The Growth Potential of Bacillus cereus in Ready-to-Reheat Vegetable Soups

: Bacillus cereus (hereafter, B. cereus ) poisoning often arises from the consumption of Ready-To-Reheat vegetable soups in which an intensive growth of the vegetative cells of B. cereus take place. The market for these soups is increasing signiﬁcantly worldwide. For the producer it is important to determine if soups can promote the growth of B. cereus , by calculating its growth potential. We can achieve this goal by carrying out an efﬁcient challenge test. In our study we have designed and performed a challenge test in three batches of an emmer ( Triticum monococcum ) and vegetable soup that undergo a second pasteurization treatment after packaging. We found out that under refrigeration conditions B. cereus is unable to multiply in the soup, instead, under conditions of thermal abuse, B. cereus can grow during 90 days of shelf life with a growth potential of 0.82 logarithms. It is essential to keep the entire production phase under control using effective GMP (Good Manufacturing Practices) and GHP (Good Hygiene Practices) measures, to ensure that the freshly produced soups contain low loads of the spores of B. cereus . In this way, the vegetative cells born from the germination of the spores cannot reach the infectious dose necessary to induce the food poisoning.


Introduction
The evolution of modern society has favored the growing popularity of Ready-To-Reheat (RTRH) foods [1,2].Soups are consumed all over the world with a wide range of ingredients [3,4] and RTRH ones have entered this market trend across the globe [5].Despite this favorable market situation, however, fresh RTRH soups are food matrices that can favor microbial growth and that can involve potential microbial hazards, especially foodborne pathogens such as B. cereus [6].Therefore, it becomes essential to keep the hygiene of soup's production process under careful control and to know the dynamic of the Bacillus spp.population, particularly of the B. cereus that can contaminate food ready for consumption.

