ARTICLE IN PRESS

International Dairy Journal 16 (2006) 4051

www.elsevier.com/locate/idairyj

Physical characteristics of yoghurts made using exopolysaccharideproducing starter cultures and varying casein to whey protein ratios
T. Amatayakula, A.L. Halmosb, F. Sherkatb, N.P. Shaha,
a

Faculty of Science Engineering and Technology, School of Molecular Sciences, Victoria University, Werribee Campus, PO Box 14428,
Melbourne City MC, Victoria 8001, Australia
b
Department of Food Science, RMIT University, City Campus, GPO Box 2476V, Melbourne 3001, Victoria, Australia
Received 12 February 2004; accepted 11 January 2005

Abstract
This study investigated the physical characteristics of set and stirred yoghurts made at 9% (w/w) total solids with various casein
(CN)-to-whey protein (WP) ratios and with exopolysaccharide (EPS)-producing starter cultures (capsular or ropy) during storage.
The yoghurt was evaluated for composition, rmness and syneresis of set yoghurt, and for the ow curve and the area of hysteresis
loop between the upward and downward curve of stirred yoghurt. Viable counts of starter bacteria and concentrations of lactic acid
and EPS in the yoghurt were also determined. EPS concentration did not decrease during storage for 28 d. Firmness and syneresis of
set yoghurt decreased when the CN-to-WP ratio was reduced from ratio 4:1 to 1:1 and when EPS starter cultures (especially ropy
EPS) were used. Stirred yoghurt with a CN-to-WP ratio of 3:1 and made using ropy EPS-producing starter cultures had a higher
shear stress and hysteresis loop area than yoghurt made using capsular EPS- or non-EPS-producing starter cultures. The results
suggested that the physical characteristics of set and stirred yoghurts can be improved by varying CN-to-WP ratio and by the use of
EPS-producing starter cultures.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Yoghurts; Exopolysaccharide; Texture; Rheology; Storage

1. Introduction
Appearance and physical characteristics are important quality parameters of yoghurt. Good quality
yoghurt should be thick and smooth with no signs of
syneresis. Set yoghurt with a high level of syneresis on
the surface may be regarded as a low quality product,
even though this is a natural phenomenon. Conventionally, syneresis is reduced by increasing the total solids of
yoghurt mix to around 14% (w/w) with dry dairy
ingredients (Tamime & Deeth, 1980) or by using
stabilizers. Dry dairy ingredients such as skim milk
powder (SMP), whey protein isolate (WPI), whey
Corresponding author. Tel.: +61 3 9216 8289;
fax: +61 3 9216 8284.
E-mail address: nagendra.shah@vu.edu.au (N.P. Shah).

0958-6946/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.idairyj.2005.01.004

protein concentrate (WPC), sodium (Na)-caseinate or

calcium (Ca)-caseinate are commonly used to increase
the solids content of the yoghurt mix. Nevertheless,
fortication with these ingredients affects production
costs. The use of stabilizers including gelatin, modied
starches, or gums may affect the consumer perception of
yoghurt. The use of stabilizers is also prohibited in some
European countries (De Vuyst & Degeest, 1999).
Yoghurts fortied with casein-based ingredients
(SMP, Na-caseinate or Ca-caseinate) showed an increase in rmness (or viscosity) and a reduction in
syneresis compared with unfortied yoghurt (Modler,
Larmond, Lin, Froehlich, & Emmons, 1983; GuzmanGonzalez, Morais, & Amigo, 2000; Remuef, Mohammed, Sodini, & Tissier, 2003). On the other hand,
there were no consistent trends between the physical
characteristics of yoghurts and the addition of whey

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

protein (WP)-based ingredients (WPI or WPC). Modler

et al. (1983) and Guzman-Gonzalez, Morais, Ramos,
and Amigo (1999) found that yoghurt supplemented
with WPC had lower apparent viscosity and rmness
than control yoghurt made without fortication. Baig
and Prasad (1996) and Bhullar, Uddin, and Shah (2002)
found that supplementation of milk with WPC improved the apparent viscosity and textural properties of
the resultant yoghurt. These differences may be due to
the variations in the composition of whey protein-based
ingredients. Remuef et al. (2003) showed that extending
the heating time of milk supplemented with WPC from 1
to 5 min at 90 1C increased the apparent viscosity of
stirred yoghurt. On the other hand, the apparent
viscosity of stirred yoghurt fortied with Na- and Cacaseinate was not affected by the increase in the heating
time. Nonetheless, fortication with WPC reduced
syneresis dramatically.
There has been an increasing trend in the use of
starter cultures that are able to produce exopolysaccharides (EPS). These EPS are heteropolysaccharides and
exist in two forms: capsular (associated with bacterial
cell surface) and ropy or slimy (secreted into the
environment) (Cerning, 1990). In some cases, bacteria
can produce both forms of EPS. EPS-producing starter
cultures are becoming increasingly popular due to their
high water binding and texture promoting abilities. Both
capsular and ropy EPS possess high water binding
ability resulting in increased water retention in yoghurt
(Wacher-Rodarte et al., 1993; Hassan, Frank, Schmid,
& Shalabi, 1996b; Jaros, Rohm, Haque, Bonaparte, &
Kneifel, 2002) and cheeses (Perry, McMahon, & Oberg,
1997). Fermented dairy products produced using EPSproducing starter cultures showed lower rmness than
yoghurts made using control culture (Hassan, Frank,
Schmid, & Shalabi, 1996a); in addition, these authors
found that each strain of EPS-producing starter culture
inuenced the rheological properties of yoghurt differently. Stirred yoghurt made with ropy EPS-producing
starter cultures had higher apparent viscosity than
yoghurts made using capsular and non-EPS-producing
starter cultures, respectively (Marshall & Rawson,
1999).
The use of EPS-producing starter cultures in yoghurt
manufacture has the potential to replace or reduce the
use of stabilizers as well as added dairy ingredients.
Recent work by Puvanenthiran, Williams, and Augustin
(2002), who studied the effects of varying casein (CN)to-WP ratios on physical characteristics of set yoghurt,
showed improvements in physical characteristics (syneresis and gel strength) when the ratio of CN-to-WP were
reduced. By combining the use of EPS-producing starter
cultures with the alteration of CN-to-WP ratio, it may
be possible to maintain comparable physical characteristics to normal yoghurts without the need for fortication of milk or the use of stabilizers. The objective of

41

this study was to investigate the combined effects of

varying CN-to-WP ratio and the use of EPS starter
cultures on the physical characteristics of set and stirred
yoghurt, with a low total solids content, throughout a
28 d storage period at 4 1C.

