Skip to main content

Development, modelling & application of time temperature integrator system to monitor chilled fish quality


Despite the increasing demand for fresh (chilled) fish in Europe and the decreasing total catch of fish for human consumption, large amounts of fish and fish products are wasted due to spoilage, usually due to mishandling. This situation could be significantly rectified through a modern quality assurance approach, which is based on prevention and requires monitoring, controlling and recording of critical parameters through the product's life cycle from production to final use. To establish such a system for chilled and minimally processed fish products, it is necessary to acquire a thorough knowledge of the relation between storage conditions and shelf life, expressed in quantitative terms i.e. effective and accurate predictive models. Further there is need to validate and develop practical systems to monitor and record these conditions from catch to consumption. Time Temperature Integrators (TTI) are such systems. Based on the aforementioned needs the main objectives of this project are:
- To compare existing empirical temperature models with spoilage data for fish species not previously studied, packed fish and minimally processed fish products.
- To develop new empirical models, if required.
- To study in detail the shelf life loss modes of specific chilled fish products, in order to establish effective kinetic shelf life models that will allow accurate correlation of fish quality to variable storage, handling conditions and the initial level of spoilage.
- To develop simple and easy-to-use Time Temperature Integrating software that allow existing and new predictive models to be combined with product temperature profiles from temperature loggers, for effective chill chain quality monitoring.
- To assess existing and develop and model new TTI devices that have the required response characteristics to serve as monitors of temperature history of the fish products.
- To apply the TTI systems assessed and developed, as efficient and inexpensive means of indicating spent and remaining shelf life of the fish products.
The results that are being presented summarise the work carried out during the first year of the project (November 1996 - November 1997).
Development and validation of proposed empirical shelf life predictive models was carried out. Shelf life testing, mainly by sensory evaluation, of various fresh and lightly preserved species (boque, seabream, whiting, haake, cold and warm water marinated shrimp) was conducted. Models that explained behaviour of relevant categories of products, based on available compiled literature data, were tested for species specificity. Validation under dynamic storage conditions of the models was performed as a prerequisite for use for TTI monitoring.
It was established that for species like the boque (Boops boops) for which the square root empirical model ?RRS=1+0.1T (RRS= shelf life at 0?C/shelf life at T) successfully models shelf life at isothermal conditions, this applies satisfactorily also for dynamic conditions and thus can be effectively used in a TTI monitoring scheme. For seabream (Sparus aurata) the empirical model was less satisfactory and this became more apparent in dynamic conditions. If higher predictive accuracy is needed then the mechanism of spoilage should be studied in detail. For whiting and hake the empirical model was sufficient especially at the 0 to the 15? C range but the Arrhenius model was better with activation energies of 21.9 and 16.3 kcal/mol. The studies gave a good indication of the kinetic values of the response of a candidate TTI. The empirical model corresponds to activation energy of 18.9 kcal/mol. The activation energy of best fit for seabream was 14 kcal/mol. The lightly preserved products that were studied, i.e. cold and warm water marinated shrimp, had significantly higher activation energy, close to 27 kcal/mol, and thus require a more temperature sensitive TTI.
Kinetic spoilage models development for the above studied fish of interest. Specific spoilage microflora (SSB) and spoilage metabolites were studied for most studied fish products at temperatures ranging from 0 to 20 oC. Correlations were established between organoleptic quality and growth of SSB and kinetic models developed. Obtained models and activation energies serve as input to the TTI software and for the evaluation and development of TTI devices.
Representative experimental data for the growth of the different bacteria comprising the microflora of boque are shown in Fig.1 along with the fitted Gompertz growth curves and the time corresponding to the end of shelf life. From the parameters of the Gompertz equation lag times, the maximum growth rate at different storage temperatures and the count of each measured population corresponding to the sensoric end of shelf life are calculated. Pseudomonads and Shewanella putrefaciens were determined as good spoilage indexes for air stored chilled boque. End of shelf life at all temperatures coincides with a level of 107 for these two bacteria.
By modelling the temperature dependence of the growth parameters of Pseudomonads and Shewanella putrefaciens by the Arrhenius or Belehradek (square root) equations kinetic shelf life predictive models of the fish are obtained.
In parallel kinetic testing, modelling and evaluation of potential of TTI devices available as industrial prototypes or products and based on different principles, based on experiments at the relevant range of constant and variable temperature conditions was conducted. Four types of triple window enzymatic VITSAB(r)-TTI (Malmo, Sweden) indicators and four models of polymer type Lifelines Fresh-Check(r)(NJ, USA) indicators were tested.

