Effect of chill storage under different icing conditions
on sensory and physical properties of canned farmed salmon (Oncorhynchus kisutch)
Alicia Rodríguez 1, Nicolás Carriles 1, and Santiago P. Aubourg 2,*
1 Department of Food Science and Chemical Technology, Facultad de Ciencias Químicas y Farmacéuticas, Universidad de Chile, Santiago (Chile)
2 Department of Food Technology, Instituto de Investigaciones Marinas (CSIC), Vigo (Spain)
* Correspondent: firstname.lastname@example.org
This work focuses the sensory and physical properties of canned farmed coho salmon (Oncorhynchus kisutch); the effect of flake ice and slurry ice as previous slaughter and chilling conditions was studied. Hydrolytic chemical changes related to sensory and physical properties were also evaluated in canned salmon. Thermal treatment led to a canned muscle showing higher firmness, lower cohesivity and colour changes (higher L* and b* values; lower a* values); filling oils showed higher turbidity scores and lower L*, a* and b* values than starting oil. Additionally, oxidised and putrid odour development in canned muscle and filling oil was low. However, previous icing condition and time (up to 9 days) provided no changes in canned muscle and filling oil, except for an increasing oxidised odour and turbidity in filling oil with chilling time. Meantime, free fatty acid formation and K value were markedly affected by previous icing system and time.
Running Head: Canned salmon sensory and physical properties
Key Words: Coho salmon, farming, flake ice, slurry ice, canning, sensory acceptance, physical properties, constituent hydrolysis
Canning is one of the most important means of fish preservation (Horner, 1997). In it, two thermal steps (cooking and sterilisation) are included, so that both enzymes and bacteria should be permanently inactivated by heat and, provided reinfection does not occur and no negative interaction with the container is produced, heat processed fish keeps for a very long time. However, owing to the thermal sensitivity of a broad number of fish chemical constituents, different breakdown and hydrolytic events have been reported on canned fish, thus showing an important detrimental effect on sensory and physical properties related to fish colour and odour, water leaching and muscle texture (toughening, drying, etc.) (Aitken & Connell, 1979; Aubourg, 2001).
Since most species used for canning occur in glut quantities, canneries require to store the raw material before it is canned. As a result, the quality of canned fish will also depend to a large extent on the adequacy of the methods used to hold the raw material, whose quality may continuously change during storage prior to processing (Slabyj & True, 1978; Aubourg & Medina, 1997). In order to slowdown the mechanisms involved in quality loss after capture, fish refrigeration can be considered one of the most employed strategies. Therefore, various preservative methods such as traditional flake ice (Whittle et al., 1990), refrigerated seawater (Kraus, 1992), and addition of chemical preservation agents (Hwang & Regenstein, 1995) have been applied.
Recently, slurry ice has been reported to be a promising technique for the preservation of aquatic fresh food products (Yamada et al., 2002; Piñeiro et al., 2004). Two relevant characteristics of slurry ice when compared to flake ice are: i) its faster chilling rate, which is a consequence of its higher heat-exchange capacity, and ii) the reduced physical damage caused to food products by its microscopic spherical particles. The fluid nature of slurry ice allows continuous processes to be carried out, and hence the automation of the processing and distribution of fresh aquatic food products guarantees a more hygienic way of handling and processing them. Finally, slurry ice systems is a versatile technique that can be combined with other preservative agents (antiseptics, antioxidants, antimelanosics, etc.).
In recent years the fishing sector has suffered from dwindling stocks of traditional species as a result of important decreases in their availability. This has prompted the fish trade to pay an increasing attention to aquaculture techniques as a source of fish and other seafood products. Among cultivated fish, coho salmon (Oncorhynchus kisutch) has received a great attention because of its increasing production in countries like Chile, Japan and Canada (FAO, 2007). Previous research related to the chilling storage of this species accounts for the development of different spoilage pathways and quality loss (Aubourg et al., 2005; Aubourg et al., 2007). The present study concerns the coho salmon canning. Its employment for this purpose would be greatly recommended since as a farmed species, slaughtering and previous storage conditions could be conveniently mastered, this leading to great advantages for canning manufacturers. The work focuses the sensory and physical properties of canned coho salmon, taking into account the effect of replacing traditional flake ice by slurry ice as previous slaughter and chilling storage conditions. The hydrolysis formation of some chemical metabolites related to sensory and physical properties changes is also undergone.
MATERIALS AND METHODS
Two different icing systems (traditional flake ice and slurry ice) were employed for the fish slaughter and chilling storage.
Flake ice (FI) was prepared from fresh water with an Icematic F100 Compact device (Castelmac SPA, Castelfranco, Italy). The temperature of the FI was –0.5ºC, being the temperature of the chilled fish in the range of 0ºC to –0.5ºC.
