Search tips
Search criteria 


Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
Appl Environ Microbiol. 2009 November; 75(22): 7060–7069.
Published online 2009 September 25. doi:  10.1128/AEM.01045-09
PMCID: PMC2786533

Design of an Experimental Viscoelastic Food Model System for Studying Zygosaccharomyces bailii Spoilage in Acidic Sauces[down-pointing small open triangle]


Within the field of predictive microbiology, the number of studies that quantify the effect of food structure on microbial behavior is very limited. This is mainly due to impracticalities related to the use of a nonliquid growth medium. In this study, an experimental food model system for studying yeast spoilage in acid sauces was developed by selecting a suitable thickening/gelling agent. In a first step, a variety of thickening/gelling agents was screened, with respect to the main physicochemical (pH, water activity, and acetic acid and sugar concentrations) and rheological (weak gel viscoelastic behavior and presence of a yield stress) characteristics of acid sauces. Second, the rheological behavior of the selected thickening/gelling agent, Carbopol 980, was extensively studied within the following range of conditions: pH 4.0 to 5.0, acetic acid concentration of 0 to 1.0% (vol/vol), glycerol concentration of 0 to 15% (wt/vol), and Carbopol concentration of 1.0 to 1.5% (wt/vol). Finally, the applicability of the model system was illustrated by performing growth experiments in microtiter plates for Zygosaccharomyces bailii at 0, 0.5, 1.0, and 1.5% (wt/vol) Carbopol, 5% (wt/vol) glycerol, 0% (vol/vol) acetic acid, and pH 5.0. A shift from planktonic growth to growth in colonies was observed when the Carbopol concentration increased from 0.5 to 1.0%. The applicability of the model system was illustrated by estimating μmax at 0.5% Carbopol from absorbance detection times.

Food structure is, next to the chemical composition and storage conditions, one of the key factors that affect microbial behavior in food products. The effects of food structure are mainly related to the mechanical distribution of water, the chemical redistribution of organic acids, and the mobility of microorganisms (55). In the case of a liquid food product, microbial growth is typically planktonic, and transport of nutrients and metabolites occurs by diffusion, resulting in a homogeneous environment. The majority of foods, however, have some degree of structure, causing microorganisms to be immobilized and constrained to grow as colonies. Within the field of predictive microbiology, where mathematical models are developed for describing microbial growth, inactivation, and survival in food (model systems), most models are based on data obtained in liquid broth media. The scarcity of predictive models that incorporate the effect of structure has been recognized as one of the most important shortcomings in this field of research (41).

Conducting experiments on a structured culture medium gives rise to several impracticalities due to the nonliquid nature of the culture medium. Starting from a liquid culture medium, a thickening or gelling agent is added to obtain a structured model system that, ideally, mimics the microstructural properties of the target food product. In order to evaluate the effect of structure on a systematic and consistent basis, the experimental setup must enable careful control and sampling methods. A widely used experimental setup is the gel cassette system (Institute of Food Research, Norwich, United Kingdom). The system, described by Brocklehurst et al. (7, 8), consists of a frame sealed with gas-permeable plastic film. This setup has been used to study growth behavior both on the surface (8, 24) and within the gel matrix (7, 10, 30, 45, 51) by applying traditional microbiological methods or noninvasive microscopical techniques (29, 46). In most of these cases, gelatin was used to induce a gelled microstructure. Other experimental setups used agar or gelatin gels in petri dishes (1, 2, 3, 48) or studied bacterial growth in oil-in-water emulsions (9, 38, 39).

The existing experimental model systems most often make use of agar or gelatin as these are widely used gelling agents within the field of microbiology. The main reasons for using agar are its stability at sterilization temperatures, high clarity, nontoxic nature, and physiologically inert behavior toward microorganisms. Gelatin has the advantage of a lower melting point (37°C compared to 85°C for agar), which facilitates sampling procedures. Main drawbacks include the possible metabolization by microorganisms and breakdown of structure during autoclaving. Although both agar and gelatin are widely used in food applications, their relevance is limited to food products with a gelled microstructure. Expanding structured food model systems to a wider range of food products, therefore, implies the use of other thickening or gelling agents. In the past, several attempts have been made to use food hydrocolloids as substitutes for agar and gelatin as solidifying agents in microbiological media (4, 19, 25, 35, 44, 53). The functional properties of a food hydrocolloid depend on its origin, preparation method, thermal processing, and environmental conditions, such as salt content, pH, and temperature (20). In choosing a food hydrocolloid for a structured food model system, the physical and chemical nature of the target food product must therefore be taken into account.

So far, most studies in the field of predictive microbiology mention only the concentration of the gelling agent as a quantitative measure of food structure. Within our research group, we introduced the use of rheological properties as a more objective way to relate structural characteristics to microbial behavior (51). This allows comparisons between food model systems based on different thickening/gelling agents and accounts implicitly for the variability in structure between different brands or processing methods of the same gelling/thickening agent. Rheology quantifies the relation between stress and flow of materials, but its concepts can also be used to analyze behavior “at rest” (31). Its methods are widely used in the food industry as they are essential tools in product development, quality control, sensory evaluation, and the design of processing equipment (49).

Among the wide range of existing foodstuffs, sauces are known for their complex microstructure and typical rheological properties. The group of acid sauces includes both emulsions, such as mayonnaise and salad dressings, and concentrated suspensions, such as ketchup. These sauces are viscoelastic, i.e., they have both viscous and elastic properties, and typically show non-Newtonian flow behavior, characterized by the presence of a yield stress. Yield stress is the minimum shear stress required to initiate flow and is an indication of the suspension abilities of a fluid. Other characteristics of acid sauces are low pH, low water activity (aw), and the presence of organic acid preservatives. Due to this harsh environment, spoilage is predominantly caused by yeasts and lactic acid bacteria (22). Within this range of microorganisms, Zygosaccharomyces bailii is particularly troublesome because of its high resistance toward organic acid preservatives and its osmophilic behavior (26).

The objective of this research is to develop an experimental viscoelastic food model system for acid sauces. In a first step, a set of requirements is formulated, taking into account the main physicochemical properties and viscoelastic characteristics of the target food product. Several thickening and gelling agents are screened and evaluated with respect to these requirements, and a suitable thickening/gelling agent is selected. As a last step, the applicability of the model system is tested by performing growth experiments for the spoilage yeast Z. bailii.


Formulation of physicochemical food model system requirements.

The development of the structured food model system started with the selection of a liquid culture medium by taking into account the target food product and its typical microflora. Sabouraud liquid medium ([SAB] 30 g/liter; CM147; Oxoid, Hampshire, United Kingdom) was chosen here because this medium is often used for the cultivation of yeasts, molds, and acidophilic bacteria.

