Model for the Effect of Carbon Dioxide on Listeria Monocytogenes in Fresh-cut Iceberg Lettuce Packaged Under Modified Atmosphere

Abstract

The fresh-cut iceberg lettuce inoculated with Listeria Monocytogenes (3 log CFU/mL) was packaged in modified atmospheres (10% CO₂/10% O₂, 30% CO₂/10% O₂, 50% CO₂/10% O₂) and then stored at 4°C, 8°C, 16°C, 24°C and 32°C separately. Modified Gompertz model was selected to estimate the growth rate and lag time. Results showed that the growth of L. monocytogens decreased with the increase of concentration of CO₂ and revised Gompertz equation fitted the growth of L. monocytogens at different test conditions. Root Square model was chosen as a secondary model to study the effects. Mathematical evaluation was also conducted with the bias factor (B_f) and accuracy factor (A_f) and residual analysis. Results showed that predictive value of the model is more accurate and reliable in level of α=0.1 and the value of B_f all is in 0.7–1.05, while the value of A_f is acceptable in general.

Introduction

Listeria monocytogenes is an important foodborne pathogen that causes listeriosis and other diseases including sepsis and meningitis, which has a high mortality rate (Mead et al. 1999; Szabo et al. 2003; Carrasco et al. 2008). As a facultative anaerobe, L. monocytogens is able to grow at temperatures ranging from −1.5 to 45°C, pH from 4.39 to 9.40, and high osmotic conditions (Te and Zwietering, 1999). Hence, L. monocytogens has raised serious concerns in minimally processed foods. The contamination of L. monocytogens in raw or processed vegetables has been frequently reported (Samelis et al. 2001; Kramarenko et al. 2016). One possible reason is the current industrial processing practice of vegetables, such as the sanitized washing, which does not provide thoroughly cleaning of the pathogen (Kramarenko et al. 2016).

Modified atmosphere packaging (MAP) is a popular strategy to inhibit the growth of microorganisms in foods to maintain their quality with extended shelf life (Xu et al. 2017). The applications of MAP have been used to protect raw and cooked meat, fish, fresh fruit and vegetables (Liamniniter et al. 2018; Parks et al. 2012; Han et al. 2010). O₂, CO₂ and N₂ are generally used in MAP, of which CO₂ is the major anti-microbial component. Low levels of O₂ with elevated CO₂ concentrations are typically applied to suppress respiration rate, retard ripening and senescence, and associated with quality loss of fruits and vegetables (Resconi et al. 2012). Fresh-cut vegetable are easily infected by L. monocytogens during MAP processing. Although the growth of L. monocytogens on lettuce has been previously reported (Kramarenko et al. 2016; Couvert et al. 2017; Koseki et al. 2005), the growth of this pathogen on fresh-cut iceberg lettuce under the MAP that is initially flushed with different concentrations of CO₂, O₂ and N₂ gases is absent from the literature. There is an urgent need to study the growth of L. monocytogens on fresh-cut lettuce processed by MAP to avoid food poisoning.

Numerous bacterial growth modes have been developed to predict the behavior of microorganisms. Highly sophisticated models are available, including Baranyi, Logistic and modified Gompertz models (Luo et al. 2015; Sheen et al. 2011). Large amount of work has been done in relation to growth models of L. Monocytogenes. Farber et al. (1996) investigated the effects of pH (5.5–6.5), temperature (4–10°C) and CO₂ (10–90%) on the growth of L. monocytogens in brain heart infusion broth. Fernándcz et al. (1997) used Baranyi and Roberts model to fit the growth curves and compared predictions of growth from the model. However, most of those models were built on nutrient broth experiments in laboratories, resulting in the overestimation of the growth of L. monocytogens in real foods and making those models not practically useful. To our knowledge, a model that is developed for predicting the effect of CO₂ on the growth of L. monocytogens in fresh-cut iceberg lettuce is not reported. The aims of this work were to study the effect of a wide range of CO₂ concentrations (10%–50%) on the growth of L. monocytogens in fresh-cut iceberg lettuce held at temperatures ranging from 4 to 32°C under MAP; and to develop a predictive model for the growth of this pathogen in MAP fresh-cut lettuce.