Slurry ice (SI) was prepared using a FLO-ICE prototype (Kinarca S.A.U., Vigo, Spain). The composition of the SI binary mixture was 40 % ice/ 60 % water, prepared from filtered seawater (salinity: 3.3 %). The temperature of the SI mixture was –1.5ºC, being the temperature of the chilled fish in the range of –1.0ºC to –1.5ºC.
Fish material, slaughter and chilling storage
Specimens (32 fish) of farmed coho salmon (Oncorhynchus kisutch) (weight range: 2.8-3.2 kg) were obtained (day 0) from an aquaculture facility (Comercial Xanquéi; Lousame, La Coruña, Spain) and slaughtered at the farm by immersion either in FI (16 fish) or SI (16 fish). Fish were kept under such conditions during 24 hours till arrival in the laboratory. At this time (day 1), four fish belonging to each icing condition (FI/ SI) were separated and studied as raw fish. The same day (day 1), four other fish of each icing condition were canned according to conditions described below. The remaining fish (8 fish of each icing condition) were placed in an isothermal room at 2ºC and were maintained in flake or slurry ice; chilled fish were taken for the canning process on days 5 and 9 of icing treatment (at each time, four fish belonging to each icing condition). Throughout the experiment, both the flake and slurry ice were renewed when required, in order to maintain a 1:1 fish to ice ratio. Both in raw and in canned samples, each fish specimen was studied separately from others to achieve the statistical study (n=4).
Previous research concerning the chilled storage of the present species (Aubourg et al., 2005; Aubourg et al., 2007) has shown that after a 10 days-period storage microbiological activity and lipid oxidation development increased largely. Accordingly, in the present experiment, three different chilling times (0, 5 and 9 days) below this time limit were chosen.
Fish were steam cooked in our pilot plant in a horizontal retort during 45 min (102-103ºC) to a final backbone temperature of 65ºC, which was measured by a set of copper-constantan thermocouples, according to Pérez-Martín et al. (1989). The fish were then cooled at room temperature (15-18ºC) for about 2 h, headed, eviscerated and skinned.
Muscle portions (90 g) from salmon specimens were placed in small flat rectangular cans (105 x 60 x 25 mm; 150 ml). Two grams of NaCl were weighted and added to each can, that was then filled with sunflower oil as filling medium. The cans were vacuum-sealed and sterilised in our pilot plant in a horizontal retort heated by means of steam (115ºC, 45 min; Fo = 7 min); such lethality value was chosen according to previous research (Banga et al., 1993). When the heating time was completed, steam was cut off and air was used to flush away the remaining steam. Cans cooling was carried out at reduced pressure.
After 3 months of storage at room temperature (15-18ºC), the cans were opened and the liquid part was carefully drained off gravimetrically, filtered by means of a filter paper and collected. Then, the resulting liquid phase was centrifuged, the oil phase separated and dried with anhydrous Na2SO4. Salmon white muscle was separated and then wrapped in filter paper. Fish white muscle and the filling oil medium were used for analyses. Initial (starting) oil and oil that was placed in cans and sterilised in the absence of salmon muscle (heated oil) were also analysed.
Canning manufacturers recommend not to open cans before a 2-3 months-period has elapsed. This time is considered necessary for an adequate homogenisation of components inside the can. According to this, a 3 months-storage period was employed in the present study.
The following descriptors were analysed in salmon muscle (raw and canned) and in oil (starting, heated and filling): Firmness (resistance of salmon muscle to be deformed), cohesivity (binding degree of myotomes in salmon muscle), oxidised odour (presence of off-odours related to rancidity development), putrid odour (presence of off-odours related to amine formation and decayed meat) and oil turbidity (loss of natural transparency and presence of small particles). The sensory analysis was conducted according to the Quality Descriptive Analysis (QDA) method by a sensory panel consisting of ten experienced judges (five females and five males). Panellists were selected and trained according to international standards in use of sensory descriptors for raw and processed fish of different quality conditions (Howgate, 1992).
At each sampling time, the fish muscle and oil portions were presented to panellists in individual trays and were scored individually. The panel members shared samples tested. The different sensory descriptors were evaluated on non-structured linear scales with numerical scores from 0 to 10. Scores among panellists were averaged. For firmness and cohesivity, score 10 corresponds to the stage where such properties are observed in their maximum value, while score 0 represents the stage where a decrease is no more noticeable. For turbidity and rancid and putrid odours, score 0 represents the stage where such attributes are not noticeable, while stage 10 corresponds to the stage where no increase is possible. For both odour descriptors, score 5.0 was considered the borderline of acceptability; scores from the remaining descriptors are comparatively discussed without considering a borderline of acceptability.