2. Materials and methods

2.1. Experimental design and statistical analysis
Set and stirred yoghurts were produced using three
types of starter cultures (non-EPS-, capsular EPS- or
ropy EPS-producing starter cultures) and four CN-toWP ratios (4:1, 3:1, 2:1 or 1:1) in triplicate giving a total
of 72 batches. All measurements on yoghurts were
carried out in triplicate at 1, 7, 14, 21 and 28 d.
2.2. Microorganisms
Non-EPS-producing starter cultures (Streptococcus
thermophilus ASCC 1342 and Lactobacillus delbrueckii
ssp. bulgaricus ASCC 1466), capsular EPS-producing
starter culture (S. thermophilus ASCC 285), and ropy
EPS-producing starter culture (S. thermophilus ASCC
1275) were used in this study. These bacteria were
previously obtained from the Australian Starter Culture
Research Centre, Werribee, Australia, and were characterized by Zisu and Shah (2003) for their EPS
production. Stock cultures were maintained at 80 1C
in 12% (w/w) sterile reconstituted skim milk (RSM) and
40% (v/v) sterile glycerol. The microorganisms were
activated in 9% (w/w) sterile RSM and for 18 h. The
process was repeated three times prior to yoghurt
manufacture. S. thermophilus and L. delbrueckii ssp.
bulgaricus were incubated at 37 and 42 1C, respectively.
Viable counts in yoghurt were enumerated separately. S.
thermophilus was enumerated aerobically in M17 agar
(Amyl Media Pty. Ltd., Dandenong, Australia) at 37 1C
for 48 h, and L. delbrueckii ssp. bulgaricus was enumerated anaerobically in MRS agar (Merck, Darmstadt,
Germany) at 42 1C for 48 h (Dave & Shah, 1998).
2.3. Manufacture of yoghurts
Low heat SMP (34% (w/w) total protein, Murray
Goulbourne Co-operative Co. Ltd., Brunswick, Australia), WPC 80 (76% (w/w) total protein, United Milk
Tasmania Ltd., Spreyton, Tasmania, Australia) and
lactose monohydrate (Merck, Darmstadt, Germany)
were blended at suitable quantities in order to vary CNto-WP ratios to 4:1, 3:1, 2:1 and 1:1. Lactose was added
to balance the level of total solids. Dry ingredients were
blended to 9% total solids (w/w) and hydrated with
distilled water overnight followed by heat treatment at
85 1C for 30 min (at natural pH of milk blends), cooled

ARTICLE IN PRESS
42

T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

(42 1C) followed by inoculation with 1% (v/v) each of

non-EPS- or EPS-producing strain of S. thermophilus
(S. thermophilus ASCC 1342: non-EPS; S. thermophilus
ASCC 285: capsular; S. thermophilus ASCC 1275: ropy)
and L. delbrueckii ssp. bulgaricus ASCC 1466. Inoculated mix (100 mL) was incubated in a closed-top plastic
container (top diameter 69 mm, bottom diameter 60 mm
and 46 mm height) at 42 1C until a pH of 4.70 was
reached; the yoghurts were then transferred to a walkin-cooler (4 1C). For making stirred yoghurt, set yoghurt
was pressed with a spoon through a sieve (pore size
1 mm2), and stored at 4 1C overnight.
2.4. Compositional analysis of liquid milk
The contents of total solids, fat, ash and protein
(Kjeldahl method) were determined according to methods described in AOAC (1995). Casein concentration
was determined by the difference between the level of
total protein and non-casein nitrogen. The level of whey
proteins was determined by the difference between noncasein nitrogen and non-protein nitrogen. Lactose was
quantied enzymatically using a sucrose/lactose/glucose
kit (Megazyme, Wicklow, Ireland). Proteins in samples
were precipitated by 12% (w/v) TCA (Sigma-Aldrich
Co., St. Louis, MO, USA) and ltered out using
Whatman no. 1 lter paper (Whatman International
Ltd. Maidstone, England). The pH of the clear ltrate
was adjusted to 4.50 using 2 M NaOH, and diluted 20
times with acetate buffer, pH 4.50. Quantication of
lactose was carried out according to the procedure
described in the test kit.

acid standard was purchased from Sigma (Sigma

Chemical Co., St. Louis, MO, USA).
2.6. EPS purification and quantification in yoghurt
Proteins in 50 mL of diluted yoghurt sample (1:1
yoghurt:Milli-Q water; Millipore Corp, Bedford, MA,
USA) were precipitated by adding 4 mL of 20% (w/v)
TCA (Sigma Chemical Co.). Precipitated proteins were
separated by centrifugation at 3313  g for 30 min at
4 1C. The supernatant was adjusted to pH 6.8 with 40%
(w/v) NaOH followed by boiling the supernatant at
100 1C for 30 min to denature whey proteins. Denatured
whey proteins were separated by centrifugation at
3313  g for 30 min at 4 1C. An equal volume (25 mL)
of cold absolute ethanol was mixed with the supernatant
to precipitate the carbohydrates from the supernatant.
The precipitation was carried out overnight at 4 1C, and
the precipitate was separated by centrifugation at
3313  g for 30 min at 4 1C. The resultant carbohydrate
pellet was completely dissolved by adding 10 mL of
Milli-Q water and the resultant suspension was subjected to sonication for 1 h at room temperature. The
resultant solution was dialysed at 4 1C in a dialysis
membrane tube with molecular mass cut-off value of
13,000 Da (Carolina Biological Supply Company, NC,
USA) against tap water over a 2-week period. Water
was changed twice a day. The EPS concentration was
quantied using the phenol-sulphuric method of Dubois, Gilles, Hamilton, Rebers, and Smith (1956) and
was expressed as glucose equivalent.
2.7. Determination of spontaneous syneresis of
undisturbed set yoghurt

2.5. Determination of lactic acid concentration of yoghurt

The concentration of lactic acid was determined by
high performance liquid chromatography (HPLC). The
HPLC system (Varian, Varian Associates, Walnut
Creek, CA, USA) consisted of a solvent delivery system
(Varian, model 9012) connecting with an autosampler
(Varian, model 9100), a variable wavelength ultraviolet
(UV)visible (VIS) light detector (Varian, model 9050)
and an organic acid analysis column (Aminex HPX87H, 300  7.8 mm, Bio-Rad Lab, Richmond, CA,
USA). The method of Shin, Lee, Pestka, and Ustunol
(2000) was followed. The mobile phase used was 0.009 N
H2SO4, and the ow rate was set at 0.6 mL min 1. The
temperature of the column was set at 65 1C. Lactic acid
was detected by UV detector at 220 nm. Five grams of
the sample were mixed with 100 mL of 15.8 N HNO3 and
5.9 mL of 0.009 N H2SO4 before centrifugation of a
1.5 mL aliquot at 11,600  g for 15 min at room
temperature. The supernatant was ltered through a
0.45 mm membrane lter (Schleicher & Schvell, Dassel,
Germany) and 50 mL of the ltrate was injected. Lactic

Spontaneous syneresis of undisturbed set yoghurt was

determined using a siphon method. The principle of the
siphon method was adapted from a method used by
Lucey (2001), who studied the amount of spontaneous
whey separated from acidied milk gels fermented in
volumetric ask. The method used by Lucey (2001) was
modied in order to prevent any effects of container
geometry, and the difference in heat transfer between
volumetric ask and the yoghurt cup, on syneresis. In
this study, a cup of set yoghurt was taken out from the
walk-in-cooler (4 1C), weighed and kept at an angle of
approximately 451 to allow the whey on the surface to
collect on the side of the cup. A needle connected to a
syringe was used to siphon the liquid whey from the
surface of the sample, and the cup of yoghurt was then
re-weighed. The siphon was carried out within 10 s to
prevent further leakage of whey from the curd. The
syneresis was expressed as the percentage weight of the
whey over the initial weight of the yoghurt sample.
Measurement of syneresis by centrifugation and drainage methods have been widely used, but were not

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

suitable for yoghurts with 9% (w/w) total solids made as

in this study. In addition, they did not represent the
actual values of syneresis of whey separation on the
surface of set yoghurt, and are likely to cause damage to
the structure of the yoghurt. Thus, the results would
represent the syneresis of broken curd.
2.8. Firmness of set yoghurt
Firmness of set yoghurt was dened as the maximum
force used in compression using TA-XT.2 Texture
Analyzer (Stable Micro System, Goaldming, UK) with
a P20 probe (diameter 20 mm) and 25 kg load cell. The
speed was set at 1 mm s 1. The ratio of the diameter of
yoghurt cup to diameter of probe ratio was 3.5:1.0.
According to Bourne (2002), it is generally accepted that
the boundary or wall effects will diminish when the
diameter of sample is at least three times higher than the
diameter of the probe. Yoghurt samples were compressed to 75% of their original height. The test was
carried out immediately after removing the samples
from the cold room.

