accordance with the higher total intensity of odor observed in the raw stored control (Fig. 3) compared to the other samples. The weighed intensities (intensity*NIF) of the detected odor-active compounds on the semipolar and polar columns were used as pre- dictors of the intensity of the odor attributes of the minces using partial least squares regression (PLS) (Fig. 4). Fishy odor, fish oil odor, marine odor, total intensity of odor, rancidity, and most odorants were associ- ated with C-3d. Many of these compounds are derived from lipid oxidation, such as propanal, hexanal, (E,Z)-2,6-nonadienal, (Z)-4-hep- tenal, (Z)-1,5-octadien-3-ol, 2,3-pentanedione, and (E,Z)-3,5-octadien- 2-one, the formation of which was retarded by the addition of SRO (1.5 g/100 g) in the mince. Ethyl-3-methylbutanoate (fruity, fresh, sweet) and two unknown compounds, A2 and A19 (Table 1), detected on the semipolar column correlated with the fruit-like/berry-like odor. Fishy odor has been acknowledged to be caused by a combination of lipid oxidation-derived aldehydes with low odor thresholds. Many of the odor-active compounds identified in the present study (Tables 1 and 2) have been previously associated with fishy odor (Sae-leaw & Benjakul, 2014; Triqui & Bouchriti, 2003; Venkateshwarlu et al., 2004). (Z)-4-heptenal, (Z)-1,5-octadien-3-one, and methional were suggested to result in fishy odor to sardine stored on ice for 2 days (Triqui & Bouchriti, 2003). They further reported that due to its low threshold, (Z)-1,5-octadien-3-one was also one of the most potent odorants in fresh sardine (stored for 0 d), and during the first two days of storage, 2,3-pen- tanedione, (E,Z)-3,5-octadien-2-one, and (E,E)-2,4-decadienal, among other secondary lipid oxidation products, also became more pronounced. Similarly, in a study by Sae-leaw and Benjakul (2014), lipid oxidation of sea bass skin during ice storage led to the development of a fishy odor, which was associated with several volatiles, including hexanal, heptanal, and 1,5-octadien-3-ol. In the present study, (Z)-1, 5-octadien-3-one, which can be produced during autoxidation of EPA (Hammer & Schieberle, 2013), was detected on both columns by all panelists in all samples. (Z)-4-heptenal, known to impart a fishy odor in combination with other volatiles such as (E,Z)-2,6-nonadienal (Ven- kateshwarlu et al., 2004), was detected at the highest frequency and intensity in the stored control mince on both columns. 4. Conclusions Side streams of berry processing were potential antioxidants in minced Baltic herring during cold storage. The analysis of PV and sec- ondary volatile oxidation compounds showed that a relatively low concentration of press residue in the case of lingonberry-bilberry retarded lipid oxidation in minced Baltic herring stored at 1 C. The addition of 1.5 g/100 g lingonberry-bilberry juice press residue or sea buckthorn residue after juice pressing and oil extraction prevented in- creases in fishy odor and fish oil odor, resulting in better preservation of the sensory quality during the 3-day storage of raw mince. Furthermore, the addition of sea buckthorn residue at a dosage of 1.5 g/100 g to fish mince kept the odor intensities low from several lipid oxidation-derived volatile compounds during storage. This study provides new insights into the potential use of berry processing side streams to improve the Fig. 4. Partial Least Squares (PLS) regression model with the odor-active compounds (Intensity*Nasal Impact Frequency) on the semipolar SPB-624 (A) or polar DB-WAX (B) column as X variables (n ˆ 26 in case of column A, n ˆ 24 in case of column B), in- tensities of GDA odor attributes as Y variables (n ˆ 7), and samples as dummy variables. Sample abbrevia- tions refer to raw minces without additions (C) or with 1.5 g/100 g sea buckthorn juice press and oil extraction residue (SRO1.5), after 0 (0d) or 3 days (3d) of storage at 1 C, and codes for odor-active compounds