UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluation of breadmaking quality from common wheat quality parameters and fundamental rheological, chemical,… Masuhara, Yasuo J. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_fall_masuhara_yasuo.pdf [ 4.18MB ]
Metadata
JSON: 831-1.0086260.json
JSON-LD: 831-1.0086260-ld.json
RDF/XML (Pretty): 831-1.0086260-rdf.xml
RDF/JSON: 831-1.0086260-rdf.json
Turtle: 831-1.0086260-turtle.txt
N-Triples: 831-1.0086260-rdf-ntriples.txt
Original Record: 831-1.0086260-source.json
Full Text
831-1.0086260-fulltext.txt
Citation
831-1.0086260.ris

Full Text

EVALUATION OF BREADMAKING QUALITY FROM COMMON WHEAT QUALITYPARAMETERS AND FUNDAMENTAL RHEOLOGICAL, CHEMICAL, ANDFUNCTIONAL PROPERTIES OF WHEAT CULTIVARSByYASUO JOHN MASUHARAB.Sc.(Cell Biol.), The University of British Columbia, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIESDepartment of Food ScienceWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAOCTOBER 1993© YASUO JOHN MASUHARA, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of Foot ScThe University of British ColumbiaVancouver, CanadaDate 0^EA- 1 r J ( 9 7 3DE-6 (2/88)iiABSTRACT The stress relaxation and deformability modulus of wheatdoughs made from five groups of wheat cultivars of differingbreadmaking ability were determined. Doughs made from poorbreadmaking cultivars had greater rates of stress relaxation thandoughs made from good breadmaking cultivars. Doughs made fromhard wheats had a greater resistance to deformation(deformability modulus) than doughs made from soft wheats. Thesolubility, sulfhydryl, and disulfide content of gluten of thefive groups of wheat cultivars were also determined. Gluten ofpoor breadmaking cultivars had greater sulfhydryl content thangluten of good breadmaking cultivars. No differences were foundin the extent of stress relaxation, gluten disulfide content, andgluten solubility of the different groups of wheat cultivars.The prediction of breadmaking quality (loaf volume) byfundamental rheological, chemical, and functional properties andcommon wheat, flour, and dough quality tests was examined. Simplelinear regression of baking tests on common wheat, flour, anddough quality tests did not result in any single test having asquared correlation coefficient greater than 0.743. Generalmultiple regression equations predicting baking test volumes weremade based on the significance of the simple linear regressions.The adjusted squared multiple correlation coefficients of theseequations were as high as 0.924. A more practical predictionliiequation (adjusted R 2=0.907) with only five independent variables(mixograph first minute slope, extensigraph extensibility toresistance ratio, sedimentaion value, thousand kernel weight, andB value) was obtained by stepwise regression analysis. Theaddition of the measurements of the fundamental rheological(stress relaxation, deformability modulus), chemical (sulfhydryland disulfide content), and functional (solubility) propertiesto the regression equations did not improve the adjusted squaredmultiple correlation coefficients of the equations.Principal component analysis reduced the data set to sevencomponents which explained 91% of the total variation.Qualitative (dough strength parameters) and quantitative (millingparameters) components were identified.ivTABLE OF CONTENTS ABSTRACT^  iiTABLE OF CONTENTS^  ivLIST OF TABLES  viLIST OF FIGURES^ viiiACKNOWLEDGEMENT  ixINTRODUCTION ^1LITERATURE REVIEW^  4Definition of Quality ^4Determinant of Breadmaking Quality  ^5Methods of Determining Breadmaking Quality ^7Baking tests ^8Protein content ^8Farinograph ^9Mixograph^  12Extensigraph  13Wheat/flour strength tests^  13Milling quality tests  17a-amylase tests^  18Prediction of Baking Tests by Quality Parameters^ 19Dough Structure/Dough Rheology Relationship^ 23MATERIALS AND METHODS^  26Wheat Samples^  26Rheological Studies^  27Dough preparation  27Force deformation/stress relaxation test^ 27Analysis^  28Functional and Chemical Studies^  31Gluten isolation^  31Solubility determination  32Determination of sulfhydryl content^  32Determination of disulfide content  34VCommon wheat, flour, and dough quality tests^ 35Statistical Analysis^  36Analysis of variance  36Multivariate statistical analysis^  36RESULTS AND DISCUSSION^  37Cultivar Comparisons of Quality Tests^  37Analysis of Variance of Rheological Tests  48Deformability modulus^  48A value^  52B value  54Analysis of Variance of Chemical and Functional Tests .... 55Solubility^  55Sulfhydryl content^  58Disulfide content  59Correlations Among Quality Parameters^  59Simple Linear Regressions^  68Multiple Regressions  72Stepwise Multiple Regressions^  86Principal Component Analysis  89CONCLUSIONS^  99REFERENCES  104viLIST OF TABLES TABLE1^Codes for parameters used in this study^  382^Baking tests of Canada Western Red Spring (CWRS), CanadaPrairie Spring (CPS), Canada Utility (CU), CanadaWestern Amber Durum (CWAD), and Soft Wheat (SW) wheatcultivars^  393 Farinograph measurements of nineteen wheat cultivars ^ 404 Mixograph measurements of nineteen wheat cultivars^ 415 Extensigraph measurements of nineteen wheat cultivars ^ 426 Wheat/flour strength of nineteen wheat cultivars^ 437^Milling quality of nineteen wheat cultivars^ 448^a-amylase tests of nineteen wheat cultivars  459 Deformability modulus of unleavened wheat flour doughs ^ 4910 Rheological properties of five groups of wheatcultivars^  5011 Stress relaxation parameters of unleavened wheat flourdoughs  5112 Solubility, sulfhydryl, and disulfide content of glutenof nineteen wheat cultivars^  5613 Chemical and functional properties of five groups ofwheat cultivars^  5714 Correlation coefficients of parameters of nineteen wheatcultivars^  6015 Significant (p<0.05) simple linear regressions of thebaking tests on the parameters of the nineteen wheatcultivars^  6916 Nonsignificant (p>0.05) simple linear regressions of thebaking tests on the parameters of the nineteen wheatcultivars^  7117 Method one prediction equations for baking tests^ 7518 Method two prediction equations for baking tests^ 77vii19 Method three prediction equations for baking tests^ 8020 Method four prediction equations for baking tests^ 8321 Method five prediction equations for baking tests^ 8522 Prediction equations for baking tests from stepwiseregression analysis of method one prediction equations. 8723 Prediction equations for baking tests from stepwiseregression analysis of method two prediction equations. 9024 Rotated principal component loadings for the sevencomponents identified^ 9225 Prediction equations for baking tests with principalcomponents^ 9526 Prediction equations for baking tests with parametersselected by principal components^ 9827 Summary of prediction equations for baking tests^ 100viiiLIST OF FIGURES FIGURE1 Modified Osborne fractionation of wheat proteins^ 62 Brabender farinograph^ 103 Representative farinogram of a wheat flour dough^ 114 Brabender extensigraph^ 145 Representative extensigram of a wheat flour dough^ 156 Brabender amylograph^ 207 Representative amylogram of a wheat flour^ 218 Model of dough structure ^ 249 Stress relaxation curve of an unleavened wheat dough^ 30ixACKNOWLEDGEMENT I would like to thank all of the members of the Departmentof Food Science for their friendship during my years as agraduate student. In particular, my supervisor, Dr. T.D.Durance, for all his help, encouragement, patience, and guidance.As well, my research committee members, Dr. S. Nakai, Dr. W.D.Powrie, and Dr. J. Vanderstoep for their input and time. Also,thanks to Dr. B.J. Skura for his administrative help during thecompletion of this thesis.Throughout my years in the Department of Food Science I amgrateful to have been able to meet such a large number of peoplefrom different backgrounds. I am thankful that they have beenable to share their experiences with me and enrich my life. Insome way they have all contributed to this thesis. Inparticular, I must express my gratitude to all of my friends whohave never given up on me. Thanks for your support and patience.A special thanks also goes to Dr. O.M. Lukow, of AgricultureCanada in Winnipeg, for the flour samples and testing. Also,Agriculture Canada for the funding of this project.Lastly, I must thank my parents Mr. and Mrs. T.F. Masuharaand my family, Frank, Doug, and Joy, for everything they haveever done for me.11.0 INTRODUCTIONBreadmaking appears to be a straightforward process. Flour,water, and the other ingredients are mixed and a dough is formed.The dough is fermented and then baked. However, upon furtherinvestigation, this process is shown to be rather complex withmany unanswered questions. It has generally been accepted thatthe component that determines the breadmaking ability of a wheatcultivar is gluten (MacRitchie, 1989), but how it does this isnot completely understood.One of the problems in understanding breadmaking ability isthe definition of breadmaking quality. Besides actually bakingthe bread, by means of a baking test, a single test has not beendeveloped that defines breadmaking quality. There are manyinstruments and tests for assessing the quality of wheat, flour,and dough, but it is not always understood what makes thesemeasurements important to breadmaking. The baking test is thestandard by which quality tests are evaluated for their abilityto determine breadmaking quality. The relationships between thequality tests and the baking test have not always beenconsistent.Two areas that have not been extensively examined for theirability to predict breadmaking quality have been the fundamentalproperties of dough rheology and chemical/structure/function2relationships. Many of the rheological tests used to predictbaking ability are empirical. In particular, instruments suchas the farinograph, mixograph, and extensigraph measure empiricalvalues. Fundamental rheological principles have not been usedvery extensively in this area.Also, the rheological propeties of dough can describe doughstructure, and functionality is determined by dough structure(Bloksma, 1990). The forces that are responsible for doughstructure include covalent bonding (disulfide bonding),hydrophobic bonding, electrostatic bonding, hydrogen bonding, andVan der Waals forces (Kaufman et al., 1986). Unfortunately, thelow solubility of gluten makes it difficult to measure theseforces. Most of the work done on these forces has been ondisulfide interchange reactions (Kaufman et al., 1986).Principal component analysis is a method of reducing a largeamount of data into a few components that will retain almost allof the variation of the original data. Each of the componentsis a linear combination of the original variables and isuncorrelated to the other components.The objectives of this study were to collect a data set ofcommon wheat, flour, and dough quality parameters from a groupof wheat cultivars with a wide range of breadmaking ability andto use this data set to predict baking quality; to measure some3fundamental rheological (deformability modulus and stressrelaxation), chemical (sulfhydryl and disulfide content), andfunctional (solubility) properties, for determining if theirinclusion in prediction equations would improve the relationshipbetween the common quality parameters and baking quality; and touse principal component analysis to reduce the above data set toa small number of components that would explain most of thevariation of the entire data set and aid in the interpretationof the data set.42.0^LITERATURE REVIEW2.1^DEFINITION OF OUALITY Quality in terms of wheat, flour, dough, and baking isdifficult to define. What is considered good quality for someuses and by some people is considered poor quality for other usesand by other people. What exactly is it about the wheat, flour,dough, or bread that makes it good or poor quality? How isquality assessed? Is there a single property or characteristicthat defines quality? Finney et al. (1987) gave a simpledefinition of wheat quality; "Wheat that is desired has goodquality and wheat that is not desired has poor quality". Perhapsthe ultimate definition of quality is that of consumer acceptance(Wrigley, 1993). Whatever the consumer accepts must be of goodquality.In reference to the comment and the three questions above,Pomeranz (1988) defined quality as "The suitability of a wheatflour for producing specific end products, such as bread, pastry,cakes, macaroni, or crackers". Pomeranz (1988) also stated thata single property cannot define quality. Quality depends on acombination of several milling, rheological, and processingcharacteristics of the wheat, flour, and dough. In this thesis,the quality of wheat, flour, and dough was related to the bakingof pan bread only.52.2^DETERMINANT OF BREADMAKING QUALITY All of the components of wheat flour are necessary for theproduction of acceptable bread. These components includeproteins, starches, lipids, and minerals. However, MacRitchie(1978) showed that it is the gluten proteins that are responsiblefor the differences in the quality of bread. It is generallyaccepted now that gluten is the determinant of breadmakingquality (MacRitchie, 1989).Flour proteins have traditionally been classified by theOsborne fractionation (Figure 1). This procedure uses varioussolvents to extract albumin (10% of total protein content),globulin (10%), gliadin (40%), and glutenin (40%) (Schofield andBooth, 1983). Gluten consists of gliadin and glutenin. Gliadincan be subdivided into about fifty monomeric protein fractions(Schofield and Booth, 1983). Glutenin subunits are linked bydisulfide bridges and form large protein aggregates. Gluteninfractions are based on molecular weight. About fifteen lowmolecular weight (31-48 kilodaltons) subfractions and three tofive high molecular weight (97-136 kilodaltons) subfractions arepresent in wheat (Schofield and Booth, 1983; Law et al., 1984).Of the wheat flour proteins, albumin and globulin areconsidered to have minor (Schofield and Booth, 1983) or no(MacRitchie, 