See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/249667618

Improvement in the Production of Smoked Trout Fillets (Salmo Gairdnerii)

Steamed with Liquid Smoke

Article in Food Science and Technology International · February 2008

DOI: 10.1177/1082013208090077

CITATIONS READS

15 317

4 authors, including:

Despina Dimitriadou Dimitris Petridis

National and Kapodistrian University of Athens International Hellenic University
2 PUBLICATIONS 15 CITATIONS 155 PUBLICATIONS 2,535 CITATIONS

SEE PROFILE SEE PROFILE

K.D.A. Taylor
University of Lincoln
52 PUBLICATIONS 3,285 CITATIONS

SEE PROFILE

All content following this page was uploaded by Dimitris Petridis on 07 May 2018.

The user has requested enhancement of the downloaded file.

 Food Science and Technology
International http://fst.sagepub.com/

Improvement in the Production of Smoked Trout Fillets (Salmo Gairdnerii) Steamed with Liquid Smoke
D. Dimitriadou, A. Zotos, D. Petridis and A.K.D. Taylor
Food Science and Technology International 2008 14: 67
DOI: 10.1177/1082013208090077

The online version of this article can be found at:

http://fst.sagepub.com/content/14/1/67

Published by:

http://www.sagepublications.com

On behalf of:

Consejo Superior de Investigaciones Científicas (Spanish Council for Scientific Research)

Additional services and information for Food Science and Technology International can be found at:

Email Alerts: http://fst.sagepub.com/cgi/alerts

Subscriptions: http://fst.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://fst.sagepub.com/content/14/1/67.refs.html

>> Version of Record - Apr 22, 2008

What is This?

Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012

 Improvement in the Production of Smoked Trout Fillets
(Salmo Gairdnerii) Steamed with Liquid Smoke
D. Dimitriadou,1 A. Zotos,1,* D. Petridis1 and A.K.D. Taylor2
1
Alexander Technological Educational Institution of Thessaloniki (ATEI)
School of Food Technology and Nutrition, Department of Food Technology
57400 Thessaloniki, P.O. Box 141, Greece
2
Faculty of Technology, University of Lincoln, Bayford Pool, Lincoln LN6 7TS, UK

A fish smoking process was applied using a combination of liquid smoke and steaming at pressures up to
1 bar above atmospheric pressure. Drying and brining prior to smoking have not shown any significant
effect on the quality of the trout fillets. The nondried and dried for 4 h trout fillets prior to processing were
assessed as slightly more acceptable products, processing yield, sensory analysis, instrumental color,
pH, available lysine, polycyclic aromatic hydrocarbons (PAHs) and preservation time were estimated.
The losses due to processing of the nondried samples were 23.7%, while 28.8% for those dried for 4 h. The
fillets processed at 1.5 and 2 bar steam pressure were assessed as highly acceptable. Lightness (L*) showed a
decreasing trend due to pressure while redness (a*), and yellowness (b*) an increasing one. The destruction
of available lysine was low (13.20%
1.01) and it was slightly dependent upon the process. No PAHs were
detected. The preservation time, studied via total viable count, lactic acid bacteria, existence of Listeria
monocytogenes, and sensory analysis, extended to more than 98 days at 4
1 8C. The drying prior to
processing seems to affect the preservation time. This processing technique is much simpler, faster,
environmental effective, and leads to high quality smoked products.

Key Words: smoked trout, steaming, color, total viable count, lactic acid bacteria

INTRODUCTION brown colored and smoke flavored than the samples

dried for 2, 4, 6, and 8 h (Siskos et al., 2005).
The use of liquid smoke has several advantages over It is well known that lysine participates with the
traditional smoking procedures having no detectable e-amino group in the early steps of Maillard reaction
levels of benzo[ ]pyrene and no mutagenic activity. pathways and the rate of the reactions increases with
Liquid smoke performs all the desired functions, allows temperature. Carbonyl compounds present in smoke or
more rigid flavor control and has the added advantages arising from lipid oxidation may react with the e-NH2
of lowering costs, less environmental damage and grea- group of lysine. The impact of these reactions is a
ter availability and variety of application methods reduction in protein utilization mainly by the reduction
(Dillon et al., 1994). of lysine through its e-amino group (Opstvedt, 1989).
Drying prior to smoking is a common process and An average reduction of available lysine at 21.1
8.4%
extensively used. Significant differences were found on was observed in steamed with liquid smoked trout fillets
flavor and color between the trout fillets that were dried and it was dependent upon the process (Siskos et al.,
at 20 8C for 2, 4, 6, 8, 16 and 24 h prior to steaming for 2005).
60 min at 2 bar pressure. The trout samples dried for 16 One of the major problems with traditional smoking
and 24 h prior to processing were assessed as more is the occurrence of polycyclic aromatic hydrocarbons,
which are well established as carcinogens (Lawrence
and Weber, 1984). It was reported a total amount of
polyaromatic hydrocarbons ranged between 200 to
480 mg/kg in liquid smoked products and 650 to
*To whom correspondence should be sent
1200 mg/kg in traditionally cold smoked fillets (Hattula
(e-mail: [email protected]).
Received 7 February 2007; revised 21 March 2007.
et al., 2001). A concentration of 0.63–3.2 ng/g polycyclic
aromatic hydrocarbons (PAHs) in steamed with liquid
Food Sci Tech Int 2008;14(1):67–77 smoke trout was detected (Siskos et al., 2005).