A shear test was used to evaluate texture in raw and canned salmon muscle. Firmness and cohesivity were determined from a stress-distance curve obtained from a Universal Testing Machine (LR-5K; Lloyd Instruments Limited, Hampshire, England, UK) including a load cell of 500 N connected to a computer, this including a Dapmat 40-0465 software data analysis (version 3.05, Lloyd Instruments Limited, Hampshire, England, UK). A Warner-Bratzler blade (knife edge 60º), 1.2 mm thick, 150 mm width and cutting at a 1 mm s-1 speed was employed at 4ºC on a 4 x 4 x 2 cm sample. The firmness (N) was regarded as the resistance maximum of the muscle fibres against transversal shearing (maximum force) and was the height of the first peak; cohesivity was measured during the upward movement of the blade and was calculated as the cohesivity (mm) at maximum peak force (Sigurgisladóttir et al., 1999). The average value of quadruplicate replicates was considered in each sample analysis.
Instrumental colour analysis (CIE 1976 L*, a*, b*) was performed by employing a tristimulus Hunter Labscan 2.0/45 colorimeter according to previous research (Ortiz et al., 2008). Measurements were made directly on the salmon muscle and by employing a quartz cuvette in the case of oils. For each sample analysis, colour scores were obtained as mean values of four measurements obtaining by rotating the measuring head 90º between duplicate measurements per position.
NaCl content in fish muscle was determined by a modification of the Volhard method, which included boiling in HNO3, neutralization of NaCl meq with excess of AgNO3, and final determination of the excess of AgNO3 meq by reverse titration with NH4SCN (AOAC, 1990). Results were calculated as g NaCl kg-1 muscle.
Trimethylamine-nitrogen (TMA-N) values were determined by means of the picrate method, this including a spectrophotometric (410 nm) assessment (Beckman Coulter DU 640, London, UK) (Tozawa et al., 1971). This involves the preparation of a 5% (w/v) trichloroacetic acid extract of fish muscle. The results were expressed as mg TMA-N kg-1 muscle.
Free fatty acid (FFA) content was determined in the lipid extract of the salmon muscle by the Lowry & Tinsley (1976) method based on complex formation with cupric acetate-pyridine followed by spectrophotometric (715 nm) assessment. Results were expressed as g FFA kg-1 lipids.
Nucleotide degradation analysis was carried out starting from 6% perchloric acid extracts from the fish muscle according to previous research (Aubourg et al., 2007). Analysis was performed by HPLC, using a Beckman device provided with the programmable solvent module 126, and the scanning detector module 167 connected to the System Gold software, version 8.1 (Beckman Coulter). Separations were achieved on a reverse-phase Spherisorb ODS-2 C18 250 x 4.60 mm column (Waters, Milford, MA, USA), with an internal particle diameter of 5 μm. Standard curves for adenosine 5’-triphosphate (ATP) and each compound involved in its degradation pathway, adenosine 5’-diphosphate (ADP), adenosine 5’-monophosphate (AMP), inosine 5’-monophosphate (IMP), inosine (INO) and hypoxanthine (Hx), were constructed in the 0-1 mM range. Results obtained for each degradation compound were calculated as mmol kg-1 muscle. The K value was calculated according to the following concentration ratio: K value (%) = 100 x (INO +Hx) / (ATP+ADP+AMP+IMP+INO+Hx).
Data (n=4) obtained from the different sensory, physical and chemical analyses were subjected to the ANOVA method to explore differences by two different ways: icing condition (FI/ SI comparison) and icing time (Statsoft Inc., Statistica, version 6.0, 2001). Comparison of means was performed using the least-square differences (LSD) test. Confidence interval at the 95% level (p<0.05) was considered in all cases.
RESULTS AND DISCUSSION
Texture changes assessment
Sensory assessment of firmness and cohesivity of the fish muscle (Table 1) indicated that the canning process has led to a product showing a higher mean firmness degree and a lower (p<0.05) cohesivity value. No effect of the icing condition during the slaughtering and chilling storage could be assessed by the sensory panel. In addition, lengthening storage in both icing conditions up to 9 days was not accompanied by any additional change in both textural properties.
Texture assessment was also carried out by physical methodology (Table 2). A firmness increase (p<0.05) could be observed as a result of the canning process, according to previous research on heated rainbow trout (Oncorhynchus mykiss) (Schubring, 2008). Further, higher values were obtained (Table 2) for fish previously kept under SI conditions than for its counterpart from flake ice. For both icing conditions, no differences could be outlined as a result of increasing the chilling time. Concerning the physical analysis of cohesivity (Table 2), canned muscle showed as for sensory assessment (Table 1), lower values (p<0.05) than raw fish muscle. No differences could be assessed by comparison between fish samples corresponding to both icing conditions, neither as a result of lengthening the chilling time.
