Você está na página 1de 7

Paper

pecorino of Appennino Reggiano cheese:


EVALUATION OF RIPENING TIME
USING selected physical properties

M. RINALDI, E. CHIAVARO* and R. MASSINI


Dipartimento di Ingegneria Industriale, Università degli Studi di Parma,
Viale G.P. Usberti 181/A, 43124 Parma, Italy
*Corresponding author: Tel. +39 0521 905888, Fax +39 0521 905705,
e-mail: emma.chiavaro@unipr.it

Abstract

Physico-chemical properties (texture, colour and image analysis) of Pecorino of Appennino Reg-
giano cheese were evaluated on six cheese samples at different ripening times (0, 80, 90, 150, 180
and 240 days). Hardness values, obtained by means of texture profile analysis and fracture test, as
well as Young’s modulus values, significantly increased during ripening. Among other textural pa-
rameters, cohesiveness decreased significantly from 0 to 80 days and then remained unchanged for
the rest of the ripening time, whereas adhesiveness and springiness did not exhibit a linear trend.
L*, b* and C* values decreased significantly with ripening, whereas the percentage area occupied by
holes remained unaltered. A statistical model was developed to predict ripening time by means of a
step-wise multiple linear regression analysis based on five textural properties (hardness, cohesive-
ness, chewiness, hardness at fracture and Young’ modulus) and three colour parameters (L*, b* and
C) (R2 = 0.984). The model was applied to a validation set of six samples from different milk lots and
successfully predicted ripening time with good reproducibility and accuracy (R2 = 0.971).

- Key words: colour, pecorino cheese, ripening time, statistical model, texture -

54  Ital. J. Food Sci. n. 1, vol. 22 - 2010


Introduction eters depending on the degree of ripening (CAST-
AGNETTI et al., 2004; PIGNEDOLI, 2005). Pecori-
Cheese is very diverse category of food prod- no of Appennino Reggiano must also be cylin-
uct and its final quality differs depending on drically shaped (the height always less than the
various parameters: milk origin, cheese-mak- radius) and should weigh 1-2.5 kg; the colour
ing process, cheese size, ripening time and con- is from white to yellow, with green reflexes that
ditions. The ripening of cheese is a very com- appear with ripening due to mould growth, and
plex, often slow, biochemical process that in- the possible presence of gas holes that are ir-
volves three primary reactions: glycolysis, lipol- regularly distributed. It can be manufactured in
ysis and proteolysis. Thus, such physical prop- different sizes (diameter x height) and weight,
erties as texture, colour and gas hole develop- according to degree of ripening: 150 mm x 60
ment may be used to differentiate among cheeses mm and about 1-1.2 kg for the soft cheese; 180
and describe their changes during ripening (CA- mm x 100 mm and about 2-2.5 kg for the semi-
SIRAGHI et al., 1989; CACCAMO et al., 2004; DU- hard cheese.
FOSSÈ et al., 2005), as they are related to milk The aim of this preliminary work was to eval-
type and/or cheese-making procedures. Evalu- uate changes in the selected physical proper-
ation of the parameters could be of great inter- ties (texture, colour and gas holes) of Pecorino
est for the protected designation of origin (PDO) of Appennino Reggiano cheese during ripening
or geographical indication (PGI) for the Europe- of a semi-hard cheese made from raw milk that
an Union (EU) in order to avoid misuse, imita- is the main commercial category. In addition, an
tion or evocation of the registered name (ROSEI- estimation model was developed on the basis of