The Production Process and Its Influence on the Soup Microbiota
Fresh RTRH industrially produced soups all have a relatively similar flow diagram [1] which is schematized in Figure 1.The ingredients are cooked in boiling water for no less than 45-60 min, usually blended, packaged and closed in bowls.In this phase, the temperature of the product remains around +80-85 • C. Packaging can take place in microbiologically aseptic chambers or in environments where there is normal air circulation.In environments with normal circulating air the soup can be newly contaminated with bacteria, yeasts and molds of environmental origin.This "secondary microbial association" would be able to develop in the product causing the spoilage of the product or making it potentially harmful for the consumer.Then, soups are subjected to a second heat treatment at +85-95 • C for 10-45 min depending on the mass of the food to be pasteurized.However, not all companies carry this out.
Hygiene 2023, 3, FOR PEER REVIEW 2 circulation.In environments with normal circulating air the soup can be newly contaminated with bacteria, yeasts and molds of environmental origin.This "secondary microbial association" would be able to develop in the product causing the spoilage of the product or making it potentially harmful for the consumer.Then, soups are subjected to a second heat treatment at +85-95 °C for 10-45 min depending on the mass of the food to be pasteurized.However, not all companies carry this out.Such a production process has remarkable influences on the microbiological characteristics of the soups.Each ingredient that enters a soup brings its own microbiota, which varies according to the ingredient.The first cooking in boiling water drastically reduces the load of the vegetative forms of bacteria, yeasts and molds.Instead, the spores of Bacillus spp.and Clostridium spp.can survive to the thermal treatment and a percentage of them, depending on the environmental conditions, are induced to germinate.The Such a production process has remarkable influences on the microbiological characteristics of the soups.Each ingredient that enters a soup brings its own microbiota, which varies according to the ingredient.The first cooking in boiling water drastically reduces the load of the vegetative forms of bacteria, yeasts and molds.Instead, the spores of Bacillus spp.and Clostridium spp.can survive to the thermal treatment and a percentage of them, depending on the environmental conditions, are induced to germinate.The second pasteurization should also guarantee the devitalization of the Bacillus spp. or Clostridium spp.cells born from the germination of their spores, ensuring a high level of absence of bacteria in the product ready for consumption [6].The evolution from inert spore to metabolically active vegetative cell involves an outgrowth phase that can take between 10 h and 40 h (or more) to complete [7].If the second heat treatment is carried out on the soups immediately after packaging, or a few hours after the former, the heat will not be able to inactivate the microbial forms in the germination phase because the process is still in its initial stages.
The most significant microbial hazards for these Ready-To-Reheat products are bacteria of the genera Bacillus spp.and Clostridium spp.whose spores can multiply in the soup reaching high loads harmful for the consumer [6].If, on the other hand, the growth involves species of toxin-producing bacteria such as B. cereus, Clostridium perfringens or Clostridium botulinum, the development of these bacteria up to high loads can make the soup concretely harmful for the consumer.More recently, other new species were identified through genetic taxonomic analyses and added to the B. cereus group, such as Bacillus gaemokensis, Bacillus manliponensis and Bacillus bingmayongensis [8][9][10][11][12].
In humans, the enterotoxigenic strains of B. cereus cause food poisoning and B. cereus sensu stricto is the most diffused pathogenic species; nevertheless, some authors have documented clinical cases of foodborne poisoning occasionally caused by B. weihenstephanensis, which seems to be able to synthesize the cereulide toxin as B. cereus does [13][14][15][16][17][18].Bacillus bacteria are ubiquitous in the environment and they can slightly contaminate all foodstuffs, starting with the raw materials.Their spores are able to adhere to any work surface within the food industries [19,20], enhancing their resistance to adverse environmental conditions if incorporated in the biofilms that many of the B. cereus strains are able to produce [21,22].B. cereus causes two forms of food poisoning in humans via two different types of toxins.Some strains of B. cereus cause a diarrheal gastroenteritis because they synthesize a series of thermolabile enterotoxic factors, including a non-hemolytic enterotoxin (NHE), an enterotoxin (entFM), cytolysin K, enterotoxin from B. cereus (bceT), a phospholipase, enterotoxin S, a sphingomyelinase and cereolysin O [14,23].Few strains are able to synthetize an emetic toxin called cereulide (CER).It is thermostable even up to +130 • C for 90 min.These clinical forms of poisoning are characterized by repeated and uncontrollable bouts of vomiting [24].The symptoms of a B. cereus poisoning appear in humans only when B. cereus exceeds very high loads in the food.In the case of the diarrheal form, there must be at least 10 5 -10 6 CFU/g of food, while in the case of the emetic form B. cereus must exceed 10 6 -10 7 CFU/g in food [25].
The Food Business Operators (FBOs) who produce foods at risk of proliferation of B. cereus, needs to know, by carrying out an appropriate challenge test, whether the food they produce allows or blocks the growth of a disease agent.Our study concerns the design and implementation of a specific challenge test in an emmer and vegetable soup, to calculate the growth potential of B. cereus at a refrigeration temperature as well as in conditions of a specific abuse temperature.

Challenge Test Design
This study was designed according to the ISO 20976-1:2019 standard [26] and only considering B. cereus.We have developed our challenge test taking a RTRH emmer and vegetable soup as the experimental model.
The vegetables present in the soup were: carrots, onions, celery, tomato, parsley, cabbage, rosemary, sage, courgettes, chard and semi-finished basil.In addition to vegetables, the soup contained emmer (Triticum monococcum) as a cereal, extra virgin olive oil and water.No preservatives were added to the soup.Furthermore, the emmer and vegetable soup we have chosen is an example of a soup that, after a first cooking at +90 • C for 45 min and after packaging, undergoes a second pasteurization reaching more than +85 • C for more than 15 min.All soup packs used in the study were purchased directly from a manufacturing company that provided us with the samples the day after production.Each bowl of emmer and vegetable soup had a net weight of 620 g and a shelf life of 90 days.