SAGE Publications 2008 The shelf life of smoked fish products depends largely
Los Angeles, London, New Delhi and Singapore
ISSN: 1082-0132 upon the initial bacterial contamination of the raw
DOI: 10.1177/1082013208090077 material; on the extent of brining and pre-drying, on the
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012 67
68 D. DIMITRIADOU ET AL.

heat treatment, on the amount of smoke components precision and 40 mL of liquid smoke condensate
that penetrate the product and on temperature, air (Nefeloudis SA) was added and diluted in 2 L tap water
humidity and oxygen levels during storage (Siskos et al., (2%). Therefore, the smoking process took place by the
2007). According to Siskos et al. (2007) total variable smoky steam produced by the liquid smoke solution.
count (TVC) values of 7  106 cfu/g, 5.1  106 cfu/g and Samples were placed at one layer with the flesh facing the
1.4  106 cfu/g reached in liquid smoked trout fillets, smoke liquid and were processed at three different
dried for 16 h prior to smoking and steamed at 2 bar processing times: 30, 45 and 60 min. Processing time
pressure for 30, 45, and 60 min, respectively, after was measured from the moment the required pressure in
48 days of storage at 4
1 8C. the steamer was achieved (12–15 min). The applied
Thus, the aim of this work was to produce smoked steam pressure conditions were: 1, 1.5, and 2 bar and the
trout fillets (Salmo gairdnerii) using a new steaming subsequent temperatures were 100, 104
1, and
process, based on liquid smoke and supported by the 113
1 8C, respectively. After each process the liquid
application of minor pressures and to assess the quality smoke solution was discarded and a new solution was
of the products via sensory characteristics, available used. After cooling at room temperatures, at sanitary
lysine, and PAHs, with a previous estimation of the conditions, the smoked products were packed in poly-
effect of drying and brining times prior to processing. ethylene bags and stored at 4
1 8C until further
Besides to estimate the preservation time at 4
1 8C of analysis. The combination of the two factors mentioned
the products via TVC, lactic acid bacteria (LAB), exis- above resulted in nine different processing levels. These
tence of Listeria monocytogenes, and sensory analysis. combined with the two drying levels (dried and non-
dried), resulted in 18 different ways of processing.

MATERIALS AND METHODS Methods

Samples Weight Loss

Drying and Brining Times during Pre-smoking All samples were weighed before and after all pro-
of Trout Fillets cessing steps. After processing fillets were left to equi-
librate at room temperature before weighting.
The initial step of this work was to find out the most
suitable drying and brining times. Trout purchased from Proximate Analysis
the fish farm (mentioned below) were brined in 20%
sodium chloride (NaCl) solution (four fillets in 1 L of Moisture content was determined by the Commission
brine), at four different brining times (1, 2, 3, and 16 h). of European Communities recommended method ISOR
The temperature of the brine was kept low (4
1 8C) in 1442 (EEC,1979). Lipid content was determined by the
order to minimize microbiological growth. Bligh and Dyer (1959) method as modified by Hanson
After brining, some samples were processed immedi- and Olley (1963). Total protein (crude protein, N x 6.25)
ately and all others were dried in an oven, at 43 8C, content was determined using the Kjeldahl method
using four different drying times (1, 2, 4, and 14 h). All according to Cowie and Mackie (1968). Salt content was
fillets were steamed (using liquid smoke solution) at the determined by the volumetric method of the AOAC
same conditions (smoking time 1 h and pressure 2 bar). (1995).
After cooling [1 h at ambient temperature 17
2 8C]
the smoked products were packed in polyethylene bags Available Lysine Analysis
and stored at 4
1 8C.
The available lysine was determined using the HPLC
Steamed Trout Fillets method with the FDNB (1-fluoro-2,4-dinitrobenzene)
reagent, described by Peterson and Warthesen (1979).
Farmed trout (S. gairdnerii) samples was purchased The HPLC was operated using: Microbondapac C18
from an aquaculture farm in Ioannina Greece. column, mobile phase 0.01M acetate buffer pH 4.0 and
Immediately after the arrival, fish samples were stored acetonitrile (80 : 20), flow rate 2.0 mL/min, detection
in ice at 0–1 8C. The weight of the samples was between at 435 nm, and injection volume 10 mL.
400 and 600 g. About 24 h after harvesting trout samples
were eviscerated, head removed, filleted and brined in Polycyclic Aromatic Hydrocarbons Analysis
20% NaCl solution for 1 h at 4
1 8C. After brining half
of the samples were processed immediately and the other Polycyclic aromatic hydrocarbons content was
half were dried for 4 h at 43 8C. The smoking process was determined using the method described by Larsson
conducted using a 10 L electrical commercial steamer, (1982). This method includes digestion in methanolic
electronically modified in the workshop to improve its solution of potassium hydroxide, partitioning in
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012
 Smoked Trout Fillets with Liquid Smoke 69