1984) importance in determining the difference inFLOURextract with 0.5N NaC1centrifugeSUPERNATANT^ PRECIPITATE6SUPERNATANT SUPERNATANT^PRECIPITATEdialyze againstwatercentrifugeextract with 70%ethanolcentrifugePRECIPITATE removealcoholfreeze dryextractwith 0.05NaceticacidcentrifugeALCOHOL SOLUBLE(GLIADINS)freezedryWATERSOLUBLE(ALBUMINS)freezedrySALTSOLUBLE(GLOBULINS)SUPERNATANT^PRECIPITATEfreezedry freezedryACETIC ACIDSOLUBLE(GLUTENINS)RESIDUE(GLUTENINS)FIGURE 1. Modified Osborne fractionation of wheat proteins(Bushuk, 1985).7breadmaking quality of wheat cultivars. Of great importance arethe gluten proteins which are mainly responsible for theviscoelastic behaviour of a dough. During the development of adough, gluten from a high quality wheat will form a matrix thatallows the expansion of gas cells which results in breads withhigh loaf volume and good texture (Kolster and Vereijken, 1993).The gliadins are associated with extensibility of the dough,while the glutenins are involved in strength and elasticity(MacRitchie, 1984).2.3 METHODS OF DETERMINING BREADMAKING QUALITY As mentioned previously, quality can be defined by amultitude of wheat, flour, and dough characteristics (Pomeranz,1988). In general, these characteristics have referred to thestrength of the wheat, flour, or dough. A strong wheat isconsidered to have good breadmaking quality while a weak wheatis considered to have poor breadmaking quality. A primeindicator of strength is protein content. Other strengthparameters are measured by the farinograph and extensigraph(Williams et al., 1988). The baking test is the best method ofevaluating breadmaking quality and is the basis for determininga test's ability to assess breadmaking quality. The tests ofwheat, flour, dough, and baking quality will now be described.82.3.1^Baking TestsSince bread is actually made in a baking test, as previouslymentioned, baking tests are the best tests for determiningbreadmaking ability. There are many variations of the formulaand procedure for a baking test, but the system is set up so thatan optimum loaf of bread can be made with regard to doughcomposition and baking procedures (Pomeranz, 1987). Optimummixing time and water absorption are used and the ingredients arenot limited (Pomeranz, 1987). Loaf volumes (cm 3 ) are the measureof the baking tests. Some common baking tests are the AmericanAssociation of Cereal Chemists (AACC) method with no bromateadded and with 10 ppm bromate added. The Grain Research Lab(GRL) remix loaf volume method is also used. This methodinvolves an extra mixing stage after fermentation to optimizeoxidation response (Kilborn and Tipples, 1981).2.3.2^Protein ContentFlour protein content is an important indicator ofbreadmaking quality. There is a direct proportional relationshipbetween flour protein content and breadmaking ability as measuredby loaf volume (Finney and Barmore, 1948). The protein qualitycan also be determined from the regression of loaf volume onflour protein content. The higher the slope of the regressionline, the higher the quality of the wheat (Finney and Barmore,9(1948). Many flour properties such as water absorption, doughdevelopment time, and mixing tolerance index are highlycorrelated with protein content (Pomeranz, 1987).Flour protein content is measured by several methodsincluding Dickey-John near-infrared reflectance. This instrumentmeasures the difference in reflected energy of six narrow wavebands in the near-infrared spectrum (Pomeranz, 1987).2.3.3 FarinographThe farinograph is an instrument (Figure 2) that recordsempirical rheological measurements of dough during the mixingstages of dough development. As the dough is mixed, theresistance of the dough to the mixing blades is recorded. Atypical farinogram is shown in Figure 3. As mixing proceeds, theresistance rises to a maximum. At this point the dough isdeveloped or is optimal. During this mixing period the flourparticles hydrate and the maximum resistance occurs when all ofthe flour particles are hydrated. Doughs that are mixed past themaximum are considered overmixed.Common measurements that are recorded by the farinographinclude dough development time which is the time it takes to mixthe dough to its maximum resistance. Absorption is the amountof water that is added to the flour until an optimum arbitraryFIGURE 2. Brabender farinograph. 1=back wall of mixer withmixing blades; 2=remainder of mixer; 3=housing of motor andgears; 4=ball race bearings; 5=levers; 6=counterweight; 7=scalehead and scale; 8=pointer; 9=pen arm; 10=recorder; 11=dashpotdamper (Bloksma and Bushuk, 1988).101 1205^10time , minutesFIGURE 3.^Representative farinogram of wheat flour dough.(Bloksma and Bushuk, 1988).150mixingtoleranceindex(5 min after peak)doughdevelopmenttimet12consistency is reached. This consistency is usually 500Brabender Units. Brabender Units are the units of resistance forthe farinograph. Mixing tolerance index is the decrease in thenumber of Brabender Units five minutes after the maximumresistance has been reached.2.3.4 MixographThe mixograph, simliarly to the farinograph, follows doughformation and records empirical rheological measurements. Themixograph contains a rotating bowl with three upright stationarypins attached to it. Four vertical pins revolve around the threestationary pins that are attached to the bowl and mix the dough.The force that rotates the bowl while the dough is being mixedis recorded with time. As the dough is developed, the forcerequired to rotate the bowl increases to a peak and thendecreases as the dough is overmixed.A number of measurements are recorded by the mixograph.Mixing development time is similar to the farinograph doughdevelopment time. This is the time required for the dough to befully developed. The peak height is the force at the peak ofdough development. Energy to peak and total energy are alsomeasured. Peak band width and band width energy are related tothe cohesiveness and elasticity of the dough. Finally, firstminute slope is the rate at which the flour particles hydrate.132.3.5^ExtensigraphThe extensigraph (Figure 4) is often used in combinationwith the farinograph. Whereas the farinograph records mixing anddough development, the extensigraph records the resistance toextension and extensibilty. These properties are importantduring fermentation when the gas cells expand and the dough mustbe strong enough to retain the gas. This is the property ofwheat gluten that enables leavened bread to be produced fromwheat flour.The dough sample for the extensigraph is mixed by thefarinograph and is shaped into cylinders. After resting, thedough cylinders are stretched until ruptured and the stretchingcurve (force vs. extension) is recorded. Figure 5 shows atypical extensigraph curve of a hard wheat flour. The length ofthe stretching curve is the extensibility. Resistance toextension is the height of the extensigraph curve at t minutes.The ratio of extensibilty to resistance is also recorded.2.3.6^Wheat/Flour Strength TestsHardness has generally reflected the strength of a wheat(Williams et al., 1988). Harder wheats are considered strongerthan softer wheats. The starch-protein bonds are weak in softwheats, the starch-protein bonds are broken before the starch14FIGURE 4.^Brabender extensigraph.^1=test piece; 2=cradle;3=clamp; 4=motor; 5=stretching hook; 6=levers; 7=scale head;8=recorder; 9=dashpot damper (Bloksma and Bushuk, 1988).maximumresistanceR altime and extensionFIGURE 5. Representative extensigram of a wheat flour dough.(Bloksma and Bushuk, 1988).1516granules are broken. With the milling of hard wheats, the strongstarch-protein bonds are not broken as easily, leaving the starchgranules to be readily broken (Mok and Dick, 1991). This resultsin higher starch damage for the harder wheats which is animportant factor in breadmaking. Damaged starch absorbs a largeamount of water and enzymes can readily break down the damagedstarch to fermentable sugars (Evers and Stevens, 1985).Hardness is measured by grinding time (Kosmolak, 1978). Ashort grinding time indicates a hard wheat, while a long grindingtime indicates a soft wheat. Starch damage is also measured.As mentioned above, high starch damage is associated with hardwheats and high water absorption, while low starch damage isassociated with soft wheats and low water absorption.The Zeleny sedimentation test (Zeleny, 1947) has been usedin breeding programs as an indicator of wheat strength. Thistest is based on the amount of acid insoluble protein in a wheatsample. Acid insoluble protein has been related to thebreadmaking ability of wheat cultivars (Orth and Bushuk, 1972).Crude flour is suspended in water in a graduated cylinder. Acidis added and the volume of the sediment after standing for fiveminutes is recorded as the sedimentation value. A value of fiveis very weak, while a value of seventy is very strong.172.3.7^Milling Quality TestsSeveral tests are involved with the milling quality ofwheat. Flour yield is a measure that is most commonly used inNorth America (Bass, 1988). It indicates milling efficiency. Thenumber of bushels of wheat required to yield one hundred poundsof flour is the usual method of defining flour yield in NorthAmerica (Bass, 1988). The amounts of wheat and flour arearbitrary. A low flour yield is associated with a high millingefficiency.Test weight is a measure of kernel soundness. Fully mature,plump, disease-free kernels will give a high test weight. Testweight is expressed as the grain weight per unit of volume. Thisis usually given as Kg/hL (Pomeranz, 1987). The average testweight in the United States is 77-83 Kg/hL (Pomeranz, 1987). Atest weight of less than 58 Kg/hL implies that the kernels arebadly shriveled (Pomeranz, 1987).Thousand kernel weight is a measure of kernel size. Itsupplements the test weight measure. Thousand kernel weight isexpressed as the weight per one thousand kernels.Ash content is an important indicator of the millingoperation. The bran contains approximately twenty times themineral content of the endosperm (Pomeranz, 1987). Thus, the ash18content of the flour indicates the effectiveness of theseparation of the bran from the endosperm.2.3.8 a-amylase Testsa-amylase is required for breadmaking since it hydrolyzesstarch and provides sugars for fermentation. In sound wheats,a-amylase is present in trace amounts (Kruger and Tkachuk, 1969).Preharvest sprouting can result in excess a-amylase which can bedetrimental to breadmaking (Buchanan and Nicholas, 1980). Damagedstarch will be broken down extensively by the excess a-amylaseand the water that is normally bound to the starch will bereleased, resulting in a slack dough (Kulp et al., 1983). Twocommon tests to measure a-amylase activity are the falling numbertest and the amylograph peak viscosity.The falling number test involves a viscometer-stirrer anda hot aqueous flour suspension. a-amylase activity decreases theviscosity of the suspension and the stirrer falls a fixeddistance. The time (seconds) that it takes the stirrer to fallthis distance is recorded as the falling number (Pomeranz, 1987).A low falling number indicates high a-amylase activity sincemore starch was degraded and thus reduced the viscosity of thesuspension.The amylograph (Figure 6) is a viscometer that measures the19viscosity of a heated flour suspension in a rotating bowl. Acurve (Figure 7) is obtained by recording the viscosity over atemperature range (usually from 30-95 °C) (Pomeranz, 1987). Thepeak viscosity of the amylograph curve will indicate a-amylaseactivity. A high peak viscosity is associated with low a-amylaseactivity.2.4 PREDICTION OF BAKING TESTS BY QUALITY PARAMETERS The relationships between the many quality parameters andbreadmaking ability are complex. As mentioned earlier, thedefinition of quality is not very clear. Each of the qualitytests intends to measure an important breadmaking property buta single test has been unable to define quality. Two studiesthat have evaluated the relationships between quality parametersand breadmaking quality were by Orth et al. (1972), and Fowlerand DeLaRoche (1975).Orth et al. (1972) measured sixteen quality parameters,including Zeleny sedimentation value, farinograph absorption,farinograph dough development time, farinograph mixing toleranceindex, flour protein content, and protein fractions of albumins,globulins, gliadins, glutenins, and residue. Zeleny sediment-ation value, farinograph dough development time, and residueprotein content were found to give the most useful informationfor predicting remix loaf volume with squared multipleFIGURE 6. Brabender amylograph. 1=helical heating coils;2=thermoregulator; 3=sample bowl; 4=pin style sensor/stirrer;5=spring cartridge; 6=strip chart recorder, 7=cooling probe;8=water-cooled cover (Shuey and Schmitz, 1982).20FIGURE 7. Representative amylogram of a wheat flour (Tipples,1982).2122correlation coefficients as high as 0.930 were obtained. Remixloaf volume was predicted by the various protein fractions,Zeleny sedimentation value, flour protein content, andfarinograph dough development time in these equations.Fowler and DeLaRoche (1975) performed twenty-nine qualitytests such as sedimentation test, seed hardness, flour proteincontent, thousand kernel weight, and mixograph and farinographmeasurements. Their findings were that bread and/or pastryquality potential could be predicted with kernel hardness,protein quantity, and mixograph peak time. These were the onlythree variables required since all of the other qualitymeasurements were related to these three variables and thus gaveredundant information. A regression of loaf volume on proteinquantity and mixograph peak time had a squared multiplecorrelation coefficient of 0.677.Other studies were by Baker and Campbell (1971) who reportedthat breadmaking quality could be determined by sedimentationvalues, nitrogen content, and centrifuge absorption. The bestsingle predictor of breadmaking potential and strength of hardwheats was found to be the sedimentation test by Greenaway et al.