Determination of the Inter-Batch Variability of Emmer and Vegetable Soup
As a preliminary action, we measured the pH and a w values of five packages taken from ten different batches of emmer and vegetable soup at the end of production.The values obtained were analyzed using the calculator "inter-batch variability" developed by the European Union Reference Laboratory [27] and confirmed with an instrument supplied by the ISO 20976-1:2019.

Selection of the Strains to Be Inoculated
We chose two B. cereus strains:

•
A reference one: ATCC 11778 [25,26]; A "wild" strain of B. cereus isolated from the same type of soup subjected to our challenge test, and from the same manufacturer.It was biochemically identified as B. cereus/Bacillus thuringiensis using the BIOLOG ® system (BIOLOG Inc. 21124 Cabot Blvd., Hayward, CA, USA).The correct identification of B. cereus was then completed by evaluating, with a microscope and Gram stain, the absence of parasporal crystals in the cytoplasm of the cells, which are typical of B. thuringiensis.

Preparation of the Suspensions of the Vegetative Cells to Be Inoculated
According to the indications of the ISO 20976-1:2019 for the preparation of the vegetative cells to be inoculated, for each strain of B. cereus used to carry out our challenge test, we prepared two successive cultures.The first culture was set up by inoculating live bacteria cells in brain-heart infusion broth (Biokar Diagnostics, Allonne, France) and was incubated at +37 • C for 18-20 h.The purpose of this first culture was to reach the end of the exponential growth phase (log-phase) of the bacteria or the start of the stationary growth phase (lag-phase) to normalize the physiological state of the microbial populations used in the challenge test.
One milliliter of the aforementioned broth was then inoculated into 9 mL of a brain-heart infusion broth whose pH and a w values matched the natural characteristics of the emmer and vegetable soup subject to the challenge test (i.e., 6.3 pH and 0.98 a w as an average of the three batches used in this challenge test).This second step in a broth culture that closely imitated the chemical characteristics of the soup served to adapt the bacteria to be inoculated to the characteristics of food substrate and at the same time constituted the worst-case scenario.This second culture was incubated at +37 • C for 96 h in order to allow the bacterial suspension to reach the end of the logarithmic increase phase and the start of the stationary growth phase.
The initial target load of the second culture was fixed at approximately 10 7 CFU of vegetative cells per ml of suspension.Based on previous analysis it was possible to estimate this load, which was then confirmed by plate sowing.It was essential to have a load of 10 7 CFU/mL because it would then allow for the inoculation levels required by the standard ISO 20976-1:2019 and so, for each test unit, allow for a load equal to at least five times the quantification limit of the enumeration method and not greater than 10 4 CFU/g.
The bacterial suspension to be inoculated in the single test unit was prepared by mixing in equal volumes the suspensions of each of the two strains of B. cereus identified previously.
To determine the load reached by B. cereus in the original suspension, we prepared a series of decimal dilutions of the original bacterial suspension in eight tubes each containing 9 mL of peptone water (from 10 −1 to 10 −8 ) and we inoculated each dilution on plates of Mannitol egg Yolk Polymyxin-Agar MYP (Biokar Diagnostics, Allonne, France).The plates were incubated at +30 • C, counting the load after 24-48 h.The test method is used to determine the quantity of the vegetative cells and of the spores of presumptive B. cereus [28].The presumptive B. cereus is usually confirmed by seeding a suspected colony on a blood agar but the ISO 20976-1:2019 standard explicitly says that during a challenge test it is not necessary to conduct a confirmation test.