43

effect as discussed by Haque, Richardson, and Morris

(2001). The use of high shear rate (1000 s 1), as used by
other researchers (Teggatz & Morris, 1990; Bhattacharya, 1999; Hassan et al., 1996a), would have severely
destroyed the structure of the stirred yoghurt resulting
in non-differentiation of ow behaviour between samples. Therefore, constructing the ow curve at low shear
rate (050 s 1) should be suitable for observing the ow
curve of yoghurt produced with low total solids. In
addition, the area of hysteresis loop between the upward
and downward curves was determined. This hysteresis
loop area represents the structural breakdown of stirred
yoghurt during shearing, as described by Halmos and
Tiu (1981), Ramaswamy and Basak (1991) and Hassan
et al. (1996a, 2003).
2.10. Statistical analysis
The data were analysed with one-way analysis of
variance with SPSS version 10.0 for Windows (SPSS
Inc., NY, USA). The comparison between means of
data was carried out using the Tukey honestly signicant difference test.

2.9. Flow curve and the area of hysteresis loop of stirred

yoghurt
3. Results and discussion
The ow curve of stirred yoghurt was constructed
using a RS 50 RheoStress (Haake Rheometer, Karlsruhe, Germany) using a coaxial measuring cell, Z20
DIN. A ow curve was constructed using the method of
Halmos and Tiu (1981), Ramaswamy and Basak (1991),
Hassan et al. (1996a) and Hassan, Ipsen, Janzen, and
Qvist (2003). The ow curve was constructed by
increasing shear rate from 10 to 50 s 1 in 200 s (upward
curve) followed by decreasing shear rate from 50 to
10 s 1 in 200 s (downward curve). The samples (10 mL)
were loaded into the measuring cell by using a spoon
and the temperature of the sample was allowed to drop
to 10 1C prior to commencing the measurement. Shear
rate in this range was chosen to prevent the slippage

3.1. Composition of liquid milk blends

Table 1 shows the total solids, protein, fat, lactose and
ash contents of the milk blends with various CN-to-WP
ratios. All constituents in milk were kept constant, but
the CN-to-WP ratio was varied. The reduction in the pH
of yoghurts made using non-EPS-producing starter
cultures with low CN-to-WP ratios (ratio 1:1) was faster
than that for yoghurts with higher ratios (ratio 4:1)
(Fig. 1). This resulted in a shorter fermentation time of
yoghurt made with low CN-to-WP ratio. Puvanenthiran
et al. (2002) reported opposite results. The difference in
fermentation time between yoghurts made from milk

Table 1
Compositional parameters of milk blends used in the manufacture of yoghurtsa
Compositional parameters

Total solids (%, w/w)

Total protein (%, w/w)
Fat (%, w/w)
Lactose (%, w/w)
Ash (%, w/w)
Casein-to-whey protein ratio

Casein-to-whey protein ratio

4:1

3:1

2:1

1:1

8.8970.01a
3.0870.07a
o 0.1a
4.9770.07a
0.4070.21a
4.3670.25a

8.8770.12a
3.0770.26a
o 0.1a
4.9970.11a
0.4570.01a
2.9870.03b

8.8570.04a
3.0970.02a
o 0.1a
4.9670.05a
0.4070.07a
2.1370.01c

9.0670.17a
3.0670.02a
o 0.1a
5.0170.09a
0.3670.12a
1.2170.02d

a
Milks were prepared by blending the desired levels of reconstituted low heat skim milk powder, whey protein concentrate (76% protein) and
lactose monohydrate to give varying casein-to-whey protein ratios. Presented values are the means of three replicate trials; 7 indicates standard
deviation from the mean. Mean values (7standard deviation) within the same row not sharing a common superscript differ signicantly (Po0.05).

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

44

blends with different ratios of CN-to-WP, as observed in

this study, may be due to the difference in the buffering
capacity of milk blends. According to Walstra and
Jenness (1984), whey proteins generally have lower
buffering capacity than casein. Therefore, lowering the
concentration of CN would result in a reduction in
buffering capacity. However, the reduction in buffering
capacity may not be the only factor contributing to the
shorter fermentation time of yoghurt made with low
CN-to-WP ratios (Fig. 1). The increase in available
nutrients from WP may partly inuence the growth of
7.0
6.5

pH

6.0
5.5
5.0
4.5
0

yoghurt bacteria, and hence shortening the fermentation

time. Baig and Prasad (1996) and Dave and Shah (1998)
reported that WP stimulated the growth of S. thermophilus.
3.2. Lactic acid concentration
The concentration of lactic acid decreased when CNto-WP ratio was reduced (Table 2). Lactic acid
concentration varied from 0.60% (w/w) in yoghurt
made from a milk blend with a CN-to-WP ratio of 1:1 to
1.00% (w/w) in yoghurt made from milk blend with a
CN-to-WP ratio of 4:1. A similar trend was observed for
yoghurts made with any of the three types of starter
cultures. This could be due to the shorter fermentation
time of yoghurts produced from blends with low CN-toWP ratios, even though the pH at the end of the
fermentation was similar. This also suggests that milks
with low CN-to-WP ratios have low buffering capacity.
In general, the types of starter cultures did not affect the
concentration of lactic acid. A slight increase in the
concentration of lactic acid was observed in all yoghurts
during the 28 d storage period.

Fermentation time (h)

3.3. Viable counts

Fig. 1. Reduction in pH of yoghurts made using non-EPS-producing

starter cultures with varying casein-to-whey protein ratios during
yoghurt manufacture. The milk blends were prepared from the desired
levels of reconstituted low heat skim milk powder, whey protein
concentrate (76% (w/w) protein) and lactose monohydrate to give
casein-to-whey protein ratios of 4:1 (J), 3:1 (K), 2:1 (B), and 1:1
(E). Each value is the mean of three replicate trials. See Section 2.2
and Table 1 for details of starter culture and blend composition.

Table 3 shows the viable counts of S. thermophilus

and L. delbrueckii ssp. bulgaricus in yoghurts made with
varying CN-to-WP ratio using the three types of starter
cultures, during the 28 d storage period. At 1 d, the
viable counts of S. thermophilus in yoghurts with
different ratios of CN-to-WP were similar, whereas the

Table 2
Concentration of lactic acid (%, w/w) in yoghurts prepared from milk blends with varying casein-to-whey protein ratios, as described in Table 1, and
made using non-exopolysaccharide producing (control) cultures, capsular exopolysaccharide-producing starter culture or ropy exopolysaccharideproducing starter culturesa
Starter culture

Casein-to-whey
protein ratio

Storage period (d)

1

Non-exopolysaccharideproducing starter cultures

Capsular exopolysaccharideproducing starter cultures

Ropy exopolysaccharideproducing starter cultures

7
g,A

14
e,A

21
d,e,A

28
d,e,A

0.9570.03d,A

4:1

0.9370.03

3:1
2:1
1:1
4:1

0.8570.04d,e,f,g,A
0.7570.06b,c,d,A
0.6970.04a,,b,c,A
0.8570.03e,f,g,A

0.8870.04d,e,A
0.8570.01c,d,A,B
0.7570.04a,b,c,A,B
1.0170.04e,B,C

0.8770.08b,c,d,e,A 0.9070.06c,d,e,A
0.7970.05a,b,c,d,A,B 0.8970.02b,c,d,e,B
0.6970.03a,A
0.8070.03a,b,c,d,A,B
1.0270.02e,C
0.9770.09e,A,B,C

