HAL Id: hal-00929519https://hal.archives-ouvertes.fr/hal-00929519

Submitted on 1 Jan 1997

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Rheological properties and maturation of New ZealandCheddar cheese

P Watkinson, G Boston, O Campanella, C Coker, K Johnston, M Luckman, NWhite

To cite this version:P Watkinson, G Boston, O Campanella, C Coker, K Johnston, et al.. Rheological properties andmaturation of New Zealand Cheddar cheese. Le Lait, INRA Editions, 1997, 77 (1), pp.109-120. <hal-00929519>

Lait (1997) 77,109-120© Elsevier/INRA

109

Original article

Rheological properties and maturationof New Zealand Cheddar cheese

P Watkinson '. G Boston 1,0 Campanella 2, C Coker 1,

K Johnston 1, M Luckman '. N White 1

1 New Zealand Dairy Research lnstitute, Private Bag 11029, Palmerston North;2 Food Technology Department, Massey University, Private Bag 11222, Palmerston North, New Zealand

Summary - Trends in the fracture strain, modulus of deformability and chemical properties as a func-tion of storage time were determined for Cheddar cheese made in the New Zealand Dairy ResearchInstitute' s pilot plant. The apparent fracture strain of Cheddar cheese increased during the first14---28days and thereafter decreased, asl-Casein levels decreased monotonically and non-protein nitro-gen levels increased with storage. Fusion of curd particles probably contributed to the initial increasein fracture strain, and the decrease in strain can be rationalized in terms of increasing proteolysis, Themodulus of deformability increased by at least a factor of two over the initial several weeks of stor-age and then increased slightly or remained constant. However, the moi sture content of Cheddar cheesechanged very little (the maximum range being 34.6-33.0% with no monotonie change over time). Theincrease in the modulus over the first 14 days was not associated with a decrease in moisture content.Differentiai scanning calorimetry indicated there was sorne crystallization of milkfat from 91 to210 days of storage, and this (together with small moisture los ses) may partly explain the smallincrease in the modulus of deformability over this period of time.

cheese maturation / rheological property / fracture property / composition / Cheddar cheese

Résumé - Propriétés rhéologiques et maturation du fromage de Cheddar de Nouvelle-Zélande.On a déterminé J'évolution, en fonction du temps d'affinage, de la déformation à la fracture, dumodule de déformabilité et des propriétés chimiques du fromage de Cheddar, fabriqué à l'usinepilote de l'Institut de recherches laitières de Nouvelle-Zélande. La déformation apparente à la frac-ture du Cheddar augmentait pendant les 14---28premiers jours, et diminuait ensuite. La teneur encaséine asl diminuait régulièrement pendant l'affinage et la teneur en azote non protéique aug-

Oral communication at the IDF Symposium 'Ripening and Quality of Cheeses', Besançon, France, February26-28, 1996.

110 P Watkinson et al

mentait. La soudure des particules de caillé contribuait vraisemblablement à l'augmentation ini-tiale de la déformation à la fracture, alors que la diminution ultérieure de ce paramètre pouvait êtreattribuée au développement de la protéolyse. Le module de déformabilité augmentait au moins d'unfacteur 2 pendant les premières semaines de maturation, puis diminuait légèrement ou restait constant.Or, la teneur en eau du fromage de Cheddar changeait très peu (la plage de variation maximumétant de 34,6-33,0 %, sans évolution régulière au cours du temps). L'augmentation du module de défor-mabilité pendant les 14 premiers jours n'était pas associée à la diminution de la teneur en eau. La calo-rimétrie différentielle à balayage a montré qu'il y avait un peu de cristallisation de la matière grasseentre 91 et 210 jours de stockage et que cela, associé à une légère perte d'eau, peut expliquer par-tiellement le faible accroissement du module de déformabilité pendant cette période.

maturation du fromage / propriété rhéologique / propriété de rupture / composition / fromagede Cheddar

INTRODUCTION

New Zealand exports most of its cheese. A largeproportion of this exported cheese is Cheddarcheese which requires a long storage time beforeit is suitable for eating. The eating and cuttingproperties of Cheddar cheese depend not onlyon conditions of curd formation and handling,but also on storage conditions. A mild NewZealand Cheddar cheese with a high longness(resistant to crumbling) requires a storage timeof about 6-12 months at a storage temperaturenear 13 "C. By comparison, a mature NewZealand Cheddar cheese with relatively highstiffness (hard to dent) and low longness(easilycrumbles) requires a storage time greater than12 months at the same storage temperature. Notonly functional properties like eating and cut-ting but also f1avour properties change duringstorage. Understanding how these changes occurcan reduce the variability of the properties ofCUITentcommercial cheese and can assist in newproduct development.