(For Figure 1 contact the Coordinator)

Figure 1. Development of the microbial association of Mediterranean fish (Boops boops) stored aerobically at 0 oC, 3 oC. The growth data were fitted with Gompertz equation. TVC:Total viable count, CFC:Pseudomonas, STAA: Brochothrix thermosphacta, IA: H2S producer bacteria (Shewanella putrefaciens) VRBG:Enterobacteriaceae. The vertical straight lines represent the end of shelf life determined by sensory evaluation. (Taoukis et al., 1997)

Further, new enzyme-substrate systems were explored and kinetically studied and modelled in search for effective TTIs for chilled fish quality monitoring. Enzymes were selected for their ability to produce a colorimetric reaction and use of enzymes in a crude form provided a cheap alternative for potential commercial application. It was concluded that pNP-based reaction systems were suitable in terms of ease of measurement and so these systems were evaluated further as TTI monitors. These systems were a Fusarium oxysporum xylosidase, an Aspergilus niger ?-galactosidase, an Aspergilus niger ? -galactosidase, and a Fusarium oxysporum esterase that had activation energies 20.5, 14.6, 12.1, and 8.9 kcal/mol respectively (Tsoka et al., 1998).
Overall the range of activation energies from 9 to 38 kcal/mol was covered. Applicability of the TTI under dynamic storage conditions and the accuracy of shelf life monitoring was demonstrated. The errors due to the difference between the TTI and the fish activation energy were assessed. Selected systems will be further studied and optimized.
Additionally, the first stages of development of the Seafood Spoilage Predictor software, a tool that will allow people in practice to apply successfully validated seafood spoilage models in conjunction with data from TTI and dataloggers, were completed.


The simultaneous systematic kinetic study of the shelf life characteristics of the fish and the response of the TTI systems has advanced, setting the required solid foundation for developing an effective and dependable quality monitoring system of the chill chain of fish products.

Work will continue on validation of shelf life models of fresh, MAP, and minimally processed fish products under dynamic conditions. Spoilage mechanisms will also be further explored and kinetically modelled. These models will be used in the "Seafood Spoilage Predictor" software and for evaluation, selection and development of the optimum TTI devices for chilled fish quality monitoring. Optimized TTI will be field tested in the later stages of the project.