(1966).More recently, the quality of Chinese steamed bread (Lukowet al., 1990) and pasta cooking quality (D'Egidio et al., 1990;23Matsuo et al., 1982) have been predicted by quality tests.2.5^DOUGH STRUCTURE/DOUGH RHEOLOGY RELATIONSHIP The structure of a dough is related to the spatialarrangement of its constituents and the forces acting between theconstituents (Bloksma, 1990). The rheological properties ofdough are dependent on dough structure (Bloksma and Bushuk,1988). Thus, functionality of dough can be described by doughstructure through its rheological properties (Bloksma, 1990).Forces that act between the dough constituents includehydrogen bonding, hydrophobic bonding, electrostatic bonding, Vander Waals forces, and disulfide linkages (Kaufman et al., 1986).The majority of the research attention has been on the disulfidelinkages and the interchange of disulfide and sulfhydryl.A well accepted model of dough structure at the molecularlevel involves aligned glutenin molecules which are unbranched(Ewart, 1968, 1977, 1979; MacRitchie, 1973, 1975). This model(Figure 8) proposes that disulfide cross links are within thechains of the glutenin molecules. Noncovalent forces such ashydrogen bonding, hydrophobic bonding, and chain entanglements,are present between the chains.^Thiol-disulfide interchangereactions are involved in changing the molecular massdistribution.disulfide cross-links--•noncovalent cross-linksFIGURE 8. Model of dough structure. Cross links within gluteninmolecules are disulfide linkages. Cross links between gluteninmolecules are noncovalent (Bloksma, 1990).2425This model can explain dough development and viscousdeformation. During dough development the glutenin molecules arealigned and the noncovalent cross links are formed. In viscousdeformation, the noncovalent cross links are broken and theglutenin chains are free to slide one along another. Reformationof noncovalent cross links at new positions after the slidingwill restore the strength of the dough. However, elasticity isnot explained by this model.Tests that measure the forces acting in dough may provideuseful information in predicting breadmaking quality. However,the insolubility of gluten makes it difficult to measure theseforces. But rheological properties of dough could provideinformation on these forces. The common rheological tests ofdough quality (farinograph, extensigraph, mixograph) areempirical. Perhaps some fundamental rheological properties ofdoughs would shed new light on breadmaking ability.263.0 MATERIALS AND METHODS 3.1 WHEAT SAMPLES Nineteen wheat cultivars were used in this study and weregrown in 1988 at Glenlea and Brandon, MB. The cultivars weremilled at the Agriculture Canada Grain Research Laboratory (GRL)in Winnipeg, MB. Approximately 1.5 kg of flour was obtained foreach sample. The flour samples were packaged in plastic bags(double-bagged) and stored until used at -35°C. The cultivarswere classified into five groups representing a wide range ofbreadmaking quality.The cultivars Neepawa, Benito, Columbus, Katepwa, Roblin,Laura, and Kenyon were representative of Canada Western RedSpring (CWRS) wheats. The cultivars HY355, HY320, and Oslo wererepresentative of Canada Prairie Spring (CPS) wheats. Thecultivars Glenlea, Bluesky, and Wildcat were representative ofCanada Utility (CU) wheats. The cultivars Coulter, Medora,Arcola, and Sceptre were representative of Canada Western AmberDurum (CWAD). Finally, the cultivars Fielder and Pitic 62 wererepresentative of soft wheats (SW). Fielder is a soft whitespring wheat while Pitic 62 is a soft red spring wheat. Sincethere was only one cultivar of each class they were combined toform one group.273.2^RHEOLOGICAL STUDIES 3.2.1^Dough PreparationDoughs were made with flour and water only.^Constantmoisture content (44%) and mixing time (15 min) were used. Alaboratory scale automatic bread making machine (WelbiltAppliance Co., New Hyde Park, NY) was used to mix the doughs.3.2.2^Force Deformation/Stress Relaxation TestForce deformation and stress relaxation curves were obtainedusing lubricated uniaxial compression (Cullen-Refai et al.,1988). Dough samples (10 g) were shaped into cylinders (26.5 mmdiam) in a cut-off syringe (60 cc, Becton Dickinson, Rutherford,NJ). The samples were coated with heavy mineral oil (StanleyDrug Products, North Vancouver, BC), to prevent dehydration, andleft to rest in a dessicator for 1 hour at room temperaturebefore testing. An Instron Universal Testing Machine (Model1122, Instron Corp., Canton, MA) was used to perform the uniaxialcompression tests. Teflon sheeting covered the metal platensurfaces and heavy mineral oil was applied to the Teflon surfacesto produce a lubricated sample-platen interface. The initialsample height was measured with electronic calipers (Model 500-115, Mitutoyo, Japan) and each sample was compressed to 50% ofthis initial height at a cross-head speed of 20 mm/min. At this28point, the cross-head was stopped and the doughs were allowed torelax. The force deformation and stress relaxation curves wererecorded in digital form using the JCL 6000 Chromatography DataSystem (Jones Chromatography Ltd., Littleton, CO) with an IBM-ATcompatible personal computer interfaced with the Instron.3.2.3^AnalysisDeformability modulus (DM) is related to Young's modulusand represents a material's overall resistance to deformation(Johnson et al., 1980). It is defined asa,DM= —=E Twhere true stressQ - F-  F[Ho-AH]m- A^AoHoand true strainHo iET=in r' Ho-AH'Here F is the force, A is the actual cross-sectional area, Ao andHo are the original cross-sectional area and the original heightof the undeformed sample, respectively, and AH is the absolute29deformation. The true stress and true strain were computed fromthe force deformation curves of the dough samples (Figure 9).The slope of the initial linear portion of the plot of truestress against true strain was recorded as the deformabilitymodulus.The stress relaxation curves were analyzed with amathematical technique developed by Peleg (1979). Y, the decayparameter, was calculated from the curves (Figure 9).y_  [ Fo F ( t) ]F0where F o = initial force, F (t) = residual force after t min. ofrelaxation. The equationt 1 t= ^+ Y ab awas used to obtain the parameters "a", which describes the levelto which stresses decayed during relaxation, and "b", whichrepresents the rate at which stresses relaxed. A plot of t/Yagainst t gave a linear relationship with slope equal to 1/a andintercept equal to 1/ab.FIGURE 9. Stress relaxation curve of an unleavened wheat dough.The force (F) of the initial portion of this curve up to the peakwas used to calculate the deformability modulus. The part of thecurve after the peak was used to compute the decay parameter, Y.Here Fo=initial force and F () =residual force after t min ofrelaxation.313.3 FUNCTIONAL AND CHEMICAL STUDIES 3.3.1 Gluten IsolationGluten isolation was based on a modification of a commercialgluten isolation process (Martin process) and that of Dreese etal. (1988), MacRitchie (1987), and AACC Method 38-10 (1962).Flour (30 g) and distilled deionized water (15 mL) were mixedwith a spatula until a ball was formed. The ball was kneaded byhand until it was free of lumps and stood at room temperature for30 min. To wash the non-gluten dough constituents out of thedough ball, the dough ball was flooded with 800-1000 mL tap water(21-24°C) and gently kneaded until the water became white(milky). The starch milk was decanted off and flooding wasrepeated 6 times. The final 2 washings concentrated on gettingas much of the non-gluten dough constituents out of the glutenball as possible. This was done by squeezing any liquid thatremained in the ball so that the starch separated from thegluten. After the final wash, very little starch was evident inthe wash water. The ball, now predominantly gluten, stood for 30min in distilled deionized water. After the washings, the glutenball was air dried for 5-10 min, wrapped in Saran wrap, andstored at 4°C.323.3.2^Solubility DeterminationThe solubility of gluten was determined by the method ofPonte et al. (1967). An 8 speed osterizer blender (SunbeamCorp., Toronto, ON) was used at low speed to disperse 5 g of wetgluten in 100 mL of 0.05M acetic acid (BDH Inc., Toronto, ON).The dispersion was centrifuged in a Sorvall centrifuge (ModelRC2-B, Sorvall Inc., Newtown, CO) at 4600 rpm for 15 min. Thesupernatant was then filtered through glass wool. The filtrateand the initial gluten sample were analyzed for nitrogen contentwith a Technicon Autoanalyzer II system (Technicon Corp.,Tarrytown, NY) following digestion by the micro-Kjeldahl methodof Concon and Soltess (1973). The percent protein was calculatedwith a conversion factor of 5.70. The percent solubility ofgluten was calculated by the equation% Protein Solubility =  protein content of supernatant ^x 100protein content of initial gluten3.3.3^Determination of Sulfhydryl ContentThe determination of the SH and SS content was by amodified method of Beveridge et al. (1974). Approximately 300 mgof wet gluten, in 1 mL Tris-Gly-EDTA buffer (pH 8.0; 10.4 g Tris,6.9 g Glycine, and 1.2 g Ethylenediaminetetraacetic acid(EDTA)/L; all from Sigma Chemical Co., St. Louis, MO) and 8M Urea- 5M GuHC1 (pH 8.0; Urea, BDH Inc., Toronto, ON; Guanidine33Hydrochloride, Sigma Chemical Co., St. Louis, MO) to a finalvolume of 15 mL, was cut into small pieces for approximately 15min with a spatula (solubilization was not complete). Afterstanding at room temperature for 30 min, the solution wascentrifuged at 6500 rpm for 10 min. The pH of the decantedsupernatant was adjusted to 8.00+0.01 (Fisher Accumet pHElectrode Meter Model 620, Fisher Scientific Co., Pittsburgh,PN). Three 3 mL aliquots of this solution were made and theabsorbance of these aliquots was measured at 412 nm (proteinblank) and at 280 nm (protein content) with a ShimadzuSpectrophotometer Model UV-160 (Shimadzu Corp., Kyoto, Japan).To each aliquot, 0.05 mL Ellman's reagent (5,5'-dithiobis-2-nitrobenzoic acid; Sigma Chemical Co., St. Louis, MO) in Tris-Gly-EDTA buffer (4 mg/mL, made fresh daily) was added. Thesolutions were mixed well and stood for 30 min at roomtemperature. The absorbance of a reagent blank that containedonly the 8M Urea - 5M GuHC1 in Tris-Gly-EDTA buffer (pH 8.0) andEllman's reagent was also measured at 412 nm. The SH content wascalculated using the following equation:73 . 53A412corrgmole/g-where A412corr = A412 (nhron's) — Reagent Blank - Protein BlankC = sample concentration (mg/mL)73.53 is 10 6 / ( 1.36 x10 4 ); 1.36 x 10 4 is the molarabsorptivity of Ellman's reagent and 10 6 is theconversion from mole/mg to gmole/gC343.3.4^Determination of Disulfide ContentThe SS content was determined by subtracting the SHcontent from the total (SS + SH) reduced content. To determinethe total reduced content, approximately 75 mg of wet gluten wassolubilized completely in 1 mL of Tris-Gly-EDTA buffer (pH 8.0),5 mL 10M Urea in Tris-Gly-EDTA buffer (pH 8.0), and 0.1 mL 2-mercaptoethanol (BDH Inc., Toronto, ON). After standing for 1hour at room temperature, 5 mL 24% Trichloroacetic acid (TCA;Sigma Chemical Co., St. Louis, MO) was added to the solution.The solution stood for another hour at room temperature and wasthen centrifuged at 6500 rpm for 10 min. The supernatant wasdecanted off and 5 mL 12% TCA was added to the pellet. Aftermixing well, the solution stood for 20 min at room temperatureand was then centrifuged at 6500 rpm for 10 min. This process ofcentrifuging, adding 5 mL 12% TCA, and standing for 20 min atroom temperature was repeated once. Distilled deionized water (5mL) was added to the pellet. The solution stood for another 20min and was then centrifiged at 7000 rpm. Following thiscentrifugation, the pellet was solubilized in 10 mL 8M Urea - 5MGuHC1 in Tris-Gly-EDTA buffer (pH 8.0). The pH of the solutionwas adjusted to pH 8.00+0.01. Three 3 mL aliquots of the samplewere made. The absorbance of the samples at 280 nm (proteincontent) and at 412 nm (protein blank) was measured. Ellman'sreagent (0.05 ml) was added to each aliquot and the solutionswere mixed well. After standing for 30 min at room temperature,35the absorbance of each aliquot at 412 nm was measured. Theabsorbance at 412 nm of a reagent blank that contained only the8M Urea - 5M GuHC1 in Tris-Gly-EDTA (pH 8.0) and Ellman's reagentwas also measured at this time. The total reduced content wascalculated from the equation above.3.4 COMMON WHEAT, FLOUR, AND DOUGH QUALITY TESTS The common wheat, flour, and dough quality tests were doneat the GRL by Dr. O.M. Lukow, who was also responsible forproviding the wheat samples. The tests include baking [GRL remixloaf volume, and American Association of Cereal Chemists (AACC)],farinograph (dough development time, mixing tolerance index, andflour absorption), extensigraph (extensibility, and resistance toextension), mixograph (peak height, mixing development time,energy to peak, first minute slope, peak band width, totalenergy, and band width energy), wheat/flour strength (proteincontent, Zeleny sedimentation test, grinding time, and starchdamage), milling quality (thousand kernel weight, test weight,flour yield, and ash content), and a-amylase (amylograph andfalling number). The procedures for these tests are given indetail in Buckley et al. (1990), Lukow et al. (1990), and Lukowet al. (1991).363.5^STATISTICAL ANALYSES 3.5.1^Analysis of VarianceAnalysis of variance (ANOVA) using an unbalanced nesteddesign was performed on the rheological, functional, and chemicaldata (SAS, 1985). Tukey's test was used to determine thedifferences between the means of the groups of cultivars for therheological, functional, and chemical tests (SAS, 1985).3.5.2^Multivariate Statistical AnalysisSystat statistical program (1991) was used to perform themultivariate statistical analyses. First, simple linearregressions of the baking tests on each of the parameters testedwere computed. Secondly, five methods, based on the significanceof