The major are a of science related to theinstrumental measurement of the functionalproperties Iike texture, cutting and melting isrheology (and fracture properties). In addition,rheological and che mi cal measurements inchee se give insights into the physico-chemicalchanges occurring du ring ripening. The effectof maturation on rheological and fracture prop-

erties has been reported for various cheesesincluding American Cheddar chee se (Creamerand OIson, 1982), Colby chee se (Crea mer et al,1988a), New Zealand Cheddar chee se (Creameret al, 1988b) and Gouda cheese (Luyten, 1988).Creamer and Oison (1982) found that compres-sion at the yield point (longness) was reducedby protein breakdown and that longness andyield force were also inf1uenced by moisturecontent and pH. Creamer et al (1988b) foundthat the force to slightly deform the chee seincreased until about 3--4 months of storage andthen changed little. Calf rennet and rennilasecoagulant gave similar curves for this force dur-ing cheese storage. Luyten (1988) explored theeffect of curd fusion and cheese pH on the rhe-ology of Gouda cheese, finding that fusionoccurred within a week in standard Goudacheese. Fracture strain (Iongness) decreasedwhen the amount of peptides and amino acidsformed from the parent caseins increased, andmodulus (stiffness) at a given pH was mainlyincreased by decreasing the moi sture contentdu ring maturation (Luyten, 1988).

The main objective of this work was to findtrends in a selection of weil defined rheologi-cal and fracture properties of New Zealand Ched-dar cheese as a function of storage time. Anotherobjective was to gain sorne insight into thephysico-chemical changes occurring duringripening by comparîng these trends with changes

Maturation and rheology of Cheddar cheese

in chemical composition. In this paper, only thefracture property fracture strain and the rheo-logical pro pert y modulus of deformability arediscussed.

Because cheese is a non-linear viscoelasticmaterial, rheological and fracture properties area function of the time scale of the experiment(strain rate, Ë) as weIl as the strain applied (Wal-stra and Peleg, 1991). Although the effect of Ëwas not examined in the CUITentstudy, the initialvalue of Ë (3.3 x 10-2 s-I) was weIl defined andwithin the range used by other workers whohave examined the effect of Ë on cheese fracturestrain and modulus (Luyten, 1988; Rohm andLederer,1991).

Cheddar cheese is eaten when it is at tem-peratures ranging from refrigerator temperaturesto those at ambient conditions. Therefore anindication of texturai (eating) properties in thisstudy was found by conducting tests at 5 and20 oc. Sensory evaluation was not carried out,but indications of the properties of stiffness(rigidity), and longness (resistance to crumbling)were obtained from the modulus of deformabil-ity and the fracture strain respectively (Visser,1991;Zoon,1991).

Fracture strain not only indicates longnessbut also influences cutting properties. If it is toosmall, cheese tends to crumble when eut.

MA TERIALS AND METHODS

Experimental designand manufacture of cheese

Two vats of Cheddar chee se were made at theNew Zealand Dairy Research lnstitute (NZDRI)pilot plant. Standardized (protein 3.5%, fat 4.9%)and heat-treated (65 "C, 15 s) milk was heldovernight before being pasteurized (72 "C, 15 s)and pumped into 375-L vats. The initial milkpH was 6.61. Then 1.5% starter bacteria (Lac-tococcus lactis subsp cremoris strains) and14 mL calf rennet/100 L (New Zealand Stan-dard Calf Rennet, New Zealand Rennet Co,

III

Eltham, New Zealand) were added, and the milkwas left to set for 40 min before cutting. The set-ting temperature was 33 "C, the cooking tem-perature was 39 "C, the draining pH was 6.25, thecheddaring time was 2 h 25 min and the saltingpH was 5.35. The cheese was made on the sameday to reduce variability from the milk supply,and was made slightly after the middle of thedairy season (in February).

Each vat yielded about 40 kg of curd, whichwas nominally split into two 20 kg lots of curd.AIl four lots received an initiallight pressing byhand into rectangular moulds. Vat 1 had oneblock of curd that was subjected to the standardtreatment of mechanical pressing (nominally of0.3 MPa) overnight, and one block that was notpressed. The two blocks from vat 2 received thestandard mechanical pressing. The two vats sub-jected to standard mechanical pressing treatmentwere regarded as replicates. After pressing, thefour nominally 20 kg blocks of cheese were eachdivided into four equal sub-blocks and werestored in vacuurn-sealed polyethylene bags at13 oc. At each of the storage times (the full rangebeing l, 14, 28, 50, 124, 209 and 394 days), anew sub-block of chee se was used. A split-split-plot design was used with an unequal numberof mainplot experimental units (vats 1 and 2) foreach mainplot treatment (pressed/unpressed).The subplot treatment, applied after splittingeach of the mainplots into equal units, was stor-age time. Data were available for aIl mainplots(vats) for storage times 14,28 and 209 days only;therefore the data set for the analyses includedonly these storage times. The sub-sub-plot treat-ment, applied after splitting each of the subplotsinto equal units, was test temperature (5 or 20 "C)for the rheological responses. Statistical analysesof the data were carried out using the generallinear models procedure in SAS (SAS [1994),Release 6.1. SAS lnc, Cary, NC, USA). Anextension of methods described by Milliken andJohnson (1984) was used due to the unbalancednature of the design.