RO et al., 2003), as consumers take these prop- these parameters by means of step-wise multi-
erties into primary consideration when making ple linear regression analysis (MLR) in order to
purchasing decisions. These important physical predict ripening time of Pecorino of Appennino
parameters could be used to develop a statistical Reggiano cheese for marketing purposes.
model for ripening prediction; such a tool could
be used to avoid misleading consumers, espe-
cially with regard to cheese that is artisanally Materials and Methods
produced using raw milk where the variability
in quality can be high (POVEDA et al. 2003; PIN- Sampling and cheese-making
HO et al. 2004; ALVARENGA et al., 2008). Ten Pecorino of Appennino Reggiano samples
Pecorino of Appennino Reggiano is an Italian manufactured under conditions specified in the
semi-hard cheese that is traditionally produced production rules were obtained from a local dairy
from whole sheep milk with the addition of se- (Succiso - Reggio Emilia); they were produced
lected starters or natural microbiological cul- the same day using the same lot of raw milk in
tures (obtained from sheep milk or whey from August 2008. Two other samples produced from
the restricted area of production). In 2000, the different lots of milk in 2009 were purchased
production process of Pecorino of Appennino from the same manufacturer.
Reggiano was established and is limited to a All the cheeses were manufactured according
restricted area of production (Comunità Mon- to the code rules. The cheese was manufactured
tana dell’Appennino Reggiano) in the region of by adding 11 mL L-1 of natural starter and 28
Emilia Romagna (northeastern Italy), by a Code mg L-1 of commercial bovine rennet to whole raw
of Rules (OFFICIAL BULLETIN OF EMILIA RO- sheep milk at 39°C. The natural starter was ob-
MAGNA REGION, 2005) in order to obtain the tained from pasteurised sheep milk of the previ-
PDO. Since 2000, Pecorino of Appennino Reg- ous day, which was incubated at 45°C until the
giano has been included in the list of the Tradi- acidity value reached 10°SH 50mL-1. The coag-
tional Food Products (TFP) by the Italian Min- ulation was completed within 20-25 min. Then,
istry of Agriculture. Moreover, the Internation- the curd was cut into maize-grain-size pieces
al Slow Food Association had established an with a standard wire cage equipped with a han-
ad-hoc presidium in the production area of this dle, subjected to several agitation steps and then
cheese (Slow Food, http://www.presidislowfood. was partially cooked at 44-46°C for 10-15 min.
it/ita) to preserve its characteristics and to pro- The curd was placed in moulds, pressed and
tect the product against imitations. dried for 5 h at 30°C. Dry salting was then car-
According to product specifications, this ried out for 48 h at 8°C. All the samples were
cheese is marketed as two commercial catego- ripened under the same temperature and hu-
ries. Soft cheese can be marketed from 30 days midity conditions (12°C and RH 70%).
after production up to 60 days of ripening and Six cheese samples, ripened for 0, 80, 90,
is produced from pasteurised milk. Semi-hard 150, 180 and 240 days, were selected as stand-
cheese can be marketed from 90 days of ripening ard cheeses to obtain the estimation model. The
up to 12 months and is generally manufactured other six were used as a validation set for the
from raw milk, although pasteurised milk can prediction model: four, ripened for 80, 150, 180
be used. These two categories obviously exhibit and 240 days, were produced on the same day
different sensory properties and quality param- as the calibration set, and the other two of 80