Subsequently, a set of requirements was formulated by considering the main physicochemical characteristics of acid sauces. Tomato ketchup contains high concentrations of sugar, resulting in a reduced aw of 0.93 to 0.95 (32). The pH is normally between 3.5 and 4.0, and the amount of vinegar added results in an acetic acid concentration of 0.8 to 1.0% (32). In the United States, the aqueous phases of mayonnaise and salad dressing typically contain 9 to 12% salt with 7 to 10% sugar and 3.0 to 4.0% salt with 20 to 30% sugar, respectively (47). For mayonnaise, the pH ranges from 3.6 to 4.0, with acetic acid representing 0.29 to 0.5% of the total product, while salad dressing has a pH of 3.2 to 3.9, with acetic acid accounting for 0.9 to 1.2% of the total product (21, 47). In Europe, the percentage of salt and sugar is mostly between 1 to 12% for mayonnaise while salad dressing typically contains 1 to 4% salt and 1 to 30% sugar (21). The pH values for mayonnaise and salad dressing range from 3.0 to 4.2, with 4.5 constituting the maximum legal value for mayonnaise in Denmark (21). Typical aw values for these sauces are 0.925 for mayonnaise and 0.929 for salad dressing (21). Based on these physicochemical characteristics, a set of environmental conditions to be reached by the model system was formulated (Table (Table1).1). Required pH and aw values were increased to higher values than usually observed in acid sauces in order to have a more complete picture. NaCl was considered as the first choice to lower the aw of the medium. However, the possibility of using other aw-lowering agents was not excluded because many thickening or gelling agents are known to be sensitive to NaCl (20).

Requirements for the acid sauce model system, covering typical features of tomato ketchup, mayonnaise, and salad dressing

Rheological measurements of acid sauces and food model systems.

Viscoelastic behavior can be characterized by performing dynamic oscillatory experiments. The values of the storage modulus, G′, and the loss modulus, G″, at low frequencies (0.01 to 0.1 rad/s) represent long-term behavior and can, therefore, be used to characterize the structure at rest. Acid sauces typically show gel-like characteristics (43); i.e., G′ predominates over G″ in the entire frequency range, as illustrated in Fig. Fig.1a.1a. Moreover, the slight frequency dependence of the moduli and the relatively large value of G″/G′ are typical of “weak” gels (43). This behavior differs from that of a “strong” gel, in which G′ is typically unaffected by the frequency, and a larger difference occurs between G′ and G″. According to the definition of Ross-Murphy (42), strong gels rupture above a critical deformation while weak gels flow without fracture under increasing deformation.

FIG. 1.
Illustration of viscoelastic behavior of a commercial ketchup (a) and determination of yield stress value using the tangent crossover method for a commercial tomato ketchup, mayonnaise, and salad dressing (b).

In this study, the frequency dependence and magnitude of G′ and G″ were evaluated for a variety of sauces and model systems. Frequency sweep tests were performed from 0.1 to 100 rad/s. Prior to these measurements, samples were tested over a range of strains to determine appropriate conditions for nondestructive testing. For this purpose, strain sweeps at a frequency of 10 rad/s were performed to determine the linear viscoelastic range.

Another characteristic of acid sauces is the presence of a yield stress, σy (Fig. (Fig.1b),1b), which is the minimum shear stress required to initiate flow. The magnitude of the yield stress can depend on the assumptions underlying the evaluation technique (49). In this study, the yield stress of the samples was determined by using the tangent crossover method (31), which is illustrated in Fig. Fig.1b1b for several acid sauces. This method is a rotational test in which the yield stress is determined as the shear stress value at which the range of reversible elastic deformation behavior ends and the range of the irreversible deformation behavior begins. The first tangent is fitted in the curve interval of the low-shear, linear-elastic deformation range, in which Hooke's law is valid. A second tangent is adapted in the measuring curve interval at higher shear stress where the sample flows. The tangent in the flow range has a clearly higher slope than the first tangent. The crossover point of both tangents occurs at the shear stress value σy, which is taken as the yield point. In this study, yield stress measurements were performed by increasing shear stress from 0 Pa up to values of 150 Pa, depending on the sample under study, with a time interval of 60 s between each measuring point.

Dynamic oscillatory and yield stress measurements were performed with a controlled stress rheometer (Physica MCR 501; Anton Paar GmbH, Graz, Austria) at 22°C. All measurements were carried out with a cone and plate geometry (50-mm 1° cone with 48-μm truncation). A set of conditions was also tested at 30°C in order to assess the effect of temperature on the rheological properties. A positive-displacement pipette (Microman M1000; Gilson Inc., Middleton, WI) was used to load the samples into the rheometer.

Formulation of rheological food model system requirements.

Starting from a liquid medium with the physicochemical characteristics summarized in Table Table1,1, a structured model system was developed that mimics the main rheological characteristics of acid sauces. Both viscoelastic behavior and the presence of a yield stress were included in the set of requirements. Requirements regarding the viscoelastic behavior are described in Table Table11 and were based on both the evaluation of several commercial acid sauces (according to the methods previously described and illustrated in Fig. Fig.1a)1a) and data obtained from literature (6, 17, 28, 34, 36, 43, 52). In literature, G′ values of approximately 100 Pa to 400 Pa have been observed in the low-frequency region (0.01 to 0.1 rad/s) for tomato ketchup (6, 28, 43, 52). For G″, values range between 10 and 100 Pa (6, 28, 43, 52). A wider range of values has been observed for mayonnaise and salad dressings, with G′ values of approximately 10 to 500 Pa and G″ values ranging between 0 and 200 Pa (17, 34).

Table Table11 also presents the requirements regarding yield stress values, which were based on a comparison of measurements performed on commercial acid sauces (according to the methods previously described and illustrated in Fig. Fig.1b)1b) with literature data obtained from different evaluation techniques (6, 11, 33, 36, 37, 40).

In addition to the physicochemical and rheological properties, transparency of the model system was also required (Table (Table1)1) in order to enable collection of microbiological data using optical density (OD) measurements and allow use of microscopical techniques.

Selection of thickening/gelling agent.

In order to develop a food model system that meets all the requirements stated in Table Table1,1, a literature search was performed to identify suitable thickening or gelling agents. In evaluating the applicability of these thickening/gelling agents toward the food model system, the following criteria were taken into account: stability toward temperature (i.e., the possibility to use autoclaving), practical use in microbiological experiments (i.e., neutral behavior toward growth of microorganisms and guaranteeing sterility during inoculation and sampling procedures), and stability toward time (i.e., no drying out of the medium and no evolution of the rheological behavior with time). Based on these criteria, several thickening and gelling agents were selected, and their suitability was screened more thoroughly with respect to the rheological properties described in Table Table11 and according to the methods previously described. This screening process was based on both dedicated data obtained at different concentrations and conditions of the thickening/gelling agents and data available in the literature. The tested gelling and thickening agents included the following: agar (0.3% [wt/vol]; Agar Technical, Oxoid), Carbopol (0.5 and 1.0% [wt/vol] carbomer 980, 0.5 and 1.0% [wt/vol] carbomer 981, 0.5 and 1.0% [wt/vol] Ultrez 20, and 1.0 and 2.0% [wt/vol] Aqua CC; Lubrizol Corporation, Wickliffe, OH), carboxymethyl cellulose (CMC) (0.5% [wt/vol] Cekol 30,000A; CP Kelco, Atlanta, GA), gellan gum (0.2% [wt/vol] with 4.0% [wt/vol] NaCl; Kelcogel F, CP Kelco), locust bean gum (0.5% [wt/vol] Genu Gum type RL-200; CP Kelco), and xanthan gum (1.0 to 4.0% [wt/vol] with 0 to 1.0% [wt/vol] NaCl; Keltrol TF; CP Kelco). All solutions were prepared according to the manufacturers' guidelines; included the required amount of SAB medium, glucose, and fructose (as stated in Table Table1);1); and were adjusted to pH 5.0.