0.8470.01b,c,d,A
0.8070.06a,b,c,A,B
0.7870.05c,d,A,B
0.8970.04c,d,A,B

3:1
2:1
1:1
4:1

0.8070.05c,d,e,f,A
0.7270.05b,c,A
0.6070.01a,A
0.8970.02f,g,A,B

0.9370.06d,e,B
0.8270.05a,b,c,d,A,B
0.7270.04a,b,A
0.9070.01d,e,A,B

0.9470.03d,e,B
0.8670.05b,c,d,e,B
0.7170.02a,b,A
0.9970.05e,B

0.8070.06a,b,c,A
0.7770.08a,b,c,A,B
0.6970.08a,A
0.8970.02c,d,A,B

3:1
2:1
1:1

0.8470.02d,e,f,g,A 0.9170.05d,e,A
0.7570.04b,c,d,e,A 0.8470.03b,c,d,A
0.6970.01a,b,A
0.7070.04a,A

0.9970.06

0.9570.07

0.9670.09

0.8670.03a,b,c,d,e,A,B
0.7770.02a,b,c,A,B
0.6970.07a,A
0.8670.08b,c,d,e,A

0.9270.06c,d,e,A
0.8970.03c,d,e,A
0.7870.09a,b,c,A 0.7970.08a,b,c,d,A
0.7970.05a,b,c,d,B 0.7270.02a,b,A,B

0.8870.04c,d,A
0.8570.04b,c,d,A
0.7270.03a,b,A,B

a
Presented values are the means of three replicate trials; 7 indicates standard deviation from the mean. For details of starter cultures, see Section
2.2 and 2.3. Mean values (7standard deviation) within the same row not sharing a common superscript (A,B,C) differ signicantly (Po0.05). Mean
values (7standard deviation) within the same column not sharing a common superscript (a,b,c,d,e,f,g) differ signicantly (Po0.05).

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

45

Table 3
Viable counts (log10 cfu g 1) of S. thermophilus (ST) and L. delbrueckii ssp. bulgaricus (LB) in yoghurts prepared from milk blends with different
casein-to-whey protein ratios, as described in Table 1, and made using different starter culturesa
Starter culture

Casein-to-whey
protein ratio

Storage period (d)

1

Non-exopolysaccharide-producing starter cultures

Capsular exopolysaccharide-producing starter cultures

Ropy exopolysaccharide-producing starter cultures

4:1
3:1
2:1
1:1
4:1
3:1
2:1
1:1
4:1
3:1
2:1
1:1

14

21

28

ST

LB

ST

LB

ST

LB

ST

LB

ST

LB

8.89
8.94
8.94
8.78
8.85
8.79
8.84
8.63
8.82
8.80
8.85
8.81

7.73
7.53
7.65
7.01
7.64
7.71
7.51
7.59
7.73
7.85
7.76
7.11

8.86
8.88
8.86
8.89
8.84
8.56
8.70
8.44
9.05
8.84
9.03
8.89

7.79
7.86
7.68
7.62
7.69
7.69
7.46
7.32
7.82
7.67
7.64
7.45

8.41
8.56
8.75
8.70
8.61
8.53
8.56
8.41
8.70
9.33
8.69
8.90

7.86
7.85
7.54
6.91
7.74
7.66
7.59
7.51
8.09
8.03
7.59
7.09

8.00
7.73
8.58
8.64
7.71
8.35
8.61
8.36
8.37
8.53
8.58
8.67

6.88
7.63
6.74
6.61
6.67
7.11
6.95
6.51
7.69
7.46
7.74
7.28

7.88
7.90
8.38
8.58
7.44
8.18
8.52
7.72
8.53
8.49
7.73
7.72

5.45
7.51
6.68
5.29
6.37
6.68
6.47
6.46
7.50
7.67
7.22
7.47

Presented values are the means of three replicate trials. For details of starter cultures, see Sections 2.2 and 2.3.

viable counts of L. delbrueckii ssp. bulgaricus decreased

as the ratio of CN-to-WP was reduced. This trend was
similar in yoghurts produced using non-EPS- or EPSproducing starter cultures. This may be explained by the
differences in the fermentation time that affects the
population of this bacterium. During yoghurt manufacture, S. thermophilus is usually active during the rst
few hours of fermentation until the pH reaches around
5.40; then, L. delbrueckii ssp. bulgaricus starts to grow
(Shah, 2003). Furthermore, during the storage, the
viable counts of S. thermophilus in the yoghurts made
with non-EPS- or ropy EPS-producing starter cultures
gradually decreased from day 1, whereas in the yoghurts
made with capsular starter cultures, the viable counts of
S. thermophilus increased slightly after 7 d of storage and
declined thereafter. The viable counts of L. delbrueckii
ssp. bulgaricus gradually decreased from day 1 in
yoghurts made with each of the starter cultures and
with CN-to-WP ratios in the range 3:1 to 1:1. The
yoghurts made with a CN-to-WP ratio of 4:1 showed a
slight increase in the population which reached the
highest counts at 14 d with the value of 7.86, 7.74 and
8.09 for those made with non-EPS-, capsular EPS- or
ropy EPS-producing starter cultures, respectively. At
28 d, the yoghurts produced using ropy EPS-producing
starter cultures showed higher viable counts than those
made using non-EPS- or capsular EPS-producing starter
cultures. This may suggest a protective effect of ropy
EPS on L. delbrueckii ssp. bulgaricus. No such protective
effects were observed for S. thermophilus.
3.4. EPS concentration
The concentration of EPS in yoghurts made using
EPS starter cultures ranged from 30 to 70 mg L 1 (Table

4). Other researchers have reported EPS concentrations

in fermented milk products ranging from 40 to
400 mg L 1 (Marshall & Rawson, 1999; De Vuyst et
al., 2003; Toba et al., 1991). There are several possible
reasons for the difference between studies, including the
use of different strains and the level of inoculation of
starter cultures, the differences in fermenting conditions
and methods of isolation, purication and quantication of EPS. The concentration of EPS in yoghurt
produced with 12% (w/w) dry matter in our preliminary
study was higher (100 mg L 1, unpublished data) than
in yoghurt produced with 9% (w/w) dry matter in the
current study. Furthermore, the concentration of EPS in
yoghurt produced using ropy EPS-producing starter
cultures was slightly higher than that in yoghurts
produced using capsular EPS-producing starter cultures
(Table 4). This could be due to the differences in their
EPS production. Furthermore, it is interesting to note
that a low concentration of EPS (10 mg L 1) was
found in yoghurt produced with non-EPS-producing
starter cultures. The low value of EPS, detected in the
yoghurt made with non-EPS-producing starter cultures,
might be due to the residue of lactose remaining after
the purication. Because the phenol-sulphuric method,
used in the determination of EPS, is not specic for the
quantication of EPS, the presence of lactose may
increase the value. However, there is a possibility that
the non-EPS-producing S. thermophilus produced small
amounts of EPS. Nonetheless, the physical characteristics of yoghurts made using non-EPS-producing
starter cultures were signicantly different from yoghurts made using EPS-producing starter cultures as
discussed below (Figs. 2 and 3). For yoghurt made using
capsular EPS-producing starter cultures, there was no
difference in EPS concentration between yoghurts with

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

46

Table 4
Concentration of exopolysaccharide (mg L 1) in yoghurts prepared from milk blends with different casein-to-whey protein ratios, as described in
Table 1, and made using different starter culturesa
Starter culture