the simple linear regressions above, were used to createprediction equations for the baking tests. Stepwise regressionanalysis was performed on some of these prediction equations aswell. The adjusted R 2 statistic was used for the multipleregression equations. The formula is 1-[(1-R 2 )(n-1)]/(n-p) wheren is the number of cases and p is the number of predictors,including the constant (Systat, 1991). Principal componentanalysis was used to reduce the data set to a small number ofcomponents. The baking tests were also regressed on theprincipal components. These techniques were applied to createregression equations that would use as few parameters as possibleto best predict the baking tests.374.0 RESULTS AND DISCUSSION4.1^CULTIVAR COMPARISONS OF OUALITY TESTS The Canada Western Red Spring (CWRS) cultivars (Neepawa,Benito, Columbus, Katepwa, Roblin, Laura, and Kenyon) werecharacterized by strong breadmaking ability (high remix loafvolume and high AACC baking test volume), strong rheologicalproperties (farinograph, mixograph, and extensigraph), highprotein content, hard kernels, moderate starch damage, moderatesedimentation value, high flour yields, low ash content, moderatethousand kernel weight, and moderate test weight (Tables 2-7.The codes for the parameters used here are given in Table 1.)Soundness of wheat was indicated by high falling number and highamylograph viscosity (Table 8). These characteristics areconsistent with the classification of CWRS cultivars as highquality milling and breadmaking wheats.The Canada Prairie Spring (CPS) cultivars (HY320, HY355, andOslo) were characterized by moderate breadmaking ability,moderate rheological properties, lower protein content and lowerstarch damage than the CWRS cultivars, with moderate kernelhardness (Tables 2-6). Flour yield and low ash content weresimilar to CWRS cultivars (Table 7). The cultivars were soundwheats as indicated by high falling number, and high amylographviscosity (Table 8). Thousand kernel weight, test weight, and38TABLE 1. Codes for parameters used in this study.Code^ ParameterAACCO^American Association of Cereal Chemists (AACC)straight dough baking method using no bromate.Result expressed as loaf volume (cm 3 ).AACC10^AACC straight dough baking method using 10 ppmbromate. Result expressed as loaf volume (cm 3 ).AMV^Micro amylograph viscosity (Brabender Units).ASH Ash content (%).BWE Mixograph; Band width energy (nm).DDT^Farinograph; Dough development time or the timeat which the curve peaks (min).E Extensigraph; Extensibility; total length (cm)of the extensigraph curve.E/R^Extensigraph; Extensibility divided byresistance.ETP Mixograph; Energy to peak (nm).FAB^Farinograph; Flour absorption (%); constantflour method)FLY Flour yield (%).FMS^Mixograph; First minute slope (nm/min).FN Falling number determination or the measuringof alpha-amylase enzyme (sec).GRT^Grinding time (min) or the measuring of kernelhardness.MDT Mixograph; Mixing development time (min) orthe time to maximum height.MTI^Farinograph; Mixing tolerance index or the dif-ference from the top of the curve at peak to thetop of curve measured at 5 min (Brabender Units).PBW^Mixograph; Peak band width (nm).PHT Mixograph; Peak height (nm).PRT Flour protein (%) determined by Dickey-John NIR.RLV^Grain Research Laboratory (GRL) remix loafvolume (ce).R Extensigraph; Resistance to extension; height ofthe extensigraph curve (cm).SD^Starch damage (%) or the measuring of thepercentage of starch granules in flourpreparations which are susceptible to hydrolysisby alpha-amylase.SED^Zeleny sedimentation which reflects differencesin quantity and quality of gluten in flour (cm 3 ).TEG Mixograph; Total energy (nm).TW^Test weight (kg/hL).TKW Thousand kernel weight (g).39TABLE^2.^Baking tests^ofCanada Prairie Spring (CPS),Amber Durum (CWAD), and SoftCanada Western Red Spring^(CWRS),Canada Utility (CU), Canada WesternWheat^(SW) wheat cultivars.Cultivar Group RLVa(cm3 )AACCO(cm3)AACC10(cm3)Neepawa CWRS 1080 910 885Benito CWRS 1070 935 985Columbus CWRS 1140 810 905Katepwa CWRS 1225 905 925Roblin CWRS 1215 830 1035Laura CWRS 1120 830 780Kenyon CWRS 1170 965 1065HY355 CPS 865 705 760HY320 CPS 880 695 795Oslo CPS 1020 740 935Glenlea CU 575 660 565Bluesky CU 650 665 655Wildcat CU 955 795 780Coulter CWAD 835 770 720Medora CWAD 805 710 645Arcola CWAD 790 705 650Sceptre CWAD 770 675 645Fielder SW 465 460 465Pitic 62 SW 615 595 585a See TABLE 1 for parameter legend.40TABLE 3. Farinograph measurements of nineteen wheat cultivars.Cultivar Groupa DDTb MTI FAB(min) (BU) (%)Neepawa CWRS 8.00 10 63.8Benito CWRS 8.00 15 61.6Columbus CWRS 8.50 10 64.0Katepwa CWRS 9.50 10 64.2Roblin CWRS 13.00 0 66.4Laura CWRS 11.00 5 64.6Kenyon CWRS 9.00 0 65.6HY355 CPS 7.50 25 58.8HY320 CPS 7.50 20 58.8Oslo CPS 9.50 20 59.4Glenlea CU 14.00 10 62.0Bluesky CU 21.00 0 62.4Wildcat CU 11.00 15 63.4Coulter CWAD 3.00 60 62.8Medora CWAD 3.50 50 64.6Arcola CWAD 3.00 55 63.0Sceptre CWAD 3.00 50 61.2Fielder SW 1.50 99 54.2Pitic 62 SW 2.00 99 56.0a See TABLE 2 for wheat group legend.b See TABLE 1 for the parameter legend.41TABLE 4. Mixograph measurements of nineteen wheat cultivars.Cultivar Group' PHTb(nm)MDT(min)ETP(nm)FMS^PBW(nm/min)^(nm)TEG(nm)BWE(nm)Neepawa CWRS 0.16 2.47 16.0 0.08 0.09 29.5 12.0Benito CWRS 0.17 2.33 16.7 0.09 0.11 32.4 12.5Columbus CWRS 0.16 2.73 19.1 0.09 0.10 30.8 13.7Katepwa CWRS 0.15 2.87 19.7 0.08 0.10 29.6 18.4Roblin CWRS 0.19 2.67 21.1 0.10 0.19 35.5 23.0Laura CWRS 0.17 3.53 24.1 0.07 0.10 28.8 24.5Kenyon CWRS 0.18 2.27 17.2 0.11 0.10 34.2 12.3HY355 CPS 0.16 2.87 17.3 0.06 0.10 27.7 18.4HY320 CPS 0.16 3.00 20.4 0.06 0.10 29.7 15.7Oslo CPS 0.15 3.07 18.7 0.07 0.10 26.7 14.1Glenlea CU 0.15 8.00 53.4 0.06 0.08 53.4 39.3Bluesky CU 0.15 7.47 53.4 0.05 0.10 58.1 44.3Wildcat CU 0.17 4.20 32.9 0.09 0.09 66.9 24.0Coulter CWAD 0.13 2.27 13.5 0.08 0.08 26.5 9.8Medora CWAD 0.15 2.33 14.8 0.09 0.09 29.0 11.1Arcola CWAD 0.13 2.07 10.6 0.07 0.08 24.7 7.8Sceptre CWAD 0.13 2.27 12.9 0.08 0.08 25.5 9.5Fielder SW 0.06 2.27 7.0 0.02 0.03 12.7 4.8Pitic 62 SW 0.09 1.47 6.6 0.04 0.05 18.9 4.8a See TABLE 2 for wheat group legend.b See TABLE 1 for parameter legend.42TABLE 5. Extensigraph measurements of nineteen wheat cultivars.Cultivar Group' Eb(cm)R(cm)E/ RNeepawa CWRS 190 141 1.35Benito CWRS 196 127 1.54Columbus CWRS 222 152 1.46Katepwa CWRS 217 135 1.61Roblin CWRS 295 137 2.15Laura CWRS 282 150 1.88Kenyon CWRS 182 105 1.73HY355 CPS 232 105 2.21HY320 CPS 245 105 2.33Oslo CPS 260 125 2.08Glenlea CU 225 175 1.29Bluesky CU 255 173 1.47Wildcat CU 257 169 1.52Coulter CWAD 182 69 2.63Medora CWAD 166 65 2.55Arcola CWAD 187 62 3.02Sceptre CWAD 174 74 2.35Fielder SW 170 43 3.95Pitic 62 SW 158 48 3.29a See TABLE 2 for wheat group legend.b See TABLE 1 for parameter legend.43TABLE 6.^Wheat/flour strength of nineteen wheat cultivars.Cultivar Groupa PRTb(%)GRT(min)SD(%)SED(cm3)Neepawa CWRS 16.3 0.61 20 54Benito CWRS 16.0 0.60 19 57Columbus CWRS 17.5 0.61 22 53Katepwa CWRS 15.6 0.63 22 60Roblin CWRS 16.5 0.62 16 74Laura CWRS 15.3 0.55 17 61Kenyon CWRS 16.8 0.62 21 51HY355 CPS 12.6 0.82 12 59HY320 CPS 12.2 0.95 11 66Oslo CPS 15.3 0.69 11 89Glenlea CU 16.0 0.50 27 52Bluesky CU 16.3 0.53 24 52Wildcat CU 15.9 0.50 23 55Coulter CWAD 14.3 0.53 48 34Medora CWAD 14.6 0.54 58 32Arcola CWAD 15.1 0.54 43 28Sceptre CWAD 14.2 0.54 48 29Fielder SW 8.2 0.99 10 22Pitic 62 SW 9.7 0.99 6 29a See TABLE 2 for wheat group legend.b See TABLE 1 for parameter legend.TABLE 7. Milling quality of nineteen wheat cultivars.Cultivar Groupa FLYb(%)ASH(%)TW(kg/hL)TKW(g)Neepawa CWRS 70.4 0.42 78.29 31.64Benito CWRS 70.1 0.42 76.50 29.80Columbus CWRS 72.0 0.42 78.97 31.66Katepwa CWRS 69.8 0.42 78.97 32.56Roblin CWRS 71.8 0.42 78.08 33.46Laura CWRS 71.6 0.46 78.92 31.92Kenyon CWRS 69.8 0.42 77.48 31.70HY355 CPS 66.2 0.44 76.91 37.06HY320 CPS 66.2 0.44 78.97 38.38Oslo CPS 71.4 0.40 77.87 37.18Glenlea CU 71.8 0.49 77.43 43.30Bluesky CU 70.0 0.50 77.88 42.70Wildcat CU 72.0 0.53 77.10 38.66Coulter CWAD 70.1 0.69 79.99 43.20Medora CWAD 60.7 0.74 79.53 44.46Arcola CWAD 65.4 0.70 77.37 47.34Sceptre CWAD 64.2 0.72 79.94 43.14Fielder SW 66.7 0.38 83.26 40.62Pitic 62 SW 62.0 0.37 77.77 33.34a See TABLE 2 for wheat group legend.b See TABLE 1 for parameter legend.4445TABLE 8.^a-amylase tests of nineteen wheat cultivars.Cultivar Group' FNb(sec)AMV(BU)Neepawa CWRS 442 1260Benito CWRS 491 1310Columbus CWRS 542 1220Katepwa CWRS 480 1320Roblin CWRS 438 1320Laura CWRS 495 1350Kenyon CWRS 536 1260HY355 CPS 431 1430HY320 CPS 444 1530Oslo CPS 337 1550Glenlea CU 404 920Bluesky CU 396 1210Wildcat CU 365 840Coulter CWAD 494 890Medora CWAD 466 820Arcola CWAD 525 880Sceptre CWAD 408 840Fielder SW 501 1140Pitic 62 SW 431 1800a See TABLE 2 for wheat group legend.b See TABLE 1 for parameter legend.46sedimentation value were higher for CPS cultivars than for CWRScultivars (Tables 6 and 7). These properties follow theclassification of CPS cultivars as medium hard, with mediumprotein content, and moderately strong characteristics andbreadmaking ability.The Canada Utility (CU) cultivars (Glenlea, Bluesky, andWildcat) had characteristics similar to both CWRS and CPScultivars. Breadmaking ability was moderate (Table 2), andrheological properties were strong (Tables 3-5). Farinographdough development time and mixograph mixing development time werevery high. Protein content and sedimentation value (Table 6)were similar to the CWRS cultivars, while thousand kernel weight,test weight, and starch damage were similar to the CPS cultivars(Tables 6 and 7). CU cultivars had harder kernels (Table 6) thanCWRS cultivars. Flour yield and ash content were similar for allthree groups of cultivars (Table 7). Falling number andamylograph viscosity were lower than the CWRS and CPS groups butstill indicated sound wheats (Table 8). These properties areconsistent with the classification of CU cultivars as hard wheatswith exceptionally strong gluten properties, and moderatebreadmaking ability.The breadmaking ability of the Canada Western Amber Durum(CWAD) cultivars (Coulter, Medora, Arcola, and Sceptre) was notas strong as the CWRS cultivars (Table 2). Rheological47properties were weak except for farinograph absorption (Tables 3-5). CWAD cultivars had moderate protein content, hard kernels,high starch damage, and high ash content (Tables 6 and 7).Sedimentation value was moderate like that of the CWRS and CUgroups (Table 6). Flour yield was similar to all of the otherfour groups of cultivars (Table 7). CPS and CU cultivars hadsimilar thousand kernel weight and test weight as CWAD cultivars(Table 7). High falling number and amylograph viscosity indicatesound wheats (Table 8). These properties define the CWAD groupas hard wheats with moderate breadmaking ability.The Soft Wheat (SW) group (Fielder and Pitic 62) was theweakest group of cultivars. Breadmaking ability was by far thepoorest of the five groups (Table 2). Rheological properties,with the exception of absorption, were very weak (Tables 3-5).Protein content was the lowest of the five groups of cultivars(Table 6). The SW cultivars had soft kernels, low starch damage,moderate sedimentation value, high flour yield, and low ashcontent (Tables 6 and 7). Thousand kernel weight and test weightwere similar to the CPS, CU, and CWAD groups (Table 7). Fallingnumber and amylograph viscosity were high, indicating soundwheats (Table 8). This group of cultivars can be described assoft wheats with weak characteristics and poor breadmakingability.484.2^ANALYSIS OF VARIANCE OF RHEOLOGICAL TESTS 4.2.1^Deformability ModulusDeformability modulus is similar to Young's modulus and isa measure of a material's overall resistance to deformation(Johnson et al., 1980). The deformability modulus of the doughsmade from the nineteen wheat cultivars are given in Table 9. Thedeformability modulus of the CU doughs was significantly (p<0.05)greater than the doughs from all of the other four groups ofcultivars (Table 10). The CU group of cultivars are hard, strongwheats as characterized by the results from the quality tests. Itappears that doughs made from hard, strong wheats have thegreatest resistance to deformation.However, it was unexpected that doughs from SW cultivars andpossibly CWAD cultivars were not less resistant to deformationthan CWRS cultivars. SW and CWAD doughs were characterized bylow extensigraph resistance, and weak rheological properties. SWcultivars had soft kernels and low protein content. CWRScultivars had hard kernels, high protein content, and formeddoughs with strong rheological properties. Also, from subjectiveevaluation of the doughs, it appeared that the SW doughs weremuch softer than the CWRS doughs.Doughs have been reported to appear stiffer as the deforma-49TABLE 9. Deformability modulus of nineteen unleavened wheatflour doughs.Cultivar Groupa DM(N) S.D.bNeepawa CWRS 1.30 0.10Benito CWRS 1.34 0.31Columbus CWRS 0.96 0.22Katepwa CWRS 1.14 0.40Roblin CWRS 1.10 0.20Laura CWRS 1.20 0.21Kenyon CWRS 1.32 0.26HY355 CPS 1.22 0.21HY320 CPS 1.34 0.07Oslo CPS 0.97 0.25Glenlea CU 1.84 0.32Bluesky CU 1.57 0.18Wildcat CU 1.19 0.25Coulter CWAD 1.13 0.08Medora CWAD 1.41 0.13Arcola CWAD 1.01 0.17Sceptre CWAD 1.39 0.14Fielder SW 0.96 0.21Pitic 62 SW 1.39 0.33a See TABLE 2 for wheat group legend.b Standard deviation.50TABLE 10.^Rheological properties of five groups of wheatcultivars.Groupz^ Rheological PropertyDeformability^A Value^B ValueModulus (N)CWRS^1.20b^0.996a^6.31bcCPS 1.16b 0.986b 6.09cdCU^1.53a^0.983bc^5.30dCWAD 1.22b 0.984b 6.94bSW^1.18b^0.969c^11.42aMeans sharing the same letter within a column are notsignificantly different at p<0.05.Z See TABLE 2 for wheat group legend.51TABLE 11.