Based on the results of the screening process, Carbopol was selected as the most suitable thickening or gelling agent for the food model system (more details are provided in the Results and Discussion sections). A thorough analysis of the rheological properties of a model system based on Carbopol 980 was subsequently performed at different pH values (4.0 to 5.0, at three equidistant levels) and different acetic acid (0 to 1.0% [vol/vol]; three equidistant levels), glycerol (0 to 15% [wt/vol]; four equidistant levels covering an aw range of 0.93 to 0.97), and Carbopol (1.0 to 1.5% [wt/vol]; two levels) concentrations. Additional measurements were performed for a Carbopol solution of 0.5% (wt/vol) Carbopol 980 and 5% (wt/vol) glycerol at pH 5.0. Glycerol was chosen as the aw-lowering agent instead of NaCl as the latter causes a breakdown of the structure and leads to a nontransparent medium. The prepared solutions contained 30 g/liter SAB medium (Oxoid), 5.5% (wt/vol) glucose (G-8270; Sigma, Steinheim, Germany), 7.5% (wt/vol) fructose (F-0127; Sigma), and the required amounts of glycerol (24388; VWR, Leuven, Belgium) and Carbopol 980 (Noveon Inc.). The amount of added glucose was lowered to 5.5% (wt/vol) as Sabouraud broth already contains 2% (wt/vol) glucose. The mixtures were vigorously stirred for at least 30 min (OST 20 basic; IKA Werke GmbH & Co. KG, Staufen, Germany) and autoclaved at 121°C for 15 min. After this, the required amount of acetic acid (818755; Merck KGaA, Darmstadt, Germany) was added, and pH values were adjusted with NaOH (Merck KGaA). pH was measured with a pH sensor (Documeter pH meter; Sartorius, Germany), and the aw was determined with an aw-kryometer Typ AWK-20 (Nagy Messsysteme GmbH, Gaufelden, Germany). Rheological measurements were performed according to the methods previously described. Prior to these measurements, samples were centrifuged to remove entrapped air bubbles.

Application of food model system in microbiological experiments.

In order to test the applicability of the developed model system, growth experiments were performed for the spoilage yeast Z. bailii. Four different concentrations of Carbopol (0, 0.5, 1.0, and 1.5% [wt/vol]) were tested at pH 5.0 and 5% (wt/vol) glycerol. As preliminary tests indicated that planktonic growth occurs at 0.5% Carbopol, the maximum specific growth rate, μmax, was estimated for this concentration, both from absorbance measurements and from viable counts, and compared to μmax at 0% Carbopol.

(i) Yeast strain and inoculum preparation.

Z. bailii (strain no. 174, culture collection of Laboratory of Food Microbiology and Food Preservation; Ghent University, Ghent, Belgium) was taken from a stock culture stored at −75°C. The strain was recovered in SAB medium by incubation at 30°C for 48 h, and afterwards it was maintained at 2°C on yeast-glucose-chloramphenicol slants (64104; Bio-Rad, Marnes La Coquette, France). In order to prepare the inoculum, cells from yeast-glucose-chloramphenicol slants were cultivated at 30°C in SAB medium for 24 h. A subculture of 5 ml was taken and grown again in 200 ml of SAB medium for 24 h at 30°C. The purity of the strain was checked by streaking on tryptic soy agar ([TSA] CM131; Oxoid) supplemented with 4% (wt/vol) fructose (F-0127; Sigma) to support the growth of Z. bailii.

(ii) Preparation of culture medium.

Growth medium containing Carbopol was prepared at a proportion of 5/3 compared to the regular levels previously described. Therefore, the medium contained 50 g/liter SAB medium, 9.17% (wt/vol) glucose, 12.5% (wt/vol) fructose, 8.33% (wt/vol) glycerol, and 0.83, 1.67, or 2.5% (wt/vol) Carbopol 980. The mixtures were vigorously stirred for at least 30 min (OST 20 basic; IKA Werke GmbH & Co. KG) and autoclaved at 121°C for 15 min. When necessary, the medium was centrifuged to remove entrapped air bubbles.

The growth medium without Carbopol was prepared with the regular amounts of SAB medium, sugar, and glycerol. The pH of the medium was adjusted to 5.0 by adding sterile HCl (Acros Organics).

(iii) Inoculation procedure for microtiter plates.

For the growth medium showing planktonic growth (i.e., at 0 and 0.5% [wt/vol] Carbopol), absorbance detection times were used to estimate the μmax. Precultures were diluted to provide an inoculation level of 2 × 104 CFU/ml, followed by five successive twofold dilutions. For the estimation of the μmax based on viable counts, a single inoculation level of 2 × 104 CFU/ml was used. For the growth medium containing 1.0 and 1.5% (wt/vol) Carbopol, precultures were diluted to provide an inoculation level of 2 × 104 CFU/ml. The inoculum densities of all media were verified by plating on TSA, supplemented with 4% (wt/vol) fructose.

All growth experiments were performed in 48-well microtiter plates (Greiner Bio-One, Austria) with a total growth medium volume of 500 μl per well. For the Carbopol medium, this volume was obtained by adding 300 μl of growth medium and 200 μl of a sterile NaOH solution to each well (hence, the factor 5/3 for growth medium preparation). The concentration of the NaOH solution was the concentration required to reach a final pH of 5.0 in the well. This was previously determined by obtaining calibration curves for each Carbopol concentration under study. The required NaOH concentrations were 0.18, 0.38, and 0.56% (wt/vol) for the growth media of 0.5, 1.0, and 1.5% Carbopol, respectively. Microtiter plate wells were filled by alternately adding 100 μl of growth medium and 100 μl of NaOH solution until the total volume of 500 μl per well was reached. This was done to reach an optimal level of mixing as Carbopol solutions immediately thicken upon addition of NaOH. For the transfer of the Carbopol medium, a positive-displacement pipette (Microman M100; Gilson Inc.) was used. Subsequently, microtiter plates were shaken for 2 min in a microplate shaker (MS 3 Digital; IKA Werke GmbH & Co. KG). Sufficient mixing was verified visually as the medium within each well became completely transparent due to the addition of NaOH.

