Effects of Onboard Refrigerated Seawater Storage of Whole Ungutted Atlantic Cod (Gadus morhua) on Quality Parameters of Fillets

ABSTRACT The study aimed to evaluate the effects of onboard refrigerated seawater (RSW) storage of whole ungutted cod on the quality parameters of fillets. The reference group was directly gutted, bled, and stored in ice, while three experimental groups were gill-cut, bled, and stored ungutted in an onboard RSW tub at −1.5°C for 24, 60, and 84 hours. The results showed a difference between groups with extended RSW storage leading to negative effects such as increased gaping, bile spots, TVB-N levels, and bacterial growth after 60 hours. Conversely, the 24-hour RSW group closely resembled the quality parameters of the reference group.


Introduction
Icelandic fish producers are increasingly producing more fresh fish products that require high quality raw materials.Notably, advancements in processing techniques and temperature control improvements have supported an increase in the export of fresh fish fillets, with extended shelf-life, improved yield, and more stable product quality (Lauzon et al. 2010;Margeirsson 2008;Margeirsson et al. 2010;Olafsdottir et al. 2006).The current study topic is focused on onboard refrigerated seawater (RSW) storage of whole-ungutted groundfish with different gutting procedures.It aims to explore the feasibility of storing ungutted fish onboard and subsequently gutting them later in onshore processing facilities.The gill-cut method can be a simpler and more effective onboard processing technique in comparison to direct gutting.The experiment was designed to evaluate the effects of RSW storage of whole gill-cut cod on fillet quality, compared to the direct gutting method.
RSW is preferred to ice as a cooling and storage medium in certain fisheries due to various advantages such as faster cooling and bulk chilling, reduced pressure on the fish, lower holding temperatures, efficient handling of large catch volumes, and decreased processing labor requirements (Graham et al. 1992).RSW systems have also proven beneficial for extended storage times for the catch, particularly in pelagic fishing.The use of RSW systems for onboard cooling and storage of pelagic fish is widespread and well-established.In Iceland, significant advancements have been made in the last decade in chilling and storing pelagic species onboard purse seiners, resulting in more valuable products (Thorvaldsson et al. 2011).For groundfish caught by trawls, however, the procedure is different.Prior to boarding, the groundfish has often been battling against the trawl net, up the ramp, and into the reception hold, where it is kept until bleeding.The waiting time in the reception hold negatively impacts fish quality due to the increased time before slaughter, leading to temperature rise and less efficient bleeding of the catch (Eliasson et al. 2019;Olsen et al. 2014).Therefore, expediting the processing from the reception is crucial to enhance the overall quality of the catch.The primary benefit of adopting a more rapid bleeding procedure, such as gill-cutting instead of direct gutting, is to increase the processing productivity, resulting in a shorter waiting time in the reception hold.Additionally, by cooling the fish below 0°C, the onset of rigor mortis can be delayed, resulting in less powerful and severe muscle contractions compared to when fish go through rigor mortis at higher temperatures, which often leads to muscle gaping (Huss 1995).However, it is important to note that RSW storage does have some common disadvantages, including excessive salt uptake, water uptake in low-fat species, protein loss, and issues with anaerobic spoilage bacteria (Graham et al. 1992).In most Northern European fisheries, gutting lean species like cod is considered mandatory, based on the assumption of reduced quality if they are not gutted.In the case of cod, it has been shown that the omission of gutting can cause a considerable quality loss and a reduction in storage life of up to five to six days, with visible discoloration of the belly area often observed after two days post-catch (Huss 1995).
The processing of whole-ungutted groundfish using onboard RSW storage offers the fishing industry opportunities for increased economic viability and sustainability.This approach has some advantages over direct onboard gutting and icing into tubs, as it facilitates faster and simpler processing and storage procedures onboard resulting in reduced processing time and labor efforts.Additionally, bringing ungutted fish onshore with the whole viscera well-chilled allows for the utilization of the entire viscera, presenting opportunities for increased side stream utilization.This, in turn, enables the utilization of typically overlooked components of the viscera, such as small-size liver, milt, and roes, the pyloric caeca and other enzyme sources, enhancing the overall efficiency and value of the catch.The industry's interests in these procedures for groundfish are therefore related to simplified onboard handling of fresh fish, shorter fishing trips, side stream utilization, and the growing emphasis of Icelandic fish producers on delivering higher quality raw material for fresh fish products.
The method of onboard RSW cooling and storage for ungutted whole fish is most common for pelagic fish in Icelandic fisheries.For pelagic fish, this requires no prior processing or bleeding before being stored in large onboard RSW tanks.Previous studies on using RSW systems as a means of storing Atlantic cod in onboard tanks are limited and outdated.Onboard RSW storage of groundfish at −1.1°C prior to processing has been carried out mainly on the Pacific coast with salmon and halibut, where it has been met with considerable success (Roach et al. 1961).Only minimal scale work has been reported on the Atlantic coast for holding cod in RSW onboard trawlers at −1.1°C or lower (MacCallum and Chan 1961).Watson (1996Watson ( , 1997) ) found that holding mackerel at −1°C in RSW treated with sodium hypochlorite reduced the rate of bacterial spoilage by over 50%.However, the same treatment did not significantly improve the quality of the whole-gutted cod.
The objective of this investigation was to evaluate the effects of onboard RSW storage of wholeungutted cod on the quality parameters of fresh fillets.

