International Journal of Food Microbiology 137 (2010) 299–302

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / i j f o o d m i c r o

Short Communication

Use of a logistic model to assess spoilage by Byssochlamys fulva in clarified apple juice
Anderson de Souza Sant´Ana a,b,⁎, Philippe Dantigny c, Ana Claúdia Tahara a,
Amauri Rosenthal d, Pilar Rodriguez de Massaguer e
a
Department of Food Science, Faculty of Food Engineering, State University of Campinas. Campinas. SP. Brazil
b
Department of Food and Experimental Nutrition, Faculty of Pharmaceutical Sciences. University of São Paulo. São Paulo. SP. Brazil
c
Laboratoire de Génie des Procédés Microbiologiques et Alimentaires, ENS.BANA, Université de Bourgogne, Dijon, France
d
Brazilian Agricultural Research Corporation—EMBRAPA. Rio de Janeiro. RJ. Brazil
e
Department of Chemical Processes, Faculty of Chemical Engineering, State University of Campinas. Campinas. SP. Brazil

a r t i c l e i n f o a b s t r a c t

Article history: The percentage P (%) of spoiled bottles (n = 40) of clarified apple juice due to Byssochlamys fulva, was modeled by
Received 20 May 2009
using a logistic model: P = Pmax
1 + expðkðτ−tÞÞ
where Pmax (%) the maximum percentage of spoiled bottles, k (h− 1) a
Received in revised form 15 November 2009
Accepted 29 November 2009 slope parameter and τ (h) the time for P =Pmax/2. Bottles of pasteurized apple juice were inoculated with B. fulva
IOC 4518 ascospores for low and high initial loads, 4.8 ± 2.3 ascospores/100 mL and 19.3± 4.6 ascospores/
Keywords: 100 mL respectively and incubated at 21 °C and 30 °C. Pmax was not significantly different from 100% except for a
Byssochlamys low initial load at 21 °C. Model parameters were estimated with a good accuracy, RMSE in the range 3.89–7.50.
Spoilage Then the model was used to determine the time for 10% bottles spoiled, t10%. This time was greater at low initial
Logistic model loads, 57.4 and 104 h at 30 and 21 °C respectively, than at high initial loads 23.9 and 75.1 h at 30 and 21 °C
Fungi respectively. This study demonstrated that even at a very low initial contamination, clarified apple juice can be
Temperature
easily spoiled by B. fulva highlighting the importance of controlling critical control steps of fruit juice processing
Ascospores
(i.e., fruit washing, juice filtration and pasteurization).
Predictive mycology
© 2009 Elsevier B.V. All rights reserved.

1. Introduction
Heat resistant fungi may play an important role in the microbiological
stability of foods and in particular of fruit juices (Tribst et al., 2009). Within
this group, Byssochlamys, Neosartorya, Eupenicillium and Talaromyces
are the most relevant genera (Pitt and Hocking, 1999). Eupenicillium,
Paecilomyces and Talaromyces have been isolated from fruit juices, but
they are not usually responsible for juice spoilage. It should be noted,
however, that Paecilomyces variotii does cause spoilage in fruit juices, as
it is the anamorph of Byssochlamys spectabilis (Houbraken et al., 2008).
Byssochlamys and Neosartorya are commonly associated with the
spoilage of pasteurized fruit juices. The spoilage caused by heat-resistant
fungi is characterized by flavor alterations, phase separation due to the
production of pectinolytic enzymes, gas production, and production of
visible mycelium (Pitt and Hocking, 1999, Jay and Anderson, 2001).
The ability of heat resistant moulds to spoil fruit juices is due mainly to
their ability to survive the heat treatment 85–105 °C for a few seconds, to
grow under low oxygen tension and to grow at low pH values (Tournas,
1994). Despite ascospores counts usually found in raw materials being
low (less than 10 ascospores/100 mL) (Hocking and Pitt, 1984; Baglioni
et al., 1999; Massaguer, 2003), heat resistant moulds can be isolated from
pasteurized juices or involved in their spoilage (Tournas, 1994; Obeta and
Ugwuanyi, 1995; Salomão et al., 2004; Houbraken et al., 2006; Salomão
et al., 2007; Salomão et al., 2008). Recently, Sant'Ana et al. (2009) have
shown that even low numbers of B. fulva ascospores in raw materials
could survive the pasteurization process.