Materials and Methods

Strain    Listeria monocytogenes CICC 21662 was obtained from the China Center of Industrial Cultural Collection (Beijing, China). A freeze-dried culture of this strain was thawed and streaked onto agar (PALCAM agar base with selective supplements) and the plates were incubated at 37°C for 48 h. Single colony from the PALCAM plate was transferred to 10 mL tryptic soy broth (TSB) tubes and incubated at 37°C for 4 h, and then transferred to another tube of fresh TSB and incubated at 37°C for 18 h, to a level about 10⁸ CFU/mL. The culture was diluted and dispersed in sterile saline solution to obtain the final concentration of 10³ CFU/mL as the concentration for inoculum.

Sample preparation    Freshly unprocessed iceberg lettuce was obtained from a local farm and transported to the lab within 1 h. The lettuce heads were processed at 7°C in a clean room. The inner leaves were individually washed using tap water twice and drained on absorbent paper. The leaves were cut into pieces (5×5 cm) by knifes. Both sides of lettuce were sterilized under UV for 20 min to receive a dose of 1.5 kJ/m² respectively to eliminate the background bacteria according to Seong et al. (2017). After sterilization, each lettuce sample (10 g) was placed into a sterilized plastic box for inoculation treatment.

Inoculation and storage of iceberg lettuce in MAP    The sample was inoculated with 1 mL suspension of L. Monocytogenes, and then shaken gently for 30 times. Once inoculated, the box was aerated using a composite modified atmosphere packaging machine. The proportions of CO₂, O₂ and N₂ in the package headspace were as the following: MA1 (10% CO₂, 10% O₂, 80% N₂), MA2 (30% CO₂, 10% O₂, 60% N₂), and MA3 (50% CO₂, 10% O₂, 40% N₂). Inoculated lettuce samples sealed in different MAP were stored at 4, 8, 16, 24 and 32°C, and samples were collected at different time intervals to be analyzed. The gases were created and adjusted every 8 h to the desired concentration by a flushing of CO₂, O₂ or N₂ into the containers, using the device previously described by Carlin et al. (1996). The time intervals between sampling varied depending on the temperature of storage and ended once the microorganisms reached stationary phase. Triplicate growth curves were performed. The samples were added with 90 mL sterile saline and homogenized for 2 min. For an enumeration, 0.1 mL of such dilute was spread on the surface of a PALCAM agar plate, and the plate was incubated at 37°C for 24 h.

Establishment of the primary and secondary models    Growth curves of L. monocytogens for each combination of the temperature and CO₂ concentration were built separately by fitting data to the Gompertz model modified by Zwietering et al. (1993) using SigmaPlot 10.0. Each combination was replicated three times to estimate two curve parameters: growth rate (µ_max) and lag time (λ).

  

where t is time (h); N_t is the number of cells at time t (CFU /g); N₀ and N_max are the initial and final number of cells (CFU / g), respectively; µ_max is growth rate (h⁻¹) and λ is lag time (h).

The secondary growth model described the dependence of the primary model parameters on environmental factors. The effect of temperature on microbial growth kinetics was usually described by square root model (Ratowsky et al. 1982), which demonstrated a linear relationship between temperature and the growth rate or the reciprocal of lag time. The model was as follows:   

  

where µ_max is the growth rate (h⁻¹); λ is lag time (h); T is temperature (°C); T_min is the conceptual minimum temperature for microbial growth; b_µmax and b_λ are the slopes of the regression coefficients. The parameters of b and T_min were estimated by searching values of T and µ_max using the Eq.1 before the effect of temperature on µ_max was obtained.

After amended by Devlieghere et al. (2001), the factor of CO₂ concentration was taken into consideration and added to the root square model as follows:   

  

where b_{µ max} and b_λ is constant, T_min is the respective estimated theoretical minimum temperature (°C), CO_{2 max} is the respective estimated theoretical maximum CO₂ concentration (mg/L), CO₂ is the concentration of dissolved CO₂ (mg/L), refers to the concentration of CO₂ without metabolic significance.

Validation of models    In order to evaluate the goodness-of-fit of the model, the bias factor (Bf) and accuracy factor (Af) proposed by Ross et al. (1996) and Mellefont et al. (2003) were adopted and calculated by the following equations:   

  

where pred was the predicted value of growth parameters using root square model; obs was the observed values of growth models; n is the number of trials.