Materials and methods
The experiment was conducted onboard a large wetfish trawler during a fishing trip, involving four trial groups.One group served as the reference group, where cod was direct gutted, bled, and placed on ice (DG-ICE).The other three groups consisted of cod stored whole and ungutted in an onboard RSW tub for varying time periods.Cod in these groups were gill-cut, bled, and stored in RSW for 24, 60, and 84 hours, respectively (groups GC-RSW24h, GC-RSW60h, and GC-RSW84).After the designated time periods, the cod from the RSW tub were gutted and placed on ice for storage.The samples were processed at an onshore fisheries company and the fillets transported to and measured at Matis in Reykjavik.The transportation distance from the processing plant to the Matis research facilities covered 400 km, utilizing a refrigerated truck.

Experimental setup
Wild Atlantic cod was caught by trawl on 29 May 2015, in the Northwest of Iceland, with a tow time of 45 minutes.The average fish size was 4.29 kg, and the total haul size was 7,145 kg (6,825 kg cod and 320 kg saithe).The day of catch, May 29 th , is defined as Day 0 in the study.The fish (n = 90) was gill-cut (one side gill-cut, GC), bled for 25 minutes in 5-6°C seawater, and then stored in a 1000 L RSW tub, where the temperature was maintained with a mechanical slurry ice refrigeration unit and a circulating pump (Figure 1).The temperature in the RSW tub was kept at an average of −1.4 ± 0.4°C throughout the sea trip storage period.Figure 3 shows the average temperature profiles for the fish core and the RSW.Maxim Integrated's iButton data loggers (San Jose, CA, USA) were used to measure the fish core temperature, with a measurement resolution of 0.0625°C and an error margin specified by the manufacturer of no more than 0.5°C for temperatures between −15°C and 65°C.However, van Marken Lichtenbelt et al. (2006) reported that during a test for error margins, these loggers exhibited a reading error of no more than 0.09°C.The RSW temperature was measured using HOBO Tidbit v2 temperature loggers placed at the bottom, middle, and top positions inside the tub.The Tidbit v2 loggers from OnSet Computer Corporation (Bourne, MA, USA) have an accuracy of ±0.2°C for temperatures between 0°C and 5°C.
The reference group DG-ICE, which followed traditional processing methods, involved direct gutting at the same time that the other three experimental groups were gill-cut.The DG-ICE group was iced in a 460 L tub and kept in the trawler storage hold.After bleeding, the experimental groups GC-RSW24h, GC-RSW60h, and GC-RSW84h were stored ungutted for 24, 60, and 84 hours, respectively, in the RSW tub before being gutted, iced in tubs, and kept in the storage hold.After landing, all groups (n = 30 fish for each group) were processed, mechanically filleted, and evaluated.The first measurements were conducted after processing on Day 4 after catch.The fillets were then packed in expanded polystyrene (EPS) boxes and stored in a 2 ± 2°C cold storage at Matis.Subsequent measurements took place on Day 7 and Day 11.Visual quality assessment of fillets for gaping, blood, and bile spots (Figure 2) was carried out immediately after fillet processing on Day 4 from catch.Water content, water-holding capacity (WHC), and microbial measurements were conducted on Days 4, 7, and 11.