For the growth media containing 0 and 0.5% Carbopol, separate microtiter plates were prepared for the absorbance measurements and viable count method. Microtiter plates used for absorbance measurements contained eight replicates of each inoculation level. For the growth media containing 1.0 and 1.5% Carbopol, 10 replicates were performed.

(iv) Growth measurements and data processing.

For growth assessment based on absorbance measurements, the microtiter plates were placed in a SpectraMax M2e microplate reader (Molecular Devices, Sunnyvale, CA) at an incubation temperature of 30°C. The OD of the medium at 600 nm was measured at regular time intervals, and the data were processed by the software package SoftMax Pro (Molecular Devices).

For the experiments at 0 and 0.5% (wt/vol) Carbopol, a single point measurement was performed in each well at a time interval of 5 min. Prior to each measurement, the microtiter plates were agitated for 20 s. OD growth curves were constructed by plotting the OD of the suspensions minus the OD at time zero versus the time of incubation.

For the media containing 1.0 and 1.5% (wt/vol) Carbopol, data collection was performed by using the well scan method, in which OD values were measured at nine different positions in a well. This method was also used to check if the medium was sufficiently mixed. If OD values at all nine positions were similar, sufficient mixing could be assumed. For each measuring point, the OD at time zero was subtracted from the OD of the suspensions, and the average of the nine values per well was calculated. This value was used to construct OD growth curves. If, at day 0, air bubbles were apparent at one (or more) of the nine measurement points of a certain well, these points were discarded. In any case, for the media with the highest levels of structure, all air entrapment disappeared after 2 days at most.

For the viable count method, the microtiter plates were separately incubated at 30°C. At each sampling time, two wells were sacrificed. Samples were appropriately diluted in physiological saline (0.85% NaCl) with peptone (0.1%) and plated on TSA supplemented with 4% (wt/vol) fructose. The petri plates were then incubated for 3 days at 30°C.

(v) Estimation of maximum specific growth rate.

Both absorbance detection times and viable counts were used to estimate the maximum specific growth rate, μmax, at 0 and 0.5% (wt/vol) Carbopol. For the absorbance measurements, μmax is estimated by using a simplified version of the method of Cuppers and Smelt (13), as described by Dalgaard and Koutsoumanis (15). In this method, μmax is calculated from plots of ln(inoculum size) against turbidity detection times by using equation 1:

equation M1

where ln(N0,i) [ln(CFU/ml)] is the natural logarithm of the initial cell population; μmax,DT (h−1) is the maximum specific growth rate determined from detection times (DT); DTi (h) is the absorbance detection time at an OD value of 0.09, and k is a constant (intercept). The suffix i refers to the different twofold serially diluted cultures.

For the viable count data, the primary model of Baranyi and Roberts (5) was used to fit the growth curves. An explicit version of this model is:

equation M2

equation M3

equation M4

where ln(N(t)) [ln(CFU/ml)] is the natural logarithm of the number of cells at time t (in hours), ln(N0) is the natural logarithm of the initial cell population; μmax,VC (h−1) is the maximum specific growth rate determined from viable counts (VC); ln(Nmax) is the natural logarithm of the maximal cell number, and q0 is a measure of the initial physiological state of the cells.

MatLab, version 6.5 (The MathWorks, Inc., Natick, MA), was used to perform model fits by linear or nonlinear regression. For the model of Baranyi and Roberts (5), nonlinear optimization of the parameter values was achieved by making use of the lsqnonlin routine of the Optimization Toolbox, version 3.0.2 (The MathWorks, Inc.) to minimize the sum of squared errors. The graphs were generated in MatLab.


Selection of thickening/gelling agent for the food model system.

This section reports on the results and the interpretation of (i) the literature study (16, 20, 42, 54) and (ii) specifically designed laboratory experiments aiming at completing the information available in the literature regarding the following thickening/gelling agents: agar, Carbopol, CMC, gellan gum, locust bean gum, and xanthan gum.

Table Table22 summarizes the rheological characteristics of aqueous solutions of the thickening/gelling agents under study (20, 42, 54). Thickening/gelling agents that do not exhibit the desired rheological characteristics are agar, alginate, CMC, gelatin, guar gum, and locust bean gum. These model systems all lack a yield stress and are either strong gels, which do not flow, or entanglement networks, in which G″ is larger than G′ at low frequencies, and a crossover occurs at increasing frequencies (Fig. (Fig.2a,2a, CMC and locust bean gum). It should be noted that some strong gels may have weak gel characteristics close to the critical gel concentration (Fig. (Fig.2a,2a, agar). Moreover, the behavior of thickening and gelling agents also depends on the origin, type, preparation method, thermal processing, and environmental conditions such as ionic strength, soluble solids content, pH, and the presence of acid. For instance, carrageenan requires the presence of specific ions, depending on the type used, and may show some degree of yield stress behavior. However, hydrolysis can occur when solutions below pH 4.3 are heat processed, which makes these solutions not suitable for autoclaving at the low-pH requirements of the model system. Pectin and starch are not of practical use either since microbiological degradation of these thickening/gelling agents can occur.

FIG. 2.
Illustration of viscoelastic behavior of various samples: growth medium based on agar, CMC, and locust bean gum (LBG) (a); growth medium containing xanthan gum (XG) (b); and growth medium containing Carbopol 980 (C980) in the presence and absence of acetic ...
Rheological characteristics of aqueous solutions of thickening and gelling agents considered for the acid sauce food model systema

Strong gellan gum gels are formed when a hot solution is cooled under quiescent conditions. Shearing during the cooling process disrupts the normal gelation and results in a pourable structured liquid or “fluid gel.” Gellan fluid gels have a weak-gel structure and a yield stress when cooled below the setting temperature, which can range from about 10°C to >50°C.

Xanthan gum is frequently used in acid sauces and is known for its typical yield stress behavior. Solutions immediately thicken upon addition of the xanthan gum powder and show a very good stability over a broad pH range (2 to 12) even in the presence of acids, such as acetic and citric acid. Another advantage is its resistance toward microbial degradation and high temperatures, the latter depending on the ion concentration. Several stock solutions of xanthan gum, containing the required amounts of SAB medium, glucose, and fructose, were tested. Figure Figure2b2b illustrates the effect of NaCl on the viscoelastic behavior of xanthan gum solutions. A strengthening effect was observed in the presence of NaCl. The storage moduli (G′), measured at 0.1 rad/s, were 4 and 14 Pa, and yield stress values were 4 and 9 Pa in the absence and presence of 2% (wt/vol) NaCl, respectively. As the values of G′ and σy at 1% are in the lower range of the requirements stated in Table Table1,1, higher concentrations were tested. A solution of 3% at pH 5.0 had a yield stress value of 14 Pa and a storage modulus G′ of 19 Pa at 0.1 rad/s (Fig. (Fig.2b),2b), which is still far below the maximum desired G′ values shown in Table Table1.1. This concentration was found to be at the limit of practical feasibility; i.e., higher concentrations were too viscous to sample the medium properly with a pipette, which limits the feasibility of inoculation and sampling procedures. Moreover, higher concentrations did not yield substantial increases in G′ (e.g., 23 Pa at 4%), indicating that there is a maximum achievable level of structure if xanthan gum is the only thickening/gelling agent used in the food model system.