^Stress relaxation parameters of nineteen unleavenedwheat flour doughs.Cultivar Groupa "A" Value S.D.b "B" Value S.D.Neepawa CWRS i.00 0.01 6.20 0.39Benito CWRS 1.01 0.01 6.47 0.40Columbus CWRS 0.99 0.01 7.30 1.00Katepwa CWRS 0.99 0.01 6.55 0.70Roblin CWRS 1.00 0.01 5.53 0.59Laura CWRS 0.99 0.01 5.21 0.43Kenyon CWRS 0.99 0.01 6.94 0.89HY355 CPS 0.97 0.01 6.28 0.43HY320 CPS 0.99 0.01 6.41 0.79Oslo CPS 0.99 0.01 5.69 0.32Glenlea CU 0.98 0.01 7.32 0.33Bluesky CU 0.98 0.00 4.15 0.27Wildcat CU 0.99 0.01 4.42 0.22Coulter CWAD 0.99 0.01 7.03 0.43Medora CWAD 0.97 0.00 7.51 0.64Arcola CWAD 0.99 0.00 6.73 0.63Sceptre CWAD 0.98 0.01 6.60 0.18Fielder SW 0.96 0.01 12.19 2.35Pitic 62 SW 0.97 0.02 10.84 1.74a See TABLE 2 for wheat group legend.b Standard deviation.52tion rate decreases (Bloksma, 1990). Perhaps the deformationrate of 20 mm/min used here was too slow for a difference indeformation to be determined between the SW and CWRS cultivars.4.2.2 A ValueA value represents the level to which stresses relax duringa stress relaxation test. If A = 0, the stress does not relaxand the material behaves as an ideal elastic solid. If A = 1.0,the stress level approaches zero and the material behaves as anideal liquid (Peleg, 1979). Table 11 contains the stressrelaxation parameters of unleavened wheat flour doughs. The CWRSgroup of cultivars had a significantly (p<0.05) greater A valuethan the other four groups of cultivars (Table 10). The CPS andCWAD groups of cultivars had significantly (p<0.05) greater Avalues than the SW group. CU and SW groups were notsignificantly different. The results show that the SW groupbehaved the most solid-like, while the CWRS group behaved themost liquid-like. The soft and weak SW cultivars were expectedto act more liquid-like than the harder and stronger CWRScultivars.The A values of all five groups of cultivars were very closeto 1.0 (0.96-0.99) suggesting that all of the doughs were closeto ideal liquids. A values between 0.4-0.7 were anticipatedsince the doughs should have a combination of elastic and plastic53properties.No oxidants were added to the doughs and the doughs wereunfermented. Previous results (Cullen-Refai et al., 1988)showed that unoxidized doughs behaved more like liquids (high AValue) than oxidized doughs, and that fermentation results indoughs becoming more solid-like (low A value). Fermentation hasan oxidizing effect on dough (Hoseney et al., 1979) which wouldexplain the solid-like behaviour. Unoxidized doughs are moreliquid-like because there are less disulfide bonds in the dough.Also, the formation of carbon dioxide bubbles during fermentationcould make the dough appear stiffer. This may partly explain whythe A values in this study were high.A constant mixing time was used here and some doughs wereundoubtably undermixed or overmixed. Cullen-Refai et al. (1988)found that overmixed doughs exhibited more solid-like (low Avalue) behaviour than optimally mixed doughs, and undermixeddoughs exhibited more liquid-like (high A value) behaviour thanoptimally mixed doughs. This may have occurred here, but with along mixing time of 15 minutes it may be more likely that thedoughs were overmixed. Another factor to be consideredconcerning the mixing time was the farinograph mixing toleranceindex or stability. Strong wheat doughs can be overmixed forsome time with little breakdown of the dough occurring, whileweak doughs will break down dramatically after only a very short54period of overmixing.4.2.3^B ValueB value represents the rate at which stresses relax duringa stress relaxation test. If B = 0, the stress does not relaxand the material behaves as an ideal elastic solid. The higherthe B Value, the faster the rate of stress relaxation (Peleg,1979). B values for the unleavened wheat flour doughs are givenin Table 11. The SW group had a significantly (p<0.05) greaterB value than all of the other four groups (Table 10). The CWADgroup had a significantly (p<0.05) greater B value than the CPSand CU groups. The CWRS group had a significantly (p<0.05)greater B value than the CU group.The poor breadmaking quality SW group had the highest rateof stress relaxation. This agrees with previous reports that aslow rate of stress relaxation is associated with good bakingquality (Launay and Bure, 1974). It was expected that CWRSwheats would have the slowest rate of stress relaxation, but CUwheats were found to be the slowest. The CWRS group had verygood baking test results, while the CU group had moderate bakingtest results. However, the CU group did have a low farinographmixing tolerance index which is indicative of a dough with stablemixing properties. Mixing tolerance index was found to becorrelated to B value in this study.55The CWAD group, which also had poor baking test results, hada higher rate of stress relaxation than the CPS group asanticipated.4.3^ANALYSIS OF VARIANCE OF CHEMICAL AND FUNCTIONAL TESTS 4.3.1^SolubilityThe gluten solubility of the wheat cultivars ranged from63.18 to 97.41% (Table 12). Table 13 shows that there were nosignificant differences in gluten solubility found among the fivegroups of cultivars. Solubility of gluten in acetic acid hasbeen found to be low for good breadmaking wheats (Orth andBushuk, 1972). The CWRS group did have the lowest solubility ofthe five groups but the difference was not found to besignificant. Gluten solubility as measured in this study doesnot appear to play a role here in differentiating breadmakingability among these groups of cultivars.A possible problem with the determination of solubility herewas that the solubility was based on a wet weight of glutenrather than a dry weight. Thus, the moisture contents of thegluten samples may have differed, resulting in initial sampleweights that were in error. Although as much starch was washedout of the gluten as possible, the starch content of the glutensamples may have also been inconsistent.56TABLE 12.^Solubility, sulfhydryl, and disulfide content ofgluten of nineteen wheat cultivars.Cultivar Groupa SOL(%)S.D.b SHCONTENT(gmole/g)S.D. SSCONTENT(gmole/g)S.D.Neepawa CWRS 67.75 12.85 1.56 0.11 103.94 17.53Benito CWRS 63.18 19.62 2.02 0.19 149.63 17.39Columbus CWRS 77.58 1.66 1.98 0.32 98.13 6.95Katepwa CWRS 68.37 2.62 2.35 0.25 138.01 5.67Roblin CWRS 69.53 7.54 1.56 0.15 145.93 12.25Laura CWRS 82.04 3.80 1.54 0.21 137.22 4.56Kenyon CWRS 97.41 2.16 1.65 0.23 148.51 22.88HY355 CPS 84.36 4.14 1.57 0.11 106.33 11.59HY320 CPS 82.53 10.37 1.35 0.38 142.47 10.39Oslo CPS 78.65 13.21 1.88 0.07 112.33 6.86Glenlea CU 81.32 14.02 1.42 0.24 150.26 14.77Bluesky CU 84.01 4.80 1.37 0.15 137.08 15.57Wildcat CU 80.56 12.91 1.22 0.15 105.87 33.06Coulter CWAD 96.36 4.74 2.13 0.15 133.09 19.27Medora CWAD 67.22 14.35 2.60 0.19 117.34 24.41Arcola CWAD 87.12 5.01 1.60 0.28 132.63 8.74Sceptre CWAD 83.61 1.49 2.02 0.06 128.62 10.50Fielder SW 87.14 4.33 2.32 0.08 148.19 4.78Pitic 62 SW 89.37 2.54 1.96 0.29 108.98 1.13a See TABLE 2 for wheat group legend.b Standard deviation.57TABLE 13. Chemical and functional properties of five groups ofwheat cultivars.Groupz^ Chemical/Functional PropertySolubility^SH Content^SS Content(%)^(limole/g) (Rmole/g)CWRS^75.05a^1.81b^131.62aCPS 81.85a 1.60bc 120.38aCU^81.71a^1.34c^131.07aCWAD 81.94a 2.09a 127.92aSW^88.03a^2.14a^128.59aMeans sharing the same letter within a column are notsignificantly different at p<0.05.Z See TABLE 2 for wheat group legend.584.3.2^Sulfhydryl ContentTable 12 includes the sulfhydryl content of the gluten ofthe nineteen wheat cultivars. Values range from 1.22 to 2.60gmole/g. The gluten sulfhydryl content of the CWAD and SW groupsof cultivars was significantly (p<0.05) greater than that of theother three groups of cultivars (Table 13). The CWRS cultivarshad significantly (p<0.05) greater gluten sulfhydryl content thanthe CU cultivars.The two flours with weak rheological properties and poorbreadmaking ability, CWAD and SW, had the highest glutensulfhydryl content, while CU, which had strong rheologicalproperties, had the lowest gluten sulfhydryl content. Flourstrength has been reported to decrease with an increase insulfhydryl groups (Tsen and Bushuk, 1968). These results areconsistent with this as the SW group had a high sulfhydrylcontent. Also sulfhydryl groups are necessary for the doughs tohave flow (Pomeranz, 1987). The strong CU cultivars, which alsohad the greatest resistance to deformation, were low in glutensulfhydryl content which indicated firm doughs.CWRS and CPS groups were not significantly different fromeach other and had similar rheological properties. Itappears that poor breadmaking wheats have high gluten sulfhydrylcontent.594.3.3^Disulfide ContentThe disulfide contents of the gluten of the wheat sampleswere much greater than the sulfhydryl contents. The range wasfrom 98.13 to 150.26 Rmole/g (Table 12). There were nosignificant differences in gluten disulfide content found amongthe five groups of cultivars (Table 13). These results suggestthat gluten disulfide content is not an important factor indifferentiating the breadmaking ability of wheat cultivars.However, there has been one report that total disulfide contentdecreased slightly with decreased mixing strength (Tsen andBushuk, 1968). A disulfide/sulfhydryl ratio of 7 has beenreported for optimum breadmaking (Kuchumova and Strelnikova,1968). The ratios in this study were all well above this, evenfor the SW and CWAD groups of cultivars.4.4 CORRELATIONS AMONG QUALITY PARAMETERS A correlation matrix of the quality parameters is given inTable 14. The farinograph, extensigraph, and mixographinstruments all had significant (p<0.05) correlations between theparameters that each instrument measured. This was expectedsince these instruments assess dough strength. Farinograph doughdevelopment time was significantly positively correlated withextensigraph resistance (p<0.001), and mixograph band widthenergy (p<0.001), energy to peak (p<0.01), and mixing determina-TABLE 14.^Correlation coefficients of parameters of nineteen wheat cultivars.Parameter' Correlation Coefficients b^(r)DDT MTI FAB E R E/R MDT BWE ETP FMS PBWDDT 1.00MTI -0.80* 1.00FAB - - 1.00E - - - 1.00R 0.89*** -0.87** - - 1.00E/R - 0.90*** - - -0.90*** 1.00MDT 0.79* - - - - - 1.00BWE 0.92*** - - - 0.80* - 0.94*** 1.00ETP 0.87** - - - 0.79* - 0.98*** 0.97*** 1.00FMS - - 0.88*** - - - - - - 1.00PBW - - - - - - - - - - 1.00PHT - -0.90*** 0.80* - - -0.79* - - - 0.82** 0.83**TEG - - - - - - - 0.81** 0.86** - -PRT - -0.83** 0.88*** - - -0.84** - - - 0.81* -SED - - - 0.80* - - - - - - -GRT - - -0.81* - - - - - - - -SD - - - - - - - - - - -TW- - - - - - - - - - -TKW - - - - - - - - - - -FLY - - - - 0.79* - - - - - -ASH - - - - - - - - - - _AMV - - - - - - - - - - -FN - - - - - - - - - - -AACCO - - - - - - - - - 0.86** -AACC10 - - - - - - - - - - _RLV - - - - - - - - - 0.80* -DM- - - - - - - - - - -AVALUE - - - - - - - - - - -BVALUE - 0.82* - - - - - - - - -SOL - - - - - - - - - - _SH - - - - - - - - - - -SS - - - - - - - - - - -a See TABLE 1 for parameter legend.b ***, **, * Significant at the 0.1%, 1.0%, 5.0% levels, respectively.TABLE 14. Correlation coefficients of parameters of nineteen wheat cultivars (cont.).Parameter° Correlation Coefficients b^(r) PHT^TEG^PROT^SED^GRT^SD^TW^TKW^FLY^ASHDDTMTIFABERE/RMDTBWEETPFMSPBWPHT 1.00TEG^-^1.00PRT 0.82**^-^1.00SED^- - -^1.00GRT - -^-0.83**^-^1.00SD -^- - - -^1.00TW^- - - -^- -^1.00TKW -^-^-^- -^- -^1.00FLY^- - - -^- -^- -^1.00ASH - - - - -^0.95***^-^- -^1.00AMV^-^-^-^-^-^-0.82** - - - -0.82**FN- - - - - -^-^-^-^-AACCO^0.79*^-^-^-^-^- - - - -AACC10 - - - - - -^-^- - -RLV^- - - -^-^- - -^-^-DM-^-^-^- - -^-^- - -AVALUE^- - - -^-^- - - -^-BVALUE -0.83**^-^- - - -^-^-^- -SOLSHSSa See TABLE 1 for parameter legend.b ***, **, * Significant at the 0.1%, 1.0%, and 5.0% levels, respectively.TABLE 14. Correlation coefficients of parameters of nineteen wheat cultivars (cont.).Parameter° ^Correlation Coefficients' jAMV^FNO^AACCO^AACC10^RLV^DM^AVALUE^BVALUE^SOL^SH^SSDDTMTIFABERE/RMDTBWEETPFMSPBWPHTTEGPRTSEDGRTSDTWTKWFLYASHAMV 1.00FN^1.00AACCO 1.00AACC10 0.88***^1.00RLV 0.90***^0.93***^1.00DM^ 1.00AVALUE 1.00BVALUE 1.00SOL 1.00SH^-^1.00SS-^1.00a See TABLE 1 for parameter legend.b *** Significant at the 0.1% level.63tion time (p<0.05). Dough development time and mixingdevelopment time measure basically the same parameter but with adifferent instrument. Farinograph mixing tolerance index wassignificantly (p<0.001) positively correlated with extensigraphextensibility to resistance ratio, and significantly negativelycorrelated with mixograph peak height (p<0.001) and extensigraphresistance (p<0.01). Farinograph flour absorption wassignificantly positively correlated with mixograph first minuteslope (p<0.001) and peak height (p<0.05). Mixograph band widthenergy and energy to peak were significantly (p<0.05) positivelycorrelated with extensigraph resistance. The extensigraphextensibilty to resistance ratio and mixograph peak height weresignificantly (p<0.05) negatively correlated to each other.Also, parameters that were measured by the same instrumentwere significantly correlated. Farinograph dough developmenttime and mixing tolerance index were significantly negatively(p<0.05) correlated. However, farinograph flour absorption wasnot significantly intercorrelated with any farinograph parameter.This was unexpected since a long dough development time indicatesthat a high quantity of water is being absorbed. Extensigraphresistance and the resistance to extensibility ratio weresignificantly (p<0.001) negatively correlated, but extensibilitywas not significantly correlated with other extensigraphparameters. Resistance and extensibilty were reported previouslyas being highly positively correlated (Singh et al., 1990).64There were many significant correlations between the mixographparameters. Mixograph energy to peak was significantlypositively correlated with band width energy (p<0.001), mixingdetermination time (p<0.001), and total energy (p<0.01). Bandwidth energy was significantly positively correlated with mixingdetermination time (p<0.001) and total energy (p<0.01). Peakheight was positively significantly (p<0.01) correlated withfirst minute slope, and peak band width. Total energy and energyto peak were significantly (p<0.01) positively correlated.Protein content was significantly correlated to some of theparameters measured by the three dough strength testinginstruments. Farinograph absoption (p<0.001), and mixographfirst minute slope (p<0.05) and peak height (p<0.01) weresignificantly positively correlated with protein content, whilefarinograph mixing tolerance index and the extensigraphextensibility to resistance ratio were significantly (p<0.01)negatively correlated with protein content. Surprisingly,farinograph dough development time was not significantlycorrelated with protein content. A high protein content isgenerally necessary for a long dough development time. Singh etal. (1990) and Preston et al. (1992) both found dough developmenttime significantly correlated with protein content.The Zeleny sedimentation test has also been used to assesswheat strength. Sedimentation value was significantly (p<0.05)65positively correlated with extensigraph extensibility.Kernel texture is an important factor in categorizing wheatsfor their end product potential (Williams et al., 1988). Aharder wheat is generally considered to be stronger. Grindingtime is an indicator of wheat hardness but was not significantlycorrelated to many of the other indicators of strength in thisstudy. Grinding time was significantly negatively correlatedwith protein content (p<0.01) and farinograph flour absorption(p<0.001). When hard wheats are milled high starch damage occurs(Bass, 1988) and water absorption increases (Evers and Stevens,1985). However, starch damage was not significantly correlatedwith grinding time, farinograph water absorption or any strengthindicators.Milling characteristics of wheat generally indicate theoverall quality of the wheat and not neccessarily that of thebread or dough strength. This was in most part true here as testweight and thousand kernel weight did not have any significantcorrelations with any other parameters. However, flour yield wassignificantly (p<0.05) positively correlated with extensigraphresistance. This result was somewhat unexpected. Also, ashcontent was positively significantly (p<0.001) correlated withstarch damage.The soundness of wheats is frequently assessed by a-amylase66tests.