Casein-to-whey
protein ratio

Storage period (d)

1

Non-exopolysaccharideproducing starter cultures

Capsular exopolysaccharideproducing starter cultures

Ropy exopolysaccharideproducing starter cultures

4:1

7
a,A

9.1072.11

14

10.3071.19

a,A

21
a,A

9.4870.09

28
a,A

10.4471.34

8.5473.71a,A

3:1
2:1
1:1
4:1

8.0370.71a,A
9.2071.04a,A,B
7.9571.20a,A
32.4170.73b,c,B

9.7470.14a,A
9.0470.51a,A,B
10.0170.26a,A
28.8970.94b,A

9.6870.97a,A
8.8670.81a,A,B
8.8571.90a,A
30.8672.96b,c,A,B

8.2770.86a,A
9.7070.83a,B
8.2872.47a,A
36.9071.00b,c,C

9.6871.40a,A
7.9670.71a,A
8.7671.69a,A
33.5170.71b,B

3:1
2:1
1:1
4:1

32.5470.31b,c,A,B
29.1672.84b,A
38.3578.56b,c,d,A
41.5778.19c,d,e,A

29.3770.93b,A
28.8371.27b,A
31.8573.80b,A
47.7873.42d,A

33.4373.92b,c,B
28.5373.43b,A
32.8574.22b,c,A
49.69713.75d,e,A

30.3471.55b,A,B
31.4870.87b,A
36.1572.17b,c,A
53.9775.63d,e,A,B

32.4071.06b,A,B
31.8675.66b,A
36.0671.80b,A
66.4575.49d,B

3:1
2:1
1:1

43.2073.04d,e,A
48.6279.99e,A,B
36.6474.15b,c,d,A

47.7272.02d,A,B
41.4874.78c,A
42.4773.81c,A,B

60.06711.79e,B,C
58.4178.79e,B
42.3974.02c,d,A,B

63.21712.47e,B,C
50.13711.59d,A,B
44.7072.72c,d,B

75.4176.70e,C
62.1372.67c,d,B
57.3372.38c,C

a
Presented values are the means of three replicate trials; 7 indicates standard deviation from the mean. For details of starter cultures, see Section
2.2 and 2.3. Mean values (7standard deviation) within the same row not sharing a common superscript (A,B,C) differ signicantly (Po0.05). Mean
values (7standard deviation) within the same column not sharing a common superscript (a,b,c,d,e) differ signicantly (Po0.05).

different CN-to-WP ratios. This may be due to the

similar viable counts of S. thermophilus (capsular EPSproducing strain) in yoghurts with different CN-to-WP
ratios. For yoghurt made using ropy EPS-producing
starter cultures, the highest concentration of EPS was
found in the yoghurt made with a CN-to-WP ratio of
3:1, whereas the lowest EPS concentration was found in
the yoghurt with a CN-to-WP ratio of 1:1. Zisu and
Shah (2003) showed that the addition of 0.5% (w/w)
WPC to 10% (w/v) RSM increased the level of EPS
produced by ropy starter cultures. The increase in EPS
concentration with the increase in concentration of WP
was also observed in the current study. It can be seen
that the reduction in fermentation time of milks with
CN-to-WP ratios ranging from 4:1 to 1:1 by EPSproducing starter cultures did not reduce the concentration of EPS in the resultant yoghurts. Furthermore, the
concentration of EPS in yoghurt made using ropy
starter cultures increased during storage. This was not
found in yoghurt made using capsular EPS-producing
starter cultures, in which the EPS concentration
remained constant during storage at 4 1C. In other
studies (Degeest & De Vuyst, 1999; Pham, Dupont,
Roy, Lapointe, & Cerning, 2000), the degradation of
EPS during prolonged fermentation (72 h) in a complex
media has been reported. These studies reported a large
spectrum of glycohydrolytic enzymes (a-D-glucosidase,
b-D-glucosidase, a-D-galactosidase, b-D-galactosidase, bD-glucuronidase and a-L-rhamnosidase) during prolonged fermentation. This might suggest that the storage
at 4 1C helps to suppress the activity of those glycohydrolytic enzymes.

3.5. Firmness and syneresis of set yoghurt

Firmness and syneresis of set yoghurts with varying
CN-to-WP ratios and three types of starter cultures are
shown in Fig. 2. The rmness of yoghurts made using
capsular EPS- (Fig. 2b) or ropy EPS- (Fig. 2c)
producing starter cultures was generally lower than that
in yoghurts made with non-EPS-producing starter
cultures (Fig. 2a). Yoghurt made using ropy EPSproducing starter cultures had the lowest rmness.
Similar results were reported by other researchers
(Marshall & Rawson, 1999; Hassan et al., 1996b; Hess,
Roberts, & Ziegler, 1997). The incompatibility between
EPS and proteins may be the explanation. The EPS and
proteins have like charges at pH values above the
isoelectric point of the protein (de Kruif & Tuinier,
2001), as occurs in yoghurt. The incompatibility between
EPS and proteins may result in depletion-induced
attraction of casein micelles by EPS, leading to the
formation of an acid milk gel lled with EPS mass (de
Kruif & Tuinier, 2001; Tolstoguzov, 1997). This in turn
may cause a difference in protein aggregation mechanisms between yoghurt made using non-EPS- and EPSproducing starter cultures, and lead to differences in the
structure of the protein networks. Hassan, Frank,
Farmer, Schmidt, and Shalabi (1995) and Hassan and
Frank (1997) observed that the microstructure of
yoghurt and rennet curd made using capsular EPSproducing starter cultures had larger spaces between the
protein matrix and bacteria compared to the corresponding curds acidied by glucono-d-lactone or nonEP-producing starter cultures. Nevertheless, the protein

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

10.0

(a)

Syneresis (%, w/w)

Firmness (N)

1.0
0.8
0.6
0.4

0
1.0

14

21

6.0
4.0
2.0

28
10.0

(b)
Syneresis (%, w/w)

Firmness (N)

(d)

8.0

0.0

0.2

0.8
0.6
0.4
0.2
0

14

21

28

Syneresis (%, w/w)

0.8
0.6
0.4
0.2
0

14

21

28

Storage period (d)

14

21

28

14

21

28

14

21

28

(e)

6.0
4.0
2.0
0.0

10.0

(c)

8.0

1.0
Firmness (N)

47

(f)

8.0
6.0
4.0
2.0
0.0
0

Storage period (d)

Fig. 2. Firmness (a, b, c) and syneresis (d, e, f) of yoghurts made using non-exopolysaccharide-producing (control) cultures (a, d), capsular
exopolysaccharide-producing starter cultures (b, e), or ropy exopolysaccharide-producing starter cultures (c, f). The yoghurts were prepared from
milk blends with varying casein-to-whey protein ratios of 4:1 (J), 3:1 (K), 2:1 (E), and 1:1 (E). Each value is the mean of three replicate trials; error
bars represent standard deviations from the mean. See details of starter culture and milk blends composition in Section 2.2 and Table 1, and details of
tests in Sections 2.7 and 2.8.