Description of experiment groups
• DG-ICE: Direct-gutted (DG), bled and iced in 460 L tub.

Chemical and microbial analysis
Total volatile basic nitrogen content (TVB-N) and spoilage bacteria measurements, total viable psychrotrophic counts (TVC), and measurements of H 2 S-producing bacteria, were done by assessing two pooled fillets (n = 2) for each sample, with measurements conducted in triplicate (n = 3).The method of Malle and Tao (1987) was used for TVB-N determination, measured by steam distillation and titration, after extracting the fish muscle with an aqueous trichloroacetic solution (7.5%).The distilled TVB-N was collected in a boric acid solution (4%) and titrated with a sulfuric acid solution (0.1 N).
For microbial analysis, fillets were aseptically minced, assessing two pooled fillets (n = 2) for each sample (n = 3).Minced flesh (25 g) was mixed with 225 mL of cooled Maximum Recovery Diluent (MRD, Oxoid, Hampshire, UK) in a stomacher for 1 minute.Successive 10-fold dilutions were performed as required.TVC (17°C, 5 days) were evaluated by spread-plating aliquots onto modified Long & Hammer agar with 1% NaCl.H 2 S-producing bacterial counts (17°C, 5 days, forming black  colonies by using sodium thiosulphate) were evaluated on spread-plated iron agar (modified from Gram et al. (1987) with 1% NaCl and no overlay).
All chemicals used for analysis in this study were of analytical grade and were purchased from Sigma-Aldrich (St. Louis, MO, USA), Sigma-Aldrich (Steinheim, Germany), and Fluka (Busch, Switzerland).

Water content and water holding capacity
Measurements of water-holding capacity (WHC), water content, and salt were performed on midsection parts of the fillets (n = 3).The pieces were minced and used for analysis.The water content of the samples was determined by measuring the weight difference of minced muscle samples after drying for 4 hours at 103°C (ISO 1999).The WHC was determined using 2 g of minced white epaxial muscle from the same location as the water content samples.The muscle was subjected to low-speed centrifugation (Heraeus Biofuge Stratos, Kendro Laboratory products, USA) as described by Eide et al. (1982), with a centrifugation force of 210 × g for 5 minutes at 4°C.The WHC was expressed as the percentage of retained weight after centrifugation, divided by the initial sample weight and multiplied by 100.The salt content of the samples was determined using the Volhard titration method (AOAC 2000).

Visual quality analysis
Two trained individuals experienced in sensory and quality evaluation of fish products conducted the visual assessments after filleting at the processing facilities.The evaluation was conducted on identical sample pools with ten fillets (n = 10) from each group.Visual assessment of the quality parameters gaping, blood, and bile spots was carried out with a 4 × 4 cm grid on the fillets, described by Margeirsson et al. (2007).The measurement method counts a defect within the grid area as one unit on the fillet.The defects were then presented as the average of a 10-fillet assessment.Nematodes were counted as the total number of nematodes present in each loin or belly flap.

Statistical analysis
Statistical analysis was performed using Microsoft Excel 2016 (Microsoft Inc.Redmond, WA, USA) and IBM SPSS Statistics v. 26 (International Business Machines, Armonk, New York, USA).Values are presented as standard error of mean (SEM).One-way analysis of variation (ANOVA) and Duncan's post hoc test were applied on all samples for each group or day, and the significance level was set to p < 0.05 for all statistical analyses.A t-test was then used to determine the significance of differences between means of two groups.