According to Ross (1999), B_f is used to evaluate the extent of under-or over-prediction by the model of the µ_max and lag time observed, when B_f is above 1, it indicates that the predicted µ_max is lower than the observed value, which means safe in practice. In model performance evaluations, if the values of B_f are between 0.9 and 1.05, the model can be considered good; if the values are between 0.7–0.9 or 1.05–1.15, the model can be considered acceptable. A_f was calculated for providing an indication of the average of estimates, the average estimates tend to be less accurate with the increase of A_f. (Lebert et al. 2000).

Although the B_f and A_f can be used to validate the models, they can't describe the systematic deviation. Therefore, Baranyi et al. (2000) suggested using the residual analysis to describe the systematic deviation, in which the residual is the difference between the predicted value and the observed value. In the current study, the residual analysis was also conducted to assess the performance of the models.

Data analysis    Data were obtained from two independent experiments and analyzed by ANOVA using SAS (Version 8.2; SAS Institute, Cary, NC, USA). The P value ≤ 0.05 was considered statistically significant.

Results

Estimation of growth parameters and primary predictive model    The growth of L. monocytogens in fresh-cut lettuce at different CO₂ concentrations from 4°C to 32°C is shown in Fig. 1. The growth curves were predicted using modified Gompertz equation. The modified Gompertz equation works well for the prediction of L. monocytogens growth in fresh-cut lettuce under tested conditions. Table 1 shows the two parameters of µ_max and λ of the growth curve in different conditions. The µ_max of L. monocytogens was lower and λ was longer at the same temperature with the increasing of CO₂ concentration (from 10% to 50%). At 32°C, when the concentration of CO₂ increased from 10% to 50%, the µ_max of L. monocytogens decreased from 0.341 (h⁻¹) to 0.195 (h⁻¹), while the λ increased from 3.792 h to 9.688 h. This indicated that CO₂ significantly inhibited to the growth of L. Monocytogenes.

Fig. 1.

Growth curves of L. monocytogens on fresh-cut lettuce at a constant temperatur. (4, 8, 16, 24 or 32°C) under modified atmosphere packaging using modified Gompertz model.

Table 1. The growth rate (µ_max) and lag time (λ) of L. monocytogens obtained using Gompertz equation.

Temperature (°C)	Concentrations of CO2 (%)	Gompertz equation
		µmax (h−1)	λ (h)
4	10	0.024	101.557
4	30	0.018	125.846
4	50	0.015	140.427
8	10	0.038	60.430
8	30	0.026	68.781
8	50	0.022	83.883
16	10	0.070	15.516
16	30	0.055	19.938
16	50	0.038	25.523
24	10	0.165	6.288
24	30	0.135	11.764
24	50	0.095	16.448
32	10	0.341	3.792
32	30	0.240	5.525
32	50	0.195	9.688

µ_max, growth rate, λ, lag time.

Establishment of the secondary predictive model    The µ_max obtained using the primary model and the corresponding temperatures were substituted into Eq. 2 and 3, respectively. Fmincon function in Matlab7.1 was applied using iterative method to fit the value of b_{µ max,} b_λ and CO_{2 max}. Equations are as follows:   

  

Model validation and verification    The model was verified at random time points under conditions of each group. Table 2 illustrates the mathematical validations of the secondary model of CO₂ inhibition of L. Monocytogenes. As it can be seen, the B_f ranged from 0.766 to 1.047. According to the critical values of B_f proposed by Mellefont et al. (2003), all B_f values obtained in this study fall in the range of acceptable level. For the data set at 10% CO₂ concentration, B_f values are all in good level, expect at 32°C with B_f value of 0.833 (considered acceptable). The A_f values are all higher than 1, and at 8°C with 10% CO₂ concentration, the value of A_f reached to 1.311, other values are all within the range of 1.0–1.3. Ross et al. (2000) reported that an acceptable A_f could be determined by considering the effect of the number of environmental parameters in a kinetic model. Therefore, the acceptable performance that might be expected from the Kinetic model encomp assing the effects of temperature and CO₂ concentration, is about 30% or an A_f of 1.3. Therefore, for the value of A_f, the model is accepted in general.

Table 2. Mathematical validation of the secondary model for CO₂ inhibition of L. monocytogenes.