dimethylsulfoxide, and a column clean up procedure. trout fillets. In this sensory analysis 10 experienced
The chromatogram was obtained using a Focus GC panellists were used every 7 days, during 98 days. They
(Thermofinigan Italia SpA) packed with a capillary 30 m were asked to estimate three attributes (presence of
column purchased from Allthech with film thickness of external moisture, firmness, and odor) of the samples
0.25 mm and ID 0.32 mm. The GC conditions were: which have been stored at 4
1 8C. They were also
initial temperature 180 8C for 3 min, rate 8 8C/min until asked, based upon the above sensory attributes, to judge
320 8C and 20 min isothermal. The injection and the suitability of the samples to be consumed.
detection temperatures were 270 8C and 350 8C,
respectively. The carrier gas was helium with a flow rate Total Viable Count and Lactic Acid Bacteria Measurement
of 1.5 mL/min.
Ten grams of each sample were diluted in 90 mL
Determination of pH peptone water 1% buffered at pH 7.5 with 9% NaCl.
Serial dilutions were made until a 105 g/mL sample was
The pH of trout flesh was measured with a Hanna obtained. One mL of each dilution was placed in a Petri
Instruments HI 8424 microcomputer pH-meter. Samples dish and 15 mL plate agar count was added for TVC
were prepared according to Cortes-Ruiz et al. (2001) by and 15 mL de Man Rogosa Sharpe (MRS) agar for
blending 2 g of trout mince with 18 mL of distilled water. LAB. Each Petri dish was carefully shaken in order to
achieve a homogenous distribution of the sample. After
Instrumental Color Measurement about 10 min all Petri dishes were inverted and placed in
an oven at 35 8C for 48 h for TVC and at 30 8C for 36 h
Color was measured directly on the flesh of the fish for LAB. All developed colonies were counted following
fillets. The color measurement was conducted using a the rules reported by Busta et al. (1984).
color difference-measuring instrument (‘Micro Color
LMC’ colorimeter, Dr Lange, Germany) which is a AQ3 Listeria Monocytogenes
tristimulus colorimeter (Siskos et al., 2005). Each mea-
surement was repeated nine times for each sample. Aseptically 25 g of the sample were added in 225 mL
L* (lightness/darkness), a* (redness/greenness), and b* of half-fraser broth (suitable for Listeria monocytogenes
(yellowness/blueness) values were measured. growth). It was stomachered and incubated for 25 h
at 30
1 8C. After incubation 1 mL of the suspension
Sensory Analysis was transferred into 10 mL of Fraser broth and it
was incubated for 25 h at 30
1 8C. Total of 500 mL
Sensory analyses were conducted at three different of the incubated samples were pipetted into the strips.
stages. The first one in order to assess the trout fillets The strips were inserted in the mini-VIDAS instrument,
regarding their different pre-treatment conditions which is capable to determine the presence or absence of
(drying and brining times). This sensory analysis was Listeria monocytogenes in the samples. The assay was
performed using 21 experienced panellists. Five samples completed within approximately 70 min.
were presented to panellists (2 days after processing)
asking to evaluate four sensory variables: color of the Statistical Analysis
flesh, saltiness, smoked flavor, and firmness. The
intensity of all four variables was recorded on a 15 cm The estimation of drying and brining times were
unstructured scale line. The left end of the line was performed by the application of a balanced incomplete
marked at 0 cm for all variables and characterized as block design (BIBD) including t ¼ 21 treatments, r ¼ 5
white color, nonsalted, nonsmoked flavor, and extre- replicates per each process with only one pair of similar
mely soft, whereas the right end was marked at 15 cm as samples l ¼ 1, k ¼ 5 treatments per panellist and b ¼ 21
brown color, extremely salted, strong smoked flavor, panellists. The samples in this experiment were 20 (five
and extremely firm. Samples were coded with three-digit different drying times  four brining times), therefore
random numbers, and served in random order. the design was increased with one more liquid smoked
The second sensory analysis was performed in order trout fillet (5  4 þ 1) in order to conform to the parti-
to assess the different processed trout fillets regarding cular plan 13.13 obtained by Cochran and Cox (1957).
their drying, pressure, and processing time. Twenty nine Adjusted sensory mean scores were deduced for the
experienced panellists were used twice apart from one 21 samples for each process and at this point the 21th
(b ¼ 57) (see below statistical analysis), testing three sample was excluded from further investigation and
samples each time. Color, flavor, and firmness (three were then analyzed using a two-way ANOVA (two fixed
variables) were assessed as well as the acceptability factors: drying and brining time).
regarding these variables. A BIBD was performed for the estimation of the
The third sensory analysis was performed in order to different processed trout fillets on factors dried/non
estimate the preservation time at 4
1 8C of the smoked dried (2 levels), pressure conditions: 1, 1.5 and 2 bar
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012
70 D. DIMITRIADOU ET AL.

(3 levels) and processing times: 30, 45, and 50 min content, indicating that small changes should be
(3 levels). Thus a combination of 2  3  3 ¼ 18 samples probably expected from farmed fish.
was achieved and the design was again increased with
one more sample of liquid smoked trout (2  3  3 þ 1) Optimisation of Drying and Brining Times
in order to conform to the particular plan 13.15a
obtained from Cochran and Cox (1957). This design The sensory analysis performed on the liquid smoked
included t ¼ 19, r ¼ 9, l ¼ 1, k ¼ 3 and b ¼ 57. Adjusted trout fillets to estimate the drying and brining times
sensory mean scores were deduced for the 19 samples revealed significant differences only in the saltiness and
for each process and at this point the 19th sample was firmness of the products. Saltiness was affected by either
excluded from further investigation. A three-way different brining ( p ¼ 0.000) or drying time ( p ¼ 0.040)
ANOVA for the three factors and their interaction while firmness only by the brining time ( p ¼ 0.007,
terms was attempted on the set of sensory variables to Figures 1A and B). The panellists detected that saltiness,
detect significant differences. as expected, was highly influenced by the brining time
The two preserved trout fillets (0 and 4 h drying time (Figure 1A). As far as the drying time concerns, only the
prior to processing) were assessed by the panellists samples that were not dried prior to processing seemed
14 times, using every time an additional newly prepared to be different and identified as less salty than all others
smoked sample (as a control) unknown to the panellists (Figure 1B). This should be expected since it was the
in order to identify possible differences. The results were samples with lower water losses due to processing, as
analyzed using a two-way ANOVA (two fixed factors: shown below. Regarding the firmness of the samples,
drying and preservation time) panellists detected that it increased with the increase of
Statistically significant differences between factor brining time, and it was not influenced by the drying
levels were tested using the Tukey test for comparison time, despite the long period used in the work (16 h).
of level mean values. However, only the samples salted for 16 h were iden-
tified as statistically significant different from all others
(Figure 1A).
Regarding smoked flavor and color no significant
RESULTS AND DISCUSSION differences were observed due to different pre-treat-
ments (drying and brining). The results for the L*, a*,
The chemical composition of the samples was and b* color parameters were similar, with no significant
relatively stable through out the study: 71.79%
0.30 differences due to different pre-treatments but they were
moisture, 23.20%
0.35 protein, and 4.23%
0.33 lipid increased due to processing (steamed with liquid smoke