Results and discussion
The temperature inside the fish loin, along with the ambient temperature of the fish in the RSW tank, is shown in Figure 3. Throughout the RSW storage period, the core temperature of the measured fish ranged from −0.9 to −0.7°C, while the ambient temperature (cooling medium or RSW temperature) of the fish ranged between −2 and −1°C during the fishing trip.The temperature mapping during the RSW storage demonstrated a consistently stable core temperature of the fish during the onboard storage.The observed change in the cooling medium temperature can likely be attributed to the repositioning of the measured fish within the RSW tub, as the temperature loggers were placed both inside the fish loin and outside the fish.These storage conditions align with similar studies conducted by Roach et al. (1961) and Longard and Regier (1971) for onboard RSW storage of cod, making them comparable.
Quality parameters evaluated immediately after fillet processing are presented in Table 1.The assessment of quality parameters after filleting revealed minimal effects of RSW storage for 24 hours compared to DC-ICE.However, the groups subjected to longer RSW storage periods of 60 and 84 hours in the RSW tub exhibited a significant increase in fillet gaping and yellow bile spots.Likely, increased fillet gaping is caused by forced water flow conditions in the RSW storage, and bile spots on the fillet (Figure 2) are due to contamination from the viscera gallbladder.The salt content (g/100 g) was measured for all groups, with the results showing a significantly (p < 0.05) lower content of 0.30 ± 0.03% for groups DC-ICE and GC-RSW24h, compared to 0.40 ± 0.03% for groups GC-RSW60h and GC-RSW84h.These findings indicate that longer RSW storage leads to some salt absorption in the fish muscle.To assess the potential migration of nematodes from the viscera into the fish muscle during prolonged storage of the whole ungutted fish, nematodes were counted in both the loin and the belly flap.Nematodes need to be removed by trimming the fillet and therefore influence the fillet utilization directly.As reported by EFSA-BIOHAZ ( 2010) there are conflicting results on nematode migration from the viscera into the fish muscle.No nematodes were found in the fish loins and, as seen in Table 1, the nematodes found in the belly flap do not exhibit a clear relationship with longer RSW storage, although group GC-RSW84h did show the highest count.
Figure 4 shows the microbial growth for all groups, measured on Days 4, 7, and 11 from catch.The TVC count and H 2 S producing bacteria both demonstrate an increase over time in all groups.The microbial measurements clearly indicate a trend of bacterial growth related to longer RSW storage, observed in both TVC and H 2 S bacteria.
A study conducted by Reynisson et al. (2010) investigated the storage of whole gutted haddock in both slurry ice and plate ice over an 8-day period.The results showed that microbial growth was delayed at early storage for both cooling methods.However, while the slurry ice provided faster initial cooling, it led to unfavorable conditions during extended storage, resulting in the dominance of the spoilage causing bacteria Photobacterium Phophereum.In a study by Digre et al. (2011), gutted cod stored in slurry ice exhibited a significantly lower difference of log 1.3 CFU/g for TVC and log 0.6 CFU/g for H 2 S bacteria compared to flake ice.However, neither of these studies involved the storage of wholeungutted cod, as in the current study, and were therefore dealing with less gut bacteria during the storage period.Olsson et al. (2007) measured the bacterial count colony-forming unit (CFU) and the ratio of H 2 S bacteria in wild caught cod fillets after different ice storage periods.The CFU results by Olsson et al. (2007) revealed log 3-4 CFU/g after 3 days, log 7-8 CFU/g after 7 days, and log 9 CFU/g after 14 days, which are comparable to or higher values than the microbial values shown in Figure 4.
The current study shows lower microbial values for gutted cod stored on ice compared to Watson (1996), who measured TVC log 5-6 CFU/g after 6 days on ice and log 7 CFU/g on Day 12.For gutted cod stored in RSW, similar to the conditions in the present study, Watson (1996) measured TVC values of log 3-4 CFU/g after 5 days in RSW and log 4-5 CFU/g after 12 days.These values are considerably lower than those observed in the current study, most likely due to the contamination of the guts during the ungutted storage period.
The results of the spoilage bacteria count and TVB-N content (Figures 4 and 5) for Days 4 and 7 align with the quality and shelf life estimations by Lauzon et al. (2010) for cod processed 3 days from catch and stored at 2°C.On Day 11, the bacteria and TVB-N results in the current study are higher, likely due to the higher storage temperature of 4°C.A study by Olafsdottir et al. (2006) on the shelf life of ice cooled cod stored at an average temperature of 2°C revealed that a shelf life limit was reached on Day 11 when the TVB-N content was 34 (mg N/100 g), the TVC reached log 8 CFU/g, and the H 2 S bacteria count Table 1.Visual assessment defects in fillets after processing, 4 days after catch.Defects per fillet (n = 10 per group), mean ± SEM; NS = no significant difference between groups (p < 0.05); different superscript letters within each defect assessment denote a significant difference (p < 0.05) between the groups.