Temperature (°C)	CO2 concentrations (%)	Trial No.	Bf	Af
4	10	6	0.928	1.222
4	30	6	1.035	1.186
4	50	6	0.862	1.23
8	10	6	0.949	1.311
8	30	6	0.766	1.127
8	50	6	0.973	1.279
16	10	6	0.956	1.024
16	30	6	0.782	1.178
16	50	6	0.886	1.232
24	10	6	0.923	1.185
24	30	6	0.991	1.217
24	50	6	0.876	1.155
32	10	6	0.833	1.288
32	30	6	0.915	1.132
32	50	6	1.047	1.264

Trial no., random trial mumbers for residum analysis, B_f, bias factor, A_f, accuracy factor

To evaluate the model more accurately, random verification experiments were conducted six times and the results are shown in Fig. 2. It demonstrated that the absolute values between the predicted and observed values in different conditions are all < 0.1, that means the predictive value of the model is liable at the level of α = 0.1 in the 95% confidence interval.

Fig. 2.

The residual analysis of model. Triangle represents the cells of L. monocytogens in the condition of MA1 at random time points. Dot represents the cells of L. monocytogens in the condition of MA2 at random time points. Square represents the cells of L. monocytogens in the condition of MA3 at random time points.

The modified root square model as the secondary model used in the current study performed well for the growth of L monocytogenes in the MAP lettuce with different temperature and CO₂ combinations.

Discussion

Generally, many studies have found that CO₂ (≥ 20%) reduce growth rate and extend lag time in the in vitro test. (Resconiet et al. 2012; Delieghere and Debevere 2000). In this study, a predictive model for relationship between CO₂ and L. monocytogens in fresh cut lettuce was developed. Consistent with the above researches, our results confirmed that the growth of L. monocytogens was inhibited with increasing concentrations of CO₂. Also, the µ_max of L. monocytogens was lower and λ was higher with the increasing of CO₂.

Predictive models offer application potential in food microbiology. In recent years, despite the increasing association of vegetables with foodborne disease outbreaks (Ross, 2000), comparing with meat and dairy products (Mcmeekin, 2007; Poirazi et al. 2007; Lindgvist et al. 2011), few studies have been conducted on predicting the growth of pathogens in fresh vegetables (Couvert et al. 2017). Therefore, there is a need for more data on the growth of foodborne pathogens in these products. Zwietering et al. (1996) compared Gompertz, Richards, Stannard, Schnute, and Logistic models that describe the s-type growth curve and concluded that only corrected Gompertz equation fits the growth pattern of L. monocytogens (Farber et al. 1996). In this study, growth curves for each temperature and CO₂ concentration were built separately by fitting data to the Gompertz model, and results showed a high correlation coefficient at all temperatures (R²>0.95), which proved that Gompertz function, as a dynamic model, could deal with time varying environmental conditions for prediction of L. Monocytogenes.

The secondary model applied is Bělehrádek or Ratkowsky Square Root model. Square root model has been applied to describe how single factor influence the growth of microorganisms. However, this model has limitations in describing how do multi-factors influence the growth of microorganisms (Devlieghere et al. 2001). In related studies, it can be replaced using responsive surface methodology, artificial neural network model or correct square root model. But square root model is commonly used for its single parameter and simple use, and can predict the growth condition well under univariate. In this study, two factors of temperature and CO₂ concentration were taken into account, and Eqs. (8) and (9) were finally got to describe the effect of temperature and CO₂ concentration on growth of L. Monocytogenes.

Conclusion

Results showed that the growth of L. monocytogens was inhibited along with the increase of concentration of CO₂, especially when the CO₂ concentration was 50%, the growth of L. monocytogens was significantly inhibited. That is to say, CO₂ has significant inhibition of the growth of L. monocytogens on fresh-cut lettuce and revised Gompertz equation has a good statistic fit to the growth of Listeria monocytogenes at different test conditions with a high correlation coefficient (R²>0.95). And Root Square model was chosen as a secondary model to study the effects of temperature and CO₂ concentration on growth kinetic parameters of Listeria monocytogenes in the primary model. Mathematical evaluation was also conducted with the B_f and A_f and residual analysis. Results showed that predictive value of the model is more accurate, and reliable in a level of α=0.1, and the value of B_f all is in 0.7–1.05, while the value of A_f is accepted in general, namely the model can be accepted. Therefore, the model can predict the growth of Listeria monocytogenes in fresh-cut iceberg lettuce successfully.

Acknowledgements    This study was supported by the project 2017YFD0400904-2.