^These tests generally do not reflect strength, butindicate wheat overall quality.^Amylograph viscosity wassignificantly (p<0.01) negatively correlated with ash content andstarch damage. Falling number was not significantly correlatedwith any other measurements.The baking tests were all highly significantly (p<0.001)positively intercorrelated, but they did not correlate with verymany dough strength parameters. AACCO was significantlypositively correlated with mixograph first minute slope (p<0.01)and peak height (p<0.05), while remix loaf volume wassignificantly (p<0.05) positively correlated with first minuteslope. There were no significant correlations between any of thebaking tests and protein content, or farinograph doughdevelopment time and mixing tolerance index, or extensigraphresistance and extensibility. The lack of correlation betweenbaking quality and protein content suggests that protein qualitywas more important to baking than to protein quantity. Previousstudies report both significant correlations between baking testand protein content (Finney and Barmore, 1948; Preston et al.,1992), and no significant correlation between baking test andprotein content (Singh et al., 1990). The sample selection inthe Preston et al. (1992) and Singh et al. (1990) studiesdiffered. Similar to this present study, Singh et al. usedwheats with a wide range of breadmaking ability, while Preston etal. selected hard breadmaking wheats only. The difference in67relationship between protein content and loaf volume in the threestudies may be due to this difference in cultivar selection.Singh et al. (1990) reported that loaf volume was highlysignificantly positively correlated with extensigraphextensibility (p<0.01) and resistance (p<0.001), as well asmixograph mixing determination time (p<0.01), while farinographdough development time was weakly significantly (p<0.05)positively correlated with loaf volume. No significantcorrelations between baking tests and farinograph doughdevelopment time and baking tests and farinograph mixingtolerance index were found by Preston et al. (1992). Thus, therelationship between strength parameters and baking quality doesnot appear to be very clear.The fundamental rheological (deformability modulus, A value,and B value), and chemical (sulfhydryl and disulfide content),and functional (solubility) properties did not appear to berelated to baking quality. There were no significantcorrelations between baking tests and any of the rheological,chemical, and functional properties. The only significantcorrelations involving any of these properties was between Bvalue and mixograph mixing tolerance index (p<0.05) and mixographpeak height (p<0.01). Thus, the rheological, chemical, andfunctional properties did not appear to strengthen or clarify therelationship between strength parameters and baking quality.684.5^SIMPLE LINEAR REGRESSIONS Simple linear regressions were computed for all of thequality parameters. The baking tests (AACCO, AACC10, and remixloaf volume) were the alternating dependent variables. They canalso be referred to as the criterion variables. The parametersthat had significant (p<0.05) regressions are shown in Table 15.Unlike the correlations, where only two dough strength parameterswere significantly (p<0.05) correlated with the baking tests,significant (p<0.05) regressions were calculated for severaldough strength parameters. However, squared correlationcoefficients (r 2) were not very high. The highest was for theregression of AACCO on mixograph first minute slope (0.743), andthe lowest for the regression of AACCO on extensigraph resistance(0.213). Most of the r 2 values were below 0.5. The regressionsof all three criterion variables on farinograph mixing toleranceindex (p<0.01), mixograph first minute slope (p<0.001), peak bandwidth (p<0.01), and peak height (p<0.001) were significant. Asreported earlier, first minute slope, peak band width, and peakheight were significantly (p<0.01) intercorrelated, and firstminute slope and peak height were significantly (p<0.05)correlated to AACCO and remix loaf volume. A significant(p<0.01) regression was also computed for AACCO regressed on theextensigraph ratio of extensibilty to resistance. Theregressions of all three baking tests on farinograph flourabsorption were significant (p<0.05), while the regression of69TABLE 15. Significant (p<0.05) simple linear regressions of thebaking tests on the parameters of the nineteen wheat cultivars.Parameter r 2AACCOa AACC10 RLVFARINOGRAPH:Flour absorption 0.569*** 0.324* 0.463**Mixing tolerance index 0.501** 0.484** 0.459**MIXOGRAPH:First minute slope 0.743*** 0.580*** 0.645***Peak height 0.626*** 0.570*** 0.577***Peak band width 0.344** 0.534*** 0.471**EXTENSIGRAPH:Extensibility / Resistance 0.484** 0.320* 0.310*Resistance 0.213* 0.173 0.175FLOUR/WHEAT STRENGTH:Protein content 0.573*** 0.400** 0.424**Sedimentation value 0.234* 0.496** 0.396**Grinding time 0.241* 0.045 0.096MILLING QUALITY:Thousand kernel weight 0.391** 0.501** 0.485**Flour yield 0.259* 0.282* 0.255*Test weight 0.244* 0.199 0.116RHEOLOGICAL:B value 0.316* 0.247* 0.275*a^***, **, * significant at the 0.1, 1, and 5% levels,respectively.70AACCO on extensigraph resistance was significant (p<0.05).Three wheat/flour strength parameters had significant(p<0.05) regressions with the criterion variables. Althoughprotein content and sedimentation value were not significantlycorrelated to any of the baking tests, the regressions of thethree baking tests on each of these parameters were significant(p<0.05). The regression of AACCO on grinding time was alsosignificant (p<0.05).Significant (p<0.05) regressions were unexpectedlycalculated for the milling quality parameters test weight,thousand kernel weight, and flour yield. In particular, theregressions of all three criterion variables on thousand kernelweight were significant at the 1% level.The only fundamental rheological, chemical, and functionalproperty to have a significant (p<0.05) regression was B value.The regressions of all three baking tests on B value weresignificant. B value was also the only fundamental rheological,chemical, and functional property to have any significant(p<0.05) correlations, although, these correlations were not withany of the baking tests.The parameters that did not have significant (p>0.05)regressions are shown in Table 16. Surprisingly, the regressions71TABLE 16. Nonsignificant (p>0.05) simple linear regressions ofthe baking tests on the parameters of the nineteen wheatcultivars.Parameter r 2AACCOa AACC10 RLVFARINOGRAPH:Dough development time 0.080 0.096 0.056MIXOGRAPH:Mixing determination time 0.023 0.037 0.071Total energy 0.039 0.007 0.001Energy to peak 0.000 0.007 0.019Band width energy 0.002 0.000 0.002EXTENSIGRAPH:Resistance 0.213* 0.173 0.175Extensibility 0.046 0.148 0.167FLOUR/WHEAT STRENGTH:Grinding time 0.241* 0.045 0.096Starch damage 0.000 0.064 0.019MILLING QUALITY:Test weight 0.244* 0.199 0.116Asha-AMYLASE:0.019 0.139 0.064Falling number 0.063 0.018 0.051Amylose viscosity 0.003 0.103 0.047RHEOLOGICAL:Deformability modulus 0.004 0.062 0.089A value 0.001 0.006 0.005CHEMICAL/FUNCTIONAL:Solubility 0.145 0.145 0.179Sulfhydryl content 0.015 0.020 0.005Disulfide content 0.000 0.000 0.009a^* significant at the 5% level.72of the baking tests on dough strength parameters such asfarinograph dough development time, extensigraph extensibility,and mixograph mixing determination time were not significant. Itis possible that these parameters had non-linear relationshipswith the baking tests. A quadratic relationship between doughdevelopment time and remix loaf volume has been reported (Orth etal., 1972). Also, the regressions of the criterion variables onmixograph band width energy, energy to peak, and total energywere not significant. None of these mixograph parameters wassignificantly correlated to mixograph first minute slope or peakheight, which were both significantly (p<0.05) correlated to thebaking tests.The regressions of the baking tests on starch damage, ashcontent, amylograph viscosity, and falling number were notsignificant. As mentioned previously, the regressions of thebaking tests on deformability modulus, A value, solubility,sulfhydryl, and disulfide content were not significant.4.6 MULTIPLE REGRESSIONS A number of multiple regression equations to predict bakingquality were obtained using five different methods. In thisstudy, twenty-nine parameters (not including the baking tests)were measured. Obviously this is too many parameters for auseful prediction equation and since there were only nineteen73cultivar samples, all of these parameters could not be used in asingle regression equation. Therefore, a reduction in parametersfor the prediction equations was required. The five methods ofcreating prediction equations were based on the significance ofthe simple linear regressions of the baking tests on the qualityparameters. The first method was to use the parameters (Table15) with significant (p<0.05) simple linear regressions. Thesecond method was to use the parameters (Table 16) withnonsignificant (p>0.05) simple linear regressions. Method threecategorized the parameters with significant (p<0.05) simplelinear regressions into three groups, consisting of doughstrength, wheat/flour strength, and milling quality parameters,and included one parameter from each group in a predictionequation. Method four categorized the parameters (Table 16)with nonsignificant (p>0.05) simple linear regressions into fourgroups, consisting of dough strength, wheat strength, millingquality, and a-amylase parameters, and included one parameterfrom each group in a prediction equation. Finally, the fifthmethod used prediction equations reported in the literature.Regression analysis of the various prediction equations was donefirst on equations that did not include any of the fundamentalrheological, chemical, and functional properties, and then onequations that did include these properties to determine if theyimproved the R 2 values of the prediction equations.The first method for creating prediction equations was to74use the parameters (Table 15) with significant (p<0.05) simplelinear regressions. The same ten parameters were used in theequations predicting AACC10 and remix loaf volume, while theseparameters and three additional parameters (test weight, grindingtime, and extensigraph resistance) were included in theprediction equation of AACCO (Table 17). The regression of allthree baking tests on the parameters that did not include therheological parameter had high adjusted squared multiplecorrelation coefficients (0.879-0.924) and low p values (<0.01).Since B value was the only fundamental rheological, chemical, orfunctional parameter to have a significant (p<0.05) simple linearregression, it was selected as the only fundamental rheological,chemical, or functional parameter to add to these predictionequations (Table 17). The addition of B value to the equationsresulted in very slight differences in the adjusted R 2 values ofthe AACCO and AACC10 equations, while the adjusted R2 value ofthe remix loaf volume equation had a small increase (0.879 to0.906). The p values of the three equations remainedapproximately the same.It is not surprising that these prediction equations wouldhave high adjusted R 2 values since all of the parameters used inthese equations had significant (p<0.05) simple linearregressions. The number of parameters used in these equations isalso high, ranging from ten to fourteen. On a practical basis,these equations would not be ideal for the prediction of bread-TABLE 17.