DG-ICE
GC-RSW24h GC-RSW60h GC-RSW84h Gaping 0.4 ± 0.2 a 0.6 ± 0.2 a 1.2 ± 0.2 b 1.2 ± 0.3 b Bile spots 0.0 ± 0.0 a 0.0 ± 0.0 a 0.4 ± 0.1 b 0.8 ± 0.2 b Nematodes in belly flap 0.0 ± 0.0 a 0.6 ± 0.3 ab 0.1 ± 0.1 a 1.2 ± 0.3 b Blood spots NS 0.2 ± 0.1 0.4 ± 0.2 0.2 ± 0.1 0.2 ± 0.1 reached log 6.6 CFU/g.The reference group in the present study, DG-ICE, exhibits lower TVB-N and spoilage bacteria compared to the experiment by Olafsdottir et al. (2006), indicating reasonable quality changes and shelf life considering the handling and storage temperature of ice cooled gutted cod.The observed trend in the bacterial growth among the groups was also noticeable in the TVB-N formation, as depicted in Figure 5, with no significant difference between groups on Days 4 and 7.However, a significant difference was observed on Day 11, with higher TVB-N content for groups GC-RSW60h and GC-RSW84h.This suggests that the increase in TVB-N occurs between 24 and 60 hours of RSW storage.The TVB-N development measured in the current study is consistent with the trimethylamine (TMA) development during the storage of cod in crushed ice and in RSW at 0°C, as reported by Roach et al. (1961).The development remained similar during the first week of storage; however, after one week, the RSW 60 h and 84 h groups showed significantly higher levels compared to DG-ICE cod and the RSW 24 h groups.Longard and Regier (1971) reported that RSW storage of gutted and ungutted whole cod resulted in slower TMA development during the first 7 days, followed by more rapid formation thereafter.Additionally, in another experiment within the same study, when the fish was subjected to fluctuating RSW temperatures from −1°C to + 1°C, exponential growth was observed in whole-ungutted cod, while the gutted group exhibited linear TMA development.
Figure 6 presents the water content and WHC for all groups.On Day 4, there is no difference between group DG-ICE and GC-RSW24h in terms of water content.However, group GC-RSW60h exhibits lower water content, and both groups GC-RSW60h and GC-RSW84h show lower WHC, indicating that longer RSW storage affects the samples resulting in a lower WHC.On Day 7, there is no difference in water content among the groups, but the WHC is significantly higher in groups GC-RSW24h and GC-RSW84h.On Day 11, there is no difference in water content among the groups, but the effects of RSW storage are evident in the WHC, as all RSW groups show significantly higher values compared to DG-ICE.This trend in WHC was also reported by both Digre et al. (2011) and Magnusson et al. (2009), where super-chilled cod stored in slurry ice showed a slight increase in WHC with storage time.GC-RSW24h: gill-cut, bled, and stored in RSW for 24 hours.GC-RSW60h: gill-cut, bled, and stored in RSW for 60 hours.GC-RSW84h: gillcut, bled, and stored in RSW for 84 hours.Different superscript letters within each sampling day denote a significant count difference (p < 0.05) between the experimental groups.Olsson et al. (2007) concluded that the relationship between the percentage of liquid loss (WHC expressed as percentage of weight of liquid release) in cod muscle and bacterial growth was broad and indirect, likely to be temporal and not causal.The WHC results in the study by Olsson et al. (2007) demonstrated WHC values of 83.2% −90.2% after 2-3 days, 75.5%−90.2%after 7-8 days, and 72.9% −89.9% after 14-17 days.This is in line with the WHC of cod in group DG-ICE, which decreases with longer storage time.Olsson et al. (2007) suggested that this observed difference could be caused by differences in autolytic or microbial degradation of the muscle structure.However, the current study's results for groups kept in RSW display an opposite trend, with WHC increasing with longer storage time in RSW.A possible reason for this is the higher salt content in the groups stored longer in RSW, indicating that a salting-in effect in the fillets becomes more apparent with storage time after processing and has a stronger effect on WHC than the microbial degradation.The results of the current study were also compared to the study by Digre et al. (2011), which demonstrated WHC values of 82.7%−85.6%and water content of 78.8%−80.3%for fresh farmed direct gutted cod (anesthetized pre and post rigor), similar to the values observed in the present study.

