MULTIVARIATE QUALITY PREDICTION OF COD (GADUS MORHUA) AND HADDOCK FILLETS (MELANOGRAMMUS
AEGLEFINUS) STORED UNDER SUPERCHILLING AND TEMPERATURE ABUSIVE CONDITIONS
Gudrun Olafsdottir*, Hélène L. Lauzon, Rósa Jónsdóttir, EmilíaMartinsdóttir,
Icelandic Fisheries Laboratories, Skulagata 4, 101 Reykjavik, Iceland
2nd Workshop "Cold-Chain-Management”. Bonn,8-9th May 2006
OBJECTIVES
to achieve better understanding better understanding of the effect of temperature effect of temperature on the complex spoilagecomplex spoilage changes of fish
volatile compounds as quality indicatorsquality indicatorsGCGC (gas chromatography) analysischaracterize the spoilage potential of specific spoilage organisms (SSO) specific spoilage organisms (SSO)
to apply an electronic noseelectronic nose as a rapid rapid technique to monitor classes of volatile compounds as indicators of quality changesquality changesduring storage of chilled fish
storage studies of chilled cod and haddock fillets – natural productsexplore the correlation of microbial, chemical, sensory and emicrobial, chemical, sensory and e--nose nose data by using multivariate models (PLSR / PCA (PLSR / PCA -- SIMCA)SIMCA)predict the quality or classify products according to sensory criteriasensory criteria.
Various factors influence the spoilage of fish
Species cod and haddock
Seasonal condition autumn
Fishing grounds Northeast and Southwest Iceland
Catching methods longline / bottom trawl
HandlingFlake ice / slurry iceSuperchilling CBC (contact blast and cooling)
Processing skinless fillets
Storage techniques styrofoam boxes
Time / temperature -1,5 to 15°C
Protein / i.e. sarcoplasmic proteins → peptides→ amino acids
Lipids / i.e. phospholipids → PUFA
Soluble substances in the muscleNucelotidesNPN non protein nitrogenous components TMAO → TMA/DMA, FA, pH ↑
Glycolysis pH↓, Lactic acid↑
Rigor mortis ATP→ Inosine →Hx →Urea
Autolysis
Endogenous enzymes lipoxygenase, cathepsin, calpain
Microbial growthSSOSpecific SpoilageOrganisms
Oxidationpro- and antioxidants
Fresh odor
Spoilage odor
Complex spoilage changes in fish
Texture firm – soft
Color changes
Identify Identify qualityquality indicatorsindicators
QUALITY INDICATORS in chilled fishSpoilage odors Volatile microbial metabolites Specific Spoilage Organisms (SSO)
Ammonia odor, dried fish, old fish
odorAmines: TMA/DMA
NH3
Putrid odor, onion like, old cabbage,
rotten egg
Sulfur compounds
SSOs Pseuodomonas spp. P. phosphoreum H2S producers
Sweet - sour, malty and fruity odors
Alcohols, aldehydes, ketones, esters
METHODS – analysis of volatile compounds
Electronic nose FreshSenseMaritech, Iceland (prototype)Electrochemical sensors
CO, NH3, H2S, SO2
2.6L sampling vesselPump Continuous sampling for 5 minutesApprox. 500 g fillets Temperature monitoring /control
Rapid techniqueA few sensors sensitive to the main classes of compounds produced in chilled fish
Comparison with gas chromatographyIdentify the main classes of compounds produced during storage to select appropriate sensorsGC-MS / GC-O
Storage study of haddock filletsComparison of the e-nose sensors and GC analysis of the main classes of volatile compounds produced
during chilled storage
0 5 10 15Days
arbi
trary
uni
ts
Alcohols, aldehydes and esters
Amines (TMA)
Sulfur compounds
0
400
800
1200
0 5 10 15Days
Res
pons
e se
nsor
s (n
A)
CO
SO2
NH3
CO sensor CO sensor alcoholsalcohols, , aldehydesaldehydes and esters and esters
SOSO22 sensorsensor sulfursulfur compoundscompounds
NHNH3 3 sensor sensor aminesamines
Electronic nose sensors Gas chromatography
In: Olafsdottir G. 2003. Developing rapid olfaction arrays for determining fish quality. In Ibtisam E Tothill(Ed) RAPID AND ON-LINE INSTRUMENTATION FOR FOOD QUALITY ASSURANCE. WoodheadPublishing Ltd, Cambridge, England, pp. 339-360
Examples of the dynamic evolution of volatile compounds in spoilage of fish (cod fillets 0°C)
0
5
10
15
20
25
30
35
40
45
50
0 2 4 6 8 10 12 14 16
Days from catch
PAR
0
100
200
300
400
500
600
700
800
900
1000
PAR
(iso
buta
nol)
ethanol
3-methyl-1-butanol
2,3-butandiol
ethyl acetate
isobutanol
End of shelf-life
0
5
10
15
20
25
30
35
40
45
50
0 2 4 6 8 10 12 14 16
Days from catch
PAR
0
50
100
150
200
250
300
PAR
(ace
toin
)
6-me-5-hepten-2-one
2-butanone
3-pentanone
acetic acid
acetoin
End of shelf-life
Alcohols, esters Ketones, acids
From: Olafsdottir, G., Jonsdottir, R., Lauzon, H.L., Luten, J., Kristbergsson, K. 2005. Characterization of volatile compounds in chilled cod (Gadus morhua) fillets by gas chromatography and detection of quality indicators by an electronic nose. JAFC, 53 (26), 10140 -10147.