According to both sensory and physical assessments of texture, a firmness increase and a cohesivity decrease have been produced as a result of the canning process, so that a more breakable fish structure and more vulnerable to mechanical damage was obtained. The texture of the fish muscle has shown to depend on numerous intrinsic biological factors related to the density of the muscle fibres, as fat and collagen content (Sigurgisladóttir et al., 1999) and sexual maturity (Reid & Durance, 1992). Previous research reports that heating treatment converts the translucent, jelly-like cellular fish mass into an opaque and firmer material, as a result of protein denaturation and water removal. Additionally, the connective tissue holding the cells together is reported to be degraded by heat and blocks of cells become readily separated from one another, so that a lower cohesivity would be observed in muscle (Aitken & Connell, 1979); collagen hydrolysis to gelatine during heating would eliminate the role of connective tissue as a supporting structure so that there is no longer any link between myotomes or between the muscle and bones (Ma et al., 1983). Resulting toughening and firmness increase in canned fish muscle can be considered the result of increased bonding between myofibrillar proteins, the denaturation of myosin and water holding capacity decrease of proteins (Aubourg, 2001; Sankar & Ramachandran, 2005). In the case of fatty fish, the role of fat in the texture changes has also been found important due to crosslinking of peptide chains by reaction with lipid oxidation products (Rzhavskaya & Fonarev, 1988; Aubourg, 2001). Texture toughening has been reported to increase in thermally treated fish as heating processing temperature (Ma et al., 1983) and canned storage temperature (Paredes & Baker, 1988) increase. However, brine pre-cooking was found to result in texture improvement in canned freshwater nase (Chondrostoma nasus) (Lazos, 1997).
Present results on texture changes as a result of the canning process agree to previous related research on canned rainbow trout and pollock (Chia et al., 1983) and sardine (Losada et al., 2006) fish species. This latest study showed that no effect could be observed on texture properties as a result of replacing FI by SI during a previous holding time in ice (Losada et al., 2006), according to the present results. Contrary to the present research, texture quality changes were observed in canned sardine (Sardinella longiceps) and mackerel (Rastrelliger kanagurta) as a result of increasing the previous holding time in traditional ice (Madhavan et al., 1970).
An additional aspect to be considered in the muscle firmness increase as a result of canning is the NaCl content change in fish muscle. In the present research, canned muscle showed NaCl contents (Table 2) included in the 10.0-14.4 g kg-1 muscle range, widely higher (p<0.05) than the value observed for the raw starting fish (0.5±0.1 g kg-1 muscle). This NaCl content increase in fish muscle agrees to the enhancement effect recognised for this compound on physical properties such as firmness (Slabyj & True, 1978; Chiralt et al., 2001). Among the different canned samples, a marked higher (p<0.05) value could be observed for canned fish previously kept under SI condition during 9 days.
Odour development assessment
Oxidised and putrid odour development was measured by the sensory panel in the canned salmon muscle (Table 1). Incipient and very low values were attributed to both parameters in all cases, so that all kinds of muscle were scored as highly acceptable. Such results indicated that under the present experiment conditions, a low development of rancidity and microbial activity was present throughout the complete canning process including slaughtering, chilling storage, cooking and sterilisation. Both odour values did not provide significant differences as a result of the previous icing time in both icing systems. Comparison between samples corresponding to the two icing conditions did not provide significant differences. Previous research related to chilling processing has shown an enhancement of sensory quality attributes in fish kept under SI condition when compared to its counterpart fish chilled under flake ice (Piñeiro et al., 2004). However, present results have shown that after a strong heat process (cooking and sterilisation) such sensory advantages were not noticeable in the canned product. Previous related research provides contradictory results concerning the effect of the previous icing time. Thus, an off-odour development was observed in different kinds of canned products prepared from white sardine (Kowala coval) (Jeyasekaran & Saralaya, 1991), Maine sardine (Slabyj & True, 1978) and mackerel (Rastrelliger kanagurta) and sardine (Sardinella longiceps) (Madhavan et al., 1970) as a result of increasing the previous chilling time under traditional flake ice conditions. Contrary, a taste panel could not distinguish between canned herring (Clupea harengus) stored at different time in ice before canning (McLay, 1968).
The oxidised and putrid odour development was also assessed in the filling oil. Results are shown in Table 3. In spite of showing increasing values (p<0.05) when compared to starting oil values, both attributes provided relatively low scores so that all kinds of canned oils were considered as largely acceptable. No effect (p>0.05) of the thermal treatment on the odour development is observed by comparing the starting oil and the heated oil values (Table 3). Accordingly, presence in filling oils of molecules responsible for the oxidised and putrid odours can be explained partially by their ability to extract them from the canned muscle and partially because of the ability of oils to interact with chemical constituents present in the fish muscle. Such interaction between both phases of the can was already observed in previous research concerning nutritional and quality aspects (Aubourg & Medina, 1997; Aubourg et al., 1997). Oxidised odour scores (Table 3) showed an increasing pattern with icing time for filling oils corresponding to both holding conditions; in addition, higher mean values were observed in the case of filling oils related to FI condition. According to the present results, an increasing rancid odour development was already observed by Madhavan et al. (1970) in filling oil corresponding to canned mackerel (Rastrelliger kanagurta) and sardine (Sardinella longiceps) as a result of increasing the previous chilling time under traditional flake ice condition.