Test Units and Control Units
For each of the three tested batches of soup we inoculated 36 test units: 21 bowls of soup to be kept regularly at +4 • C for the duration of the test and another 15 inoculated packages for the thermal abuse test (Section 2.6).In each test unit we inoculated (with a sterile syringe) 5 mL of the suspension of B. cereus prepared beforehand, so as not to excessively modify the a w value of the food subject to the challenge test.At the same time, the quantity of 5 mL allowed for the inoculation levels required by the standard ISO 20976-1:2019: in the range between five times the quantification limit of the enumeration method and 10 4 CFU/g.The challenge could be considered valid only if the food did not originally contain B. cereus.As a check, we added 12 control units to the 36 test units, and they were also used to measure the pH, water activity (a w ) and redox values of the food at the beginning and at the end of the challenge test, as well as to quantify the Total Viable Count (TVC) of the soup as "background microflora".
Consequently, in the whole challenge test (3 batches of soup analyzed) we set up 108 test units inoculated with B. cereus and 36 control units, for a total of 144 soup packs.

Storage Conditions
For the soup analyzed, normal storage conditions require the food to be kept at temperatures below +4 • C for the entire duration of its shelf life.Moreover, in our specific case, the tests in thermal abuse were set up as follows: • Up to the 20th test day the soups were stored at +4 • C and this situation represented storage from the manufacture to retail where there are reasonably no temperature abuses;

•
From the 21st to the 35th day the bowls were kept at +8 • C to mimic retail storage;

•
From the 36th day the samples were moved to +20 • C for 4 h and then to +10 • C for the rest of the test.This last situation represented the moment of transport from retail to the consumer's home and then storage by the consumer.
A diagram of the storage condition is in Figure 2.

Analytical Parameters
The analyses were performed on the 1st, 20th, 35th, 45th, 51st, 60th and 90th days of testing.The moments of analysis were chosen considering the standard written in the ISO 20976-1:2019, considering the temperature changes during thermal abuse and considering the company and laboratory historical data.The units subjected to thermal abuse were analyzed starting from day 20.Three test units kept at +4 °C were considered in each of the seven steps of analysis and three test units kept in thermal abuse were considered in each of the five steps of analysis.In the test units the determination of the load of B. cereus according to the ISO 7932:2020 [28] standard was carried out.In each of the control units the determination of the TVC as "background" microbial flora was carried out using the

Analytical Parameters
The analyses were performed on the 1st, 20th, 35th, 45th, 51st, 60th and 90th days of testing.The moments of analysis were chosen considering the standard written in the ISO 20976-1:2019, considering the temperature changes during thermal abuse and considering the company and laboratory historical data.The units subjected to thermal abuse were analyzed starting from day 20.Three test units kept at +4 • C were considered in each of the seven steps of analysis and three test units kept in thermal abuse were considered in each of the five steps of analysis.In the test units the determination of the load of B. cereus according to the ISO 7932:2020 [28] standard was carried out.In each of the control units the determination of the TVC as "background" microbial flora was carried out using the standard ISO 4833-1:2013 [29].Every load was expressed as colony forming unit per gram or CFU/g.Furthermore, on the same control units the pH and redox values (with Crison GLP 22 pH-meter, Barcelona, Spain) and the water activity values (with Novasina ® LabMaster, Lachen, Switzerland) were measured.

Statistical Analysis
For each sampling point (T0, T1, T2, T3, T4, T5, T6) all the data intended as colony forming units per gram were converted into the base 10 logarithm.Then, the arithmetic mean of the three loads of B. cereus belonging to the test units and the relative standard deviations (s.d.) were calculated.The effect of time was analyzed using the One-way ANOVA model.Post hoc pairwise comparisons were performed using the Bonferroni correction.
The growth potential (∆) of B. cereus was evaluated by subtracting the logarithm of the initial average load from the highest average load found during the challenge test (Log 10 max-Log 10 i); this was carried out for both storage conditions.If the highest load value detected during the challenge test was that of the initial load, the growth potential was reported to be equal to zero.This means that during the shelf life of that food, the inoculated bacterium had no chance to grow.Indeed, its load probably decreased rather than increased.