(A) (B)

11
8.5

10
8.0
Average sensory scores

9
Saltiness

7.5
8

7.0
7

6 6.5

5
6.0
0 2 4 6 8 10 12 14 16 0 1 2 4 14
Brining time (h) Salting time (h)
Figure 1. (A) Effect of the different brining time prior to processing on sensory scores of saltiness (m) and
firmness (g) of the steamed trout fillets with liquid smoke at 2 bar for 1 h. (B) Effect of the different drying time prior
to processing on sensory scores of saltiness of the steamed trout fillets with liquid smoke at 2 bar for 1 h.
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012
 Smoked Trout Fillets with Liquid Smoke 71

at 2 bar for 1 h, Table 1). The high increase of a* and b* under different steam pressure conditions ( p ¼ 0.002).
values due to smoking indicated the extension of brown- As the pressure increased, yield losses were also
ing reactions occurred in the process, leading in well- increased. During smoking at atmospheric pressure
browned (yellow/red) smoked products. Temperature yield losses were not more than 20% and became
and time of heating are known to strongly influence higher (more than 30%) when more intense conditions
the color properties of food, which develops faster as the were used (Figure 2). Although, yield losses seemed to
product surface temperature increases (Maga 1988). be depended on the different applied pressures, no sig-
These results indicate that the drying process prior to nificant differences were identified between the trout
steaming is not an important factor and it does not fillets treated at 1.5 bar and 2.0 bar pressure conditions.
influence the quality regarding color and flavor of the The mean losses for the fillets that have not been pre-
steamed (with liquid smoke solution) products. Thus, viously dried were 23.7% due to processing (smoking)
the 16 h drying process prior to smoking reported by while the pre-dried fillets showed an average of 28.8%
Siskos et al. (2005) for better colored smoked trout which came from either the drying (9%) or the smoking
fillets it should be probably attributed to the homo- (19.8%) processes. These results indicate that the non-
genation technique used to measure the color of the dried prior to processing trout fillets lead to better
samples. quality regarding yield losses smoked products.
Subsequently, evaluating the above results and con-
sidering the hygiene of the final smoked products, the Effect of Processing on Moisture, Protein
1 h in 20% NaCl solution was decided as a more appro- and Lipid Content
priate brining time. As far as drying time concerns, no
significant differences was observed between the samples Moisture content in all samples varied between 63.5%
produced at different times and therefore two different and 69.5% (Table 2). Moisture was influenced only
drying times (0 and 4 h) were chosen. The drying time of by the drying prior to processing, where statistically
4 h was chosen for a further study on the effect of significant differences were observed ( p ¼ 0.016).
drying, while the 0 h drying time prior to smoking leads
in a quick and economic process, with probably better
quality final products. The salt content of these samples 35
was 1.79% for the nondried fillets and 2.12% for the 4 h
dried ones. Low salt content (1–2%) has been reported
30
Weight losses (%)

that improve the yield and the liquid holding capacity of

fish muscles (Sigurgisladottir et al., 2000).
25
Yield Losses due to Drying (0, 4 h), Pressure and
Processing Time 20

During the brining process, fillets were soaked in a

liquid solution so that water and NaCl diffused in the 15
samples; therefore the increase of 2.9
1.2% in weight 30 45 60
found it should be expected. These results were similar Smoking time (h)
with those reported by Siskos et al. (2005). They found Figure 2. Weight losses of steamed trout fillets with
an increase of 1.3
1% in weight due to brining of the liquid smoke respect to time of processing and
trout fillets. Statistical analysis revealed significant pressure conditions regardless the drying process
differences between the dried and nondried fillets prior to processing. Pressure (bar): 1.0 (f), 1.5 (g),
( p ¼ 0.010) and between those that were processed and 2.0 (^).

Table 1. The effect of drying and brining times prior to processing at 2 bar for 1 h on L*, a*
and b* values of steamed trout fillets with liquid smoke.
Drying time (h)
Color
parameter Unprocessed 0 1 2 4 14
L* 41.36 (3.22) 53.29 (3.14) 54.00 (2.29) 55.58 (3.26) 53.95 (3.37) 48.95 (4.38)
a* 3.76 (0.97) 5.03 (1.29) 4.78 (3.11) 3.81 (1.41) 3.23 (1.86) 5.47 (2.75)
b* 5.00 (0.97) 21.87 (1.18) 22.40 (1.18) 21.62 (1.41) 20.36 (2.50) 21.50 (2.62)

The values are means of 36-fold (4 brining times  9 measurements) fold determinations. SD are shown in parentheses.

Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012

72 D. DIMITRIADOU ET AL.

Thus, a decrease of 8.90% in moisture content was by the drying prior processing ( p ¼ 0.004), reversely to
detected in the pre-dried for 4 h samples while the protein content. Increase of lipid content due to pro-
reduction in moisture content in the samples processed cessing was also reported by Garcia-Arias et al. (1994).
without prior drying was only 5.65%. Neither the Steamed samples showed a higher lipid content mainly
smoking time nor the applied pressure was shown to due to the water loss during steaming. However, the
have any significant effect in moisture content. The increase lipid content in steamed samples cannot be only
combined treatments used in this study led to the explained by water loss, without ruling out other pos-
production of trout fillets with relatively high moisture sible factors in sample composition, such as loss of
content. A moisture reduction of 21.2% was reported by nitrogen as amino acid and other nitrogen compounds.
Goulas and Kontominas (2005), during hot smoking The increase of lipid content in this work should be
(30 min at 40 8C, 30 min at 50 8C, 30 min at 55 8C, arisen by either water loss or mainly the reduction of
30 min at 60 8C and 30 min at 70 8C) of club mackerel protein content due to processing.
(Scomber japonicus) fillets. These results confirm the
results of the yield losses and the effectiveness of the
Effects of Processing on pH
process and particularly of the nondried prior to
processing trout fillets. The initial pH of the untreated fillets was 6.50
0.14.
Protein content was also significantly influenced The pH values of the samples were not significantly
( p ¼ 0.013) by the drying time. Higher proteic material influenced by the different processing conditions
reduction (16.14%) was observed in the trout fillets (Table 3). The slight decrease in pH observed was pro-
processed with 4 h drying time prior to processing, while bably due to reduction of moisture content, to brining
this reduction in the samples processed without any pre- prior to processing as well as to the high temperature of
drying treatment was 10.42% (Table 2). Castrillon et al. the process. These slight changes in pH values were in
(1996) found a reduction in proteic material of 8.5% in agreement with those reported by Espe et al. (2002).
steamed and canned sterilized for 55 min tuna samples. Similar differences have been also noticed by Goulas
Additionally, when the canning process was extended to and Kontominas (2005) which were ascribed to the
90 min the protein content reduction was 13.6%. These lower moisture content of the smoked samples and to
results indicated the significance of the processing time, the brining process.
which is a crucial factor of protein loss.
The percentage lipid content increased inversely to
protein content due to processing. The average lipid Effect of Processing on Available Lysine
content in all samples was 19.83 with 16.16% minimum
and 23.71% maximum (on dry weight free salt basis, The reduction in lysine availability was low (Table 4)
Table 2). Lipid content was significantly affected only indicating a slight nutritional effect in the samples. The
reduction showed a possible slight trend due to all
different parameters considered in the study. However,
Table 2. Protein, lipid content (on dry weight free salt no statistical differences were identified between the
basis), and moisture content, of the unprocessed and samples and the average value was 13.20
1.01%.
processed at different conditions steamed with liquid Losses in available lysine have been reported at 50 to
smoke trout fillets. 63% due to hot smoking of mackerel (Zotos et al. 1995).
Unprocessed Nondried Dried
Losses of 21.1%
8.4 in available lysine were also
reported by Siskos et al. (2005). They also found that the
Moisture (%) 71.79 (0.30) 67.73 (1.17) 65.40 (1.63)
drying time prior to processing significantly affected the
Protein (%) 82.89 (0.35) 74.25 (5.60) 69.50 (2.10)
Lipid (%) 15.11 (0.33) 18.53 (1.88) 21.11 (1.83) destruction of available lysine. This once more confirms
the effectiveness of these processes and particularly of
The values are means of 3-fold determinations for the unprocessed samples the nondried prior to processing trout fillets in the
and 27-fold determinations for the processed ones. SD are shown in
parentheses. quality of the smoked products.

Table 3. pH changes of steamed trout fillets with liquid smoke trout due to different pressures (1, 1.5, and 2 bar)
processing times (30, 45, and 60 min) and drying (0, 4 h) at 43 8C prior to processing.
Pressure (bar)
1 1.5 2
Sample 30 (min) 45 (min) 60 (min) 30 (min) 45 (min) 60 (min) 30 (min) 45 (min) 60 (min)
Nondried 6.40 (0.14) 6.50 (0.07) 6.57 (0.40) 6.40 (0.14) 6.58 (0.23) 6.40 (0.14) 6.30 (0.03) 6.28 (0.35) 6.24 (0.10)
Dried 4 h 6.53 (0.01) 6.39 (0.08) 6.28 (0.06) 6.37 (0.08) 6.30 (0.03) 6.43 (0.06) 6.33 (0.06) 6.22 (0.06) 6.22 (0.07)

The values are means of three fold determinations. SD are shown in parentheses.

Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012

 Smoked Trout Fillets with Liquid Smoke 73

Effects of Processing on PAHs not statistically significant different. The effect of pres-
sure on smoked flavor and color of trout fillets was also
No detectable amounts of any PAHs were found reported by Siskos et al. (2005). They found that all
either in smoke condensates or as expected in the fillets that were processed at 2.0 bar showed more
steamed (with liquid smoke solution) trout fillets using intense smoked flavor and were browner in color. This
an FID detector. should be expected since the higher the applied pressure
the higher the temperatures and consequently the higher
Sensory Analysis and Instrumental Color the production of carbonyl and phenolic compounds
that led to more colored and flavored final products,
The steam pressure was the only factor that influenced respectively.
all determined variables ( p50.05) (Figure 3A and B). The drying factor was significant only for firmness of
Drying prior to smoking process and smoking time the final products. The dried trout fillets were more the
influenced only firmness and firmness acceptability. nondried ones. Firmness and firmness acceptability were
The significance of the main effects was tested through also influenced by the time of processing. The firmer
a three-way ANOVA in all measured variables. Thus, smoked products were those that processed for 45 min,
the smoked color intensity and acceptability, the while the smoked products processed for 30 min and
smoked flavor intensity and acceptability, and the firm- 60 min were quite similar in firmness as assessed by the
ness and its acceptability seemed to increase with the panellists. The trout fillets processed for 45 min at 2 bar
increase of steam pressure conditions (Figures 3A and B). steam pressure were assessed as the most acceptable
However, the differences between the trout fillets products regarding color, flavor, and firmness by the
processed at 1.5 and 2.0 bar pressure conditions were panellists.

Table 4. Available lysine content and losses of steamed with liquid smoke trout fillets dried at 0 and 4 h
at 43 8C prior to processing.
Time (min) Pressure (bar)
Unprocessed 30 45 60 1 1.5 2
Protein (mg/g) 70.80 (0.85) 62.31 (3.63) 61.44 (3.47) 60.60 (3.50) 62.29 (4.90) 61.12 (3.66) 60.94 (0.70)
Loss (%) 12 13.2 14.4 12.0 13.7 13.9

Data are means of 6- fold determinations. SD are shown in parenthesis.