Ital. J. Food Sci. n. 1, vol. 22 - 2010  55


and 150 days of ripening, were obtained from The elastic modulus was then calculated by
other milk lots in a different period. fitting the following mathematical model to the
Proximate composition at time 0 of ripening whole true tension–true strain curves (DEL NO-
(moisture 41.3%, protein 22.4%, fat 30.3%, salt BILE et al., 2007) according to the following
1.7% and ash 3.2%) was confirmed using Stand- equations:
ard Association of Official Analytical Chemists
(AOAC) methods (1995). σ T (ε T ) = EC ⋅ ε T ⋅ exp(ε T ⋅ K ) (2)

Texture analysis where EC is the elastic modulus (i.e. the tan-
Texture evaluation was performed using a gent to the stress strain curve at the origin); K
TA.XT2 Texture Analyzer equipped with a 25 is a constant and is considered to be a fitting
kg load cell (Stable Micro Systems, Goldalm- parameter. Measurements were carried out on
ing, U.K.) and Texture Expert for Windows soft- 5 slices of each cheese, 3 samples (30 mm x 30
ware (version 1.22) for data analysis. The load mm x 30 mm) were extracted from each slice.
cell calibration was performed daily according All samples were equilibrated at room tempera-
to the TA.XT2 manual (Stable Micro Systems). ture (25°C) for at least 30 min before each test.
A TPA (Texture Profile Analysis) test was per-
formed. The compression test was performed Colour analysis
with a 35 mm diameter aluminium cylindrical Colour determinations were carried out on
probe and cheese samples were compressed to three central slices (30 mm thickness) of each
approximately 67% of the original height with cheese using a Minolta Colorimeter (CM 2600d,
a constant speed of 2 mms-1 and a 5 s delay Minolta Co., Osaka Japan) on a CIE L* a* b* sys-
between the first and second bite. Five slices tem equipped with a standard illuminant D65 on
were cut from each cheese sample and three 10 pre-selected locations for each slice. L* [light-
cylindrically-shaped samples were obtained ness: black = 0, white = 100], a* [redness: a* >
from each slice (approximately 25 mm diameter 0, green < 0], b* [yellowness: b* > 0, blue <0], C
and 15 mm thick). The TPA parameters consid- [Chroma or colour saturation = (a*2 + b*2)1/2: 0
ered were: hardness (H, peak force of the first completely unsaturated (neutral grey, black or
compression cycle in N), cohesiveness (CO, ra- white), 100 maximum saturation] were quanti-
tio between the positive force area during the fied on each sample by using the 2nd position of
second compression over that during the first the standard observer. The instrument was cali-
compression, dimensionless), springiness (SP, brated before each analysis with white and black
ratio between the time duration of force input standard tiles. A total of 30 determinations were
during the second compression over that dur- performed at each ripening time.
ing the first compression, dimensionless), ad-
hesiveness (AD, negative force area for the first Image analysis
bite representing the work necessary to pull the For each cheese sample, digital pictures of
plunger away from the food sample, N x mm) both faces of 2 central slices (30 mm thickness)
and chewiness (CH, hardness multiplied by were taken using a flatbed scanner (Model Scan-
cohesiveness multiplied by springiness in N) jet 8200, HP, Cupertino, CA, USA), with a res-
(BOURNE, 1978). olution of 600 dots per inch (dpi) and convert-
Fracture tests were obtained using a Guillo- ed from true colour to 256 level grey scale. The
tine blade (3 mm thick) compressing cheese slic- images were calibrated, standardized and opti-
es (15 mm thickness) until the breaking point mised applying appropriate filters to measure
at a constant speed of 2 mms-1. The maximum pore number using the Image-Pro Plus 4.5 (Me-
force of the force/distance peak was taken as dia Cybernetics Inc., Bethesda, MD, USA) soft-
a measure of hardness of the sample (Hf in N). ware with a procedure previously reported (IN-
Measurements were carried out on 5 slices from NOCENTE and CORRADINI, 1998). Cheese hole
each sample. distribution was characterized by means of per-
The elastic modulus or Young’s modulus (Y) cent ratio between the area occupied by holes
was calculated according to DEL NOBILE et al. and total slice area, as previously reported (CAC-
(2007), as the slope of the stress-strain curve, CAMO et al., 2004). A total of 4 determinations
converting the instantaneous values of specimen were performed at each ripening time.
height H(t) and compression charge applied F(t)
into engineering ( σ E ), true ( σ T ) stress vs. engi- Statistical analysis
neering ( ε E ) and Hencky ( ε T ) strain (MANCINI et Means and standard deviations (SD) were cal-
al., 1999) according to the following equations: culated with STATISTICA (Release 5.5, StatSoft,
Tulsa, OK, USA) statistical software. STATISTI-
F (t ) H 0 − H (t ) CA was used to perform one-way-analysis of var-
σE = ,    ε E = , iance (ANOVA) and Least Significant Difference
S0 H0 test (LSD) at a 95 % confidence level (p≤0.05)
to identify differences among samples and re-
σ T = σ E ⋅ (1 − ε E ),    ε T = − ln(1 − ε E ). (1) gression analysis. Step-wise multiple linear re-

56  Ital. J. Food Sci. n. 1, vol. 22 - 2010


gression analysis (MLR), which combines the
backward elimination and forward selection re-
gression methods (ALVARENGA et al., 2008) was
used to estimate the length of the ripening peri-
od of Pecorino of Appennino Reggiano. The lev-
el of significance of an independent variable for
entry and stay in the calibration model was set
at 0.05 during the execution of step-wise vari-
able selection.