matrix of yoghurt containing EPS was more compact

(Hassan et al., 1995; Hassan & Frank, 1997). These
observations suggested that EPS affected the microstructure of yoghurt and, thereby, affect the physical
characteristics yoghurt.
Increasing the CN-to-WP ratios from 1:1 to 4:1
resulted in higher rmness values. Tamime, Kalab, and
Davies (1984) reported that yoghurts made using milk
with a CN-to-WP ratio of 4.62 were rmer than those
made using milk with CN-to-WP ratios of 3.20 to 3.40.
A reduction in the rmness with a decrease in CN-toWP ratio was observed in yoghurts made with the three
types of starter cultures evaluated (Fig. 2ac). The
reduction in rmness as the CN-to-WP ratio was
reduced contradicts to the results reported by Puvanenthiran et al. (2002), who found that the gel strength
(representing yield point, or the rst fracture point
during the rst bite in the compression test) increased as
the CN-to-WP ratio was reduced. The difference in
rmness or gel strength between the current study and

that of Puvanenthiran et al. (2002) could be due to the

difference in sizes of protein aggregates as a result of
heating of milks at different pH values. Puvanenthiran
et al. (2002) reported the increase in size of particles in
milk heated at pH 7.0 when the CN-to-WP ratio
decreased. They found a relationship between the
increase in particle size and gel strength, and explained
that heating of milk at pH 7.0 promoted the formation
of dissociated k-caseinwhey protein aggregates, whey
proteinwhey protein aggregates and conglomerates of
whey protein aggregates as the CN-to-WP ratios were
decreased. Vasbinder and de Kruif (2003) reported that
heating milk at pH above 6.55 promoted the formation
of soluble denatured whey protein aggregates as
compared with heating at pH 6.35, at which all
denatured whey protein coated on the surface of casein
micelles. This may imply that heating milks with various
CN-to-WP ratios at natural pH had different effects on
the size of protein aggregates compared with heating
milk at pH 7.00. Furthermore, Cho, Singh, and Creamer

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

48

200

18
Loop area (Pa s-1)

Shear stress (Pa)

(a)
14
10
6

150
100
50

2
10

20

30

40

(d)

50

1
200

18

14

21

28

14

21

28

14

21

28

(e)

Loop area (Pa s-1)

Shear stress (Pa)

(b)
14
10
6

100
50
0

2
10

20

30

40

50
200

18

(c)
Loop area (Pa s-1)

Shear stress (Pa)

150

14
10
6
2

(f)

150
100
50
0

10

20

30

40

50

-1

Shear rate (s )

Storage period (d)

Fig. 3. Shear stress as a function of increasing and decreasing shear rate (a, b, c) for yoghurts that were prepared from milk blends with varying
casein-to-whey protein ratios of 4:1 (J), 3:1 (K), 2:1 (B), and 1:1 (E). Loop area as a function of the corresponding areas of shear stress/shear rate
hysteresis loops (d, e, f) for yoghurts that were prepared from milk blends with varying casein-to-whey protein ratios of 4:1 (T), 3:1 (), 2:1 ( ), and
1:1 ( ). Yoghurts were made using non-exopolysaccharide-producing (control) starter cultures (a, d), capsular exopolysaccharide-producing starter
cultures (b, e), or ropy exopolysaccharide-producing starter cultures (c, f). Each value is the mean of 3 replicate trials; error bars represent standard
deviations from the mean. See details of starter culture and milk blend composition in Section 2.2 and Table 1, and details of tests in
Section 2.9.

(2003) studied on the effect of heat treatment on the

mixtures of k-casein and b-lactoglobulin at pH 6.70
using polyacrylamide gel electrophoresis and reported
that the size of protein aggregates decreased as the ratio
of k-casein to b-lactoglobulin was decreased. This may
also suggest that the reduction of the CN-to-WP ratio of
milk blends in our study promotes the formation of
small protein aggregates after heat treatment. According
to the correlation between the increase in particle size
and gel strength of set yoghurt, as observed by
Puvanenthiran et al. (2002), the small protein aggregates
formed as a result of heat treatment of a milk blend with
a low CN-to-WP ratio (ratio 1:1) at natural pH would
result in a lower rmness in set yoghurt than in yoghurt
made from milk with a high CN-to-WP ratio (ratio 4:1).
Guzman-Gonzalez et al. (1999) reported that a decrease
in the ratio of denatured whey protein-to-total protein
(kept constant at 4.3%, w/w) was positively correlated

with the decrease in the apparent viscosity of yoghurts

made from milks supplemented with WPC. In our study,
the decrease in the level of whey protein denaturation in
milk blended to low CN-to-WP ratio (ratio 1:1) or
increased proportion of WP might have contributed to
the decrease in the rmness of set yoghurt. During
storage, the rmness of yoghurts made with EPSproducing starter cultures did not change signicantly.
On the contrary, the rmness of yoghurts produced with
non-EPS-producing starter cultures (especially from
milks with a CN-to-WP ratio of 4:1 or 3:1) increased
during the rst few weeks of storage and thereafter
remained constant.
In general, syneresis in yoghurts decreased when the
ratio of CN-to-WP was reduced, regardless of the type
of starter cultures (Fig. 2df). No syneresis was found in
yoghurt with a CN-to-WP ratio of 1:1. This result was
similar to that reported by Puvanenthiran et al. (2002).

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

Guzman-Gonzalez et al. (1999) reported that the

addition of WPC to milk reduced the level of syneresis
in yoghurt. Bhullar et al. (2002) and Remuef et al. (2003)
reported similar results. Harwalkar and Kalab (1986)
found that the level of syneresis in yoghurt decreased as
the density of protein matrix increased. Puvanenthiran
et al. (2002) suggested that an increase in the compactness of yoghurt microstructure, as the CN-to-WP ratio
was reduced, led to immobilization of a high level of free
water. Further, whey proteins and EPS are known to
have high water binding capacity (Walzem, Dillard, &
German, 1996; Cerning, 1990; De Vuyst & Degeest
1999). Whey proteins and EPS may act synergistically in
retaining water in the gel structure. Our results showed
that yoghurts made using EPS-producing starter cultures (both capsular and ropy) had a lower level of
syneresis than yoghurt produced with non-EPS-producing starter cultures (Fig. 2df). Similar results have also
been reported by others (Marshall & Rawson, 1999;
Wacher-Rodarte et al., 1993). Apart from the water
binding property of EPS, the modication of yoghurt
microstructure by EPS may also partly affect syneresis
levels. Yoghurt made using capsular EPS-producing
starter cultures had the lowest level of syneresis,
especially when the CN-to-WP ratio was 4:1. Interestingly, there appeared to be no difference in the syneresis
between yoghurts made with capsular EPS- and ropy
EPS-producing starter cultures at CN-to-WP ratios of
2:1 and 1:1. It seems that the water binding ability of
WP at CN-to-WP ratios of 2:1 and 1:1 surpassed waterbinding ability of capsular EPS and ropy EPS. During
storage, the level of syneresis between yoghurts did not
appear to change, regardless of the type of starter
cultures or the CN-to-WP ratios used.
3.6. Flow curve and hysteresis loop area of stirred
yoghurt
The ow curves and hysteresis loop area of the different
stirred yoghurts are shown in Fig. 3. Similar ow curves
were observed for stirred yoghurts made using the different
three types of starter cultures. Stirred yoghurt made with
ropy EPS-producing starter cultures (Fig. 3c) had a higher
shear stress (on increasing shear rate, corresponding to the
upward curve) than yoghurts made using the other types
of starter cultures. Several researchers have found that
yoghurt made with ropy EPS-producing starter cultures
had higher apparent viscosity than that made with nonEPS- or capsular EPS-producing starter cultures (WacherRodarte et al., 1993; Hassan et al., 1996a, 2003). Stirred
yoghurts made using non-EPS-producing starter cultures
(Fig. 3a) had a similar shear stress to those made using
capsular EPS-producing starter cultures (Fig. 3b). For
yoghurts made with EPS- or non-EPS-producing starter
cultures, a CN-to-WP ratio of 1:1 gave a shear stress that
was signicantly lower than that found at the higher CN-