Linear regression analyses were performedusing Minitab (Minitab [1994], Release 10.1,Minitab Inc, State College, PA, USA).

112 P Watkinson et al

Another two vats of Cheddar cheese weremade just before the middle of another dairyseason (in November) to allow measurernent ofdifferential scanning calorimetry (DSC). Themanufacturing procedures were similar to thoseused during the main manufacturing in February,and measurements were taken at storage times of28,91 and 210 days.

Sample preparationfor rheological measurements

For uniaxial compression and relaxation exper-iments, a core borer with an inner diameter of20.0 mm mounted on a drill press was used.Cores were taken parallel to the direction ofpressing of the original block at points half-waybetween the centre and the edge of the top sur-face. The core borer was lubricated with paraf-fin oil (high viscosity type, 340-360 Sayboltuniversal seconds at 37.8 "C) to make it easier toeut out a sample. The cores were placed in atemplate and eut into cylinders 25.0 mm inheight by a wire-cutting apparatus. The sampi eswere wrapped in polyethylene film and allowedto equilibrate to the test temperature.

Uniaxial compression

Uniaxial compression experiments were per-formed on a T AHD compression tension testinstrument (Stable Micro Systems, Haslemere,UK) with a 50 kg load cell with a resolution of1 g and an accuracy of 0.025%. The distancemeasurement had a resolution of 0.001 mm. TheTAHD was connected to a personal computerwith a data transfer rate of force, displacementand time data triplets of 50 Hz. Temperaturewas controlled by placement of the instrumentand sample in a controlled temperature room.Two parallel Teflon plates were used. Sampleswere placed between plates that had been lubri-cated with paraffin oil, and a crosshead speedof 0.83 mm/s was used to compress samples to80% Cauchy strain. The number of measure-ments taken for each chee se at each storage time

and test temperature was typically six for uni-axial compression.

The experimental data were initially anal-ysed using XTRAD software (Stable Micro Sys-tems) and the appropriate data were exportedinto software (Master Work Software, Tawa,New Zealand) written in J, a functional pro-gramming language (lverson Software Inc,Toronto, Canada) (Iverson, 1991; McIntyre,1991). This software (using J) calculated Henckystrain as suggested by Peleg (1977), and hereafteris abbreviated to strain. The software calculatedstress assuming a constant volume during com-pression. The assumption of a negligibledecrease in sample volume upon compressionwas a reasonable one for Cheddar cheese(Calzada and Peleg, 1978), being on average a9% volume reduction. The equations were:

Ëo= v / ho

Ec Sh, / ho

Eh -In (1- Ec)

cr (IOOOFt / 1trÔ)( 1 - E c)

[1]

[2]

[3]

[4]

speed of compression (mm/s); .initial sample height (mm);displacement of crosshead at time t(mm);force from lubricated compression atlime t (N);initial radius of sample (mm);

initial strain rate (s-I);

Cauchy strain (-);

Hencky strain (-);stress from lubricated compression test(kPa).

The apparent fracture strain was the strain atthe local maximum for stress in the stress versusstrain curve. The apparent modulus of deforma-bility was the slope of the stress versus straincurve at low strain (typically below 0.03) where

Maturation and rheology of Cheddar cheese

the curve was close to a straight line. Althoughrheological and fracture properties are depen-dent on E, the term apparent, used to denote thisdependence, is omitted for the sake of brevity.

Chemical measurementsand proteolysis measurements

The procedures described previously (NewZealand Oairy Board, 1993) were used for mea-suring fat (Babcock type), moi sture (16 h oyendry at 105 "C, gravimetrie measurement), salt(Volhard method), pH (measured electrometri-cally against two reference buffer solutions) andcalcium (titration using EOT A and Patton andReeders reagents). To measure non-proteinnitrogen (NPN) the sample was first dissolved in1 mol/L NaOH. Trichloroacetic acid (TCA)(15%) was added until the acid strength of thesolution was 12% to precipitate the proteins.The filtrate (NPN soluble in 12% TCA, accord-ing to the lOF [1993]) was tested for nitrogencontent by an automated Kjeldahl procedurewith a Kjel-Foss automatic 16200 (A/S N FossElectric, Hillerod, Oenmark). The fraction sol-uble in 15% TCA contained mainly urea, freeamino acids and peptides.