^Method one prediction equations for baking tests.Prediction equations^ AdjustedR2pEquations did not include BVALUEAACCO=1708.64+19.47(FAB)-1.97(MTI)-104.82(E/R)-2.74(R)+1705.03(FMS)-992.70 0.886 0.007(PHT)-557.00(PBW)-8.35(TKW)+11.56(FLY)-19.46(TW)-26.99(PRT)+1.02(SED)-156.87(GRT)AACC10=2075.32-28.71(FAB)-3.71(MTI)+142.79(E/R)+7220.02(FMS)-1979.96(PHT)+ 0.924 <0.001659.66(PBW)-12.25(TKW)-0.74(FLY)+27.53(PRT)+1.85(SED)RLV=-210.72+22.54(FAB)-2.63(MTI)+145.756(E/R)+6671.29(FMS)-1781.38(PHT)- 0.879 <0.0011367.49(PBW)-18.21(TKW)-3.24(FLY)+0.45(PRT)+5.72(SED)Equations included BVALUEAACCO=1596.73+21.09(FAB)-1.66(MTI)-111.22(E/R)-2.74(R)+2025.16(FMS)-1598.16 0.887 0.016(PHT)-513.15(PBW)-8.78(TKW)+12.52(FLY)-18.71(TW)-25.44(PRT)+0.55(SED)-46.00(GRT)-14.81(BVALUE)AACC10=2393.11-29.54(FAB)-3.88(MTI)+146.14(E/R)+7705.70(FMS)-2754.01(PHT)+ 0.922 <0.001754.40(PBW)-13.52(TKW)-0.73(FLY)+24.28(PRT)+1.50(SED)-12.81(BVALUE)RLV=759.65+20.01(FAB)-3.133(MTI)+155.98(E/R)+8154.33(FMS) -4144.94(PHT) - 0.906 0.0011078.20(PBW)-22.08(TKW)-3.20(FLY)-9.48(PRT)+4.65(SED)-39.11(BVALUE)a See TABLE 1 for parameter legend.Cri76making quality since there are a high number of parameters to bemeasured. The addition of the fundamental rheological parameter,B value, to the equations did not appear to greatly improve theadjusted R 2 values.The second method of creating prediction equations was to usethe parameters (Table 16) with nonsignificant (p>0.05) simplelinear regressions. Due to their very low r 2 values, parameterswhich had r 2 values less than 0.01 for the simple linearregressions, were not included in the prediction equations here.Thus, depending on which baking test was used as the dependentvariable, mixograph energy to peak, mixograph band width energy,mixograph total energy, starch damage, and amylograph viscositywere not included in the prediction equations.The regression of remix loaf volume on the qualityparameters that did not include any of the fundamentalrheological (deformability modulus) and functional (solubility)properties had the greatest adjusted R2 value (0.875) of themethod two equations that did not include deformability modulusand solubility (Table 18). The regressions of AACC10 and remixloaf volume on the quality parameters that did not includedeformability modulus and solubility were significant (p<0.05).However, the regression of AACC10 on the quality parameters thatincluded deformability modulus and solubility was no longersignificant.TABLE 18.^Method two prediction equations for baking tests.Prediction equationa AdjustedR2Equations did not include DM, SOL, or SHAACCO=675.77+33.63(DDT)-0.69(E)-96.01(MDT)+3.55(TEG)+1.91(ASH)+0.49(FN)- 0.422 0.0570.07(AMV)AACC10=4161.37+21.77(DDT)+0.50(R)+1.34(E)-107.20(MDT)-36.73(TW)-1921.86 0.759 0.006(ASH)-0.11(FN)+15.62(SD)+75.71(GRT)-0.09(AMV)RLV=1696.71+4.38(DDT)+2.89(R)+2.17(E)-172.12(MDT)+6.53(ETP)-12.54(TW)- 0.875 0.0011572.29(ASH)+15.34(SD)-206.13(GRT)+0.38(FN)+0.12(AMV)Equations included DM, SOL, and/or SHAACCO=883.01+29.41(DDT)-0.81(E)-86.00(MDT)+3.43(TEG)+64.29(ASH)+0.63(FN) 0.358 0.140-0.03(AMV)-2.93(SOL)-43.21(SH)AACC10=4205.04+18.63(DDT)+0.94(R)+1.81(E)-117.47(MDT)-37.88(TW)-2855.27 0.715 0.054(ASH)+25.39(SD)+236.76(GRT)-0.40(FN)-0.09(AMV)+5.41(DM)+3.54(SOL)-48.76(SH)RLV=1913.13+12.39 (DDT) +3.93 (R) +2.64 (E) -131.45 (MDT) -6.01 (ETP) -21.08 (TW) - 0.866 0.0102055.42 (ASH) +21.47 (SD) +28.02 (GRT) +0.17 (FN) 0.04 (AMV) +108.14 (DM) +3.69(SOL)a See TABLE 1 for parameter legend.78Since the simple linear regressions of the parameters usedin these equations were not significant, it was a bit of asurprise that the adjusted R2 values of some of these equationswere this high. In particular, the prediction equation of remixloaf volume (adjUsted R 2=0.875) had a similar adjusted R 2 valueas the prediction equation of remix loaf volume (Table 17) frommethod one (adjusted R 2=0.879) . It appears that baking qualitycan be predicted by a group of parameters that would individuallyappear to have weak relationships with baking quality.The addition of the fundamental rheological and functionalparameters to the prediction equations resulted ina reduction inthe adjusted R 2 values and in the case of the prediction ofAACC10, the regression was nonsignificant.Categorizing the parameters used in the first method (Table17) into dough strength (farinograph, extensigraph, and mixographmeasurements), wheat/flour strength (protein content,sedimentation value, and grinding time), and milling qualityparameters (thousand kernel weight, flour yield, and test weight)was part of the third method of creating prediction equations.Equations included one parameter from each of the three groups.The parameters with the higher r 2 values from the simple linearregressions were selected for the equations. A regression wascalculated for a number of equations including either farinographflour absorption, farinograph mixing tolerance index,79extensigraph extensibility to resistance ratio, mixograph firstminute slope, mixograph peak height, or mixograph peak bandwidthas the dough strength parameter. Protein content orsedimentation value was selected as the wheat/flour strengthparameter depending on which baking test was used as thedependent variable. Thousand kernel weight was selected as themilling quality parameter for all of the prediction equations.The equations with the highest adjusted R 2 values are shown inTable 19.All of the equations created had significant regressions,with p<0.001 for all but one of the equations. The adjusted R 2values of all of the equations ranged from 0.616 to 0.885. Mostof the equations had adjusted R 2 values greater than 0.70. Theprediction equations with the greatest adjusted R 2 values (0.973-0.885) had adjusted R 2 values that were only a small amount lessthan those of the equations from the first method (0.879-0.924).However, only three parameters were used in the equations here,while up to fourteen parameters were used in the equations fromthe first method. The prediction of AACC10 by mixograph firstminute slope, sedimentation value, and thousand kernel weight didnot have an extremely high adjusted R 2 value, but consideringthat only three parameters need to be measured, it could be auseful equation for predicting baking quality.TABLE 19.^Method three prediction equations for baking tests.Prediction equationa AdjustedR2Equations did not include BVALUEAACC10=726.01+4258.83(FMS)+3.37(SED)-11.45(TKW) 0.885 <0.001RLV=879.10+6552.27(FMS)+3.13(SED)-16.22(TKW) 0.877 <0.001AACCO=685.78+3262.54(FMS)+12.28(PRT)-9.29(TKW) 0.873 <0.001Equations included BVALUEAACC10=537.87+4693.83(FMS)+4.07(SED)-9.97(TKW)+9.65(BVALUE) 0.881 <0.001RLV=996.05+6281.87(FMS)+2.70(SED)-17.14(TKW)-6.00(BVALUE) 0.869 <0.001AACC0=837.79+3333.56(FMS)+6.39(PRT)-9.60(TKW)-8.72(BVALUE) 0.872 <0.001a See TABLE 1 for parameter legend.81The equations with first minute slope as the dough strengthparameter had the greatest adjusted R 2 value. In general, theequations with mixing tolerance index, peak band width, or peakheight as the dough strength parameter had the smallest adjustedR 2 value (results not shown). Although sedimentation value wasthe wheat/flour strength parameter that was included in theequations with the two greatest adjusted R 2 values, the overalltrend was that the equations with protein content as thewheat/flour strength parameter had greater adjusted R 2 valuesthan those equations with sedimentation value as the wheat/flourstrength parameter.The addition of B value to the prediction equations resultedin slight decreases in adjusted R 2 values, and the p valuesremained the same (Table 19). Again the addition of afundamental rheological parameter did not improve the adjusted R 2values of the prediction equations.Similarly to the third method of creating predictionequations, the fourth method involved categorizing parameters,but this time the parameters involved were those used in thesecond method (Table 18). Dough strength (farinograph,extensigraph, and mixograph measurements), wheat strength(grinding time and starch damage), milling quality (test weightand ash content), and a-amylase parameters (falling number andamylograph viscosity), made up the four groups of parameters.82One parameter from each of the four groups was included in aprediction equation. Equations included either farinograph doughdevelopment time, extensigraph resistance, extensigraphextensibility, or mixograph mixing determination time as thedough strength parameter. Grinding time or starch damage was thewheat strength parameter, test weight or ash content was selectedas the milling quality parameter, and either falling number oramylograph viscosity was the a-amylase parameter depending onwhich baking test was used as the dependent variable. Theregressions with the greatest adjusted R2 values are shown inTable 20.Two-thirds of the equations analyzed had nonsignificantregressions including all of the equations predicting AACCO andremix loaf volume (results not shown). All of the adjusted R 2values were low, ranging from 0.000 to 0.653. Most of theregressions that were significant had p values above 0.01. Theseresults were not unexpected since only three or four parameterswere used in these equations and all of the parameters used hadnonsignificant simple linear regressions.The parameters selected from the fundamental rheological andchemical/functional groups were deformability modulus andsolubility, respectively. The addition of these two parametersresulted in equations with large increases and decreases inadjusted R 2 values over the equations that did not include defor-TABLE 20.^Method four prediction equations for baking tests.Prediction equationa AdjustedR2Equations did not include DM, SOL, and/or SHAACC10=1819.80-56.92 (MDT) -1082.24 (GRT) -899.91 (ASH) +0.24 (AMV) 0.653 0.001AACC10=1848.67-42.41 (MDT) -914.60 (GRT) -11.86 (TW) +0.49 (AMV) 0.483 0.009AACC10=1550.93-13.04 (DDT) -1156.39 (GRT) -797.18 (ASH) +0.40 (AMV) 0.415 0.019Equations included DM, SOL, and/or SHAACC10=1846.31-56.19 (MDT) -1063.35 (GRT) -880.71 (ASH) +0.24 (AMV) +0.07 (DM) - 0.597 0.0060.62 (SOL)AACC10=2527.53-27.91 (MDT) -797.41 (GRT) -16.57 (TW) +0.44 (AMV) -122.17 (DM) - 0.441 0.0352.67 (SOL)AACC10=1771.81-3.95 (DDT) -958.76 (GRT) -589.49 (ASH) +0.34 (AMV) -220.67 (DM) - 0.444 0.0342.31 (SOL)8 See TABLE 1 for parameter legend.84mability modulus and solubility (Table 20). However, most of thelarge increaseswere with the equations that initially had lowadjusted R 2 values. Thus, the addition of the fundamentalrheological and functional parameters increased the adjusted R2values of the equations, but the additions did not appear to beenough to make the prediction equations useful.The fifth method of producing prediction equations was touse equations from the literature. Orth et al. (1972) predictedremix loaf volume with Zeleny sedimentation test, proteincontent, and farinograph dough development time. Also, Fowlerand DeLaRoche (1975) predicted remix loaf volume with proteincontent and mixograph peak time.Table 21 contains the prediction equations with the greateradjusted R2 values from the two sets of equations created bymethod five. The adjusted R 2 values of these equations were notvery high (0.573-0.720), but were highly significant (all pvalues<0.001). The prediction of AACCO with protein content andmixograph mixing determination time (Fowler and DeLaRoche) hadthe greatest adjusted R2 value (0.720). Two equations from Orthet al. had lower adjusted R 2 values (0.679 and 0.676).The adjusted R 2 value (0.677) of the prediction of remixloaf volume with protein content and mixograph mixingdetermination time (result not shown) was the same as the R2TABLE 21.^Method five prediction equations for baking tests.Prediction equationa^ Reference^Adjusted^pR2Equations did not include BVALUEAACCO=172.85+46.88(PRT)-32.84(MDT)AACC10=-36.79+40.39(PRT)-17.12(DDT)+7.17(SED)RLV=-262.30+64.75(PRT)-27.66(DDT)+8.83(SED)Equations included BVALUEAACCO=296.76+42.41(PRT)-34.24(MDT)-7.94(BVALUE)Fowler + DeLaRoche (1975) 0.720 <0.001Orth et al. (1972)^0.679 <0.001Orth et al. (1972)^0.676 <0.001Fowler + DeLaRoche (1975) 0.709 <0.001AACC10=-324.86+48.41(PRT)-15.48(DDT)+7.67(SED)^Orth et al. (1972)^0.675 <0.001+19.41 (BVALUE)RLV=-356.28+67.37(PRT)-27.12(DDT)+8.99(SED)+^Orth et al. (1972)^0.654 <0.0016.33 (BVALUE)a See TABLE 1 for parameter legend.86value reported by Fowler and DeLaRoche of 0.677. Orth et al.reported R 2 values as high as 0.930 for the prediction of remixloaf volume. The regressions here did not have adjusted R 2values very close to this number (0.676 and 0.677), but Orth etal. included protein fractions in their equations and theirequations did not include the exact same parameters as in thisstudy.B value was again selected as the rheological parameter toadd to the prediction equations. The adjusted R 2 valuesdecreased slightly for the equations that included B value (Table21).4.7 STEPWISE MULTIPLE REGRESSIONS The method one and method two equations described earlierwere analyzed by stepwise regression in an attempt to reduce thenumber of parameters in the prediction equations. Stepwiseregression analysis was computed for both the equations thatincluded and did not include the fundamental rheological,chemical, and functional parameters. The alpha value to enterand remove variables was 0.150.The number of parameters in the method one predictionequations was reduced considerably with the stepwise regressioncomputation (Table 22). The prediction equation of AACCO hadTABLE 22. Prediction equations for baking tests from stepwise regression analysis of methodone prediction equations.Prediction equation' AdjustedR2General prediction equations did not include BVALUEAACCO=1048.83+3356.00(FMS)-174.22(GRT)-11.34(TKW) 0.874 <0.001AACC10=516.54+5062.59(FMS)+51.52(E/R)+4.30(SED)-11.60(TKW) 0.905 <0.001RLV=-463.924+20.51(FAB)+5037.38(FMS)+89.00(E/R)+4.60(SED)-18.34(TKW) 0.906 <0.001General prediction equations included BVALUEAACCO=1051.11+4096.38(FMS)-809.93(PBW)-10.58(TKW)-18.04(BVALUE) 0.887 <0.001AACC10=516.54+5062.59(FMS)+51.52(E/R)+4.30(SED)-11.60(TKW) 0.905 <0.001RLV=1059.75+6772.82(FMS)+102.93(E/R)+2.76(SED)-21.24(TKW)-30.72(BVALUE) 0.907 <0.001a See TABLE 1 for parameter legend.88only three parameters, compared to thirteen for the correspondinggeneral regression equation. The AACC10 and remix loaf volumeprediction