are the same as in Tables 1 and 2. T. Kakko et al. LWT 184 (2023) 115044 10 stability of fish or fish products during cold storage. CRediT authorship contribution statement Tanja Kakko: Conceptualization, Methodology, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Funding acquisition. Annelie Damerau: Conceptualization, Writing – original draft, Writing – review & editing, Supervision. Claudia Mejia Rios: Methodology, Formal analysis, Investigation. Oskar Laaksonen: Conceptualization, Writing – original draft, Writing – review & editing, Supervision. Baoru Yang: Concep- tualization, Methodology, Resources, Writing – review & editing, Su- pervision, Project administration, Funding acquisition. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Data will be made available on request. Acknowledgements This work was funded by the Finnish Operational Programme of the European Maritime and Fisheries Fund (project 33338) and a personal working grant for Tanja Kakko from the Finnish Cultural Foundation (grant 00200455). The authors want to thank Kiantama Oy and Arom- tech Oy Ltd for kindly providing the lingonberry-bilberry and sea buckthorn press residues, respectively. Kalaset Oy is acknowledged for providing the Baltic herring used in the second part of the study. Kris- tiina Aarnio is thanked for her assistance in the sensory evaluations. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.lwt.2023.115044. References Acree, T. & Arn, H. Flavornet and human odor space, Gas chromatography - olfactometry (GCO) of natural products. https://www.flavornet.org.. Ahonen, E., Damerau, A., Suomela, J.-P., Kortesniemi, M., & Linderborg, K. M. (2022). Oxidative stability, oxidation pattern and α-tocopherol response of docosahexaenoic acid (DHA, 22:6n–3)-containing triacylglycerols and ethyl esters. Food Chemistry, 132882. https://doi.org/10.1016/j.foodchem.2022.132882 Aitta, E., Marsol-Vall, A., Damerau, A., & Yang, B. (2021). Enzyme-assisted extraction of fish oil from whole fish and by-products of Baltic herring (Clupea harengus membras). Foods, 10, 1811. https://doi.org/10.3390/foods10081811 Aro, T., Tahvonen, R., Mattila, T., Nurmi, J., Sivonen, T., & Kallio, H. (2000). Effects of season and processing on oil content and fatty acids of Baltic herring (Clupea harengus membras). Journal of Agricultural and Food Chemistry, 48, 6085–6093. https://doi.org/10.1021/jf000389‡ Aydin, I., & Gokoglu, N. (2014). Effects of temperature and time of freezing on lipid oxidation in anchovy (Engraulis encrasicholus) during frozen storage. European Journal of Lipid Science and Technology, 116, 996–1001. https://doi.org/10.1002/ ejlt.201300450 Baron, C. P., KjÆrsgård, I. V. H., Jessen, F., & Jacobsen, C. (2007). Protein and lipid oxidation during frozen storage of rainbow trout (Oncorhynchus mykiss). Journal of Agricultural and Food Chemistry, 55, 8118–8125. https://doi.org/10.1021/jf070686f Bellucci, E. R. B., dos Santos, J. M., Carvalho, L. T., Borgonovi, T. F., Lorenzo, J. M., & Silva-Barretto, A. C. da (2022). Açaí extract powder as natural antioxidant on pork patties during the refrigerated storage. Meat Science, 184, Article 108667. https:// doi.org/10.1016/j.meatsci.2021.108667 Cavonius, L. R., & Undeland, I. (2017). Glazing herring (Clupea harengus) fillets with herring muscle press juice: Effect on lipid oxidation development during frozen storage. International Journal of Food Science and Technology, 52, 1229–1237. https:// doi.org/10.1111/ijfs.13399 Cropotova, J., Mozuraityte, R., Standal, I. B., & Rustad, T. (2019). Assessment of lipid oxidation in Atlantic mackerel (Scomber scombrus) subjected to different antioxidant and sous-vide cooking treatments by conventional and fluorescence microscopy methods. Food Control, 104, 1–8. https://doi.org/10.1016/j. foodcont.2019.04.016 Dai, Y., Wu, H., Liu, X., Liu, H., Yin, L., Wang, Z., Xia, X., & Zhou, J. (2021). Antioxidant activities and inhibitory effects of blueberry