The relative amounts of usl-casein, usi-Icasein and ~-casein were measured using alka-line urea-polyacrylamide gel electrophoresis(PAGE). The Bio-Rad Mini-Protean II systemwas used with ten slot gels of 0.75 mm thick-ness and a Bio-Rad model 1000/500 power sup-ply according to previously described methods(Creamer, 1991). Following destaining, the gelswere photographed and scanned. Modificationsto these methods were as follows. Chee se(0.500 g) was dispersed in 25 mL of samplebuffer. The samples were heated to 40 "C andheld at that temperature for 1 h, to transform fatinto the liquid phase prior to blending with anUltra-Turrax T25 (Janke and Kunkel, IKA-Labortechnik, supplied by Labsupply Pierce,New Zealand) at approximately 24000 rey/min

113

for 20 s. The warm sampi es were then cen-trifuged at 10 000 rpm at 4 "C for 10 min tosolidify and separate the fat. The aqueous sub-natant (2 mL) was treated with 2-mercap-toethanol (20 ilL/mL) and held for 18 h prior toloading 5 ul, of the mixture into the gel slab. Arennet casein standard was prepared by dis-solving 12.0 mg of rennet casein in 6 mL ofsample buffer. After stirring for 1 h, the stan-dard was diluted to give a final concentrationof 1 mg/mL. A trim milk standard was preparedby diluting 0.1 mL of trim milk with 3.9 mL ofurea sample buffer. Each standard was treatedwith 2-mercaptoethanol (20 ilL/mL) as weil asbromophenol blue and held for 18 h prior toloading 10 ul, of the mixture into the gel slab.

One chemical and PAGE measurement wastaken for each cheese at each storage time.Chemical measurements were made soon aftersampling. At each storage time a sample ofcheese was also stored below -40 "C until ailthe samples had been collected. These sampleswere then analysed at the same time for NPNand by PAGE.

Differentiai scanning calorimetry

Samples were vacuum-packed and kept at thestorage temperature until just prior to the OSCtest. A cylinder 2 mm in diameter was takenwith a core borer, and eut to about 1 mm inlength. Il was immediately put into a pre-weighed metal cup, sealed and reweighed.Shortly afterwards the sample and eup wereplaced in a OSC 7-PC (Perkin-Elmer Corp, Nor-walk, USA) and the following temperature timehistory was used. Samples were cooled fromnear 20 "C to -40 "C at-10°C/min. They wereheld at-40 "C for 5 min, and then heated whilemeasurements were taken of the energy requiredto heat the sam pIe from -40 "C to 70 "C at5 OC/min. The onset of melting and the energyper mass of the water and two milkfat fractionswere measured.

114 P Watkinson et al

RESUL TS AND DISCUSSION

Chemical composition

The fat and moisture contents of the main exper-imental cheese made in February (table 1) variedlittle over the storage time. However, from 124to 394 days for vat 2, the moi sture contentdecreased slightly. Visible exudate from thecheese surface together with the use of samplesfor moisture measurement that had very little ofthe original cheese surface probably caused thisdecrease in moisture content. Polyethylene bagssealed under a partial vacuum provide a water-tight container for the maturing cheese. There-fore the moi sture content of the Cheddar cheeseduring maturation did not decrease significantlydue to evaporation of chee se moisture. Asexpected, the pH initially decreased (table 1)bec au se of the continued metabolism of theremaining lactose to lactic acid by the starterbacteria. During maturation, the usl-casein and

~-casein levels decreased and the usI-I case inlevels increased, reached a maximum near 28days and then decreased (fig 1). The NPN levelsincreased with time (table 1).

The composition of the cheeses made inNovember was slightly different from that ofthe cheeses in the main trial in February. In par-ticular, the moi sture in the non-fat substance(MNFS) at 1 day for the cheese made in Novem-ber was higher (52.6-53.2%) than that for thecheese made in February (50.1-51.0%).

Fracture strain

There were two different regions on the frac-ture strain (E f, or longness) versus storage timegraph (fig 2). The first region showed an initialincrease in Ef from 1 to 14 or 28 days, and thesecond region a decrease in Ef after 28 days.This trend during storage was similar for bothtest temperatures, but E f was lower at 5 "C thanat 20 oc. Temperature had a significant effect

Table I.Chemical properties and NPN during maturation of Cheddar chee ses made in February.Propriétés chimiques et NPN déterminés pendant la maturation des fromages de cheddar fabriqués en février.