Colour and turbidity changes analysis
Concerning the colour assessment, the physical tristimulus method was applied to canned muscle (Table 4). According to the raw fish values, the canning process led to a very different product where higher L* and b* values were obtained, while a* values reflected a substantial decrease (p<0.05). Present results on L* and a* values agree to previous research concerning fish species thermal treatment (Choudhry, 1977; Schubring, 2008), this including a good correlation between a* value and carotenoid pigment content (Choudhry, 1977). In addition, an a* value loss has been reported to correlate with hemoglobin-mediated lipid oxidation and to provide an inverse relationship with secondary lipid oxidation development (Wetterskog & Undeland, 2004). On the other hand, previous research has shown that heat processed fish were always more yellow (higher b* value) than raw fish (Choudhry, 1977), according to the present results. Additionally, an important relationship between b* value and the formation of polymerised Schiff bases and fluorescent compounds (tertiary lipid oxidation compounds) has been observed (Undeland et al., 2003)
Although some differences could be observed in the present research as a result of the chilling time and the kind of ice employed, a general pattern could not be concluded concerning both effects on the colour properties of the canned fish muscle. In this sense, previous research (Ortiz et al., 2008) did not afford differences in L*, a* and b* values in chilled rainbow trout as a result of employing different icing systems (FI and SI). Contrary, an increasing previous holding temperature and time in ice led to an increase in L* value and a decrease in a* value, while no effect on b* value could be observed for canned skipjack tuna (Little, 1972).
The present research accounts also for the colour analysis of the filling oil (Table 5). The canning process has led to marked decreases for all colour parameters in filling oils, a* and b* values attaining green and blue colour scores, respectively. Meantime, comparison between starting and heated oil showed an increase for L* and b* values and a decrease for a* value as a result of the thermal treatment. Some differences as a result if the icing condition and the chilling time could be assessed in the filling oils; however, great fish-to-fish differences could be observed in some cases, so that a marked pattern could not be concluded about the effect of such factors on L*, a* and b* values. Contrary to the present results, a browning development (b* value increase) was observed in filling oil corresponding to sardine (Sardinella longiceps) and mackerel (Rastrelliger kanagurta) muscle by increasing the previous holding time in ice of chilled fish (Madhavan et al., 1970).
Significant increases (p<0.05) in turbidity were detected in filling oils when compared to starting oil and heated oil (Table 3). Such increase could be mostly explained as a result of water loss of the salmon muscle inside the can. Water loss of fish muscle is considered the result of protein denaturation and accordingly, water holding capacity decrease (Aubourg, 2001; Schubring, 2008). Higher mean values were obtained for turbidity in filling oil corresponding to FI-treated fish when compared to its counterpart from SI system; additionally, an increasing pattern could be observed with storage time for both holding conditions.
Formation of hydrolysis compounds
Different hydrolysis and breakdown events concerning chemical constituents that are known to be related to sensory and physical changes in canned fish (Aitken & Connell, 1979; Aubourg, 2001) were analysed in the present research. Such analyses were carried out on the salmon muscle.
Trimethylamine (TMA) is known to be produced as a result of microbial activity during the chilled storage (Whittle et al., 1990; Olafsdóttir et al., 1997) and as a result of thermal breakdown of trimethylamine oxide (TMAO) during cooking and sterilisation steps (Madhavan et al., 1970; Gallardo et al., 1990; Losada et al., 2006). TMA presence has shown a great effect on off-odours development related to putridity. In the present research (Figure 1), great differences (p<0.05) were found between the raw fish TMA-N values and those from canned fish from day 1, this leading to the conclusion that thermal treatment (cooking and sterilisation) has exerted a marked effect on the TMA formation. Such result agrees to previous research where albacore tuna was canned under different sterilization conditions (Gallardo et al., 1990); in all cases, a marked TMAO content decrease was accompanied by an important TMA formation during cooking and sterilisation steps. However, no effect of the icing time or the icing system employed could be outlined in the present study, according to a previous experiment on canned sardine where both previous icing systems (FI and SI) were comparatively checked (Losada et al., 2006). Contrary, previous research (Madhavan et al., 1970; Slabyj & True, 1978) showed that an increasing holding time in traditional ice would lead to an increase in the TMA-N content in the canned product. Present results on TMA formation agree to putrid odour development in both muscle (Table 1) and filling oil (Table 3) in the sense that a marked increasing effect (p<0.05) is observed for thermal treatment, while no effect (p>0.05) can be attributed to holding conditions (icing system and time).