Conclusions
The results of this investigation show the negative effects of prolonged onboard RSW storage of whole-ungutted cod on various quality parameters of cod fillets.Specifically, longer RSW storage periods of 60 and 84 hours resulted in significantly higher levels of gaping and bile spots compared to both the reference group and the 24-hour RSW storage group.Moreover, bacterial growth, as indicated by the measurements of spoilage bacteria (TVC and H 2 S), was more pronounced with prolonged RSW storage.The TVB-N content as a measure of spoilage did not show significant difference between the groups on Days 4 and 7.However, on Day 11, the 60 and 84 hours RSW storage groups exhibited higher TVB-N levels.The WHC of the cod fillets were also influenced by RSW storage, likely due to microbial degradation for DG-ICE and salting-in effect for groups GC-RSW60h and GC-RSW84h, where higher salt content was observed.These findings suggest that the spoilage effects of prolonged RSW storage of ungutted cod on fillet quality become more apparent over time, particularly after 60 and 84 hours.In contrast, the difference in quality parameters between the 24-hour RSW group and the DG-ICE group was minimal.Based on these results, it is advisable to limit the RSW storage of whole-ungutted cod to 24 hours to mitigate the negative impacts on fillet quality.

Highlights
• Onboard RSW storage of whole-ungutted cod was studied and compared to iced cod • Prolonged onboard storage in RSW resulted in decreased quality • RSW storage for 60 hours and more led to a significant growth of spoilage bacteria and TVB-N • RSW storage for over 60 hours was a deciding factor for quality parameters • Difference in quality between the 24 hours RSW group and iced cod was minimal

Acknowledgments
This research was conducted as part of the research project "Redesign of demersal wetfish trawler processing decks" supported by the Technology Development Fund (grant number 142667-0611) and the AVS R&D Fund of Ministry of Fisheries and Agriculture in Iceland (grant number R 15 068-14).Thanks to the crew on board the trawler for their cooperation with the experiment and to the staff of Utgerdarfelag Akureyrar for helping with processing the samples and the quality evaluation.Thanks also to the Matis staff participating in sample storage and chemical measurements.

Figure 1 .
Figure 1.1000 L tub for onboard RSW trial.

Figure 2 .
Figure 2. Bile spot on fillet belly flap marked in red circle on the figure (yellow discoloration).

Figure 3 .
Figure 3. Temperature profiles of a fish loin (solid line) and the RSW cooling medium (dotted line) during the fishing trip.

Figure 4 .
Figure 4. Development of microbial count for all groups on days 4, 7, and 11 from catch.DG-ICE: direct gutted, bled, iced in tub.GC-RSW24h: gill-cut, bled, and stored in RSW for 24 hours.GC-RSW60h: gill-cut, bled, and stored in RSW for 60 hours.GC-RSW84h: gillcut, bled, and stored in RSW for 84 hours.Different superscript letters within each sampling day denote a significant count difference (p < 0.05) between the experimental groups.

Figure 5 .
Figure 5. TVB-N measurements (mean ± SEM, n = 3) for all groups on days 4, 7, and 11 from catch.Different superscript letters within each sampling day denote a significant difference (p < 0.05) between the experimental groups.

Figure 6 .
Figure 6.RSW experiment water and WHC (mean ± SEM, n = 3) for all groups on days 4, 7, and 11 from catch.Different superscript letters within each sampling day denote a significant difference (p < 0.05) between the experimental groups.