Dalgaard, P. (1996): Predictive modelling and time-temperature integration. In Bordeaux Aquacul-ture: Proceedings of Meeting of Commission C2 (March 20-22, 1996), Inter-national Institute of Refrigeration, Bordeaux, France. pp 409-419.
Dalgaard, P. and Huss, H.H. (1997) Mathematical modelling used for evaluation and prediction of microbial fish spoilage. In Seafood safety, processing and biotechnology. Eds.Kramer, D., Shahidi, F. and Jones, Y. Technomic pub. Co. Inc., Lancaster, PA, Canada, pp 73-89.
Dalgaard, P.(1998) Modelling of seafood spoilage. Proceeding from "Predictive Microbiology Applied to Chilled Food Preservation" 16-18 June Quimper, France. P and M 'in press'
Dalgaard, P. Predictive microbiological modelling and seafood quality. In: Seafood from producer to Consumer, Integrated approach to quality. Eds. J.B. Luten, T. Borresen and J. Oehlenschlager. Amsterdan: Elseviers, pp 431-443.
Dalgaard, P. Predictive microbiology - practical applications. Presentation at Minimally processed foods, 29 October 1997, Copenhagen, Danish Society of Food Science and Technology.
Dalgaard, P., Manfio, G.P. and M. Goodfellow (1997) Classification of Photobacteria associated with spoilage of fish products by numerical taxonomy and pyrolysis mass spectrometry. Zentralblatt fue Bakteriologie. 285, 157-168.
Dalgaard, P., Mejlholm, O. and H.H. Huss (1997) Application of an iterative approach for development of a microbial model predicting the shelf-life of packed fish. International Journal of Food Microbiology. "in press"
Dalgaard, P., Mejlholm, O., Christiansen, T.J. and H.H. Huss (1997) Importance of Photobacterium phosphoreum in relation to spoilage of MAP fish products. Lett. Appl. Microbiol. 24, 373-378.
Huss, H.H., Dalgaard, P. and Gram, L. Microbiology of fish and fish products. In: Seafood from producer to Consumer, Integrated approach to quality. Eds. J.B. Luten, T. Borresen and J. Oehlenschlager. Amsterdam: Elseviers, pp 413-430.
Koutsoumanis K., Trifinopoulou, P., Taoukis, P.S. & G-J. E. Nychas (1997). Storage of Mediterranean fresh tsipoura fish (Sparus aurata ) under aerobic or modified atmospheres at 0, 3, 7 and 10°C. Third Main Meeting, Process Optimization and Minimal Processing of Foods, Copernicus Project CIPA-CT94-0195. Leuven, Belgium, 24-25 October 1997
Olafsdottir. G., Martinsdottir, E., Oechenschlager, J., Dalgaard, P., Jensen,B., Luten, J., Undeland, I., Mackie, I., Henehan, G., Nielsen, J. and H. Nilsen (1997) Methods to evaluate fish freshness in research and industry. Trends in Food Science and Technology. 8, 258-265.
Taoukis P.S. (1997) Time-Temperature Indicators for Quality Assurance of the Distribution Chain of Chilled, Minimally Processed and Frozen Foods. Plenary lecture. Third Main Meeting, Process Optimisation and Minimal Processing of Foods, Copernicus Project CIPA-CT94-0195. Leuven, Belgium, 24-25 October 1997.
Taoukis P.S. and Labuza T.P. (1998) Chemical Time-Temperature Integrators as Quality Monitors in the Chill Chain. Proc. Of the International Symposium Quimper Froid'97. Predictive Microbiology of chilled foods. June 16-18, (1997).
Taoukis, P.S., Koutsoumanis, K. and G.J.E. Nychas (1998) ?odelling of spoilage microflora of boque (?oops boops) as a basis for chilled distribution monitoring with time-temperature indicators. Proc. of the International Symposium Quimper Froid'97. Predictive Microbiology of chilled foods. June 16-18, (1997).
Tsoka, S. , Taoukis P.S., Christakopoulos, P., Kekos, D., Macris B.J. (1998). Time temperature Integration for Chilled Food Shelf life Monitoring using Enzyme-Substrate Systems. Food Biotechnology (In press).
Tsoka, S. , Taoukis P.S., Christakopoulos, P., Kekos, D., Macris B.J. (1997). Development of Enzyme -Substrate systems as Time- Temperature Integrators for the Chilled Foods Chain. Third Main Meeting, Process Optimisation and Minimal Processing of Foods, Copernicus Project CIPA-CT94-0195. Leuven, Belgium, 24-25 October 1997.
To systematically pursue the above objectives work, within the project has been organized in two main Tasks.
TASK 1: Establishment of accurate shelf life models for different chilled fish products by:
Evaluation of empirical shelf life models for different fish products and species and validation and development of kinetic spoilage models, for prediction of shelf life of real fish products under expected range of storage and packaging conditions.

TASK 2: Development and application of Time Temperature Integrators for fish spoilage.
This includes development of Time Temperature Integrating software for application of predictive spoilage models, testing and assessment of existing TTI and development of TTI prototypes with required response to simulate modelled fish shelf life behaviour. Evaluation of the potential of tested and developed TTI as "active" shelf life labels by field tests and optimisation of systems will follow.

Funding Scheme

CSC - Cost-sharing contracts


National Technical University of Athens
Iroon Polytechniou 5, Zografou Campus
15780 Athens

Participants (3)

Rue De L'universite, 6
29191 Quimper
Agricultural University of Athens
Iera Odos 75
11855 Athens
Danish Institute for Fisheries Research
Soltofts Plads, C/o Technical University Of Denmark, Bygning 221
2800 Lyngby