Cheese vat Storage time Fat content Moisture content MNFS pH NPN(days) (0/0) (0/0) (0/0)

1 14 32.5 34.2 50.7 5.20 0.191 28 33.5 33.8 50.8 5.27 0.341 124 33.0 33.9 50.6 5.25 0.711 209 34.0 33.8 51.2 5.45 0.821 (unpressed) 1 32.5 34.4 51.0 5.27 0.201 (unpressed) 14 35.0 34.6 53.2 5.17 0.281 (unpressed) 28 32.0 34.1 50.1 5.28 0.421 (unpressed) 209 35.0 34.3 52.8 5.52 0.782 1 33.5 33.3 50.1 5.21 0.242 14 33.0 34.5 51.5 5.07 0.322 28 33.0 33.8 50.4 5.08 0.362 50 32.0 33.9 49.9 5.09 0.432 124 34.5 34.0 51.9 5.23 0.742 209 34.5 33.7 51.5 5.32 0.882 394 35.5 33.0 51.2 5.28 1.15

MNFS: moisture in the non-fat substance; NPN: non-protein nitrogen.

Maturation and rheology of Cheddar cheese

on Er (P = 0.0001). In addition, Er was slightlylower when unpressed at storage limes from 1 to28 days, but was slightly higher when unpressedat 209 days. Pressing had a significant effect onEr that depended on storage time (P = 0.0025).The graph of ErOver time (fig 2) shows the vari-ability between the two replicates (vat 1 and 2)and the properties of vat 1 unpressed. The coef-ficient of variation (cov) of Er for a given stor-age time, vat and temperature indicates themethod variability plus the variability of thematerial within a replicate. Half of the covs of Erwere below 8.2%. The full range for these covswas 3.3 to 24%.

Luyten (1988) reported a study of the changesin rheological and fracture properties of Goudaduring early maturation. During about the firstweek, curd fusion and diffusion of salt and waterincreased the viscous-Iike nature of the cheese,which was partly shown by an increase in Er.At later storage limes, Er was decreased byincreased splitting of caseins into small peptidesand amino acids. The breakdown of lXsl-casein

115

to large fragments in itself was probably notenough for Gouda cheese to decrease in Er.

In this study, there was a similar trend in Erduring the maturation of Cheddar cheese. Theinitial increase in Er was probably related tocurd fusion. Curd fusion starts at the interfaceof curd particles. Increasing the interfacial areaby pressing promotes greater curd fusion. Thehigher values of Er for pressed cheese comparedwith unpressed chee se can be explained by theincreased curd fusion induced by greater inter-facial area. In addition, the eut surfaces of the1-day old cheese were not as smooth as thoseof older cheese, further indicaling that curd par-ticles were less fused at 1 day than at later stor-age limes. Curd fusion in this Cheddar cheeseprobably occurred within the first 14 days, butmore experiments during early maturation areneeded to accurately characterise this time.

The current work showed an associationbetween Er and NPN (which includes peptidesand amino acids), which is consistent with thehypothesis reported by Luyten (1988) that a

0.7 ,i.1800 ~~

~ 1600 .. ::;:: 0.6

" 1400 • ~s ... 0a c:.c

1200 " 0.5'5 ;S

f 1"1000 • t "i • ü

.ê 0.4ë 800 âi

" • t c:§, 600 ... .~.g: • iii 0.3c:

400 ••• . ..."31~ • • 0<; 200 1 0.2c •.~ • • •"o 0 100 200 300 400

100 300 400200

Storage time (days)

Fig I.Changes in <1,;,-casein, <1,;,-1casein and f3-caseinduring maturation of three Cheddar cheeses; .: as ,-1casein; .: asl-casein: .A.: f3-casein; one data pointrepresents one cheese val.Évolutions de la caséine as!' asrl et de la caséine f3pendant l'affinage de trois fromages de type ched-dar.

500 Slorage lime (days)

Fig 2. Changes in strain at fracture during maturationof three Cheddar cheeses at two test tempe ratures;.: vat l,5°C; .: vat 1 unpressed, 5 oC; .A.: vat 2,5 "C; ... : vat 1, 20 "C; +: vat 1 unpressed, 20 "C;o :vat 2, 20 oc.Évolution de la déformation à la fracture pendantl'affinage de trois fromages de type cheddar à deuxtempératures de test.

116 P Watkin son et al

decrease in Efis caused by an increase in smallpeptides and ami no acids. Using the values of Efthat showed a decrease over time (omitting dataat 1 day), the CUITentresults for E f and NPNshowed a good fit using linear regression equa-tions with a significant negati ve slope(P < 0.0005 for both temperatures, r2(adj) = 0.949for 5 -c, r2(adj) = 0.895 for 20 -ci If ail datawere used (fig 3), the fit to the linear regressionequations (significant slope, P < 0.0005,r2(adj) = 0.680 for 5 "C, r2(adj) = 0.589 for 20 "C)was not as good as the fit with 1 day dataexcluded. This observation is consistent withthe hypothesis that the early maturation periodhas processes affecting the association betweenEf and NPN that are different from those in thelater main maturation period.