References

•  Baranyi,  J.,  Pin,  C., and  Ross,  T. (1999). Validating and comparing predictive models. Int. J. Food Microbiol., 48, 159-166.
•  Carlin,  F.,  Nguyen-the,  C.,  Abreu,  D.S.A., and  Cochet,  C. (1996). Effects of carbon dioxide on the fate of Listeria monocytogenes, of aerobic bacteria and on the development of spoilage in minimally processed fresh endive. Int. J. Food Microbiol., 32, 159-172.
•  Carrasco,  E.,  Pérez-Rodríguez,  F.,  Valero,  A.,  Garcı A-Gimeno,  R.M., and  Zurera,  G. (2008). Growth of Listeria monocytogenes on shredded, ready-to-eat iceberg lettuce. Food Control, 19, 487-494.
•  Couvert,  O.,  Guégan,  S.,  Hézard,  B.,  Huchet,  V.,  Lintz,  A., and  Thuault,  D., (2017). Modeling carbon dioxide effect in a controlled atmosphere and its interactions with temperature and ph on the growth of L. monocytogens and p. fluorescens. Food Microbio., 68, 89-96.
•  Devlieghere,  F. and  Debevere,  J. (2000). Influence of dissolved carbon dioxide on the growth of spoilage bacteria. Lebensm. Wissenschaft Techno., 33, 531-537.
•  Devlieghere,  F.,  Geeraerd,  A.H.,  Versyck,  K.J.,  Vandewaetere,  B,.  Van Impe,  J.,  Debevere,  J. (2001). Growth of Listeria monocytogenes in modified atmosphere packed cooked meat products: a predictive model. Food Microbiol., 18: 53-66.
•  Farber,  J.M.,  Cai,  Y.,  Ross,  W.H. (1996). Predictive modeling of the growth of Listeria monocytogenes in CO₂ environments. I Int. J. Food Microbiol., 32: 133-144.
•  Fernándcz,  P.S.,  George,  S.M.,  Sills,  C.C., and  Peck,  M.W. (1997). Predictive model of the effect of CO₂, pH, temperature and NaCl on the growth of Listeria monocytogenes. Int. J. Food Microbiol., 37, 37-45.
•  Han,  C.,  Yue,  X.,  Lu,  F., and  Sun,  B. (2010). Effect of high oxygen modified atmosphere packaging on fresh-cut onion quality. Proceedings of the 17th iapri world conference on packaging, Tianjin, pp.124-128.
•  Koseki,  S. and  Isobe,  S. (2005). Growth of Listeria monocytogenes on iceberg lettuce and solid media. Int. J. Food Microbiol., 101, 217-225.
•  Kramarenko,  T.,  Roasto,  M.,  Keto-Timonen,  R.,  Mäesaar,  M.,  Meremäe,  K., and  Kuningas,  M. (2016). Listeria monocytogenes in ready-to-eat vacuum and modified atmosphere packaged meat and fish products of Estonian origin at retail level. Food Control, 67, 48-52.
•  Lebert,  I.,  Robles-Olvera,  V., and  Lebert,  A. (2000). Application of polynomial models to predict growth of mixed cultures of Pseudomonas spp. and Listeria in meat. Int. J. Food Microbiol., 61, 27-39.
•  Liamnimitr,  N.,  Thammawong,  M.,  Techavuthiporn,  C.,  Fahmy,  K.,  Suzuki,  T., and  Nakano,  K. (2018). Optimization of bulk modified atmosphere packaging for long-term storage of ‘fuyu’ persimmon fruit. Postharvest Biol. Technol., 135, 1-7.
•  Lindqvist,  R. and  Lindblad,  M. (2011). Time to growth and inactivation of three STEC outbreak strains under conditions relevant for fermented sausages. Int. J. Food Microbiol., 145, 49-56.
•  Luo,  K.,  Hong,  S.S., and  Oh,  D.H. (2015). Modeling the effect of storage temperatures on the growth of listeria monocytogenes on ready-to-eat ham and sausage. J. Food Prot., 78, 1675.
•  McMeekin,  T.A. (2007). Predictive microbiology: quantitative science delivering quantifiable benefits to the meat industry and other food industries. Meat Sci., 77, 17-27.
•  Mead,  P.S.,  Slutsker,  L.,  Dietz,  V.,  McCaig,  L.F.,  Bresee,  J.S.,  Shapiro,  C.,  Griffin,  P.M., and  Tauxe,  R.V. (1999). Food-related illness and death in the United States. Emerging Infect. Dis., 5, 607-625.