49

to-WP ratios. For stirred yoghurts made using EPSproducing starter cultures (both capsular and ropy), the
highest shear stress was found at a CN-to-WP ratio of 3:1.
The area of hysteresis loop between the up and down
shear rate versus shear stress curves was determined. The
hysteresis loop area of stirred yoghurts made with ropy
EPS-producing starter cultures (Fig. 3f) was higher than
those made with non-EPS- (Fig. 3d) and capsular EPSproducing starter cultures (Fig. 3e). As all stirred yoghurts
had the same level of total solids and were made using the
same procedure and given the incompatibility between
EPS and milk proteins (de Kruif & Tuinier, 2001), the
presence of ropy EPS could be solely responsible for the
results observed. According to Morris (1995), most
polysaccharides exist in solution as random coils and can
form entangled networks depending on their numbers
(proportion to concentration) and molecular volume (size).
The entangled network causes an increase in the viscosity
of solution (Sworn, 2004).
Hassan, Frank, and Qvist (2002) observed the
microstructure before and after stirring of milk fermented with a single strain of L. delbrueckii ssp. bulgaricuss
RR (ropy starter culture). They reported that the EPS
segregated from the proteins and the EPS formed a
more extensive structure in the stirred gel compared with
the set gel. They suggested that stirring promoted the
interactions between molecules of ropy EPS. However,
they did not specify whether the interactions are
chemical or physical (entanglement) type. However, it
is likely that the interactions, as stated by Hassan et al.
(2002), are physical types or the entanglement of
polysaccharides. The entangled networks of ropy EPS
are also expected to occur in our study.
Theoretically, the area of hysteresis loop between
upward and downward ow curves represents the
structural breakdown of stirred yoghurt during shearing
(Halmos & Tiu, 1981; Ramaswamy & Basak, 1991;
Hassan et al., 1996a, 2003). The formation of entangled
networks of ropy EPS, as an additional structure in
stirred yoghurt, may explain the increase in the area of
hysteresis loop of products made using ropy EPSproducing starter cultures. It also implies that there was
no entangled network of EPS formed in the products
made using capsular EPS-producing starter cultures as
the value of hysteresis loop was comparable to that in
yoghurt made with non-EPS-starter cultures.
The area of hysteresis loop of stirred yoghurts made
using non-EPS- or capsular EPS-producing starter
cultures decreased as the CN-to-WP ratio was decreased. As the presence of capsular EPS did not
contribute to the hysteresis loop area, the results can
be interpreted as a decrease in structural breakdown of
proteins during shearing or an increase in the damage of
protein structure of initial stirred yoghurts made with
low CN-to-WP ratios compared with yoghurts with high
ratios. In this study, stirred yoghurt was produced by

ARTICLE IN PRESS
50

T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051

pressing the gel of set yoghurt with a spoon through a

sieve. Applying the same procedure to the soft gel (set
yoghurt made at a CN-to-WP ratio of 1:1) could result
in higher structural damage of the protein structure than
that with a rm gel (set yoghurt made at a CN-to-WP
ratio of 4:1). Stirred yoghurt made with a CN-to-WP
ratio of 3:1 and using ropy EPS-producing starter
cultures had the highest loop area. During storage, a
slight increase in the area of hysteresis loop was
observed in stirred yoghurts. The change in loop area
of the products made with ropy EPS-producing starter
cultures correlated with the increase in its EPS
concentration (Table 4). This relationship was not
observed in those made with non-EPS- or capsular
EPS-producing starter cultures. In addition, after 14 and
21 d, the loop area decreased slightly in stirred yoghurts
made using either non-EPS- or EPS-producing starter
cultures (both capsular and ropy). However, as the
phenomenon was also observed in stirred yoghurt made
using non-EPS-producing starter cultures, the decrease
in loop area may be due to changes in milk proteins.

4. Conclusions
A decrease in the ratio of CN-to-WP of set yoghurt
reduced the level of syneresis and rmness. The use of
EPS-producing starter cultures also reduced the level of
syneresis and gel rmness. In stirred yoghurt made using
ropy EPS-producing starter cultures, there was an
increase in shear stress and the area of hysteresis loop
between the up and down shear rate versus shear stress
curves; the increase in loop area represents an additional
structure of EPSEPS in stirred yoghurt. The results of
this study suggest that there is a possibility of reducing
the level of added dairy ingredients and stabilizers in the
manufacture of stirred yoghurt by using ropy EPSproducing starter cultures. For set yoghurt, a CN-toWP ratio of 3:1 and the use of non-EPS-producing
starter cultures gave optimal rmness and syneresis
levels.

Acknowledgments
The authors would like to thank Mr. Micheal
Kakoullis, Laboratory Manager, Department of Food
Science, RMIT University, City Campus, Melbourne,
Australia for the assistances in using texture analyser
and rheometer.
References
AOAC. (1995). Official methods of analysis of AOAC international
(16th ed.). Arling, VA, USA: AOAC International.

Baig, M. I., & Prasad, V. (1996). Effect of incorporation of cottage

cheese whey solids and Bifidobacterium bifidum in freshly made
yogurt. Journal of Dairy Research, 63, 467473.
Bhattacharya, S. (1999). Yield stress and time-dependent rheological
properties of mango pulp. Journal of Food Science, 64, 10291033.
Bhullar, Y. S., Uddin, M. A., & Shah, N. P. (2002). Effects of
ingredients supplementation on textural characteristics and microstructure of yoghurt. Milchwissenschaft, 57, 328332.
Bourne, M. (2002). Food texture and viscosity: Concept and measurement (2nd ed.). New York: Academic Press.
Cerning, J. (1990). Exocellular polysaccharides produced by lactic acid
bacteria. FEMS Microbiology Reviews, 87, 113130.
Cho, Y., Singh, H., & Creamer, L. K. (2003). Heat-induced
interactions of b-lactoglobulin A and k-casein B in a model
system. Journal of Dairy Research, 70, 6171.
Dave, R. I., & Shah, N. P. (1998). Ingredient supplementation effects
on viability of probiotic bacteria in yogurt. Journal of Dairy
Science, 81, 28042816.
Degeest, B., & De Vuyst, L. (1999). Indication that the nitrogen source
inuences both amount and size of exopolysaccharides produced
by Streptococcus thermophilus LY03 and modelling of the bacterial
growth and exopolysaccharide production in a complex medium.
Applied and Environmental Microbiology, 65, 28632870.
de Kruif, C. G., & Tuinier, R. (2001). Polysaccharide protein
interactions. Food Hydrocolloids, 15, 555563.
De Vuyst, L., & Degeest, B. (1999). Heteropolysaccharides from lactic
acid bacteria. FEMS Microbiology Reviews, 23, 153177.
De Vuyst, L., Zamr, M., Mozzi, F., Adriany, T., Marshall, V. M.,
Degeest, B., & Vaningelgem, F. (2003). Exopolysaccharide-producing Streptococcus thermophilus strains as functional starter
cultures in the production of fermented milks. International Dairy
Journal, 13, 707717.
Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., & Smith, F.
(1956). Colorimetric method for determination of sugars and
related substances. Analytical Chemistry, 38, 350356.
Guzman-Gonzalez, M., Morais, F., & Amigo, L. (2000). Inuence of
skimmed milk concentrate replacement by dry dairy products in a
low-fat set-type yoghurt model system. II: Use of caseinates, coprecipitate and blended dairy powders. Journal of the Science of
Food and Agriculture, 80, 433438.
Guzman-Gonzalez, M., Morais, F., Ramos, M., & Amigo, L. (1999).
Inuence of skimmed milk concentrate replacement by dry dairy
products in a low fat set-type yoghurt model system. I: Use of whey
protein concentrates, milk protein concentrates and skimmed milk
powder. Journal of the Science of Food and Agriculture, 79,
11171122.
Halmos, A. L., & Tiu, C. (1981). Liquid foodstuffs exhibiting yield
stress and shear-degradability. Journal of Texture Studies, 12,
3946.
Haque, A., Richardson, R. K., & Morris, E. R. (2001). Effect of
fermentation temperature on the rheology of set and stirred yogurt.
Food Hydrocolloids, 15, 593602.
Harwalkar, V. R., & Kalab, M. (1986). Relationship between
microstructure and susceptibility to syneresis in yoghurt made
from reconstituted nonfat dry milk. Food Microstructure, 5,
287294.
Hassan, A. N., & Frank, J. F. (1997). Modication of microstructure
and texture of rennet curd by using a capsule-forming non-ropy
lactic cultures. Journal of Dairy Research, 64, 115121.
Hassan, A. N., Frank, J. F., Farmer, M. A., Schmidt, K. A., &
Shalabi, S. I. (1995). Formation of yogurt microstructure and
three-dimensional visualization as determined by confocal scanning
laser microscopy. Journal of Dairy Science, 78, 26292636.
Hassan, A. N., Frank, J. F., Schmid, K. A., & Shalabi, S. I. (1996a).
Rheological properties of yogurt made with encapsulated nonropy
lactic cultures. Journal of Dairy Science, 79, 20912097.