FFA are known to be produced as a result of endogenous and microbial enzymes activity on large-size lipid molecules (TG and PL classes, namely) during the chilled storage (Olafsdóttir et al., 1997), and as a result of large-size lipid molecules breakdown during the thermal treatment (Aubourg et al., 1997); further, FFA formation during the canning process has shown a great dependence on sterilisation temperature (Aubourg et al., 1997). FFA presence has shown an important role in fish muscle texture changes (Aubourg, 2001) and by enhancing the lipid oxidation development and accordingly, the formation of rancid off-odours (Rzhavskaya & Fonarev, 1988). In the present research (Figure 2), canned fish corresponding to day 1 showed an important FFA content increase (p<0.05) when compared to raw fish (thermal breakdown effect). In addition, an important effect of icing time could be observed for both icing conditions, since increasing FFA contents could be observed by increasing the previous icing time for both FI and SI conditions, according to previous related research (Madhavan et al., 1970; Aubourg & Medina, 1997). Finally, higher mean values were obtained in all cases for fish that had been previously stored under FI system when compared to its counterpart from SI condition. This result agrees to previous research on chilled material where, a lower FFA formation was obtained in fish kept under SI condition when compared to its counterpart under FI (Piñeiro et al., 2004). A good correlation value (r2 = 0.91 and 0.94, for FI and SI conditions, respectively) was obtained in this study between FFA values in canned muscle and oxidised odour scores in filling oils, according to the above mentioned lipid hydrolysis-oxidation relationship (Rzhavskaya & Fonarev, 1988; Aubourg, 2001). Meantime, lower correlation values were obtained between the FFA formation in muscle and the oil turbidity scores (r2 = 0.81 and 0.70, for FI and SI conditions, respectively).
During post-mortem fish storage, muscle nucleotides are known to degrade in a series of stages as a result of endogenous biochemical changes; the level of major adenine nucleotides and their related compounds (K value assessment) has been utilised extensively as an index of freshness of fish muscle (Olafsdóttir et al., 1997). In addition, some of the molecules included in the K value assessment (IMP and Hx) have been reported to exert an important effect on sensory properties of the fish muscle (Hughes & Jones, 1966; Kawai et al., 2002). In the present research (Figure 3), no effect of the thermal treatment (cooking and sterilisation) could be observed on the K value since, canned fish samples corresponding to day 1 did not reveal differences (p>0.05) when compared to raw fish values. However, an important effect of icing time and icing condition could be outlined. Thus, an increasing K value could be observed with icing time, and a lower autolysis behaviour could be concluded in canned fish corresponding to previous SI system storage. A good correlation value (r2 = 0.93 and 0.95, for FI and SI conditions, respectively) could be obtained between the K value in canned fish and the oxidised odour scores in filling oils; however, poor correlation values were obtained between the K value and the putrid odour development in fish muscle and in filling oil. In previous research (Piñeiro et al., 2004), a lower K value has been reported for chilled fish kept under SI system when compared to its counterpart stored under FI condition. Previous research carried out on K value analysis in canned fish is scarce, but shows that the K value assessment would not be likely to be influenced by the thermal treatment (Hughes & Jones, 1966; Vázquez-Ortiz et al., 1997) in agreement with present results. Accordingly, this quality index could be considered a profitable tool in order to assess the freshness degree of the raw material employed for canning.
The different steps included in the canning process have led to important sensory and physical properties changes in the fish product. Most changes (texture and colour, specially) can be related to the thermal treatment (cooking and sterilisation steps), while previous holding conditions studied (icing system and time) were found to have short effects on sensory and physical properties changes (namely, oxidised odour and turbidity in filling oil) of the canned product. According to oxidised and putrid odour values (fish muscle and filling oil assessments), it could be concluded that coho salmon has provided a widely acceptable canned product under the present conditions checked (slaughter and icing storage up to 9 days).
A greater effect of icing condition and time could be observed by analysing the hydrolysis formation of some chemical metabolites. Thus, FFA formation and K value in canned muscle showed a marked influence of the previous icing system employed and the icing time, while TMA formation showed a similar pattern than most sensory and physical properties. A good correlation value was obtained for the oxidised odour in filling oil with FFA content and K value in canned muscle.
The authors thank Mr. Marcos Trigo for technical assistance, Kinarca S.A.U. for providing the slurry ice equipment, Comercial Xanquéi (Lousame, La Coruña, Spain) for providing the salmon fish and Justo López Valcarcel S.A. for lending its equipment to elaborate the canned product. The work was carried out in the Chilean University (Chile)-Consejo Superior de Investigaciones Científicas (Spain) Program (Project 2006 CL 0034) and supported by the Secretaría Xeral de I+D (Xunta de Galicia, Spain) (Project PGIDIT05TAL00701CT).
Aitken, A. & Connell, J. (1979). Fish. In: Effects of heating on food stuffs (edited by R. Priestley). Pp. 219-254. London (UK): Applied Science Publishers Ltd.
AOAC (1990). Official Methods of Analysis of the Association of Analytical Chemistry (edited by K. Helrich), 15th edition. P. 870. Arlington, VA (USA): Association of Official Chemists, Inc.
Aubourg, S. & Medina, I. (1997). Quality differences assessment in canned sardine (Sardina pilchardus) by fluorescence detection. Journal of Agricultural and Food Chemistry, 45, 3617-3621.