•  Mellefont,  L.A.,  Mcmeekin,  T.A., and  Ross,  T. (2003). Performance evaluation of a model describing the effects of temperature, water activity, pH and lactic acid concentration on the growth of Escherichia coli. Int. J. Food Microbiol., 82, 45-58.
•  Parks,  A.R.,  Brashears,  M.M.,  Woerner,  W.D.,  Martin,  J.N.,  Thompson,  L.D., and  Brooks,  J.C. (2012). Spoilage characteristics of ground beef with added lactic acid bacteria and rosemary oleoresin packaged in a modified-atmosphere package and displayed at abusive temperatures. J. Anim. Sci., 90, 2054-2060.
•  Poirazi,  P.,  Leroy,  F.,  Georgalaki,  M.D.,  Aktypis,  A.,  De Vuyst,  L., and  Tsakalidou,  E. (2007). Use of artificial neural networks and a gamma-concept-based approach to model growth of and bacteriocin production by Streptococcus macedonicus ACADC 198 under simulated conditions of Kasseri cheese production. Appl. Environ. Microbiol., 73, 768-776.
•  Ratkowsky,  D.A.,  Olley,  J.,  McMeekin,  T.A., and  Ball,  A. (1982). Relationship between temperature and growth rate of bacterial cultures. J. Bacterio., 149, 1-5.
•  Resconi,  V.C.,  Escudero,  A., and  Beltran.  J.A. (2012). Color, lipid oxidation, sensory quality, and aroma compounds of beef steaks displayed under different levels of oxygen in a modified atmosphere package. J. Food. Sci., 77, S10-S18.
•  Ross,  T. (1996). Indices for performance evaluation of predictive models in food microbiology. J. Appl. Bacteriol., 81, 501-508.
•  Ross,  T. (1999). Predictive food microbiology models in the meat industry (MSRC.003). Meat and Livestock, Australia.
•  Ross,  T.,  Dalgaard,  P., and  Tienungoon,  S. (2000). Predictive modelling of the growth and survival of Listeria in fishery products. Int. J. Food Microbiol. 62, 231-245.
•  Samelis,  J.,  Sofos,  J.N.,  Kendall,  P.A., and  Smith,  G.C. (2001). Fate of Escherichia coli o157:h7, salmonella typhimurium dt 104, and listeria monocytogenes in fresh meat decontamination fluids at 4 and 10 degrees c. J. Food Prot., 64, 950.
•  Seong,  Ji-Yeong,  Kwon,  Ki-Hyun, and  Oh,  Se-Wook. (2017). Combined effect of aerosolized malic acid and uv-c for the inactivation of escherichia coli o157:h7, salmonella typhimurium, and listeria monocytogenes on fresh-cut lettuce. J. Food Saf, e12359.
•  Sheen,  S.,  Hwang,  C.A., and  Juneja,  V.K. (2011). Modeling the impact of chlorine on the behavior of Listeria monocytogenes on ready-to-eat meats. Food Microbiol., 28, 1095-1100.
•  Szabo,  E.A.,  Simons,  L.,  Coventry,  M.J., and  Cole,  M.B. (2003). Assessment of control measures to achieve a food safety objective of less than 100 CFU of Listeria monocytogenes per gram at the point of consumption for fresh precut iceberg lettuce. J. Food Prot., 66, 256-264.
•  Te,  G.M. and  Zwietering,  M.H. (1999). Validation of predictive models describing the growth of Listeria monocytogenes. Int. J. Food Microbiol. 46, 135-149.
•  Xu,  L.,  Jia,  F.,  Luo,  C.,  Yu,  Q.,  Dai,  R., and  Li,  X. (2017). Unravelling proteome changes of chicken egg whites under carbon dioxide modified atmosphere packaging. Food Chem., 239, 657.
•  Zwietering,  M.H.,  Rombouts,  F.M., and  Van'T Riet,  K. (1993). Some aspects of modelling microbial quality of food, Food control, 4, 89-96
•  Zwietering,  M.H.,  De Wit,  J.C., and  Notermans,  S. (1996). Application of predictive microbiology to estimate the number of Bacillus cereus in pasteurized milk at the point of consumption. Int. J. Food Microbiol., 30, 55-70.