Results and Discussion
In the first part of our research, we focused our attention on the design of the challenge test in emmer and vegetable soup to calculate the growth potential of B. cereus during the shelf life of soup maintained in two different temperature conditions.For the soup analyzed, normal storage conditions require the food to be kept at temperatures below 4 • C for the entire duration of its shelf life.We know that the cold chain of perishable products can suffer interruptions, especially during transport and storage in the consumers' refrigerator.It was therefore advisable to keep a part of the test units in conditions that could mimic the conditions of thermal abuse that the soup could undergo during its shelf life.In the second part of the research, we implemented the challenge test by analyzing each of the inoculated test units following a specific previously programmed period.
We have developed our challenge test, taking an emmer and vegetable soup as an experimental model for two reasons:

•
It is one of the best-selling and most-consumed products in Italy [5,30,31]; Cereal makes the soup potentially more at risk than others in promoting the growth of B. cereus.This foodborne pathogen has a high prevalence of diffusion in foods rich in proteins and/or in starch, such as emmer [32][33][34].
Based on the results obtained in the preliminary analysis using the calculator "interbatch variability" and the instrument supplied by the ISO 20976-1:2019 (Section 2.2.), there was significant variability between the different batches of the emmer and vegetable soup.We therefore planned the challenge test in three different batches of emmer and vegetable soup for B. cereus.
The search for B. cereus gave negative results in all control units, thus the results obtained in the entire challenge test could be considered valid.
As regards the results recorded via the single control units for the pH, redox and a w values and for the TVC determined as background microflora, the data obtained (Table 1) can be summarized as follows:

•
The control units of the 3 analyzed batches showed variable pH values from one batch to another (pH 6.12 in batch 1, pH 6.49 in batch 2, pH 6.34 in batch 3).In the inoculated test units and maintained regularly at 4 • C, we detected only a very slight increase of the pH value in all three batches analyzed.In the test units maintained in conditions of thermal abuse, on the other hand, we found a rather marked increase in pH in the test units of batch 1 while the pH remained practically constant in batches 2 and 3;

•
The redox potential always showed positive values, indicating the aerobic condition of the substrate.The recorded value tended to decrease during storage, especially in batch 1;

•
The a w value of the soups subject to the challenge test remained practically unchanged (0.98-0.99) throughout the test in all three batches of soup;