(A) (B)
12
11

10 11

9 10
Average sensory scores

8
9

7
8
6
7
5
6
4
5
3

2 4
1.0 1.5 2.0 1.0 1.5 2.0
Processing pressure (bar) Processing pressure (bar)

Figure 3. (A) Changes sensory scores of color (f), smoked flavor (g), and firmness (^) of the steamed trout
fillets with liquid smokedue to increase of steam pressure. (B) Changes in their acceptability regarding color (f),
smoked flavor (g), and firmness (^).
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012
74 D. DIMITRIADOU ET AL.

Table 5. Changes in L*, a*, and b* color values of steamed trout fillets with liquid smoke due to
different pressures (1, 1.5, and 2 bar), processing times (30, 45, and 60 min), and drying (0, 4 h)
at 43 8C prior to processing.
Pressure (bar)
1 1.5 2
30 (min) 45 (min) 60 (min) 30 (min) 45 (min) 60 (min) 30 (min) 45 (min) 60 (min)
L*
Nondried 63.27 (3.34) 64.38 (1.84) 61.28 (3.97) 57.46 (6.48) 51.87 (3.45) 55.59 (3.01) 57.33 (3.80) 55.25 (2.26) 54.46 (4.05)
Dried 4 h 60.85 (4.12) 64.16 (4.80) 63.25 (5.61) 56.62 (3.95) 51.60 (3.42) 53.04 (4.57) 55.55 (2.29) 50.99 (3.86) 54.71 (3.08)
a*
Nondried 2.56 (0.72) 1.76 (0.92) 0.52 (1.00) 1.04 (3.90) 5.94 (3.22) 3.85 (3.34) 1.66 (4.14) 3.670 (3.34) 5.51 (3.14)
Dried 4 h 3.58 (1.56) 2.00 (0.87) 1.22 (1.15) 1.26 (2.01) 3.56 (2.73) 2.03 (2.74) 2.66 (3.02) 5.85 (3.09) 4.15 (2.45)
b*
Nondried 12.93 (2.45) 14.48 (2.83) 15.58 (2.50) 15.09 (6.73) 23.44 (3.99) 19.30 (3.13) 18.33 (3.14) 23.24 (4.92) 21.81 (3.31)
Dried 4 h 10.64 (3.61) 12.65 (3.56) 11.20 (2.54) 17.64 (3.38) 16.31 (5.59) 18.27 (2.76) 22.82 (4.28) 24.65 (5.21) 22.97 (5.39)

The values are means of nine fold determinations. SD are shown in parentheses.

The different steam pressure conditions also affected observed (25 cfu/g). This was mainly because the
the instrumentally measured color parameters. L* temperature was between 105 and 113
1 8C, too high
values decreased with the increase of pressure 1 to 1.5 for all Gram-positive cocci predominant on smoked
or 2 bar (Table 5). Conversely a* and b* values were fillets to survive. This microbiological flora remained
increased (Table 5). However, no significant differences stable in all samples for 7 days of storage at 4
1 8C.
were observed in all color parameters between the A slight increase at 6  102 cfu/g of TVC was observed
smoked trout fillets processed at 1.5 and 2.0 bar on the 14th day in the samples processed without the
pressure. As far as a* parameter concerns, it was also drying pre-treatment.
affected by the different smoking times ( p ¼ 0.020). However, this increase disappeared on the 21st day
Redness (a*) of the samples became higher at the indicating that it came from a probable contamination
processing time of 45 min and was similar with the a* during determination. An exactly similar contamination
values at the processing time of 60 min. was also found for the LAB.
The L*, a*, and b* values found in this work were A slight increase in TVC values from 35 cfu/g to
different from those reported by Siskos et al. (2005). 1.1  102 cfu/g was observed only in the smoked fillets
They reported a decrease in L* values while no signi- that were dried for 4 h at 43 8C prior to smoking after 91
ficant differences were detected for a* and b* values. days storage at 4
1 8C. The TVC values in the smoked
This should be probably attributed to the different trout fillets produced without any drying pre-treatment
applied technique by the authors. In this work the color were still low (35 cfu/g), even after 91 d at 4
1 8C.
was measured directly on the flesh, while they measured On the 98th day, the microbial numbers were only
color after homogenation of the flesh in order to avoid 40 cfu/g for the smoked trout fillets produced without
flesh porosity, however, this could be possibly involve any drying pre-treatment and 5.5  102 cfu/g for those
air bubbles in the flesh, leading in higher L* values. that dried for 4 h at 43 8C prior to processing. The high
temperatures (105 to 113
1 8C), and pressure (2 bar)
Estimation of Preservation Time at 4
1 8C used in this investigation reduced the initial microflora
and the spoilage that occurred was mainly due to con-
The samples processed without any previous drying tamination after processing. The higher TVC values
and those that were dried for 4 h at 43 8C prior to observed in the dried prior to processing samples should
smoking, were processed at 2 bar steam pressure (shown be probably attributed to possible bacteria contamina-
to give better smoked color, flavor, and firmer products tion occurred during the drying process.
and more acceptable regarding these attributes) for Very high TVC values were reported much earlier
45 min smoking time (which gave products similar to (max. 48 d) for the same products (trout fillets) by Siskos
those processed for 60 min), preserved at 4
1 8C and et al. (2007). They were processed their products with a
weekly studied for their microbiological and sensory prior drying time of 16 h at ambient temperature. These
changes for up to 98 days. results confirm that the drying process prior to pro-
cessing leads to a high bacteria contamination that
Total Viable Counts significantly affects the preservation time of the smoked
products.
Average TVC of fresh samples was 5.9  105 cfu/g. Storage of smoked fish has been already investigated
After processing a TVC reduction in all samples was by many researchers. Dondero et al. (2004) studied
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012
 Smoked Trout Fillets with Liquid Smoke 75
2
changes in quality of vacuum-packed, cold-smoked with drying pre-treatment increased at 1.8  10 cfu/g.
salmon (Salmo salar) during storage at different These results are in accordance with the results found in
temperatures. Total aerobic count at the beginning of sensory analysis and confirmed the results found for
storage ranged between 150 and 174  103 cfu/g. TVC. The results found in literature are totally different
According to sensory analysis salmon had a shelf life of with those found in this work. Leroi et al. (1998)
20 days at 4 8C. As it can be seen, the microbial loads reported that farmed salmon, salted in dry salt for 3 h
were much higher than those found in the present work. dried for 3 h at 22 8C and smoked at 22 8C for 3 h at
Longer shelf life was observed by Kolodziejska et al. relative humidity 70%, had immediately after smoking
(2002). They produced mild hot smoked Atlantic very low count of LAB. During the first week LAB
mackerel (57 8C for 85 min) from defrosted raw material increased to 15  105 cfu/g and reached a maximum
which was immersed in 200 g/L NaCl at 7 8C for 2.5 after 4 weeks of storage.
to 3 h. Drying took place at ambient temperature for
30 and 60 min. The samples stored at two different Listeria Monocytogenes
temperatures 2 and 8 8C. The initial microbial count was
5104 cfu/g. Storage of the smoked fish at 2 8C did not Listeria monocytogenes was not detected in any
affect the microbiological quality for 21 days. The sample. This is in accordance with the results reported
bacterial growth in the meat of such product kept at by Dondero et al. (2004) who also did not find Listeria
8 8C is not sufficiently retarded even during 7 days. monocytogenes during storage of cold-smoked salmon.
Comparing the results found in this work with those in However, Basti et al. (2006) while investigating bac-
literature it can be concluded that this simple and quick terial pathogens in fresh, smoked, and salted Iranian
smoking method, based upon one hygienic processing fish, detected Listeria monocytogenes in fresh, cultivated
method (steaming), followed by the absence of any pre- fish, smoked fish, and salted fish. The occurrence of
drying treatment could lead in high quality smoked Listeria monocytogenes in a farm environment was due
products with a preservation time up to 100 days. to the faeces of cow which commonly used for fertilizing
Iranian fish farms. They found the presence of the
Lactic Acid Bacteria organism in 51% of the cow faeces