pomace and wine pomace crude extracts on oxidation of oil in water emulsion and fish mince. Journal of Food Processing and Preservation, 45, Article e15509. https://doi.org/10.1111/jfpp.15509 Damerau, A., Kakko, T., Tian, Y., Tuomasjukka, S., Sandell, M., Hopia, A., & Yang, B. (2020). Effect of supercritical CO2 plant extract and berry press cakes on stability and consumer acceptance of frozen Baltic herring (Clupea harengus membras) mince. Food Chemistry, 332, Article 127385. https://doi.org/10.1016/j. foodchem.2020.127385 EFSA Panel on Food additives and Nutrient Sources added to Food (ANS). (2015). Scientific Opinion on the re-evaluation of tocopherol-rich extract (E 306), α-tocopherol (E 307), γ-tocopherol (E 308) and δ-tocopherol (E 309) as food additives. EFSA Journal, 13, 4247. https://doi.org/10.2903/j.efsa.2015.4247 Farvin, K. H. S., Grejsen, H. D., & Jacobsen, C. (2012). Potato peel extract as a natural antioxidant in chilled storage of minced horse mackerel (Trachurus trachurus): Effect on lipid and protein oxidation. Food Chemistry, 131, 843–851. https://doi.org/ 10.1016/j.foodchem.2011.09.056 Fu, X., Xu, S., & Wang, Z. (2009). Kinetics of lipid oxidation and off-odor formation in silver carp mince: The effect of lipoxygenase and hemoglobin. Food Research International, 42, 85–90. https://doi.org/10.1016/j.foodres.2008.09.004 Gomez-Cortes, P., Sacks, G. L., & Brenna, J. T. (2015). Quantitative analysis of volatiles in edible oils following accelerated oxidation using broad spectrum isotope standards. Food Chemistry, 174, 310–318. https://doi.org/10.1016/j. foodchem.2014.11.015 Granato, D., Fidelis, M., Haapakoski, M., dos Santos Lima, A., Viil, J., Hellstrom, J., Ratsep, R., Kaldmae, H., Bleive, U., Azevedo, L., Marjomaki, V., Zharkovsky, A., & Pap, N. (2022). Enzyme-assisted extraction of anthocyanins and other phenolic compounds from blackcurrant (Ribes nigrum L.) press cake: From processing to bioactivities. Food Chemistry, 391, Article 133240. https://doi.org/10.1016/j. foodchem.2022.133240 Hamilton, R. J., Kalu, C., McNeill, G. P., Padley, F. B., & Pierce, J. H. (1998). Effects of tocopherols, ascorbyl palmitate, and lecithin on autoxidation of fish oil. Journal of the American Oil Chemists’ Society, 75, 813–822. https://doi.org/10.1007/s11746- 998-0231-4 Hammer, M., & Schieberle, P. (2013). Model studies on the key aroma compounds formed by an oxidative degradation of ω-3 fatty acids initiated by either copper(ii) ions or lipoxygenase. Journal of Agricultural and Food Chemistry, 61, 10891–10900. https://doi.org/10.1021/jf403827p Jacobsen, C. (1999). Sensory impact of lipid oxidation in complex food systems. Lipid/ Fett, 101, 484–492. https://doi.org/10.1002/(SICI)1521-4133(199912)101: 12<484::AID-LIPI484>3.0.CO;2-H Jiarpinijnun, A., Geng, J.-T., Thanathornvarakul, N., Keratimanoch, S., Üçyol, N., Okazaki, E., & Osako, K. (2022). Identifying volatile compounds in rabbit fish (Siganus fuscescens) tissues and their enzymatic generation. European Food Research and Technology, 248, 1485–1497. https://doi.org/10.1007/s00217-022-03977-1 Joaquin, H. J. F., Tolasa, S., Oliveira, A. C. M., Lee, C. M., & Lee, K. H. (2008). Effect of milk protein concentrate on lipid oxidation and formation of fishy volatiles in herring mince (Clupea harengus) during frozen storage. Journal of Agricultural and Food Chemistry, 56, 166–172. https://doi.org/10.1021/jf072460i Kakko, T., Aitta, E., Laaksonen, O., Tolvanen, P., Jokela, L., Salmi, T., Damerau, A., & Yang, B. (2022). Baltic herring (Clupea harengus membras) protein isolate produced using the pH-shift process and its application in food models. Food Research International, 158, Article 111578. https://doi.org/10.1016/j.foodres.2022.111578 Kakko, T., Damerau, A., Nisov, A., Puganen, A., Tuomasjukka, S., Honkapaa, K., Tarvainen, M., & Yang, B. (2022). Quality of protein isolates and hydrolysates from Baltic herring (Clupea harengus membras) and Roach (Rutilus rutilus) produced by pH-shift processes and enzymatic hydrolysis. Foods, 11, 230. https://doi.org/ 10.3390/foods11020230 Lee, H., Kizito, S.a., Weese, S.j., Craig-Schmidt, M.c., Lee, Y., Wei, C.