Predictive models can be very useful for predicting the extent of
spoilage of food products by fungi as a function of time, kind of product,
type of microorganism and environmental conditions. Primary kinetic
models such as the logistic functions have been used to determine the
time of growth and the time of toxin production by Clostridium
botulinum (Whiting and Call, 1993; Whiting and Oriente, 1997), the
proportion of tubes showing evidence of growth and toxin production
by C. botulinum (Daifas et al., 2003) and the germination time of several
fungi (Dantigny et al., 2002, Dantigny et al., 2007). This study aimed at
identifying the parameters of a logistic model and predicting the time at
which 10% of the apple juice bottles were spoiled by B. fulva under
common storage temperatures (21 °C and 30 °C).
2. Material and Methods
2.1. Mould

⁎ Corresponding author. Current Address: University of São Paulo. Av. Prof. Lineu
Byssochlamys fulva IOC 4518 was isolated from a clarified Brazilian
Prestes, 580, Bloco 14, CEP: 05508-900, São Paulo, SP, Brazil. Tel.: +55 11 3091 2191. apple juice (Salomão et al., 2008). This strain was identified according
E-mail address: assantana@usp.br (A. de Souza Sant´Ana). to the microscopic and macroscopic characteristics as described by

0168-1605/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijfoodmicro.2009.11.029
300 A. de Souza Sant´Ana et al. / International Journal of Food Microbiology 137 (2010) 299–302
Pitt and Hocking (1999). This organism was the most heat resistant
amongst three Byssochlamys strains surviving to heat shocks for up to
95 °C for 5 min (Sant'Ana et al., 2009).
2.2. Preparation of the ascospores suspension
Ascospores of B. fulva IOC 4518 were prepared as previously described
(Sant'Ana et al., 2009). The mould was grown at 30 °C for 30 days in Roux
bottles that contained 200 mL malt extract agar (MEA) sterile formula-
tion (20 g malt extract; 20 g glucose; 20 g agar-agar and 1 g bacteriolog-
ical peptone). Then, the ascospores were collected by scraping the
biomass from each Roux bottle. The suspension was filtered for removing
hyphal fragments, homogenized, centrifuged and ultra-sounded to
separate the ascospore clusters (Splittstoesser and Splittstoesser, 1977).
The final suspension was standardized to 107 ascospores/mL and stored
into sterile glass bottles containing glass balls at 2 °C±0.2 °C until used.
2.3. Apple juice preparation
Concentrated clarified apple juice (70ºBrix; pH = 3.74± 0.01 and
1.26 ± 0.04 % malic acid) was tested for the presence of heat resistance
fungi using the protocol described by Beuchat and Pitt (2001). The
concentrated clarified apple juice that did not exhibit any microorgan-
ism per 100 mL of juice was diluted to 11ºBrix with sterile distilled
water. Before inoculation, 150 mL of the diluted juice was dispensed in
glass bottles closed with polypropylene tops and sterilized at 105 °C for
10 min in an autoclave (Sant'Ana et al., 2009). Temperature was allowed
to stabilize in incubators at 21 °C or 30 °C for 48 h before inoculation.
2.4. Inoculation of the apple juice
Before apple juice inoculation, ascospores were heat activated at
75 °C for 5 min and enumerated according to the methodology described
by Beuchat and Pitt (2001). Optimum ascospores activation conditions
were previously determined as described in Sant'Ana et al. (2009). Low
and high initial loads of ascospores were 4.8 ± 2.3 ascospores/100 mL
and 19.3 ± 4.6 ascospores/100 mL, respectively. A total of 4 × 40 bottles
containing pasteurized apple juice (11°Brix) were inoculated under
aseptic conditions (disinfected laminar flow cabinet) with B. fulva IOC
4518 ascospores at low and high initial loads and incubated at 21 ± 1 °C
and 30 ± 1 °C for a maximum duration of 90 days (i.e., pasteurized apple
juice shelf-life) according to a full factorial design. For each of the four
different experimental conditions five non-inoculated bottles were
incubated as control. The experiments were repeated twice for each of
the conditions studied, a total of 320 bottles. Temperature storage
conditions of 21 °C±1 °C and 30 °C±1 °C represent an approximation of
the mean annual temperatures registered in the sub-tropical and
tropical regions of Brazil, (Climate Forecast and Climatic Studies Center,
CPTEC—www.cptec.inpe.br). Therefore, these conditions gave a pano-
rama of temperature conditions that B. fulva ascospores that survived
pasteurization would be submitted to, during transport, storage and
retailing.