Aubourg, S. (2001). Review: Loss of quality during the manufacture of canned fish products. Food Science and Technology International, 7, 199-215.
Aubourg, S., Gallardo, J. & Medina, I. (1997). Changes in lipids during different sterilising conditions in canning albacore (Thunnus alalunga) in oil. International Journal of Food Science and Technology, 32, 427-431.
Aubourg, S., Quitral, V., Larraín, MªA., Rodríguez, A., Gómez, J., Maier, L. & Vinagre, J. (2007). Autolytic degradation and microbial activity in farmed coho salmon (Oncorhynchus kisutch) during chilled storage. Food Chemistry, 104, 369-375.
Aubourg, S., Vinagre, J., Rodríguez, A., Losada, V., Larraín, MªA., Quitral, V., Gómez, J., Maier, L. & Wittig, E. (2005). Rancidity development during the chilled storage of farmed coho salmon (Oncorhynchus kisutch). European Journal of Lipid Science and Technology, 107, 411-417.
Banga, J., Alonso, A., Gallardo, J. & Pérez-Martín, R. (1993). Mathematical modelling and simulation of the thermal processing of anisotropic and non-homogeneous conduction-heated canned foods: Application to canned tuna. Journal of Food Engineering, 18, 169-387.
Chia, S., Baker, R. & Hotchkiss, J. (1983). Quality comparison of thermoprocessed fishery products in cans and retortable pouches. Journal of Food Science, 48, 1521-1526.
Chiralt, A., Fito, P., Barat, J., Andrés, A., González-Martínez, C., Escriche, I. & Camacho, M. (2001). Use of vacuum impregnation in food salting process. Journal of Food Engineering, 49, 141-151.
Choudhry, M. (1977). Effect of thermal processing, size, and storage on the carotenoid pigment content and other quality factors of channel catfish. Dissertation Abstracts International B, 37, 69.
FAO (2007). Fishery statistics. Aquaculture production. In: Food and Agriculture Organization of the United Nations. P. 73. Rome, Italy: Yearbook 2005. Vol. 100/2.
Gallardo, J. Pérez-Martín, R., Franco, J., Aubourg, S. & Sotelo, C. (1990). Changes in volatile bases and trimethylamine oxide during the canning of albacore (Thunnus alalunga). International Journal of Food Science and Technology, 25, 78-81.
Horner, W. (1997). Canning fish and fish products. In: Fish Processing Technology (edited by G. Hall). Pp. 119-159, 2nd edition. London (UK): Blackie Academic and Professional, Chapman and Hall.
Howgate, P. (1992). Codex review on inspection procedures for the sensory evaluation of fish and shellfish, CX/FFP, 92/14.
Hughes, R. & Jones, R. (1966). Measurement of hypoxanthine concentration in canned herring as an index of the freshness of the raw material with a comment on flavour relations. Journal of the Science of Food and Agriculture, 17, 434-438.
Hwang, K. & Regenstein, J. (1995). Hydrolysis and oxidation of mackerel (Scomber scombrus) mince lipids with NaOCl and NaF treatments. Journal of Aquatic Food Products Technology, 4, 19-30.
Jeyasekaran, G. & Saralaya, K. (1991). Effect of chilled seawater storage of white sardine (Kowala coval) on its canned product quality. Journal of Food Science and Technology, 28, 23-26.
Kawai, M., Okiyama, A. & Ueda, Y. (2002). Taste enhancements between various amino acids and IMP. Chemical Senses, 27, 739-745.
Kraus, L. (1992). Refrigerated seawater treatment of herring and mackerel for human consumption. In: Pelagic fish: The resource and its exploitation (edited by J. Burt, R. Hardy & K. Whittle). Pp. 73-81. Aberdeen, Scotland, UK: Fishing News Books.
Lazos, E. (1997). Freshwater nase (Chondrostoma nasus): thermally processed as a potential source. Journal of Aquatic Food Product Technology, 6, 45-63.
Little, A. (1972). Effect of pre- and post-mortem handling on reflectance characteristics of canned skipjack tuna. Journal of Food Science, 37, 502-502.
Losada, V., Rodríguez, A., Ortiz. J. & Aubourg, S. (2006). Quality enhancement of canned sardine (Sardina pilchardus) by a preliminary chilling treatment. European Journal of Lipid Science and Technology, 108, 598-605.
Lowry, R. & Tinsley, I. (1976). Rapid colorimetric determination of free fatty acids. Journal of the American Oil Chemists’ Society, 53, 470-472.
Ma L., Deng, J., Ahmed, E. & Adams, J. (1983). Canned shrimp texture as a function of its heat history. Journal of Food Science, 48, 360-363.
Madhavan, P., Balachandran, K. & Choudhuri, D. (1970). Suitability of ice stored mackerel and sardine for canning. Fishery Technology, 7, 67-72.
McLay, R. (1968). The effect of storage prior to processing on the quality of canned herring. Journal of Food Technology, 3, 361-365.