-I., & An, H. (2003). Analysis of headspace volatile and oxidized volatile compounds in DHA-enriched fish oil on accelerated oxidative storage. Journal of Food Science, 68, 2169–2177. https://doi.org/10.1111/j.1365-2621.2003.tb05742.x Lehtonen, M., Kemmo, S., Lampi, A.-M., & Piironen, V. (2011). Effects of esterification on the formation and decomposition of steryl hydroperoxides. European Food Research and Technology, 232, 255–264. https://doi.org/10.1007/s00217-010-1384-6 Linderborg, K. M., Lehtonen, H.-M., Jarvinen, R., Viitanen, M., & Kallio, H. (2012). The fibres and polyphenols in sea buckthorn (Hippophae rhamnoides) extraction residues delay postprandial lipemia. International Journal of Food Sciences & Nutrition, 63, 483–490. https://doi.org/10.3109/09637486.2011.636346 Maqsood, S., Benjakul, S., & Balange, A. K. (2012). Effect of tannic acid and kiam wood extract on lipid oxidation and textural properties of fish emulsion sausages during refrigerated storage. Food Chemistry, 130, 408–416. https://doi.org/10.1016/j. foodchem.2011.07.065 Mattje, L. G. B., Tormen, L., Bombardelli, M. C. M., Corazza, M. L., & Bainy, E. M. (2019). Ginger essential oil and supercritical extract as natural antioxidants in tilapia fish burger. Journal of Food Processing and Preservation, 43, Article e13942. https://doi. org/10.1111/jfpp.13942 Nieva-Echevarría, B., Manzanos, M. J., Goicoechea, E., & Guillen, M. D. (2017). Changes provoked by boiling, steaming and sous-vide cooking in the lipid and volatile profile of European sea bass. Food Research International, 99, 630–640. https://doi.org/ 10.1016/j.foodres.2017.06.043 Noble, A. C., & Nawar, W. W. (1975). Identification of decomposition products from autoxidation of methyl 4,7,10,13,16,19-docosahexaenoate. Journal of the American Oil Chemists’ Society, 52, 92–95. https://doi.org/10.1007/BF02633045 T. Kakko et al. LWT 184 (2023) 115044 11 Ozalp Ozen, B., Eren, M., Pala, A., Ozmen, _I., & Soyer, A. (2011). Effect of plant extracts on lipid oxidation during frozen storage of minced fish muscle. International Journal of Food Science and Technology, 46, 724–731. https://doi.org/10.1111/j.1365- 2621.2010.02541.x Packer, J. E., Slater, T. F., & Willson, R. L. (1979). Direct observation of a free radical interaction between vitamin E and vitamin C. Nature, 278, 737–738. https://doi.org/ 10.1038/278737a0 Pihlajamaki, M., Asikainen, A., Ignatius, S., Haapasaari, P., & Tuomisto, J. T. (2019). Forage fish as food: Consumer perceptions on Baltic herring. Sustainability, 11, 4298. https://doi.org/10.3390/su11164298 Puganen, A., Kallio, H. P., Schaich, K. M., Suomela, J.-P., & Yang, B. (2018). Red/green currant and sea buckthorn berry press residues as potential sources of antioxidants for food use. Journal of Agricultural and Food Chemistry, 66, 3426–3434. https://doi. org/10.1021/acs.jafc.8b00177 Püssa, T., Pallin, R., Raudsepp, P., Soidla, R., & Rei, M. (2008). Inhibition of lipid oxidation and dynamics of polyphenol content in mechanically deboned meat supplemented with sea buckthorn (Hippophae rhamnoides) berry residues. Food Chemistry, 107, 714–721. https://doi.org/10.1016/j.foodchem.2007.08.090 Puupponen-Pimia, R., Nohynek, L., Suvanto, J., Salminen, J.-P., Seppanen-Laakso, T., Tahtiharju, J., Honkapaa, K., & Oksman-Caldentey, K.