Carbopol is a high-molecular-weight carbomer (a polymer of acrylic acid) that is often used as a model yield stress liquid in rheological research (14). Most types, such as Carbopol 980 and 981 and Ultrez 20, have a pH in the range of 2.5 to 3.5 when dissolved in water, and addition of a base is required to reach the typical yield stress and weak-gel behavior. Maximum thickening/gelling occurs within a pH range of 5 to 9, but lower pH values are possible if higher Carbopol concentrations are used. For all types, transparency and rheological behavior are affected by pH, salt, and acid concentration. For this study, Carbopol 980 is found to be the most suitable type as it is able to reach a higher level of structure than Carbopol 981 and Ultrez 20 (results not shown).

An extensive analysis of the rheological behavior of Carbopol 980 solutions was performed at different pH levels and different acetic acid, glycerol, and Carbopol concentrations. Figure Figure2c2c compares the viscoelastic behavior of a Carbopol solution of 1.0% (wt/vol) containing 1.0% (vol/vol) acetic acid at pH 4.0 with a solution of 1.5% Carbopol 980 and 0% acetic acid at pH 5.0. This figure shows that the Carbopol solutions exhibit a similar weak-gel behavior as the ketchup sample shown in Fig. Fig.1a.1a. G′ is larger than G″ in the entire frequency window under study, and both viscoelastic quantities show only a slight dependence on the imposed frequency of oscillation. Tables Tables33 and and44 show the results of the rheological characterization of the Carbopol 980 media. Carbopol concentration, acetic acid concentration, and pH all show a pronounced effect on the rheological properties and transparency of the model system. Higher levels of structure and transparency are obtained at decreasing acetic acid concentrations and increasing Carbopol concentrations and pH. The most critical set of conditions studied here, i.e., 1% Carbopol and 1% acetic acid at pH 4.0, still resulted in a clear solution. Lower pH values did not yield clear solutions in the presence of acetic acid. As indicated in Tables Tables33 and and4,4, the addition of glycerol did not significantly affect the rheological properties. In order to evaluate the effect of temperature, a set of conditions was also tested at 30°C. The selected conditions included combinations both at the edges and at intermediate values of the experimental design. Table Table55 indicates that neither G′ nor σy changed when the temperature was increased from 22°C to 30°C. However, a small increase in G″ was observed, indicating a slightly higher viscous contribution.

Summary of storage modulus G′ values at 0.1 rad/s and 22°C
Summary of yield stress σy values determined by the tangent crossover method at 22°C
Comparison of rheological properties at 22°C and 30°Ca

The aw values were also determined for all conditions under study. Carbopol concentration, acetic acid concentration, and pH did not affect the aw of the medium (results not shown). The only factor accounting for a change in aw was glycerol, with aw values of 0.982, 0.968, 0.950, and 0.932 at 0, 5, 10, and 15% (wt/vol) glycerol, respectively.

Application of food model system in microbiological experiments.

A shift from planktonic growth to growth in colonies was observed when the Carbopol concentration was increased from 0.5% to 1.0%. Rheologically, this corresponded to an increase of G′ at 0.1 rad/s from 39 Pa to 291 Pa and an increase of σy from 2.8 Pa to 65 Pa. Figure Figure33 shows a microtiter plate well containing the 1.0% Carbopol growth medium. Due to the presence of colonies, a smoothly developing OD growth curve could not be observed at 1.0 and 1.5% Carbopol (Fig. 4a and b). However, growth could be detected under these conditions as a clear detection time could be observed (17.8 ± 2.0 h and 22.8 ± 1.7 h at 1.0 and 1.5% Carbopol, respectively), and the average OD value obtained from nine different positions in a well increased clearly during the experiment.

FIG. 3.
Picture of Z. bailii colonies in a microtiter plate well in a growth medium at 1.0% Carbopol and 5% glycerol and pH 5.0. Colonies on the surface and submerged colonies are visible.
FIG. 4.
Illustration of growth curves obtained in four representative wells at 1.0% Carbopol 980 (a) and 1.5% Carbopol 980 (b).

Figure Figure5a5a shows the OD growth curves of a series of twofold dilutions at 0.5% Carbopol. Air entrapment at the beginning of the experiment resulted in a high variability of the OD baseline. However, air bubbles disappeared before the OD threshold value of 0.09 was reached. For each dilution, the average detection time of eight replicates was calculated, and the hypothesis of a linear relation with inoculum size was investigated (Fig. (Fig.5b).5b). For both liquid and Carbopol media, linear regression resulted in high r2 values. The μmax was calculated as the slope of the regression line and compared to the μmax obtained from viable count experiments and based on the model of Baranyi and Roberts (5) (Fig. (Fig.66 and Table Table6).6). For both media, absorbance measurements resulted in slightly higher μmax values (Table (Table6).6). Compared to liquid medium, lower values of μmax were obtained in the medium with Carbopol.

FIG. 5.
Series of OD growth curves at 0.5% Carbopol used for determination of absorbance detection times (a) and determination of μmax from linear regression between inoculum size and absorbance detection time in liquid medium (○) and ...
FIG. 6.
Growth curves of Z. bailii in liquid medium (○) and at 0.5% Carbopol (+), with primary fits of the model of Baranyi and Roberts (5).
Comparison of values of μmax (h−1) of Z. bailii cultures determined from viable counts and absorbance detection times for liquid medium and at 0.5% Carbopola


In the field of predictive microbiology, the small amount of research on the effect of food structure has been performed by using a structured food model system instead of a real food product. The approach adopted in this study, i.e., to develop a viscoelastic food model system starting from a target food product, has, as far as the authors are aware, not been carried out before.

Model system selection and analysis.

In order to develop a model system with the required rheological properties (Table (Table1),1), a suitable thickening/gelling agent had to be selected. Thickening/gelling agents are often added to food products to emulsify, stabilize, and modify the thickening properties (20, 54). However, the major contribution to the consistency and rheological characteristics does not come from the thickening/gelling agents but from the main constituents of the food products. For example, the consistency of tomato ketchup is predominantly determined by the composition of the raw material (maturity of tomatoes, particle size, particle interactions, and insoluble and total solid content) and processing conditions (temperature) (18). Thickening/gelling agents such as xanthan gum and guar gum are only added to adapt, rather than to create, the level of structure. Furthermore, combinations of thickening and gelling agents are often used in food applications in order to combine their functional properties or benefit from their possible synergistic interactions. For instance, xanthan gum is known to interact synergistically with galactomannans such as guar gum and locust bean gum. In this study, however, we opted to develop a food model system that could be of considerable practical use in microbiological studies, and, therefore, only a single thickening/gelling agent was considered.