•
As regards the TVC, in the three batches of soup analyzed, at the beginning of the challenge we recorded values below the detection limit of the method (<10 CFU/g).After 90 days of testing, in the samples kept regularly at +4 • C the TVC still was <10 CFU/g in batches 2 and 3.It is likely that storing the test units at a constant refrigeration temperature limited the growth of B. cereus and that some of the inoculated cells died or sporulated during the test.This justifies the reduction of the load of B. cereus recorded in these two batches.As expected, the scenario was different in the samples subjected to thermal abuse storage where the TVC grew modestly, reaching 7.5 × 10 2 CFU/g (2.88 Log 10 CFU/g) in batch 2 and 3.2 × 10 3 CFU/g (3.51 Log 10 CFU/g) in batch 3, respectively.
As for batch 1, the TVC measured after 90 days in the samples kept regularly at +4 • C had increased moderately (up to 1.6 × 10 3 CFU/g or 3.20 Log 10 CFU/g) while we observed a marked increase (up to 1.8 × 10 6 CFU/g or 6.26 Log 10 CFU/g) in the samples kept under conditions of thermal abuse.This increase in the background microbial load could have a correlation to the increase of the pH value in batch 1 and probably also to the increase of the load of B. cereus in the test units.
As for B. cereus, the levels of inoculated vegetative cells were all within the range mentioned in Section 2.4 and then better described in Section 2.5.For each single analytical step we considered the arithmetic average among the three values coming from the counts of the three test units.We collected all the data of the test and control units in tables (Tables 2-4) after transforming them into the logarithmic value (Log 10 CFU/g) required by the ISO 20976-1:2019 standard.The graphs are in Figures 3-5.
In batch 2 and 3, we recorded a regular and progressive decrease in the B. cereus load from the start to the end of the test, even beyond the 90th day (end of the shelf life).The load reduction in batch 2 and 3 was more marked in the samples kept at +4 • C, while it was slower in the samples subjected to the thermal abuse test at +8 • /+10 • C. It is likely that storing the test units at constant refrigeration temperatures discouraged the growth of B. cereus and that some of the inoculated cells died during the test or that they have sporulated.This justifies the reduction of the load of B. cereus recorded in these two batches of soup.1.10 ± 0.17 b 1.72 ± 0.12 b T4 (day 51) <1 ± 0.00 c <1 ± 0.00 c T5 (day 60) <1 ± 0.00 c <1 ± 0.00 c T6 (day 90) <1 ± 0.00 c <1 ± 0.00 c In batch 1 we recorded a load increase between the 20th and 51st test day (up to 1200 CFU/g at the 51st day), and only in the test units kept at +8 • C and then at +10 • C.This concentration peak of viable cells of B. cereus could be due to a multiplication of viable cells, followed by a new reduction of the specific load, due to the death of viable cells or to a sporulation of these last ones.Nonetheless, if the number of spores increases, it is not dangerous for the production of toxins since they are inert.
For each batch of test units, the growth potential of the inoculated bacterium was calculated as reported in the statistical analysis in Section 2.8.At the end of the challenge test, we obtained three values corresponding to the individual growth potentials calculated for each of the three food batches tested.In Table 5 we summarized them.1.10 ± 0.17 b 1.72 ± 0.12 b T4 (day 51) <1 ± 0.00 c <1 ± 0.00 c T5 (day 60) <1 ± 0.00 c <1 ± 0.00 c T6 (day 90) <1 ± 0.00 c <1 ± 0.00 For each batch of test units, the growth potential of the inoculated bacterium was calculated as reported in the statistical analysis in Section 2.8.At the end of the challenge test, we obtained three values corresponding to the individual growth potentials calculated for each of the three food batches tested.In Table 5 we summarized them.In batches 2 and 3, the estimated growth potential was always equal to zero, because the highest value we recorded in the tests turned out to be the initial load.This means that in these two batches of emmer and vegetable soup B. cereus was not able to multiply.
In the test units of batch number 1, instead, we recorded a growth peak of B. cereus in the samples kept under conditions of thermal abuse, a peak that is not found in the test units regularly maintained at refrigeration temperature.The load increase occurred between the 45th and the 60th day of storage, i.e., in the last third of the product's shelf life.The growth potential calculated is 0.82 logarithms.
In our opinion, the different growth potentials recorded between the three batches subjected to challenge test are due to the small but inevitable differences in chemicalphysical and microbiological characteristics that occur between one batch and another of the same product.
As a rule, established by the ISO 20976-1:2019 standard, the final estimated growth potential of the bacterium inoculated in the food must be the greater value of the three obtained.In our study it is 0.82 logarithms.

Conclusions
As far as we know, no challenge test results have been published in the scientific literature on the dynamics of B. cereus in vegetable and cereal-based soups and our research is the first experiment conducted with this aim.In particular, this is the first challenge test for B. cereus in vegetable soups that has been designed and conducted according to the guidelines dictated by the ISO 20976-1:2019 standard.
Based on the data obtained, we can establish that not complying to the recommended storage of a soup made of cereals and vegetables (shelf life of 90 days) could result in an increase of 0.82 logarithms of the initial load of B. cereus.It should occur mainly between the 30th and the 51st day of storage, and only in conditions of temperature abuse.
The different growth potentials that we found between the three batches of soup tested lead us to believe that the growth dynamics of B. cereus are probably influenced by the growth of any residual microflora that may have remained in the soup, probably made up of other species of spore-forming bacteria whose spores are induced to germinate after the second pasteurization treatment.
The changes in the pH value that we detected in the test units stored under conditions of thermal abuse are probably due to the residual TVC metabolism that developed in some of the control units compared to others.