Although the asl-casein breakdown to asl-Icasein probably does not in itself cause adecrease in Ef, as1-casein breakdown influences

0.8 -,-~~~~~~~~~~~~~-----,

0.7

o

v ...~

~ ... ~::;:: 06

~cE. 0.5i"-§.ê 0.416c

~ 0.3

0.2

0.1 +-~---'-~-.~--'-~----'~~r-~-,------l0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Non-protein nitrogen (% wlw)

Fig 3. Influence of non-protein nitrogen levels onstrain at fracture during maturation of three Cheddarcheeses at two rheological test temperatures, Datapoints at adjacent storage limes are joined .• : vat l ,5 oC;.: vat 1 unpressed,5 "C; À: vat 2, 5 "C: T: vatl, 20 oC; +: vat 1 unpressed, 20 -c. 0: vat 2, 20 oc.Influence des niveaux d'azote non protéinique sur ladéformation à la fracture pendant l' affinage de troisfromages de type cheddar dans deux températures detest rhéologique. Les points à des temps de stockagevoisins sont réliés.

the subsequent breakdown to peptides and aminoacids. The later breakdown to peptides andamino acids probably does cause a decrease inE r- Thus there is likely to be an associationbetween asl-casein levels and Ef. A quadraticregression curve fitted the E f versus as I-caseindata with a local maximum at 14-28 days, andhad an intercept, linear coefficient and quadraticcoefficient that were significant (P < 0.0005).This quadratic curve had a reasonable fit to theEf versus asl-casein data (r2(adj) = 0.796 for5 "C, r2(adj) = 0.722 for 20 OC). However, a lin-ear regression equation had a po or fit to thesedata (slope P < 0.055, r2(adj) = 0.197 for 5 "C,r2 (adj) = 0.1 17 for 20 "C). Whereas a linear modeldid not adequately explain the relationshipbetween ast-casein and Ef, a quadratic regres-sion curve did.

The small increase in Ef in this work as thetemperature increased from 5 to 20 "C has beenfound by others (Luyten, 1988) and can be partiyexplained by an increase in viscous properties(Roefs, 1986).

Modulus of deformability

The modulus of deformability (Ed, or stiffness)increased markedly (by at least a factor of two)from 1 to 14 days, increased less steeply from 14to 28 days and thereafter increased slightly orremained about constant (fig 4). There was nogeneral decrease in Ed with time as found forE r- The trends in Ed during maturation weresimilar for both test temperatures, but Ed washigher at the lower test temperature of 5 oc.Temperature had a significant effect on Ed thatdepended on storage time (P = 0.046). Pressingdid not have a significant effect on Ed (P = 0.14)(although unpressed cheese had slightly lowermean values of Ed than pressed cheese [fig 4]).

Half the covs of Ed for a given storage time,vat and temperature were below 9.8%. The fullrange of these covs was 1.8 to 27%.

Luyten (1988) reported that Ed increased dur-ing initial storage and then decreased betweentwo and six days of storage (for standard Gouda,

Maturation and rheology of Cheddar cheese 117

5000

.. 4000~~.i5..

3000§S'""U

0'" 2000=>"5"8:2

1000

•00

'0

300 400100 200

Starage lime (days)

Fig 4. Changes in modulus of deformability duringmaturation of three Cheddar cheeses at two test tem-peratures .• : vat l,5°C; .: vat 1 unpressed, 5 "C;.À.: vat 2, 5 "C; T: vat 1,20 "C; +: vat 1 unpressed,20 -c. 0: vat 2, 20 -c.Évolution du module de déformabilité pendant l'affi-nage de trois fromages de type cheddar à deux tem-pératures de test.

but with salt added to the cheese milk insteadof by brining). Ed generally increased thereafterfor standard Gouda cheese. The decrease in Edmay have been caused by the increase in thespatial homogeneity of the water and the caseinin the curd and/or by proteolysis. The generalincrease in Ed found for standard Gouda cheeseafter about a week or more was attributed mainlyto a decrease in moi sture content or MNFS(induced in this case by evaporation of water inthe cheese) and also to changes in ionie strength.Later changes in Ed were not significantly influ-enced by proteolysis.

The trend toward an initial increase in Ed inthis study was similar to that in Luyten's work(Luyten, 1988). However, there were insuffi-cient measurements in this study over this earlymaturation period to determine if Ed decreasedafter an initial increase, as found by Luyten(1988). The large increase in Ed from one to14 days in this study was not associated with a

decrease in chee se moi sture content (table 1) orMNFS. There was little evidence in this studyfrom the effect of pressing to support the ideathat curd fusion in itself caused the large initialincrease in Ed. It is not clear from this work whatdid in fact cause this initial large increase in Ed .