A variety of frequently used thickening and gelling agents were screened and compared with respect to the requirements stated in Table Table1.1. Thickening and gelling agents are highly sensitive to processing procedures and environmental conditions. In order to select a suitable thickening/gelling agent for the model system, the physicochemical nature of the target food product had to be taken into account. Based on Table Table2,2, the most promising thickening/gelling agents were gellan gum, xanthan gum, and Carbopol. The gel structure of a gellan gum is created when a hot solution is cooled below its setting temperature. This implies that microbial cells should preferably be inoculated above the setting temperature. Because of the complexity and impracticalities related to this preparation method, gellan gum was not selected for the model system under study. The main limitation of xanthan gum is that the maximum achievable level of structure, i.e., a storage modulus G′ of around 20 Pa (Fig. (Fig.2b),2b), is still far below the maximum desired value (500 Pa). Results obtained at 1% xanthan gum suggest that a higher level of structure could be reached by the addition of salt. This result is in agreement with Ma and Barbosa-Cánovas (27), who observed a significant increase in G′ when 0.01 M Na+ was added to xanthan gum solutions of 0.5 and 1.0%. However, the strengthening effect also implies that the concentration of xanthan gum must be adapted in order to reach comparable levels of structure in the presence of NaCl as an aw-lowering agent. An additional drawback of the xanthan gum model system is its limited practical use. Solutions immediately thicken upon addition of the xanthan gum powder, which complicates the inoculation procedure for microbiological experiments. Microbial cells cannot be added to the solution prior to the addition of the xanthan gum powder as the latter is not sterile. This implies that the solution containing cells and the xanthan gum solution must be prepared separately and aseptically mixed in the appropriate proportion.

Based on the aforementioned considerations, Carbopol was selected as the most suitable thickening or gelling agent for the food model system. Among the different types of Carbopol, Carbopol 980 reaches the widest range of structures and was therefore chosen for the model system. An extensive analysis of the rheological properties of the Carbopol 980 model system was performed at different pH values (4.0 to 5.0) and different acetic acid (0 to 1.0% [vol/vol]), glycerol (0 to 15% [wt/vol], covering an aw range of 0.932 to 0.982), and Carbopol (1.0 to 1.5% [wt/vol]) concentrations. Compared to the requirements formulated in Table Table1,1, the Carbopol model system covers the whole range of aw values for acid sauces but does not fully reach the requirements regarding pH and acetic acid concentration. A minimum pH of 4.0 was chosen because lower pH values did not yield clear solutions in the presence of acetic acid. The maximum concentration of acetic acid was limited to 1.0% for the same reason. However, in both cases a higher level of transparency can be obtained by increasing the Carbopol concentration.

As shown in Tables Tables33 and and4,4, adapting the Carbopol concentration also affects the rheological properties. The microstructure of Carbopol solutions typically consists of water-swollen microgels (12). Above a critical concentration, these gel particles are closely packed together and form a sample-spanning network structure, characterized by yield stress behavior (12, 50). By adapting the concentration of Carbopol, a wide range of yield stress values and gel strengths can therefore be achieved. A solution of 1.5% Carbopol and 0% acetic acid at pH 5.0 has a G′ value near the upper limit of the required range (Table (Table3).3). The same solution, however, has a yield stress value above the desired range. This shows that it is not possible to reach an identical rheological fingerprint as in acid sauces, and the composition of the Carbopol mixtures should therefore be chosen with respect to the specific rheological property and target food product under study.

pH has the most pronounced effect on the rheological properties. The addition of NaOH enhances the formation of hydrogen bonds between the swelling microgels and water on the inside and outside, thus stabilizing and strengthening the network structure (50). Maximum thickening and gelling efficiency occurs in the pH range of 5 to 9, in which rheological properties do not significantly change. Below this range, a nonlinear decrease in gel strength and yield stress can be observed. Tables Tables33 and and44 show that the rheological properties at pH 4.5 and pH 5.0 are comparable, but an abrupt drop can be observed if pH is further lowered to a value of 4.0. The maximum G′ and σy at pH 4.0 and 0% acetic acid is around 200 Pa and 40 Pa, respectively. These values lie within the required range but could be increased by adapting the Carbopol concentration.

Acetic acid has a destructive effect on the structure of the medium. Similar to pH, a linear effect cannot be observed as the drop in G′ and σy is more pronounced when the acetic acid concentration is increased from 0 to 0.5% than when it is increased from 0.5 to 1.0% (Tables (Tables33 and and44).

Glycerol was chosen as an aw-lowering agent instead of NaCl as the latter is known to destabilize Carbopol solutions. Due to the anionic character of Carbopol polymers, precipitates are formed with large monovalent and polyvalent cations. This could be avoided by using higher Carbopol concentrations, but in combination with required concentrations of acetic acid an acceptable degree of structure could not be reached at the Carbopol concentrations used in this study. As shown in Table Table33 and Table Table4,4, glycerol did not have a significant effect on the rheological properties. Moreover, glycerol was the only factor accounting for a change in aw. By altering the glycerol concentration, the aw value of the medium can therefore be changed in a straightforward way.

Temperature is an important factor with regard to the structural and rheological characteristics of a material. The temperature values considered for this study were 22 and 30°C, which are relevant values with respect to the assessment of shelf stability. Moreover, 30°C is the optimal growth temperature for Z. bailii (23).

As shown in this study, both the rheological behavior and transparency of Carbopol growth medium depend to a great extent on pH and acetic acid concentration. However, by adapting the Carbopol concentration, the effect of these factors can be compensated so that comparable levels of structure can be reached at different pH values and acetic acid concentrations. Moreover, the range of experimental conditions can be extended to lower pH values and higher acetic acid concentrations, provided that the Carbopol concentration is sufficiently high.

Application of a food model system in microbiological experiments.

From a practical point of view, the close connection between the thickening/gelling properties of Carbopol and pH makes it possible to use pH as the structure-inducing factor in microbiological experiments. In a first step, microbial cells are added to a nearly liquid solution of Carbopol 980. After this, a basic solution (e.g., NaOH) is added to adapt the pH of the medium to the required value, thereby inducing the required level of structure.

The transparency of the model system is an important advantage as it enables the use of microscopical techniques and turbidity measurements, as with a Bioscreen system, which are cheaper and less time-consuming than traditional plating methods. In this study, the applicability of the model system with respect to collection of microbiological data was illustrated by performing growth experiments in a medium with planktonic growth. The linear relation between ln(inoculum size) and absorbance detection time at 0.5% Carbopol (Fig. (Fig.5b)5b) demonstrates the possibility to determine μmax from absorbance measurements. Moreover, good agreement was found with μmax values obtained from viable counts, with a value for the ratio Rmax,VCmax,DT) of 0.96 ± 0.03 (Table (Table6).6). Rmax,VCmax,DT) values for both liquid and Carbopol media (Table (Table6)6) correspond to those of Dalgaard and Koutsoumanis (15), who estimated an average R value of 0.96 based on the dilution method for various microorganisms. It should be noted, however, that these authors used the log-transformed four-parameter logistic model to estimate the growth rate from viable counts. The results shown in this study extend the practical use of the dilution method to structured media with planktonic growth.