Olafsdóttir, G., Martinsdóttir, E., Oehlenschläger, J., Dalgaard, P., Jensen, B., Undeland, I., Mackie, I., Henehan, G., Nielsen, J. & Nilsen, H. (1997). Methods to evaluate fish freshness in research and industry. Trends in Food Science and Technology, 8, 258-265.
Ortiz, J., Palma, Ó., González, N. & Aubourg, S. (2008). Lipid damage in farmed rainbow trout (Oncorhynchus mikiss) after slaughtering and chilling storage. European Journal of Lipid Science and Technology, 110, 1127-1135.
Paredes, M. & Baker, R. (1988). Physical, chemical and sensory changes during thermal processing of three species of canned fish. Journal of Food Processing and Preservation, 12, 71-81.
Pérez-Martín, R., Banga, J., Sotelo, C., Aubourg, S. & Gallardo, J. (1989). Prediction of precooking times for albacore (Thunnus alalunga) by computer simulation. Journal of Food Engineering, 10, 83-95.
Piñeiro, C., Barros-Velázquez, J. & Aubourg, S. (2004). Effects of newer slurry ice systems on the quality of aquatic food products: a comparative review versus flake-ice chilling methods. Trends in Food Science and Technology, 15, 575-582.
Reid, R. & Durance, T. (1992). Textural changes of canned chum salmon related to sexual maturity. Journal of Food Science, 57, 1340-1342.
Rzhavskaya, F. & Fonarev, N. (1988). The effect of lipid content on the quality of canned mackerel. Rybnoe Khozyaistvo, 5, 82-84.
Sankar, T. & Ramachandran, A. (2005). Thermal stability of myofibrillar protein from Indian major carps. Journal of the Science of Food and Agriculture, 85, 563-568.
Schubring, R. (2008). Comparative study of the DSC pattern, color texture and water-binding capacity of rainbow trout muscle during heating. Journal of Food Processing and Preservation, 32, 190-218.
Sigurgisladóttir, S., Hafsteinsson, H., Jonsson, A., Nortvedt, R., Thomasses, M. & Torrisen, O. (1999). Textural properties of raw salmon fillets as related to sampling method. Journal of Food Science, 64, 99-104.
Slabyj, B. & True, R. (1978). Effect of pre-process holding on the quality of canned Maine sardines. Journal of Food Science, 43, 1172-1176.
Tozawa, H., Erokibara, K. & Amano, K. (1971). Proposed modification of Dyer’s method for trimethylamine determination in codfish. In: Fish Inspection and Quality Control (edited by R. Kreuzer). Pp. 187-190. London (UK): Fishing News Books Ltd.
Undeland, I., Hultin, H. & Richards, M. (2003). Aqueous extracts from some muscles inhibit hemoglobin-mediated oxidation of cod muscle membrane lipids. Journal of Agricultural and Food Chemistry, 51, 3111-3119.
Vázquez-Ortiz, F., Pacheco-Aguilar, R., Lugo-Sánchez, M. & Villegas-Ozuna, R. (1997). Application of the freshness quality index (K value) for fresh fish to canned sardines from north-western Mexico. Journal of Food Composition and Analysis, 10, 158-165.
Wetterskog, D. & Undeland, I. (2004). Loss of redness (a*) as a tool to follow hemoglobin-mediated lipid oxidation in washed cod mince. Journal of Agricultural and Food Chemistry, 52, 7214-7221.
Whittle, K., Hardy, R. & Hobbs, G. (1990). Chilled fish and fishery products. In: Chilled foods. The state of the art (edited by T. Gormley). Pp. 87-116. New York (USA): Elsevier Applied Science.
Yamada, M., Fukusako, S. & Kawanami, T. (2002). Performance analysis on the liquid-ice thermal storage system for optimum operation. International Journal of Refrigeration, 25, 267-277.
Figure 1: Trimethylamine-nitrogen (TMA-N) content (mg kg-1 muscle)* in canned salmon muscle that was previously stored under different icing conditions**
* Mean values of four independent determinations (n=4). Standard deviations are indicated by bars. Raw fish values (FI and SI, respectively): 0.5±0.2 and 0.5±0.1.
** Previous icing conditions: FI (flake ice) and SI (slurry ice).
Figure 2: Free fatty acid (FFA) content (g kg-1 lipid)* in canned salmon muscle that was previously stored under different icing conditions**
* Mean values of four independent determinations (n=4). Standard deviations are indicated by bars. Raw fish values (FI and SI, respectively): 1.7±0.1 and 1.6±0.7.
** Previous icing conditions: FI (flake ice) and SI (slurry ice).
Figure 3: K value (%) assessment* in canned salmon muscle that was previously stored under different icing conditions**
* Mean values of four independent determinations (n=4). Standard deviations are indicated by bars. Raw fish values (FI and SI, respectively): 6.4±0.4 and 7.3±0.7.
** Previous icing conditions: FI (flake ice) and SI (slurry ice).