ARTICLE IN PRESS
T. Amatayakul et al. / International Dairy Journal 16 (2006) 4051
Hassan, A. N., Frank, J. F., Schmid, K. A., & Shalabi, S. I. (1996b).
Textural properties of yogurt made with encapsulated nonropy
lactic cultures. Journal of Dairy Science, 79, 20982103.
Hassan, A. N., Frank, J. F., & Qvist, K. B. (2002). Direct observation
of bacterial exopolysaccharide in dairy products using confocal
scanning laser microscopy. Journal of Dairy Science, 85,
17051708.
Hassan, A. N., Ipsen, R., Janzen, T., & Qvist, K. B. (2003).
Microstructure and rheology of yogurt made with cultures differing
only in their ability to produce exopolysaccharides. Journal of
Dairy Science, 86, 16321638.
Hess, S. J., Roberts, R. F., & Ziegler, G. R. (1997). Rheological
properties of nonfat yogurt stabilized using Lactobacillus delbrueckii ssp. bulgaricus producing exopolysaccharide or using
commercial stabilizer systems. Journal of Dairy Science, 80,
252263.
Jaros, D., Rohm, H., Haque, A., Bonaparte, & Kneifel, W. (2002).
Inuence of the starter cultures on the relationship between dry
matter content and physical properties of set-style yogurt.
Milchwissenschaft, 57, 325326.
Lucey, J. A. (2001). The relationship between rheological parameters
and whey separation in milk gels. Food Hydrocolloids, 15, 603608.
Marshall, V. M., & Rawson, H. L. (1999). Effects of exopolysaccharide-producing strains of thermophilic lactic acid bacteria on the
texture of stirred yoghurt. International Journal of Food Science and
Technology, 34, 137143.
Modler, H. W., Larmond, M. E., Lin, C. S., Froehlich, D., &
Emmons, D. B. (1983). Physical and sensory properties of yogurt
stabilized with milk proteins. Journal of Dairy Science, 66, 422429.
Morris, E. R. (1995). Polysaccharide rheology and in-mouth perception. In A. M. Stephen (Ed.), Food polysaccharides and their
applications (pp. 517546). New York: Marcel Dekker, Inc.
Perry, D. B., McMahon, D. J., & Oberg, C. J. (1997). Effect of
exopolysaccharide-producing cultures on moisture retention in low
fat mozzarella cheese. Journal of Dairy Science, 80, 799805.
Pham, P. I., Dupont, I., Roy, D., Lapointe, G., & Cerning, J. (2000).
Production of exopolysaccharide by Lactobacillus rhamnosus R and
analysis of its enzymatic degradation during prolonged fermentation. Applied and Environmental Microbiology, 66, 23022310.
Puvanenthiran, A., Williams, R. P. W., & Augustin, M. A. (2002).
Structure and visco-elastic properties of set yoghurt with altered
casein to whey protein ratios. International Dairy Journal, 12,
383391.
Ramaswamy, H. S., & Basak, S. (1991). Rheology of stirred yogurts.
Journal of Texture Studies, 22, 231241.
Remuef, F., Mohammed, S., Sodini, I., & Tissier, J. P. (2003).
Preliminary observations on the effects of milk fortication and

51

heating on microstructure and physical properties of stirred yogurt.

International Dairy Journal, 13, 773782.
Shah, N. P. (2003). Yogurt: The product and its manufacture. In B.
Caballero, L. C. Trugo, & P. M. Finglas (Eds.), Encyclopedia of
food science and nutrition (Vol. 10. 2nd ed., pp. 62526259). UK:
Academic Press.
Shin, H. S., Lee, J. H., Pestka, J. J., & Ustunol, Z. (2000). Growth and
viability of commercial Bifidobacterium spp. in skim milk containing oligosaccharides and insulin. Journal of Food Science, 65,
884887.
Sworn, G. (2004). Hydrocolloid thickeners and their application. In P.
A. Williams, & G. O. Phillips (Eds.), Gums and stabilisers for the
food industry 12 (pp. 1222). UK: The Royal Society of Chemistry.
Tamime, A. Y., & Deeth, H. C. (1980). Yogurt: Technology and
biochemistry. Journal of Food Protection, 43, 939977.
Tamime, A. Y., Kalab, M., & Davies, G. (1984). Microstructure of setstyle yoghurt manufactured from cows milk fortied by various
methods. Food Microstructure, 3, 8392.
Teggatz, J. A., & Morris, H. A. (1990). Changes in the rheology and
microstructure of ropy yoghurt during shearing. Food Structure, 9,
133138.
Toba, T., Uemura, H., Mukai, T., Fuji, T., Itoh, T., & Adachi, S.
(1991). A new fermented milk using capsular polysaccharideproducing Lactobacillus kefiranofaciens isolated from ker grains.
Journal of Dairy Research, 58, 497502.
Tolstoguzov, V. (1997). Proteinpolysaccharide interactions. In S.
Damodaran, & A. Paraf (Eds.), Food proteins and their application
(pp. 176177). New York: Marcel Dekker, Inc.
Vasbinder, A. J., & de Kruif, C. G. (2003). Caseinwhey protein
interactions in heated milk: The inuence of pH. International
Dairy Journal, 13, 669677.
Wacher-Rodarte, C., Galvan, M. V., Farres, A., Gallardo, F.,
Marshall, V. M., & Garcia-Garibay, M. (1993). Yogurt production
from reconstituted skim milk powders using different polymer and
non-polymer forming starter cultures. Journal of Dairy Research,
60, 247254.
Walstra, P., & Jenness, R. (1984). Dairy chemistry and physics
(pp. 194196). New York: Wiley.
Walzem, R. L., Dillard, C. J., & German, J. B. (2002). Whey
components: Millennia of evolution create functionalities for
mammalian nutrition: What we know and what we may be
overlooking. Critical Reviews in Food Science and Nutrition, 42,
354375.
Zisu, B., & Shah, N. P. (2003). Effects of pH, temperature,
supplementation with whey protein concentrate, and adjunct
cultures on the production of exopolysaccharides by Streptococcus
thermophilus 1275. Journal of Dairy Science, 86, 34053415.