Growth of Z. bailii at higher concentrations of Carbopol, i.e., 1.0 and 1.5%, caused a shift to growth in colonies (Fig. (Fig.3).3). For these media, the experimental setup used in this study, i.e., combination of a structured medium with OD measurements in microtiter plates (Fig. 4a and b), can be applied to growth/no-growth studies as in this case a binary decision, clearly visible detection or not, is sufficient. In order to estimate kinetic parameters from growth in colonies, image analysis techniques could be used. For instance, in a study by Skandamis et al. (46), growth rates were estimated from changes in the surface area of individual colonies.

To the authors' knowledge, no research has been conducted on the effect of Carbopol on microbial behavior so far. Estimations of μmax obtained for 0.5% Carbopol showed a lower value than obtained in liquid medium (Table (Table6).6). At higher Carbopol concentrations, growth was detected as an increase in OD, with a detection time of 17.8 ± 2.0 h and 22.8 ± 1.7 h at 1.0 and 1.5% Carbopol, respectively (Fig. (Fig.4).4). These results indicate that, within the concentration range tested and in the presence of 7.5% (wt/vol) glucose and 7.5% (wt/vol) fructose, Carbopol does not have a significant growth-promoting or toxic effect, which would prevent its use in microbiological media. Thus, the model system developed in this study may provide a useful means to investigate spoilage characteristics in acid sauces more closely. The authors believe that the systematic development of the model system starting from the target food product, combined with the assessment of structural properties by means of rheological properties, is a novel contribution within the field of food microbiology.


This project is supported by the Fund for Scientific Research—Flanders (FWO—Vlaanderen, project G.0565.06), by the Belgian Program on Interuniversity Poles of Attraction initiated by the Belgian Federal Science Policy Office, and by the K. U. Leuven Research Council (EF/05/006 [Center-of-Excellence Optimization in Engineering] and OT/03/30).


[down-pointing small open triangle]Published ahead of print on 25 September 2009.