Further work to investigate the physical andchemical causes of the large initial increase in Edwould provide more insight into the physico-chemical changes occurring in young Cheddarcheese.

The trend toward no large increases in Ed inthis study after about 28 days is not the same asthat found by Luyten (J 988). Increases in Edcannot he explained by changes in moi sture con-tent or the MNFS as in Luyten's study (1988).Unlike Luyten's brine-salted Gouda left to evap-orate in the air, the Cheddar chee se made in thisstudy was sealed in polyethylene bags. Con se-quently, beyond sample and method variability(and one case of exudate), no definite trends inmoi sture content or MNFS were observed. Therewas no significant correlation between Ed andmoi sture content (r2(adj) = 0.035 for 5 "C tests,r2(adj) < 0.0005 for 20 "C tests) and sIopes atboth temperatures were not significant(P> 0.05). However, the data at 1 day were out-liers; when these were removed, there was a cor-relation between Ed and moisture contentwith a significant (P < 0.05) negative slope(r2(adj) = 0.678 for 5 "C tests, r2(adj) = 0.525 for20 "C tests). There was no significant correlationbetween Ed and MNFS (r2(adj) < 0.05%,P> 0.3) either for ail data or when the outliers at1 day were removed. This lack of correlationmay partI y be a reflection of the increased vari-ability introduced by the fat measurement usedto calcuJate MNFS.

The increase in Ed (particularly after the ini-tiaI maturation period) may have been partlycaused by a slight decrease in the amount ofwater in the casein, induced by proteolysis. Pep-tide bond cleavage causes more ionie groups,which compete for water in the chee se, to beformed (Creamer and OIson, 1982). Thus thereis Jess water available for solvation of the casein.Less plasticization of casein by water will

118 P Watkinson et al

Table II. Melting properties during maturation of Cheddar cheeses made in November (from differential scan-ning calorimetry).Propriétés de fonte déterminées par calorimétrie différentielle à balayage pendant la maturation des fromagesde cheddar fabriqués en novembre.

Cheese vat Storage time Onset of melting, Onset of melting, Energy/mass. Energy/mass.(days) low temp milkfat high temp milkfat low temp milkfat high temp milkfat

fraction (oC) fraction (OC) fraction (J/g) fraction (J/g)

1 91 7.7 26.7 1l.8 6.92 91 7.8 26.6 12.3 5.31 210 4.8 23.8 6.9 10.22 210 4.4 23.9 5.6 1l.9

increase Ed. The moi sture content of the cheesemay not change much during proteolysis in asealed polyethylene bag, but the water availablefor solvation of the casein may decrease. Theseideas need to be confirmed by measurements ofsolvation of the casein in cheese. Although pep-tide cleavage uses up water, the cheese mois-ture content did not generally decrease withtime (with the exception of vat 2 from 124to 394 days).

The small increases in Ed from 91 to 210days may be partI y explained by an increase incrystallinity of the milkfat in the cheese.Although the casein matrix mainly determinesthe solid nature of cheese, the amount and mod-ulus (stiffness) of the milkfat influences the mod-ulus of cheese (Visser, 1991). Prentice (1987)reported that sorne glycerides in cheese slowlycrystallized if the processing temperature washigher than the storage temperature. This require-ment for crystallization was met for the CUITentcheese processing temperature (near 30 "C) andstorage temperature (13 "C),

Table Il shows changes in fat melting prop-erties from DSC on the Cheddar cheese madein November. Although these results apply tothe cheese made in November and cannot becorrelated with the Ed values from the Februarychee se, the principles probably apply to bothcheeses. These results gave an indication that

from 91 to 210 days of storage there was sornecrystallization of milkfat. Over this period oftime, the energy per mass of the low temperaturemilkfat fraction decreased and the energy permass of the high temperature milkfat fractionincreased by a slightly lower amount. There wastherefore a larger proportion of the higher rnelt-ing temperature milkfat fraction in the cheeseat the longer storage time.

From 91 to 210 days of storage, the onset ofmelting of both the lower melting point milkfatfraction and higher melting point milkfat fractiondecreased (table Il). These onset of melting tem-peratures indicated that at a test tempe rature of5 "C both milkfat fractions were nominally solid(bec au se both milkfat fractions had an onset ofmelting near 5 "C or more). Similarly, at a testtemperature of 20 "C only the higher meltingmilkfat fraction was nominally solid (becauseonly the high temperature milkfat fraction had anonset of melting near 20 "C or more). These areguidelines only because the exact melting tem-peratures of the milkfat in the cheese will dependon the temperature history and rates of change oftemperature used for the DSC measurement.