Results and Discussion

TPA parameters and hardness at fracture were


measured during ripening time and the results
are summarized in Tab. 1 for standard cheeses.
Ripening significantly increased the TPA hard- Fig. 1. Changes in Young’s modulus of standard cheeses dur-
ing ripening. Error bars represent +/- 1 standard deviation
ness of cheese and also significantly decreased (n= 1, sample size = 15). Histogram bars with the same let-
the cohesiveness from 0 to 80 days. On the con- ter are not significantly different (p ≤ 0.05).
trary, adhesiveness and springiness remained
substantially unaltered during ripening. Hard-
ness at fracture significantly increased from 0 Table 2 - Colour parameter results of standard cheeses at
each ripening time.
to 240 days of ripening.
Biochemical changes of cheese during ripen- Days L* a* b* C
ing such as proteolysis and/or water availabil-
ity to bond components were found to be relat- 0 87.8a (0.6) -2.5d (0.1) 16.4a (0.6) 16.5 a (0.6)
ed to cheese texture, although the mechanism 80 80.4b (1.5) -2.4c (0.1) 16.2a (0.7) 16.4a (0.7)
of this relation was not completely understood 90 76.5c (1.2) -1.9b (0.2) 12.3b (0.7) 12.5 b (0.8)
(Lucey et al., 2003). In particular, the formation 150 72.5d (1.4) -2.0bc (0.3) 12.5b (1.0) 12.6 b (1.1)
of ionic groups generated by peptide bond cleav- 180 68.0e (1.6) -2.2cd (0.3) 11.8b (0.9) 12.2 b (0.9)
age that could compete with protein for bonding 240 57.7f (2.0) -1.4a (0.2) 7.5c (1.1) 7.6 c (1.1)
available water were related to the increase of
Values with same letters a, b, c, d, e, f within each column are not sig-
both cheese hardening and resistance to defor- nificantly different (p≤0.05), standard deviation given in parenthe-
mation during ripening (LAWRENCE et al., 1986). ses. (n= 1, sample size = 30).
Changes of Young’s modulus (Y) during ripen-
ing are reported in Fig. 1. Values of Y increased
significantly with ripening until 150 days and
continued to increase until the end of ripening was previously reported in the literature (DEL NO-
time. An increase of Y with ripening of cheese BILE et al., 2007: GONZÀLEZ VIŇAS et al., 2007);
it was related to the decrease in moisture con-
tent and the strengthening of the protein matrix
Table 1 - TPA and fracture test results of standard cheeses due to maturation (GONZÀLEZ VIŇAS et al., 2007).
at each ripening time. Colorimetric indices for Appenino Reggiano
cheese during ripening are summarized in
Days H (N) CO SP AD CH (N ) Hf (N)
Tab. 2. Cheese samples became less bright (L*)
0 57.69d 0.37a 0.58a 1.43a 12.36b 10.67d and yellow (b*) as ripening time increased. A
(10.17) (0.06) (0.06) (0.31) (3.67) (1.25) loss of colour saturation (C) was also observed
80 158.58c 0.19b 0.39b 1.72a 11.76b 21.92cd during ripening.
(27.01) (0.02) (0.03) (0.29) (2.22) (7.78) Ripening was previously found to induce
90 239.86b 0.21b 0.46ab 1.86a 23.17a 33.26bc changes in the colour of cheeses produced with
(47.10) (0.02) (0.06) (0.85) (5.07) (6.39) raw cow (ROHM and JAROS, 1996) and goat milk
150 242.72b 0.20b 0.46ab 0.95a 22.27a 50.83a (BUFFA et al., 2001). In particular, the L* values
(48.47) (0.03) (0.10) (1.40) (7.83) (9.38)
were found to decrease (ROHM and JAROS, 1996;
180 234.41b 0.19b 0.50ab 1.08a 22.29a 39.44ab
(29.44) (0.02) (0.19) (0.75) (3.81) (16.86) BUFFA et al., 2001) whereas the b* values gen-
240 303.63a 0.17b 0.46ab 0.96a 23.73a 40.25ab erally increased during cheese-ripening (ROHM
(35.35) (0.02) (0.06) (0.57) (3.92) (12.67) and JAROS, 1996) but a slight decrease after a
long ripening time (more than 60 days) was also
Values with same letters a, b, c, d within each column are not signif- reported for ewe cheese (PINHO et al. 2004).
icantly different (p≤0.05), standard deviation given in parenthe- The description of cheese appearance at 0
ses (n= 1, samples size = 15 for TPA parameters; n= 1, sample and 240 days is reported in Fig. 2a and 2b, re-
size = 5 for fracture test). aAbbreviations: H, hardness; CO, cohe- spectively. The Pecorino samples examined in
siveness; SP, springiness; AD, adhesiveness; CH, chewiness; Hf,
hardness at fracture. this study exhibited an appearance typical of
this cheese, i.e. a compact structure with few

Ital. J. Food Sci. n. 1, vol. 22 - 2010  57


Young’s modulus and hardness at fracture)
and three colour parameters (L*, b* and C) were
useful for predicting ripening time of Pecorino
cheese, as based on the high R2 value and the
estimation error of calibration. The following
equation was used:

RT = 631.4 + 0.001 H – 121.3 CO + 0.75 CH -


19.4 Y + 0.5 Hf – 5.6 L* - 0.2 b*- 5.9 C (3)

with R2 = 0.984 and an estimation error of 8.2


days.
The estimation error of ripening time was
similar to that previously reported in the litera-
ture for a similar statistical model developed by
means of textural and colour variables on other
types of cheeses (POVEDA et al., 2003; PINHO et
al., 2004; ALVARENGA et al., 2008).
This model was applied to a validation set of
six samples, four produced on the same day as
the standard cheeses and ripened for 80, 150,
180 and 240 days, the other two samples were
obtained from another lot of milk (80 and 150
days ripened); the effect that the variability in
the composition of the milk would have on the
predictive model was considered.
The texture and colour parameters of the val-
Fig. 2. Image of standard cheeses obtained at 0 (A) and 240 idation set of cheeses were similar to those of
(B) days of ripening. standard samples (data not shown). Small dif-
ferences between the predicted and the real rip-
ening times (about 10 days) were obtained. The
holes. A low percentage of the area was occu- correlation coefficient obtained between the pre-
pied by holes at 0 time of ripening (approximate- dicted and the real ripening times was slight-
ly 0.2%) and this value remained substantially ly lower than that of the calibration set of sam-
unchanged during ripening. Thus, cheese ap- ples (R2 = 0.971).
pearance may not be representative of change These results show that the textural and col-
during ripening. our parameters considered in this study can
A step-wise multiple linear regression (MLR) give practical and useful information to de-
analysis using textural and colour data was per- scribe changes in Pecorino of Appennino Reg-
formed to develop a statistical model that would giano semi-hard cheese during ripening. This
allow the ripening time of Pecorino of Appenni- information can be used to regulate the produc-
no Reggiano to be estimated. A summary of the tive process in order to obtain the PDO recogni-
step-wise analysis is reported in Tab. 3. tion of the cheese.
The MLR analysis showed that five variables The predictive model based on these selected
from the texture determinations (hardness, co- physical parameters can differentiate samples of
hesiveness and chewiness from the TPA test, Pecorino of Appennino Reggiano according to rip-

Table 3 - Summary of MLR with textural and colourimetric data as independent variables.

Step Regression equation R2

1 RT = -30.2 + 0.73 H 0.715


2 RT = 69.1 + 0.52 H – 257.3 CO 0.737
3 RT = 93.1 + 0.35 H – 362.0 CO + 1.8 CH 0.748
4 RT = 74.1 + 0.08 H – 276.8 CO + 1.2 CH + 80.3 Y 0.892
5 RT = 32.7 + 0.09 H – 156.2 CO + 0.49 CH + 80.0 Y + 1.1 Hf 0.914
6 RT = 576.9 - 0.10 H – 158.7 CO + 0.83 CH + 7.6 Y + 0.8 Hf – 6.0 L* 0.975
7 RT = 619.7 - 0.04 H – 128.0 CO + 0.75 CH - 11.9 Y + 0.6 Hf – 5.8 L* - 4.7 b* 0.981
8 RT = 631.4 + 0.001 H – 121.3 CO + 0.75 CH - 19.4 Y + 0.5 Hf – 5.6 L* - 0.2 b*- 5.9 C 0.984
a
Abbreviations: RT, ripening time; H, hardness; CO, cohesiveness; CH, chewiness; Y, Young’s modulus; Hf, hardness at fracture; L*, lightness;
b*, yellowness; C*, chroma; R2, adjusted coefficient of determination.