1. Antwi, M., K. Bernaerts, J. F. Van Impe, and A. H. Geeraerd. 2007. Modelling the combined effects of structured food model system and lactic acid on Listeria innocua and Lactococcus lactis growth in mono- and coculture. Int. J. Food Microbiol. 120:71-84. [PubMed]
2. Antwi, M., A. H. Geeraerd, K. M. Vereecken, R. Jenné, K. Bernaerts, and J. F. Van Impe. 2006. Influence of a gel microstructure as modified by gelatin concentration on Listeria innocua growth. Innov. Food Sci. Emerg. Technol. 7:124-131.
3. Antwi, M., T. E. Theys, K. Bernaerts, J. F. Van Impe, and A. H. Geeraerd. 2008. Validation of a model for growth of Lactococcus lactis and Listeria innocua in a structured gel system: effect of monopotassium phosphate. Int. J. Food Microbiol. 125:309-329. [PubMed]
4. Babbar, S. B. 2006. Xanthan gum: an economical substitute for agar in microbial culture media. Curr. Microbiol. 52:287-292. [PubMed]
5. Baranyi, J., and T. A. Roberts. 1994. A dynamic approach to predicting bacterial growth in food. Int. J. Food Microbiol. 23:277-294. [PubMed]
6. Bayod, E., E. P. Willers, and E. Tornberg. 2008. Rheological and structural characterization of tomato paste and its influence on the quality of ketchup. LTW Food Sci. Technol. 41:1289-1300.
7. Brocklehurst, T. F., G. A. Mitchell, Y. P. Ridge, R. Seale, and A. C. Smith. 1995. The effect of transient temperatures on the growth of Salmonella typhimurium LT2 in gelatin gel. Int. J. Food Microbiol. 27:45-60. [PubMed]
8. Brocklehurst, T. F., G. A. Mitchell, and A. C. Smith. 1997. A model experimental surface for the growth of bacteria on foods. Food Microbiol. 14:303-311.
9. Brocklehurst, T. F., M. L. Parker, P. A. Gunning, H. P. Coleman, and M. M. Robins. 1995. Growth of food-borne pathogenic bacteria in oil-in-water emulsions: II. Effect of emulsion structure on growth parameters and form of growth. J. Appl. Bacteriol. 78:609-615. [PubMed]
10. Brocklehurst, T. F., R. B. Piggott, A. C. Smith, and D. C. Steer. 1996. An apparatus for studying the effect of transient temperature on growth of bacteria in a gel matrix. Food Microbiol. 13:109-114.
11. Campanella, O. H., and M. Peleg. 1987. Determination of the yield stress of semiliquid foods from squeezing flow data. J. Food Sci. 52:214-215 and 217.
12. Carnali, J. O. 1992. The use of dilute-solution viscometry to characterize the network properties of Carbopol microgels. Colloid Polym. Sci. 270:183-193.
13. Cuppers, H. G. A. M., and J. P. P. M. Smelt. 1993. Time to turbidity measurement as a tool for modeling spoilage by Lactobacillus. J. Ind. Microbiol. 12:168-171.
14. Curran, S. J., R. E. Hayes, A. Afacan, M. C. Williams, and P. A. Tanguy. 2002. Properties of Carbopol solutions as models for yield-stress fluids. J. Food Sci. 67:176-180.
15. Dalgaard, P., and K. Koutsoumanis. 2001. Comparison of maximum specific growth rates and lag times estimated from absorbance and viable count data by different mathematical models. J. Microbiol. Methods 43:183-196. [PubMed]
16. Fleet, G. 1992. Spoilage yeasts. Crit. Rev. Biotechnol. 12:1-44. [PubMed]
17. Gallegos, C., M. Berjano, and L. Choplin. 1992. Linear viscoelastic behavior of commercial and model mayonnaise. J. Rheol. 36:465-478.
18. Hayes, W. A., P. G. Smith, and A. E. J. Morris. 1998. The production and quality of tomato concentrates. Crit. Rev. Food Sci. Nutr. 38:537-564. [PubMed]
19. Hotchin, J. E. 1955. Use of methyl cellulose gel as a substitute for agar in tissue-culture overlays. Nature 175:352. [PubMed]
20. Imeson, A. 1999. Thickening and gelling agents for food. Springer, New York, NY.
21. International Commission on Microbiological Specifications for Foods. 2005. Microorganisms in foods 6: microbial ecology of food commodities. Kluwer Academic/Plenum Publishers, New York, NY.
22. Jay, J. M., M. J. Loessner, and D. A. Golden. 2005. Modern food microbiology. Springer, New York, NY.
23. Jermini, M. F. G., and W. Schmidt-Lorenz. 1987. Cardinal temperatures for growth of osmotolerant yeasts in broths at different water activity values. J. Food Prot. 50:473-478.
24. Koutsoumanis, K. P., P. A. Kendall, and J. N. Sofos. 2004. A comparative study on growth limits of Listeria monocytogenes as affected by temperature, pH and aw when grown in suspension or on a solid surface. Food Microbiol. 21:415-422.
25. Lines, A. D. 1977. Value of K+ salt of carrageenan as an agar substitute in routine bacteriological media. Appl. Environ. Microbiol. 34:637-639. [PMC free article] [PubMed]
26. Loureiro, V., and A. Querol. 1999. The prevalence and control of spoilage yeasts in foods and beverages. Trends Food Sci. Technol. 10:356-365.
27. Ma, L., and G. V. Barbosa-Cánovas. 1997. Viscoelastic properties of xanthan gels interacting with cations. J. Food Sci. 62:1124-1128.
28. Mackley, M. R., R. T. J. Marshall, J. B. A. F. Smeulders, and F. D. Zhao. 1994. The rheological characterization of polymeric and colloidal fluids. Chem. Eng. Sci. 49:2551-2565.
29. Malakar, P. K., T. F. Brocklehurst, A. R. Mackie, P. D. G. Wilson, M. H. Zwietering, and K. van't Riet. 2000. Microgradients in bacterial colonies: use of fluorescence ratio imaging, a non-invasive technique. Int. J. Food Microbiol. 56:71-80. [PubMed]
30. Meldrum, R. J., T. F. Brocklehurst, D. R. Wilson, and P. D. G. Wilson. 2003. The effects of cell immobilization, pH and sucrose on the growth of Listeria monocytogenes Scott A at 10 °C. Food Microbiol. 20:97-103.
31. Mezger, T. G. 2006. The rheology handbook: for users of rotational and oscillatory rheometers, 2nd ed. Vincentz Network, Hannover, Germany.
32. Michels, M. J. M., and W. Koning. 2000. Mayonnaise, dressings, mustard, mayonnaise-based salads, and acid sauces, p. 807-835. In B. M. Lund, T. C. Baird-Parker, and G. W. Gould (ed.), The microbiological safety and quality of food, vol. 1. Aspen Publishers, Inc., Gaithersburg, MD.
33. Missaire, F., C. G. Qiu, and M. A. Rao. 1990. Yield stress of structured and unstructured food suspensions. J. Texture Stud. 21:479-490.
34. Muñoz, J., and P. Sherman. 1990. Dynamic viscoelastic properties of some commercial salad dressings. J. Texture Stud. 21:411-426.
35. Nene, Y. L. 1996. Tapioca—a potential substitute for agar in tissue culture media. Curr. Sci. 70:493-494.
36. Ofoli, R. Y., R. G. Morgan, and J. F. Steffe. 1987. A generalised rheological model for inelastic fluid foods. J. Texture Stud. 18:213-230.
37. Paredes, M. D. C., M. A. Rao, and M. C. Bourne. 1988. Rheological characterization of salad dressings. 1. Steady shear, thixotropy and effect of temperature. J. Texture Stud. 19:247-258.
38. Parker, M. L., T. F. Brocklehurst, P. A. Gunning, H. P. Coleman, and M. M. Robins. 1995. Growth of food-borne pathogenic bacteria in oil-in-water emulsions. I. Methods for investigating the form of growth. J. Appl. Bacteriol. 78:601-608. [PubMed]
39. Prachaiyo, P., and L. A. McLandsborough. 2003. Oil-in-water emulsion as a model system to study the growth of Escherichia coli O157:H7 in a heterogeneous food system. J. Food Sci. 68:1018-1024.
40. Rani, U., and G. S. Bains. 1987. Flow behavior of tomato ketchups. J. Texture Stud. 18:125-135.
41. Ross, T., T. A. McMeekin, and J. Baranyi. 2000. Predictive microbiology and food safety, p. 1699-1710. In R. K. Robinson, C. A. Batt, and P. D. Patel (ed.), Encyclopedia of food microbiology, vol. 3. Academic Press, San Diego, CA.
42. Ross-Murphy, S. B. 1995. Structure-property relationships in food biopolymer gels and solutions. J. Rheol. 39:1451-1463.
43. Sharoba, A. M., B. Senge, H. A. El-Mansy, H. Elm Bahlol, and R. Blochwitz. 2005. Chemical, sensory and rheological properties of some commercial German and Egyptian tomato ketchups. Eur. Food Res. Technol. 220:142-151.
44. Shungu, D. 1983. Gelrite as an agar substitute in bacteriological media. Appl. Environ. Microbiol. 46:840-845. [PMC free article] [PubMed]
45. Skandamis, P., E. Tsigarida, and G.-J. E. Nychas. 2000. Ecophysiological attributes of Salmonella typhimurium in liquid culture and within a gelatin gel with or without the addition of oregano essential oil. World J. Microbiol. Biotechnol. 16:31-35.
46. Skandamis, P. N., T. F. Brocklehurst, E. Z. Panagou, and G.-J. E. Nychas. 2007. Image analysis as a mean to model growth of Escherichia coli O157:H7 in gel cassettes. J. Appl. Microbiol. 103:937-947. [PubMed]
47. Smittle, R. B. 1977. Microbiology of mayonnaise and salad dressing: a review. J. Food Prot. 40:415-422.
48. Stecchini, M. L., M. Del Torre, L. Sarais, O. Saro, M. Messina, and E. Maltini. 1998. Influence of structural properties and kinetic constraints on Bacillus cereus growth. Appl. Environ. Microbiol. 64:1075-1078. [PMC free article] [PubMed]
49. Steffe, J. F. 1996. Rheological methods in food process engineering. Freeman Press, East Lansing, MI.
50. Taylor, N. W., and E. B. Bagley. 1974. Dispersions or solutions—mechanism for certain thickening agents. J. Appl. Polym. Sci. 18:2747-2761.
51. Theys, T. E., A. H. Geeraerd, A. Verhulst, K. Poot, I. Van Bree, F. Devlieghere, P. Moldenaers, D. Wilson, T. Brocklehurst, and J. F. Van Impe. 2008. Effect of pH, water activity and gel micro-structure, including oxygen profiles and rheological characterization, on the growth kinetics of Salmonella Typhimurium. Int. J. Food Microbiol. 128:67-77. [PubMed]
52. Valencia, C., M. C. Sanchez, A. Circuelos, A. Latorre, J. M. Franco, and C. Gallegos. 2002. Linear viscoelasticity of tomato sauces products: influence of previous tomato paste processing. Eur. Food Res. 214:394-399.
53. Watson, N., and D. Apirion. 1976. Substitute for agar in solid media for common usages in microbiology. Appl. Environ. Microbiol. 31:509-513. [PMC free article] [PubMed]
54. Williams, P. A. 2000. Handbook of food hydrocolloids. Woodhead Publishing, Cambridge, United Kingdom.
55. Wilson, P. D. G., T. F. Brocklehurst, S. Arino, D. Thuault, M. Jakobsen, M. Lange, J. Farkas, J. W. T. Wimpenny, and J. F. Van Impe. 2002. Modelling microbial growth in structured foods: towards a unified approach. Int. J. Food Microbiol. 73:275-289. [PubMed]

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)