Very similar to TVC, the LAB values in all samples Firmness

were very low (30 cfu/g). During storage at 4
1 8C the
count of these microorganisms remained almost stable. The firmness of all samples (Figure 4) gradually
On the 98th day the LAB in the smoked fillets produced increased as expected, probably due to dehydration
Moisture * Storage Firmness* Storage

11
5

4 10

3
9
2
Average sensory scores

8
1

0 25 50 75 100 0 25 50 75 100

Odor * Storage Color* Storage

11

10
10
9

8 9
7

6 8
0 25 50 75 100 0 25 50 75 100
Storage time (days)
Figure 4. Changes of surface moisture, firmness, odour, and smoked color due to preservation at 4
1 8C of
trout fillets steamed at 2 bar for 45 min of the nondried and dried for 4 h before processing.
Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012
76 D. DIMITRIADOU ET AL.

occurred during storage at 4
1 8C. The highest values ACKNOWLEDGMENTS

in firmness observed after 70 days of storage. Changes in
firmness due to storage were also reported by Leroi et al. The authors would like to thank the Ministry of
(1998). Smoked salmon slices, after 13 days of storage Education and Religious Affairs of Greece and the
were more pasty and fatty in texture. European Union (EU), which co-funded this work by 25
and 75%, respectively.
Smoked Odor