2.5. Data treatment
Bottles were examined daily except for a high level of ascospores at
30 °C (twice a day). Development of fungal mycelium was assessed
visually by means of incident light. Following the concept of rejection
time introduced by Horner and Anagnostopoulos (1973) a bottle was
considered spoiled when at least one colony (diameter ≥3 mm) was
observed. The morphology of the colonies in all bottles were tested for
consistency and compared to a standard colony of the B. fulva strain used.
Additionally, at least 3 randomly chosen positive bottles per experimen-
tal condition were sampled for identification of the colonies. Prior to
pooling the data of the repeated experiments, a Student's t test was
performed to compare the variance and the mean (Judet et al., 2008).
2.6. Model fitting
The logistic model originally described by Whiting and Call (1993)
was used to represent the percentage of spoiled bottles P (%) as a
function of time t (h).
Pmax
P= ð1Þ
1 + expðkðτ  tÞÞ
Where: Pmax (%) the maximum percentage of spoiled bottles, k (h− 1)
a slope parameter and τ (h) the time for P =Pmax/2. t10% was chosen
based on Whiting and Call (1993), which mentioned that the time to
reach 0.1 was a better estimate than the common calculation of time for
half of samples to become toxic that were used for estimating the risk of
botulism. This time would allow industries to recall their products when
the first spoiled bottles were detected.
Rearranging Eq. (1):
h i
Ln Pmax
P 1
t = τ ð2Þ
k
 
P
Ln max 1
10 ð3Þ
for P = 10%; t10% = τ 
k
Non linear regressions were performed by using SlideWrite 5.0
(Advanced Graphics Software, Inc., Carlsbad, CA, USA) as described
previously (Dantigny, 1998). This software was based upon the
Levenberg-Marquardt Algorithm. The goodness of fit was evaluated
by means of the root mean square error, RMSE, (Ratkowsky, 2004)
and extracted from ANOVA tables.
3. Results and Discussion
B. fulva IOC 4518 was inoculated in pasteurized apple juice at two
inoculum levels (4.8 ± 2.3 and 19.3 ± 4.6 ascospores/100 mL) then
stored at 21 °C and 30 °C to assess spoilage. The low inoculation level
corresponded to the contamination load of heat resistant fungi that
may often be found in raw materials or in pasteurized fruit juices
(Hocking and Pitt, 1984; Baglioni et al., 1999; Massaguer, 2003). The
higher level was used to simulate a greater contamination of the raw
material. In all cases the maximum percentage of spoiled bottled was
not significantly different from 100%, with the notable exception of
experiments carried out at 21 °C for a low inoculum, Table 1. At 21 °C,
the slope parameter obtained at low inoculum was less than that at
high inoculum, but this observation was not significant at 30 °C. The
least and the greatest time constants τ = 49.6 h and τ = 145 h was
exhibited for high inoculum at 30 °C and for low inoculum at 21 °C,
respectively. However the time constants obtained at 21 °C for a high
inoculum and at 30 °C for a low inoculum were not significantly
different. The good quality of fit of the data by the logistic model was
obtained as supported by the RMSE values reported in Table 1. A
visual inspection of the fitting was also shown on Fig. 1. The curve
obtained for a low inoculum at 21 °C would suggest that after a longer
incubation time than 400 h, all bottles would have been spoiled, but
this was not observed experimentally. Overall the 95% confidence
intervals were rather small as indicated by great t-values (data not
shown), thus the model parameters were estimated with a good
accuracy. The parameters were then used to estimate t10% values,
Table 1. The time at which 10% of bottles were spoiled was the
smallest for high inoculum at 30 °C and the largest for low inoculum at
21 °C. In contrast with the estimation of τ, t10% for a high inoculum at
21 °C was greater than that for a low inoculum at 30 °C, although the
confidence intervals for t10% were not assessed.
Under marginal environmental conditions some spores are unable to
initiate the formation of the germination tube, leading to percent
 A. de Souza Sant´Ana et al. / International Journal of Food Microbiology 137 (2010) 299–302
Table 1
Results ±95% confidence intervals of the fit to the logistic model to experiments carried out at low and high initial loads (mean ± sd) at different temperatures.