Potential quality indicators in cod fillets (0°C)
0
50
100
150
200
250
300
350
400
450
0 2 4 6 8 10 12 14 16
Days from catch
PAR
- TV
B-N
- se
nsor
s
6,2
6,3
6,4
6,5
6,6
6,7
6,8
6,9
7,0
pH
TMATVB-NCO sensorNH3 sensoracetoinpH
End of shelf-life
1
2
3
4
5
6
7
8
9
0 2 4 6 8 10 12 14 16
Days from catch
log
CFU
/g
TVC
H2S counts
Pseud
Pp
End of shelf-life
Microbial counts: TVC (total viable counts)H2S producersPseudomonas spp.P. phosphoreum (predominating)
Volatiles: TMA (trimethylamine), TVB-Nacetoin (3-hydroxy-2-butanone)
E-nose: CO and NH3 sensor
Quantification and identification of the main classes of compounds in cod fillets during storage (0°C)
by GC-MS and GC-O
0
10
20
30
40
50
60
70
80
90
100
110
120
Ketones TMA Alcohols Acids Aldehydes Esters Sulfurcompounds
PAR
Day 4 Day 7 Day 10 Day 12
3-methyl-butanal (sweet, malty odor)hexanal (grass),heptanal (boiled potato),nonanal, decanal (fresh, floral, sweet)
acetoin, 3-pentanone, (sweet, butter-like, sour)
Fishy odorDMS, DMDS, DMTS: onion like
rotten, sulfur, cabbage
Sickenly sweet odor
33%
29%
4% 3%
15%
<1%
ethanol, 3-methyl-1-butanol, 2-methyl propanol
No odor detected
Early spoilage Early spoilage changeschangesKetonesKetonesAlcoholsAlcoholsAldehydesAldehydes
Late spoilage changesLate spoilage changesTMA TMA EstersEstersAcidsAcids(Sulfur(Sulfur compounds)compounds)
Sensory rejection on day 12
Storage studies - METHODSFive storage studies on cod and haddock fillets packed in styrofoam boxes
from two factories in Icelanddifferent temperatures (-1,5 to 15°C)
Traditional reference methodsSensory analysis
Torry scoreMicrobial counts
TVC (total viable counts) (15°C)SSO (specific spoilage organisms)
– Pseudomonas spp– H2S-producers– Photobacterium phosphoreum
Chemical analysis TVB-N (total volatile basic nitrogen)
Electronic nose
Data analysis
Multivariate modelsPLSR / PCA /SIMCA
Unscrambler Camo
Shelf-life determined by sensory analysis
3
4
5
6
7
8
9
10
0 2 4 6 8 10 12 14Days from catch
Torr
y sc
ore
0°C0°C a0°C+abuse7°C15°C
Sensory rejection Torry = 5,5
Influence of different storage temperature on the shelf-life of haddock fillets
From: Olafsdottir G, Lauzon HL, Martinsdottir E, Kristbergsson K. 2005. Influence of storage temperature on microbial spoilage characteristics of haddock fillets (Melanogrammus aeglefinus) evaluated by multivariate quality prediction. In press.
Storage conditions:
Study 1: 0°C, 7°C, 15°C
Study 2: 0°C, 0°C + abuse
2
3
4
5
6
7
8
0 2 4 6 8 10 12 14
days from catch
Pseu
dom
onas
spp
log
cfu/
g
15°C
7°C
0°C+ abuse0°C
0°C
0
100
200
300
400
500
600
700
800
0 2 4 6 8 10 12 14
days from catch
Res
pons
e C
O s
enso
e (n
A)
0°C
15°C 7°C
0°C
0°C+ abuse
Microbial counts Pseudomonas spp.
0
20
40
60
80
100
120
0 2 4 6 8 10 12 14
days from catch
TVB-
N m
g N
/100
g
7°C
15°C0°C+ abuse
0°C
0°C
E- nose: CO sensor
Influence of different storage temperature (0°C, 7°C, 15°C and 0°C, 0°C + abuse)
on the rate of spoilage changes haddock fillets
From: Olafsdottir G, Lauzon HL, Martinsdottir E, Kristbergsson K. 2005. Influence of storage temperature on microbial spoilage characteristics of haddock fillets (Melanogrammus aeglefinus) evaluated by multivariate quality prediction. In press.