Figure 1 .
Figure 1.A generic example of a flowchart for making soup (the phases that we evaluated as Critical Control Points valid for keeping the different hazards under control are highlighted in red).

Figure 1 .
Figure 1.A generic example of a flowchart for making soup (the phases that we evaluated as Critical Control Points valid for keeping the different hazards under control are highlighted in red).

Figure 2 .
Figure 2. A diagram showing the design of the storage conditions.

Figure 2 .
Figure 2. A diagram showing the design of the storage conditions.

Figure 3 .Figure 3 .
Figure 3. Graph representing the loads of B. cereus in batch number 1 during the challenge test period.By convention, 0.95 Log10 CFU/g was used to represent data <1 Log10 CFU/g.Table3.Average value (µ) ± s.d. of the loads of B. cereus detected in batch 2. The loads are reported for the test units maintained regularly at +4 °C as well as for the test units maintained in the conditions of thermal abuse at +8° and +10 °C.Superscripted letters mean significant different values for p < 0.05.

Figure 4 .
Figure 4. Graph representing the loads B. cereus in batch number 2 during the challenge test period.By convention, 0.95 Log10 CFU/g was used to represent data <1 Log10 CFU/g.

Figure 4 .
Figure 4. Graph representing the loads B. cereus in batch number 2 during the challenge test period.By convention, 0.95 Log 10 CFU/g was used to represent data <1 Log 10 CFU/g.

Figure 5 .
Figure 5. Graph representing the loads of B. cereus in batch number 3 during the challenge test period.By convention, 0.95 Log10 CFU/g was used to represent data <1 Log10 CFU/g.

Figure 5 .
Figure 5. Graph representing the loads of B. cereus in batch number 3 during the challenge test period.By convention, 0.95 Log 10 CFU/g was used to represent data <1 Log 10 CFU/g.
1.2.Bacillus Cereus as Foodborne Disease Agent What microbiologists once called Bacillus cereus is now called the Bacillus cereus group because within this species there are various sub-species (B.cereus sensu stricto, B. thuringiensis, B. anthracis, B. mycoides, B. pseudomycoides, B. weihenstephanensis and B. toyonensis).

Table 1 .
Values of pH, redox, a w and TVC found at the beginning (T0) and at the end (T6) of the challenge test.The reported values are the average (µ) of the three values obtained by the individual test units analyzed for each analytical phase.

Table 2 .
Average value (µ) ± s.d. of the loads of B. cereus detected in batch 1.The loads are reported for the test units maintained regularly at +4 • C as well as for the test units maintained in the conditions of thermal abuse at +8 • and +10 • C. Superscripted letters mean significant different values for p < 0.05.

Table 3 .
Average value (µ) ± s.d. of the loads of B. cereus detected in batch 2. The loads are reported for the test units maintained regularly at +4 • C as well as for the test units maintained in the conditions of thermal abuse at +8 • and +10 • C. Superscripted letters mean significant different values for p < 0.05.

Table 4 .
Average value (µ) ± s.d. of the loads of B. cereus detected in batch 3. The loads are reported for the test units maintained regularly at +4 • C as well as for the test units maintained in the conditions of thermal abuse at +8 • and +10 • C. Superscripted letters mean significant different values for p < 0.05.

Table 5 .
The growth potentials (δ Log 10 ) of B. cereus calculated for each batch of soup analyzed.The loads are referred for the test units maintained regularly at +4 • C as well as for the test units submitted to a thermal abuse test at +8 • and +10 • C.

Table 5 .
The growth potentials (δ Log10) of B. cereus calculated for each batch of soup analyzed.The loads are referred for the test units maintained regularly at +4 °C as well as for the test units submitted to a thermal abuse test at +8° and +10 °C.