Experimental conditions
Initial load (spores/100 ml) Temperature (°C)
4.8 ± 2.3 21
4.8 ± 2.3 30
19.3 ± 4.6 21
19.3 ± 4.6 30
RMSE: root mean square error. t10% was the time for P = 10%.
germinations below 100% (Dantigny, 2004). In the present study, at low
inoculum and 21 °C, 3 bottles out of 80 (3.75%), failed to exhibit one
colony within 90 days. This observation was unlikely to be due to these
bottles not being inoculated by any ascospore. For a low number of
ascospores, the distribution of the number of spores inoculated in each
bottle can be described by a Poisson law as suggested by the mean
4.8 spores/100 mL was approximately equaled to the variance
2.32 = 5.3. For 150 mL, the mean was η = 7.2 spores. The probability of
having a bottle not inoculated was equal to e− η = 0.075% (Box et al.,
1979). Therefore, the percentage of bottles exhibited no colony at 21 °C,
could not be explained by the low inoculum. In addition, for the same
inoculum all bottles were spoiled at 30 °C. It can be concluded that an
inhibitory effect on the germination of the ascospores was exhibited at
21 °C. However, this effect cannot be highlighted at the higher level of
inoculation because a bottle was spoiled as soon as 1 colony was visible
in the bottle.
It was shown in a previous study that the lag time for growth was
dependent on the temperature and the spore load (Dantigny et al.,
2002). Accordingly, it was shown in the present study that τ values
were reduced by 62 and 50% with increasing the temperature from 21
to 30 °C. The increase in ascospores level resulted in the decrease of τ
values: 68% and 55% at 21 °C and 30 °C respectively. It was demon-
strated that a larger number of spores can form visible mycelia earlier
than when less spores are inoculated (Sautour et al., 2003). In the
latter study, spores were inoculated at the same place, whereas
ascospores were inoculated randomly in the bottles in the present
study. It is suggested that the τ and t10% values were smaller for a high
inoculum than for a lower one, because the probability of having a fast
germinating ascospore was increased for a larger population.
In fact, the time for visible growth was dependant upon the ger-
mination time of the ascospores. The ascopores were not germinating at
the same time; there is some variability that depends on the physiologic
state of the spores (Judet et al., 2008) and many factors such as water
Fig. 1. Percentage of bottles (n = 40) spoiled by Byssochlamys fulva in clarified apple juice
versus time for low and high initial loads at different 5 temperatures. (○) 4.8 ± 2.3 spores/
▪
100 mL, 21 °C; (□) 4.8 ± 2.3 spores/100 mL, 6 30 °C; (●) 19.3± 4.6 spores/100 mL, 21 °C;
( ) 19.3± 4.6 spores/100 mL, 30 °C.
301
Parameter values
Pmax (%) k (h− 1) τ (h) RMSE t10% (h)
92.1 ± 2.1 0.051 ± 0.008 145 ± 4 3.89 104
99.5 ± 8.4 0.066 ± 0.020 90.6 ± 5.6 5.33 57.4
96.4 ± 7.5 0.089 ± 0.041 99.3 ± 5.7 7.50 75.1
101 ± 8 0.086 ± 0.025 49.6 ± 3.4 4.36 23.9
activity and temperature. The variability of the germination time
amongst a population of spores increased as environmental conditions
differed from the optimum conditions for germination (Judet et al.,
2008). Accordingly, a greater variability in the germination time of the
ascospores would have occurred at 21 °C than at 30 °C (i.e., optimum
temperature for B. fulva). A greater variability in the germination time
was synonymous for a decrease of the slope term of the germination
curve (Judet et al. 2008). By analogy to this observation, it can be
hypothesized that the variability in the time for visible growth was
greater at 21 °C than at 30 °C as suggested by the slopes of the spoilage
curves being less at 21 °C than at 30 °C.
The predictive model can only be applied to juice in clear containers,
such as glass bottles or clear PET plastic bottles. It would be less effective
for containers made from translucent plastics such as polypropylene or
polyethylene, and would be ineffective for containers such as aseptically
filled box-type products such as Tetra-Pak. This is a major problem for
application of the predictive model, as a significant proportion of
beverages that may potentially be affected by heat resistant mould
spoilage are packaged in such containers.