TVB-N
Quality indictors do not always agree on the spoilage rate of sample groups
Partial Least Squares Regression (PLSR) modelPartial Least Squares Regression (PLSR) modelto explore the correlation of the variablesto explore the correlation of the variables
Selection of quality indicators [X] to predict sensory quality [Y] of haddock fillets (stored at different temperatures 0°C, 7°C, 15°C - 0°C, 0°C + abuse)
Quality indicators [X]H2S sensor - Tacc
NH3sensor – TVB-N - pH
CO sensor
Pseudomonas spp.H2S producersP. phosphoreumTVC
(Total viable microbial counts)
Different models were explored to select the best quality indicators Best models when all significant variables were used Of interest to use rapid techniques like the e-nose sensors (CO, H2S and NH3)
Sensory Quality [Y]Torry score
Partial Least Squares Regression (PLSR) model Partial Least Squares Regression (PLSR) model for haddock samples for haddock samples (0°C, 7°C, 15°C (0°C, 7°C, 15°C -- 0°C, 0°C + abuse0°C, 0°C + abuse)
Models based on pseudomonads counts and the CO, NH3 H2S sensors, and Tacc (time / temp) gave best results to predict sensory scores
(r2=0.94; RMSEP: 0,49)
5% error between predicted sensory scores and experimental values when using a subset of the data
Low temp. (0°C)
High temp. ( 7-15 °C)
PC1 TIME
PC2 TEMPERATURE
Storage studies on cod fillets at low temperature -1.5 (superchilling) to 0.5°C and temperature abuse
Factory 1: Process 1 Traditional (no cooling of fillets) + abuseProcess 2 Superchilling (CBC)
Factory 2: Process 3 Traditional (cooling of fillets)
Important to study natural products Catching, handling and storage conditions were not the sameThe aim was to select the best quality indicators (microbial, chemical, e-nose) to predict the sensory quality
-2-10123456789
10
D A B C E G F I H
Tem
pera
ture
°C
Initial temperature Average temperature of fillets
B 11.5dA 12 d
C 12.5 d
I 10.5 d
H 13.5 d
Process 3 (ice)
E 14.5 dG 15 d F 15+d
Process 2 (ice slurry)
Factory I Factory II
Traditional – no cooling
Superchilling - CBC
!16°C 8h
-1.5°C0.5 °C
0.5°C
!16°C 8h
!RT 16h
0.5 °C
!
17°C 16h
Process 1 (ice)
0.5 °C
CBC
D 8d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Day
s fro
m c
atch
Temperature of fillets in experimental groups
Traditional - cooling
Influence of different processes and temperature conditions (-1.5 °C superchilling, 0.5 °C and abuse) on the shelf-life of cod fillets
Partial Least Squares Regression (PLSR) modelto explore the potential of the quality indicators (X)
to predict sensory quality (Y)
Cod fillets stored at -1.5 (superchilling) to 0.5°C + abuse– different factories and processes
Quality indicators [X]CO sensorTVB-N
P.phosphoreumTVCpseudomonadsH2S-producers
Sensory Quality [Y]Torry score
From: Olafsdottir G, Lauzon HL, Martinsdottir E, Oehlenschläger J, Kristbergsson K. 2006. Evaluation of shelf-life of superchilled cod (Gadus morhua) fillets and influence of temperature fluctuations on microbial and chemical quality indicators. J. Food Sci 71 (2): 97-109.
PLSR model for chilled and superchilled cod fillets
Multiple quality criteria based on 5 variables, the SSOs (P. phosporeum,H2S-producers, pseudomonads), CO sensor and TVB-N was needed when using a global model to classify samples from different factories stored at different temperatures
Process 3 – Cooling: H, I
Process 2 – Superchilling: E, F, G
Process 1 – Temp. fluctuations: A, B, D, C
• 85% correct classification based on sensory criteria (Torry score 7)
Volatile compoundsVolatile compounds contributed by different SSOsSSOs give information about the spoilage processes in fish
Electronic noseElectronic nose is a multimulti--indicator indicator device selective sensorsselective sensors for ketonesketones, amines, alcohols, , amines, alcohols, aldehydesaldehydes, acids, , acids, esters and sulfur compoundsesters and sulfur compoundse-nose can give more information than a single reference measurement for example TVB-N.
Establishment of quality Establishment of quality criteracritera or fixed values to determine the end of shelf-life or the quality of fish based on electronic nose responses, microbial counts and TVB-N reference methods will have to be developed for each product and the respective storageconditions.
CONCLUSIONS
FUTURE PERSPECTIVESImplementation of the electronic nose technique for quality
monitoring of fish products? The fast development of sensor technologies and data processing techniques will improve the possibility of quality monitoring in the food industry.
On-line monitoring ? Handheld devices ? “Smart” sensors in packaging?
samplingtemperature control exclusion of background odors and contaminantssensitive / selective sensors for quality indicating compoundsrapid detection of SSOs specific applications - establishment of quality criteria for different products and temperature conditions
AcknowledgementsFunding:
Icelandic Centre for ResearchAVS Research Fund of the Ministry of Fisheries
Participating companies:MaritechTrosTangiSkaginn (www.skaginn.is/)