The smoked odor intensity gradually decreased

during storage (Figure 4). The smoked odor remained REFERENCES
nearly constant until 42 days of storage. It is important
AOAC (1995). Salt (Chlorine as Sodium Chlorine). Official
that even after 98 days of storage the panellists did not Method 937.09. In: Seafood. Volumetric Method. Arlington,
detect any off-odors. The samples were in a good VA: AOAC.
condition and according to the panellists, suitable for Basti A.A., Taghi A.M., Salehi Z. and Kamkar A. (2006). Bacterial
consumption. pathogens in fresh smoked and salted Iranian fish. Food Control
17: 183–188.
Bligh E.G. and Dyer W.J. (1959). A rapid method of total lipid
Color oxidation and purification. Canadian. Journal of Biochemistry
and Physiology 37: 911–917.
The brown color of all samples was gradually Busta F.F., Peterson E.H., Adams D.M. and Jonson M.G. (1984).
Colony count methods. In: Speck M.L. (ed.), Compendium of
increased during the preservation of the trout fillets at Methods for the Microbiological Examination of Foods.
4
1 8C (Figure 4). This increase in brownish occurred Washington: American Public Health Association, pp. 62–83.
until the 77th day. This should be probably due to the Castrillon A.M., Navaro M.P. and Garcia-Arias M.T. (1996).Tuna
further development of browning reactions between protein quality changes after canning. Journal of Food Science 61:
1250–1253.
either carbonyls from the smoke condensates or Cochran W.G. and Cox G.M. (1957). Incomplete Latin Squares.
oxidized lipids and proteins during storage. Experimental designs. New York: Wiley & Sons, pp. 528 and 542.
Cortes-Ruiz J.A., Pacheco-Aguilar R., Garc|a-Sanchez G. and Lugo-
Sanchez M.E. (2001). Functional characterization of a protein
Presence of Surface Moisture concentrates from bristly sardine made under acidic conditions.
Journal of Aquatic Food Product Technology 10: 5–23.
The presence of external surface moisture was Cowie W.P. and Mackie I.M. (1968). Examination of the protein
observed after 42 days of storage at 4
1 8C (Figure 4). extractability method for determining cold storage protein
From the 42nd day until the 98th a slight gradually denaturation in cod. Journal of Food Science and Agriculture
19: 696–700.
increase was observed quite similar with the increase Dillon R., Pattel T. and Martin A.M. (1994). Microbiological control
observed in firmness, confirming a gradual dehydration of smoking operations. In: Martin A.M. (ed.), Fisheries proces-
but quite slow. These results also indicate probable sing, Memorial University of Newfoundland. St. Johns’ Canada:
Chapman and Hall, pp. 51–81.
protein changes due to storage of the smoked trout
Dondero M., Cisternas F., Carvajal L. and Simpson R. (2004).
fillets. A much faster increase in external surface mois- Changes in quality of vacuum-packed salmon (Salmo salar) as a
ture during storage of trout fillets at 4
1 8C was also function of storage temperature. Food Chemistry 87: 543–550.
observed by Siskos et al. (2005) followed by an increase EEC. (1979). Commission of european communities. Method – ISQ
in TVC and a rejection of the samples. This should be 1442–1973.
Espe M., Nortvedt R., Lie O. and Hafsteinsson H. (2002). Atlantic
probably ascribed to the drying prior to smoking salmon (salmo salar) as raw material for the smoking industry. II:
process, which was 16 h. Effect of different smoking methods on losses of nutrients and on
Thus high quality smoked products regarding their the oxidation of lipids. Food Chemistry 77: 41–46.
color, flavor, firmness, weight loss, reduction of avail- Garcia-Arias M.T., Sanchez-Muniz J.F., Castrillon A.M. and Navaro
M.P. (1994). White tuna canning, total fat, and fatty acid changes
able lysine, PAHs, and preservation time were produced during processing and storage. Journal of Food Composition and
in this work using the steaming technique, well known Analysis 7: 119–130.
for its beneficial effect on cooking food. It was observed Goulas E.A. and Kontominas G.M. (2005). Effect of salting and
that the nondried prior to processing smoked trout smoking-method on the keeping quality of chub mackerel
(Scomber japonicus): biochemical and sensory attributes. Food
fillets were either better in quality or prolong their Chemistry 93: 511–520.
preservation (up to 100 days at 4
1 8C); therefore the Hanson S.W.F. and Olley J. (1963). Application of the bligh and dyer
drying process should be avoided since it does not method of lipid extraction to tissue homogenates. Biochemical
Journal 89: 101–102.
have any effect on the sensory characteristics of the
Hattula T., Elfving K., Mrouch U.M. and Luoma T. (2001). Use of
products. This processing technique is much simpler liquid smoke flavoring as an alternative to traditional flue gas
and faster (without drying prior to processing) than smoking of rainbow trout fillets (Oncorhynchus mykiss).
the traditional methods of smoking, it could be Lebensmittel -Wissenschaft und Technologie 34: 521–525.
home made and leads to high quality smoked products Kolodziejska I., Niecikowska C., Januszewska E. and Sikorski Z.E.
(2002). The microbial and sensory quality of mackerel hot
and environmentally effective since no smoke was smoked in mild conditions. Lebensmittel-Wissenschaft und
generated. Technologie 35: 87–92.

Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012

 Smoked Trout Fillets with Liquid Smoke 77
Larsson B.K. (1982). Polycyclic Aromatic Hydrocarbons in Peterson W.R. and Warthesen J.J. (1979). Total available lysine
Smoked Fish. Lebensmittel Unters Forschung 174: 101–107. determinations using high-pressure liquid chromatography.
Lawrence J.F. and Weber D.F. (1984). Determination of polycyclic Journal of Food Science 44: 994–997.
aromatic hydrocarbons in some Canadian commercial fish, Sigurgisladottir S., Sigurdardottir M., Tirrisen O., Vallet J. and
shellfish and meat products by liquid chromatography with Hafsteisson H. (2000). Effects of different salting and smoking
confirmation by capillary gas chromatography-mass processes on the microstructure, the texture and the yield of
spectrometry. Journal of Agricultural and Food Chemistry 32: Atlantic salmon (Salmo Salar). Food Research International 33:
789–794. 847–855.
Leroi F., Jorffaud J.J., Chevalier F. and Cardinal M. (1998). Study of Siskos I., Zotos A. and Taylor K.D.A. (2005). The effect of drying,
the microbiological ecology of cold-smoked salmon during pressure and processing time on the quality of liquid smoked
storage at 8 8C. International Journal of Food Microbiology 39: trout (Salmo gairdnerii) fillets. Journal of the Science of Food and
111–121. Agriculture 85: 2054–2060.
Maga J.A. (1988). Factors influencing rate of color formation. In: Siskos I., Zotos A., Melidou S. and Tsikritzi R. (2007). The effect of
Maga J.A. (ed.), Smoked and Food Color. Florida, Boca Raton: liquid smoking of fillets of trout (Salmo gairdnerii) on sensory,
CRC Press Inc., pp. 103–106. microbiological and chemical changes during chilled storage.
Opstvedt J. (1989). Influence of drying and smoking on protein Food Chemistry 101: 458–464.
quality. In: Burt J.R. (ed.), The Effect of Smoking and Drying on Zotos A., Hole M. and Smith G. (1995). The effect of frozen storage of
the Nutritional Properties of Fish. London and New York: mackerel (Scomber scombrus) on its quality when hot-smoked.
Elsevier Applied Science, pp. 23–36. Journal of the Science of Food and Agriculture 67: 43–48.

Downloaded from fst.sagepub.com at Technological Educl Inst on May 15, 2012

View publication stats