It is very difficult to assess losses attributable to heat resistant fungi. In
the fruit juice industry these losses varied greatly depended on season,
type of product and method of processing. A rough estimate of these losses
would be less than 1% of packages in a lot. It was shown in the present
study that the presence of about 5 spores/100 ml led to a contamination
of more than 90% bottles within 100 h. According to an analysis carried out
in our laboratory, the levels found in fruit juices susceptible to spoilage
were in the range 1 to 5 spores per 100 ml. This result suggested that the
great majority of commercialized fruit juice bottles contained no or less
than 5 ascospores, otherwise these bottles would have been contaminat-
ed. This is also in accordance with the statement of Pitt and Hocking
(1999) that a level higher than 5 ascospores/100 mL in samples taken just
before pasteurization are considered highly critic.
The present study highlighted that juice packages can show early
signals of spoilage, one to 4 days after processing depending on storage
condition and initial level of ascospores. This is of great concern in
addition to the fact that B. fulva ascospores can survive to continuous
apple juice pasteurization process adequately designed to cause 5
decimal reductions even when low counts are present in the raw
material (Sant'Ana et al., 2009). Thus, even if the heat resistant mould
count was homogenously distributed, a great number of packages in a
single batch would present contamination with the ascospores
(b101/100 mL). The aforementioned information make possible to
conclude that the reduced number of spoiled packed fruit juices caused
by heat resistant fungi can be considered as the result of a low contam-
ination (number of ascospores/100 mL) of the juice before pasteuriza-
tion, the heterogeneity of fungal contamination in raw material and food
products and also the occurrence of heat sensitive strains that prevented
the ascospores to survive pasteurization of the juice. Therefore, attention
should be paid to all the steps of the processing chain because apple juice
is not stored at refrigeration temperatures.
Acknowledgements
The authors thank to Conselho Nacional de Desenvolvimento Científico
e Tecnológico (CNPq), to Fundo de Apoio ao Ensaio, à Pesquisa e à Extensão
302 A. de Souza Sant´Ana et al. / International Journal of Food Microbiology 137 (2010) 299–302
(Faepex) (Processes 282/06 and 129/07) and to Prodetab (Embrapa)
(Process 042-01/01) for providing funds for this research.
References
Baglioni, F., Gumerato, H.F., Massaguer, P.R., 1999. Occurrence of heat resistant molds in
tomato pulp packed aseptically. Ciência e Tecnologia Alimentos 19, 258–263.
Beuchat, L.R., Pitt, J.I., 2001. Detection and enumeration of heat-resistant molds. In:
Downes, F.P., Ito, K. (Eds.), Compendium of the methods for the microbiological
examination of foods. APHA, Washington DC, pp. 217–222.
Box, G.E., Hunter, W.G., Hunter, J.S., 1979. Statistics for experimenters. An introduction
to design, data analysis, and model building. John Wiley and Sons, New York, USA.
653 pp.
Daifas, D.P., Smith, J.P., Blanchfield, B., Cadieux, B., Sanders, G., Austin, J.W., 2003. Effect
of ethanol on the growth of Clostridium botulinum. Journal of Food Protection 66,
610–617.
Dantigny, P., 2004. Predictive Mycology. In: McKellar, R.C., Lu, X. (Eds.), Modelling
microbial responses infoods. CRC Press, Boca Raton, pp. 313–320.
Dantigny, P., 1998. Dimensionless analysis of the microbial growth rate dependence on
sub-optimal temperatures. Journal of Industrial Microbiology and Biotechnology
21, 215–218.
Dantigny, P., Mansur, C.S., Sautour, M., Tchobanov, I., Bensoussan, M., 2002. Relationship
between spore germination kinetics and lag time during growth of Mucor racemosus.
Letters in Applied Microbiology 35, 395–398.
Dantigny, P., Marín, S., Beyer, M., Magan, N., 2007. Mould germination: data treatment
and modelling. International Journal of Food Microbiology 114, 17–24.
Hocking, A.D., Pitt, J.I., 1984. Food Spoilage fungi. II. Heat-resistant fungi. CSIRO Food
Research Quaterly 44, 73–82.