-M. (2021). Natural antimicrobials from cloudberry (Rubus chamaemorus) seeds by sanding and hydrothermal extraction. ACS Food Science and Technology, 1, 917–927. https://doi. org/10.1021/acsfoodscitech.0c00109 Ross, C. F., & Smith, D. M. (2006). Use of volatiles as indicators of lipid oxidation in muscle foods. Comprehensive Reviews in Food Science and Food Safety, 5, 18–25. https://doi.org/10.1111/j.1541-4337.2006.tb00077.x Rundblad, A., Holven, K. B., Ottestad, I., Myhrstad, M. C., & Ulven, S. M. (2017). High- quality fish oil has a more favourable effect than oxidised fish oil on intermediate- density lipoprotein and LDL subclasses: A randomised controlled trial. British Journal of Nutrition, 117, 1291–1298. https://doi.org/10.1017/S0007114517001167 Sae-leaw, T., & Benjakul, S. (2014). Fatty acid composition, lipid oxidation, and fishy odour development in seabass (Lates calcarifer) skin during iced storage. European Journal of Lipid Science and Technology, 116, 885–894. https://doi.org/10.1002/ ejlt.201300381 Sampels, S., Åsli, M., Vogt, G., & Mørkøre, T. (2010). Berry marinades enhance oxidative stability of herring fillets. Journal of Agricultural and Food Chemistry, 58, 12230–12237. https://doi.org/10.1021/jf1017862 Sarkki, S., & Pihlajamaki, M. (2019). Baltic herring for food: Shades of grey in how backcasting recommendations work across exploratory scenarios. Technological Forecasting and Social Change, 139, 200–209. https://doi.org/10.1016/j. techfore.2018.11.001 Silveira Alexandre, A. C., Corre^a Albergaria, F., dos Santos Ferraz e Silva, L. M., Carneiro Fernandes, L. A., de Sousa Gomes, M. E., & Pimenta, C. J. (2022). Effect of natural and synthetic antioxidants on oxidation and storage stability of mechanically separated tilapia meat. LWT, 154, Article 112679. https://doi.org/10.1016/j. lwt.2021.112679 Sveinsdottir, H. I., Karlsdottir, M. G., Arason, S., Stefansson, G., Sone, I., Skåra, T., Rustad, T., Larsson, K., Undeland, I., & Gudjonsdottir, M. (2020). Effect of antioxidants on the sensory quality and physicochemical stability of Atlantic mackerel (Scomber scombrus) fillets during frozen storage. Food Chemistry, 321, Article 126744. https://doi.org/10.1016/j.foodchem.2020.126744 Tomic, O., Luciano, G., Nilsen, A., Hyldig, G., Lorensen, K., & Næs, T. (2009). Analysing sensory panel performance in a proficiency test using the PanelCheck software. European Food Research and Technology, 230, 497. https://doi.org/10.1007/s00217- 009-1185-y Triqui, R., & Bouchriti, N. (2003). Freshness assessments of moroccan sardine (Sardina pilchardus): Comparison of overall sensory changes to instrumentally determined volatiles. Journal of Agricultural and Food Chemistry, 51, 7540–7546. https://doi.org/ 10.1021/jf0348166 Venkateshwarlu, G., Let, M. B., Meyer, A. S., & Jacobsen, C. (2004). Modeling the sensory impact of defined combinations of volatile lipid oxidation products on fishy and metallic off-flavors. Journal of Agricultural and Food Chemistry, 52, 1635–1641. https://doi.org/10.1021/jf0351321 Vieira, S. A., Zhang, G., & Decker, E. A. (2017). Biological implications of lipid oxidation products. Journal of the American Oil Chemists’ Society, 94, 339–351. https://doi.org/ 10.1007/s11746-017-2958-2 Wang, T., Li, Z., Yuan, F., Lin, H., & Pavase, T. R. (2017). Effects of brown seaweed polyphenols, α-tocopherol, and ascorbic acid on protein oxidation and textural properties of fish mince (Pagrosomus major) during frozen storage. Journal of the Science of Food and Agriculture, 97, 1102–1107. https://doi.org/10.1002/jsfa.7835 Wang, W., Xiong, P., Zhang, H., Zhu, Q., Liao, C., & Jiang, G. (2021). Analysis, occurrence, toxicity and environmental health risks of synthetic phenolic antioxidants: A review. Environmental Research, 201, Article 111531. https://doi. org/10.1016/j.envres.2021.111531 Yanishlieva, N. V., & Marinova, E. M. (2001). Stabilisation of edible oils with natural antioxidants. European Journal of Lipid Science and Technology, 103, 752–767. https://doi.org/10.1002/1438-9312(200111)103:11<752::AID-EJLT752>3.0.CO; 2-0 Zhang, J., Abdollahi, M., Alminger, M., & Undeland, I. (2022). Cross-processing herring and salmon co-products with agricultural and marine side-streams or seaweeds produces protein isolates more stable towards lipid oxidation. Food Chemistry, 382, Article 132314. https://doi.org/10.1016/j.foodchem.2022.132314 T. Kakko et al.