The value of Ed at 5 "C was much higher (byat least a factor of three) than at 20 "C because thereduction in temperature caused an increase inthe fraction of milkfat in the solid phase. Anhy-drous milkfat has about 60% solid fat at 5 "C

Maturation and rheology of Cheddar cheese

and about 20% solid fat at 20 "C (MacGibbonand McLennan, 1987). Thus the modulus of themilkfat increased with decreasing temperatureand in this case caused the Ed of the cheese(which may indicate stiffness) to increase withdecreasing temperature. Over the range from 14to 26 "C, the rheological properties of Goudachee se were changed mostly by the change incrystallinity of the milkfat (Visser, 1991).

CONCLUSIONS

The trends in E f during the maturation of Ched-dar cheese made at the NZORI were an increasefrom 1 to 14 or 28 days and a monotoniedecrease thereafter to 209 days. Curd fusion wasprobably one factor related to the initial increasein E f' The decrease in E f after 14-28 days wasassociated with an increase in NPN which wasconsistent with the hypothesis that the amount ofproteolysis shown by the appearance of peptidesand amino acids influenced E f' The Ed increasedmarkedly (by at least a factor of two) over thefirst 14-28 days, and thereafter remained con-stant or increased slightly. The initial increase inEd over the first 14 days was not associated witha decrease in moisture content. Further investi-gations of the causes of this initial increase inEd would give more insights to the physico-chemical changes occurring during the earlymaturation period of Cheddar cheese. The tem-perature of the cheese over the range 5 to 20 "C(corresponding to a range of temperatures atwhich cheese is eaten) has a small influence onEf (longness) and a large influence on Ed (stiff-ness). OSC indicated there was sorne crystal-lization of milkfat from 91 to 210 days of stor-age, and this (together with small moisturelosses) may parti y explain the small increase inEd over this period of time.

ACKNOWLEDGMENTS

The authors wish to thank LK Creamer for valu-able discussions. We wish to thank the CheeseTechnology Section of the NZORI for cheese-

119

making and the Analytical Chemistry Sectionof the NZORI for compositional analysis.

REFERENCES

Calzada JF, Peleg M (1978) Mechanical interpreta-tion of compressive stress-strain relationships ofsolid foods. J Food Sei 43, 1087-1092

Creamer LK (1991) Electrophoresis of cheese. BullIntDairyFed26l,14-28

Creamer LK, Oison NF (1982) Rheological evalua-tion of maturing Cheddar cheese. J Food Sei 47,631-636,646

Crea mer LK, Gilles J, Lawrence RC (1988a) Effectof pH on the texture of Cheddar and Colby cheese.NZ J Dairy Sei Technol23, 23-35

Crea mer L, Aston J, Knighton D (1988b) Sorne dif-ferences between Cheddar cheeses made using calfrennet and a microbial coagulant (Rennilase 46L).NZ J Dairy Sci Technol23, 185-194

IDF (1993) Milk determination of nitrogen content.4. Determination of non-protein nitrogen content.IDF Standard 20B, lnt Dairy Fed, Brussels

lverson KE (1991) The ISI Dictionary of J. Version3.3. Iverson Software lnc, Toronto, Canada

Luyten H (1988) The rheological and fracture prop-erties of Gouda cheese. PhD thesis, WageningenAgric Univ, Wageningen, The Netherlands

MacGibbon AKH, McLennan WD (1987) Hardness ofNew Zealand patted butter: seasonal and regionalvariations. NZ J Dairy Sei Technol Tl; 143-156

Mc1ntyre DB (1991) Language as an intellectual tool:from hieroglyphics to APL. IBM Systems J 3D,554-581

Milliken GA, Johnson DE (1984) Analysis of MessyData. Vol 1. Designed Experiments. Van NostrandReinhold, New York, 378-384

New Zealand Dairy Board (1993) NZTM3 NewZealand Dairy Industry Chemical Methods Man-ual. NZ Dairy Board, Wellington

Peleg M (1977) Operational conditions and the stressstrain relationship of solid foods. Theoretical eval-uation. J Texture Stud 8,283-295

Prentice JH (1987) Chee se rheology. ln: Cheese:Chemistry, Physics and Microbiology. Val 1. Gen-eral Aspects, 1st edn (Fox PF, ed) Elsevier,London

Roefs SPFM (1986) Structure of acid casein gels. PhDthesis, Wageningen Agric Univ, Wageningen, TheNetherlands

120 P Watkinson et al

Rohm H, Lederer H (1992) Uniaxial compressionof Swiss-type chee se at different strain rates.lnt Dairy J 2, 331-343

Visser J (1991) Factors affecting the rheological andfracture properties of hard and semi-hard cheese.Bull/Ill Dairy Fed 268, 49-61

Walstra P, Peleg M (1991) General considerations.Bull/Ill Dairy Fed 268,3-4

Zoon P (1991) The relation between instrumental andsensory evaluation of the rheological and fractureproperties of cheese. Bull/Ill Dairy Fed 268,30-35