Horner, K.J., Anagnostopoulos, G.D., 1973. Combined effects of water activity, pH and
temperature on the growth and spoilage potential of fungi. Journal of Applied
Bacteriology 36, 427–436.
Houbraken, J., Samson, R.A., Frisvad, J.C., 2006. Byssochlamys: significance of heat resistance
and mycotoxin production. In: Hocking, A.D., Pitt, J.I., Samson, R.A., Thrane, U. (Eds.),
Advances in Food Mycology. Advances in Experimental Medicine Biology, vol. 571.
Springer, New York, pp. 211–224.
Houbraken, J., Varga, J., Rico-Munoz, E., Johnson, S., Samson, R.A., 2008. Sexual reproduc-
tion as the cause of heat resistance in the food spoilage fungus Byssochlamys spectabilis
(anamorph Peacilomyces variotii). Applied and Environmental Microbiology 74,
1613–1619.
Jay, S., Anderson, J., 2001. Fruit juice and related products. In: Moir, C.J., Andrew-Kabilafkas,
C., Arnold, G., Cox, B.M., Hocking, A.D., Jenson, I. (Eds.), Spoilage of processed foods:
causes and diagnosis. Australian Institute of Food Science and Technology Inc,
Waterloo, NSW Australia, pp. 187–197.
Judet, D., Bensoussan, M., Perrier-Cornet, J.-M., Dantigny, P., 2008. Distributions of the
growth rate of the germ tubes and germination time of Penicillium chrysogenum
conidia depend on water activity. Food Microbiology 25, 902–907.
Massaguer, P.R., 2003. Partial report of research project “Microbiological safety of fruit
juices and drinks asseptically filled”. Research project nº035-02/98 Prodetab.
Obeta, J.A.N., Ugwuanyi, J.O., 1995. Heat-resistant fungi in Nigerian heat-processed fruit
juices. International Journal of Food Science and Technology 30, 587–590.
Pitt, J.I., Hocking, A.D., 1999. Fungi and Food Spoilage. Aspen, Gaithersburg.
Ratkowsky, D.A., 2004. Model fitting and uncertainty. In: McKellar, R.C., Lu, X. (Eds.),
Modeling microbial responses in food. CRC Press, Boca Raton, FL, pp. 152–196.
Salomão, B.C.M., Costa, C.A., Massaguer, P.R., Aragão, G.M.F., 2004. Influence of distinct
pH values in the thermal resistance of Neosartorya fischeri isolated from apple
nectar processing. Alimentos e Nutrição 15, 7–10.
Salomão, B.C.M., Slongo, A.P., Aragão, G.M.F., 2007. Heat resistance of Neosartorya fischeri in
various juices. LWT—Food Science and Technology 40, 676–680.
Salomão, B.C.M., Massaguer, P.R., Aragão, G.M.F., 2008. Isolation and selection of heat
resistant moulds in the production process of apple nectar. Ciência e Tecnologia de
Alimentos 28, 116–121.
Sant'Ana, A.S., Rosenthal, A., Massaguer, P.R., 2009. Heat resistance and effects of conti-
nuous pasteurization on the inactivation of Byssochlamys fulva ascospores in clarified
apple juice. Journal of Applied Microbiology 107, 197–209.
Sautour, M., Dantigny, P., Guilhem, M.-C., Bensoussan, M., 2003. Influence of inoculum
preparation on the growth of Penicillium chrysogenunm. Journal of Applied Microbiology
95, 1034–1038.
Splittstoesser, D.F., Splittstoesser, C.M., 1977. Ascospores of Byssochlamys fulva compared
with those of a heat resistant Aspergillus. Journal of Food Science 42, 685–688.
Tournas, V., 1994. Heat-resistant fungi of importance to the food and beverage industry.
Critical Reviews in Microbiology 20, 243–263.
Tribst, A.A.L., Sant'Ana, A.S., Massaguer, P.R., 2009. Review: microbiological quality and
safety of fruit juices: past, present and future perspectives. Critical Reviews in
Microbiology 35, 310–339.
Whiting, R.C., Oriente, J.C., 1997. Time-to-turbidity model for non-proteolytic type B
Clostridium botulinum. International Journal of Food Microbiology 36, 49–60.
Whiting, R.C., Call, J.E., 1993. Time of growth model for proteolytic Clostridium botulinum.
Food Microbiology 10, 295–301.