58  Ital. J. Food Sci. n. 1, vol. 22 - 2010


ening time and the level of quality. The applica- ing of ovine and caprine cattle (SIPAOC), Siena, Italy, p.
254. September 29th- 2nd October.
tion of these predictive models is of great inter-
est for Pecorino of Appennino Reggiano, as well Del Nobile M.A., Chillo S., Falcone P.M., Laverse J., Pati
S. and Baiano A. 2007. Textural changes of Canestrel-
as for other similar cheeses made with raw milk, lo Pugliese Cheese measured during storage. J. Food
where the addition of natural autochthonous mi- Eng. 83: 621.
croflora starters obtained from milk, the environ- Dufossé L., Galaup P., Carlet E., Flamin C. and Valla A. 2005.
ment and artisanal cheese-making procedures Spectrocolorimetry in the CIE L*a*b* color space as use-
ful tool for monitoring the ripening process and the quali-
and tools can lead to variable quality. ty of PDO red-smear soft cheeses. Food Res. Int. 38: 919.
These initial findings must be confirmed by González Viňas M.A., Ballesteros C., Martín-Alvarez P.J. and
analysing a larger set of semi-hard samples that Cabezas L. 2007. Relationship between sensory and in-
takes into consideration the effect of different strumental measurements of texture for artisanal and in-
dustrial Manchego cheeses. J. Sensory Studies 22: 462.
producers, manufacturing conditions and time
Innocente N. and Corradini C. 1998. Impiego di una tecni-
of production on reproducibility of the model. ca di analisi dell’immagine nel controllo di qualità del for-
The model must also be improved by analys- maggio Montasio. Sci. Tecn. Lattiero-Casearia 49: 8294.
ing soft Pecorino of Appennino Reggiano chees- Lawrence R.C., Creamer L.K. and Gilles J. 1986. Sympo-
es in order to differentiate between samples of sium: cheese ripening technology. Texture development
during cheese ripening. J. Dairy Sci. 70: 1748.
the two commercial categories according to rip-
Lucey J.A., Johnson M.E. and Horne D.S. 2003. Invited re-
ening time. view: perspectives on the basis of the rheology and tex-
ture properties of cheese. J. Dairy Sci. 86: 2725.
Mancini M., Moresi M. and Rancini R. 1999. Mechanical
References properties of alginate gels: empirical characterisation. J.
Food Eng. 39: 369.
Alvarenga N., Silva P., Garcia J.R. and Sousa I. 2008. Esti- Official Bulletin of Emilia Romagna Region of 30 March
mation of Serpa cheese ripening time using multiple lin- 2005. Code of Rules of Pecorino of Appenino Reggiano.
ear (MLR) considering rheological, physical and chemi- Off. Bull. Emilia Romagna Region 59: 194.
cal data. J. Dairy Res. 75: 233. Pignedoli S. 2005. Caratterizzazione del formaggio “Pecorino
AOAC. 1995. “Official Methods of Analysis” 16th ed., Asso- dell’Appennino Reggiano”. Doctoral Fellowship Thesis in
ciation of Official Analytical Chemistry, Washington, DC. Food Biotechnology, University of Bologna, Bologna, Italy.
Bourne M.C. 1978. Texture profile analysis. Food Tech- Pinho O., Mendes E., Alves M.M. and Ferreira I.M.P.L.V.O.
nol. 32: 62. 2004. Chemical, physical and sensorial characteristics of
“Terrincho” ewe cheese: change during ripening and in-
Buffa M.N., Trujillo A.J., Pavia M and Guamis B. 2001. travarietal comparison. J. Dairy Sci. 87: 249.
Changes in textural, microstructural, and colour charac-
teristics during ripening of cheese made from raw, pas- Pollard A., Sherkat F., Seuret M.G. and Halmos A.L. 2003.
teurized or high-pressure-treated goats’milk. Int. Dairy Textural changes of natural cheddar cheese during the
J. 11: 927. maturation process. J. Food Sci. 68: 2011.
Caccamo M., Melilli C., Barbano D.M., Portelli G., Marino G. Poveda J.M., García A., Martín-Alvarez P.J. and Cabezas L.
and Licita G. 2004. Measurement of gas holes and me- 2004. Application of partial least squares (PLS) regres-
chanical openness in cheese by image analysis. J. Dairy sion to predict the ripening time of Manchego cheese.
Sci. 87: 739. Food Chem. 84: 29.
Casiraghi E., Lucidano M. and Pompei C. 1989. Correla- Rohm H. and Jaros D. 1997. Colour of hard cheese. 2. Fac-
tion among instrumental texture, sensory texture and tor of influence and relation to compositional parame-
chemical composition of five Italian cheeses. Ital. J. Food ters. Zeit. Leben. Unter. Forsch. 204: 259.
Sci. 1: 53. Roseiro L.B., Wilbey R.A. and Barbosa M. 2003. Serpa
Castagnetti G.B., Arlotti G., Bertolini L. and Pignedoli S. cheese: Technological, biochemical and microbiological
2004. Caratterizzazione e valorizzazione del formaggio characterisation of a PDO ewe’s milk cheese coagulated
Pecorino dell’Appenino Reggiano. Proceedings of XVI Na- with Cynara cardunculus L. Lait 83: 469.
tional Congress of Italian Society of pathology and breed- Slow Food. http://www.presidislowfood.it/ita

Revised paper received June 25, 2009 Accepted September 10, 2009

Ital. J. Food Sci. n. 1, vol. 22 - 2010  59


Copyright of Italian Journal of Food Science is the property of Chiriotti Editori S.P.A. and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.