equations were reduced by six and five parameters,respectively. Despite this decrease in the number of parameters,the adjusted R 2 value of the remix loaf volume equation (did notinclude B value) was greater than the adjusted R 2 value of thecorresponding method one general regression equation (Table 17).The adjusted R 2 values of the AACCO and AACC10 equations wereslightly less than the adjusted R 2 values of their correspondinggeneral regression equations. The p values of all of thestepwise equations were <0.001. Although the adjusted R 2 values(0.874-0.906) were not extremely high (i.e. >0.98), theseequations may be useful for predicting baking quality because ofthe small number of parameters that they contain, and their highlevel of significance.Addition of B value to the method one general equations hadlittle affect on the adjusted R 2 values of the stepwiseregression analysis of these equations (Table 22). The stepwiseprediction equation of AACCO had a small increase in adjusted R 2value from its corresponding general prediction equation, whilethe stepwise prediction equations of AACC10 and remix loaf volumehad no changes in the adjusted R 2 values from their correspondinggeneral prediction equations. The p values remained the same forall three equations. B value was not retained by the stepwiseanalysis of the prediction equation of AACC10. Therefore, the89stepwise prediction equation of AACC10 and its correspondinggeneral prediction equation were the same.Stepwise analysis of the method two general equations alsoresulted in a large decrease in the number of parameters from thegeneral equations (Table 23). The adjusted R 2 values of thestepwise equations were slightly greater than those of thecorresponding general equations (Table 18). The p values of thestepwise prediction equations of AACC10 and remix loaf volumewere less than those of their corresponding general equations.The equations that resulted from the stepwise analysis ofthe general equations that did and did not include deformabilitymodulus, solubility, and sulfhydryl content were the same (Table23), except for the prediction equation of AACCO which wasnoncomputable.4.8 PRINCIPAL COMPONENT ANALYSIS Principal component analysis was used to reduce the qualityparameter data set to a small number of components that explainedmost of the variation of the data set. The three baking tests(AACCO, AACC10, and remix loaf volume) were not included in thetwenty-nine parameters that made up the data set for theprnicipal component analysis. Multiple regression of theprincipal components on the baking tests was computed and thus,TABLE 23. Prediction equations for baking tests from stepwise regression analysis of methodtwo prediction equations.Prediction equations Adjusted^pR2General prediction equation did not include DM, SOL, or SHAACCO^ noneAACC10=693.99+28.99(DDT)-128.35(MDT)+2.25(R) 0.774 <0.001RLV=639.03+27.75(DDT)-185.59(MDT)+4.84(R)+3.36(SD) 0.883 <0.001General prediction equation included DM, SOL, and/or SHAACCO=1143.34-4.82(SOL) 0.084 0.121AACC10=693.99+28.99(DDT)-128.35(MDT)+2.25(R) 0.774 <0.001RLV=639.03+27.75(DDT)-185.59(MDT)+4.84(R)-3.36(SD) 0.883 <0.001a See TABLE 1 for parameter legend.91the ommission of the baking tests from the data set.An eigenvalue of greater than 1.00 was the determinant ofthe number of components to be chosen. Seven components wereobtained and explained 91% of the total variance of the twenty-nine quality parameters (Table 24).Varimax orthogonal rotation was used to spread theparameters with high loadings among the seven componentsidentified, and thus improve the interpretability of the dataset. The rotated component loadings, eigenvalues, and percent oftotal variance are given in Table 24. Component one was loadedwith mainly dough strength parameters such as mixograph firstminute slope, farinograph absorption, mixograph peak height,farinograph mixing tolerance index, extensigraph extensibility toresistance ratio, mixograph peak band width, and extensigraphresistance. Protein content, grinding time, and test weight werealso included. This component appears to be linked withqualitative parameters (except protein content and test weight).The second component contained the quantitative parametersash content, starch damage, and thousand kernel weight.Amylograph viscosity, sedimentation value, and grinding time werealso associated with component two. This component containedboth wheat/flour strength and milling quality parameters. How-ever, the milling quality parameters had the greater loadings.92TABLE 24. Rotated principal component loadings for the sevencomponents identified.Parametera Component1 2 3 4 5 6 7FMS 0.95FAB 0.91PRT 0.90PHT 0.90MTI -0.78E/R -0.74 0.57PBW 0.71 0.55GRT -0.69 0.53BVALUE -0.69R 0.53 0.75TW -0.52 -0.53ASH 0.98SD 0.94TKW 0.84AMV -0.84SED -0.53 0.60MDT 0.98ETP 0.97BWE 0.92TEG 0.80DDT 0.80DM 0.72SH -0.53 -0.64SS 0.92SOL 0.65E 0.71FN -0.57AVALUE 0.72FLY 0.54Eigenvalues 7.88 4.67 7.28 1.40 1.32 2.32 1.54% of totalvariance27.2 16.1 25.1 4.8 4.6 8.0 5.3% cumulativeproportion27.2 43.3 68.3 73.1 77.7 85.7 91.0a See TABLE 1 for parameter legend.Loadings less than 0.5 in absolute value were omitted.93Component three included sulfhydryl content and deformabil-ity modulus, but was mostly concerned with mixograph parameters.Mixograph mixing determination time, energy to peak, band widthenergy, and total energy were all linked to component three aswell as other dough strength parameters such as farinograph doughdevelopment time, and extensigraph resistance. This componentwas also a qualitative component.The chemical and functional parameters were related tocomponents four and five. Disulfide content had a high loadingon component four, while sulfhydryl content and solubility wereassociated with component five. Test weight was also linked tocomponent five which was quantitative.Component six involved extensigraph extensibility,sedimentation value, falling number, and mixograph peak bandwidth. Finally, A value and flour yield were associated withcomponent seven.Examining the percent of the total variance explained byeach component shows that seventy-six percent of the totalvariation of the data set is explained by the first threecomponents plus component six. Principal component analysis hasreduced the number of parameters from twenty-nine to four whilestill keeping seventy-six percent of the total variation and fromtwenty-nine to seven while keeping ninety-one percent of the94total variation.Table 25 includes the prediction equations of the bakingtests with the seven principal components identified. Theadjusted R 2 values for all three of the prediction equations werehigh and the equations were all highly significant (p<0.001).The prediction equation of remix loaf volume had the greatestadjusted R2 (0.937) of the three equations.The prediction equations from stepwise analysis are alsoshown in Table 25. These equations had greater adjusted R 2values than the equations with all seven components, andcontained fewer components.The regression analysis of the baking tests on the principalcomponents avoided multicolinearity. Multicolinearity occurswhen there is a large amount of linear correlation among thepredictor variables. Inflated standard errors of the regressioncoefficients may result and will invalidate the accuracy of thecomputation (Systat, 1991; Nakai and Li-Chan, 1993). This wasnot a problem here since each of the principal components wereuncorrelated.Multicolinearity may be present in some of the predictionequations created from the five methods previously. In particu-lar, method one equations contained several parameters that hadTABLE 25.^Prediction equations for baking tests with principal components.Prediction equationa b^AdjustedR2pGeneral prediction equationsAACCO=755.79+115.38(PC1)-30.32(PC2)-12.49(PC3)-0.86(PC4)+1.35(PC5)-12.18 0.865 <0.001(PC6)+6.34(PC7)AACC10=777.90+134.69(PC1)-78.02(PC2)-34.44(PC3)+5.04(PC4)+1.88(PC5)+31.47 0.875 <0.001(PC6)+8.88(PC7)RLV=907.63+193.57(PC1)-85.28(PC2)-57.88(PC3)-4.05(PC4)-23.20(PC5)+36.71 0.937 <0.001(PC6)+20.26(PC7)Stepwise prediction equationsAACCO=755.79+115.38(PC1)-30.32(PC2) 0.882 <0.001AACC10=777.90+134.69(PC1)-78.02(PC2)-34.44(PC3)+31.47(PC6) 0.897 <0.001RLV=907.63+193.57(PC1)-85.28(PC2)-57.88(PC3)-23.20(PC5)+36.71(PC6)+20.26 0.941 <0.001(PC7)a See TABLE 1 for parameter legend.b PC = Principal component.l90196high loadings on principal component 1. The tolerances of theseparameters from the regression analysis were also low. These areboth indicators of the presence of multicolinearity. Although,the adjusted R 2 values of the method one equations were high(adjusted R2>0.87) and most of the equations were highlysignificant (p<0.001), the accuracy of these equations may bequestionable.Regression analysis of the prediction equations of thebaking tests with parameters that had high loadings on principalcomponents 1, 2, and 3 was performed. One parameter from each ofthe above principal components was included in the equations.Principal components 1, 2, and 3 were used because they had thehighest eigenvalues and together explained 68.3% of the totalvariation of the data set. As mentioned previously, principalcomponents 1 and 3 were linked with qualitative parameters, whileprincipal component 2 was associated with quantitativeparameters, thus these equations had both a qualitative and aquantitative component.The parameters included in the prediction equations all hadhigh loadings on their respective principal components. Proteincontent (principal component 1), ash content, thousand kernelweight, starch damage (principal component 2), mixograph mixingdevelopment time, energy to peak, and band width energy(principal component 3) were included in the equations.97Table 26 shows the equations with the three highest adjustedR 2 values. The range of the adjusted R 2 values of all of theregressions computed here was 0.687-0.853. All of the equationswere highly significant (p<0.001) and only included threeparameters.The prediction equations from the stepwise analysis of themethod one equations included a similar number of parameters (3-5), but had greater adjusted R 2 values (0.874-0.907) than theadjusted R2 values of the equations here.TABLE 26.^Prediction equations for baking tests with parameters selected by principalcomponents.Prediction equationa^ Adjusted^pR2AACC10=122.41+66.97(PRT)-6.15(SD)-56.13(MDT) 0.853 <0.001RLV=-6.57+91.17(PRT)-6.55(SD)-83.36(MDT) 0.847 <0.001RLV=818.61+64.94(PRT)-18.59(TKW)-52.37(MDT) 0.847 <0.001a See TABLE 1 for parameter legend.995.0^CONCLUSIONS A number of wheat, flour, and dough quality parameters weremeasured and the relationships between these parameters and thebaking quality of bread were examined. A clear understanding ofthe relationship between these quality parameters and bakingability was not obtained. Several prediction equations withdifferent numbers and types of parameters were significant andhad similar adjusted R2 values. It did not appear that therewere any consistent patterns between the parameters and thebaking tests.A few of these equations are given in Table 27.^Theprediction equation of AACC10 with farinograph absorption,farinograph mixing tolerance index, extensigraph extensibility toresistance ratio, mixograph first minute slope, mixograph peakheight, mixograph peak band width, thousand kernel weight , flouryield, protein content, and sedimentation value had one of thehighest adjusted R 2 values (0.924) of all the equations.However, there may be a multicolinearity problem with thisequation. Two equations with fewer parameters were theprediction of remix loaf volume with mixograph first minuteslope, extensigraph extensibility to resistance ratio,sedimentation value, thousand kernel weight, and B Value(adjusted R2=0.907) and the prediction of AACC10 with mixographfirst minute slope, extensigraph extensibility to resistanceTABLE 27. Summary of prediction equations for baking tests.Prediction equation' AdjustedR2p TableA1CC10=2075.32-28.71(FAB)-3.71(MTI)+142.79(E/R)+7220.02(FMS)-1979.96 0.924 <0.001 17(PHT)+659.66(PBW)-12.25(TKW)-0.74(FLY)+27.53(PRT)+1.85(SED)RLV=1059.75+6772.82(FMS)+102.93(E/R)+2.76(SED)-21.24(TKW)-30.72 0.907 <0.001 22(BVALUE)AACC10=516.54+5062.59(FMS)-51.52(E/R)+4.30(SED)-11.60(TKW) 0.905 <0.001 22See TABLE 1 for parameter legend.101ratio, sedimentation value, and thousand kernel weight (adjustedR 2=0.905). These latter two equations could be useful forpredictive purposes since they have a small number of parametersand relatively high adjusted R 2 values.None of the prediction equations with the individualparameters had high R 2 values (<0.743). Some unexpectedparameters such as thousand kernel weight and flour yield hadsignificant relationships with baking tests while other doughstrength parameters (farinograph dough development time,extensigraph extensibility, and mixograph mixing determinationtime) that were expected to have significant relationships withbaking tests did not. Only two of the quality parameters(mixograph first minute slope, and mixograph peak height) weresignificantly (p<0.05) correlated to the baking tests.Principal component analysis reduced the data set to sevencomponents which explained 91% of the total variation. Thecomponents could be identified as qualitative (dough strengthparameters) and quantitative (milling quality/strengthparameters). The prediction of baking tests with the principalcomponents resulted in equations with relatively high adjusted R 2values (R 2=0.865-0.937) and avoided multicolinearity. Theprediction equations of the baking tests with parameters that hadhigh loadings on principal components 1, 2, and 3 did not havehigh adjusted R 2 values (0.687-0.853) but were highly significant102(p<0.001).The fundamental rheological (deformability modulus andstress relaxation parameters), chemical (sulfhydryl and disulfidecontent), and functional (solubility) properties did not appearimportant to the relationships with baking quality. The stressrelaxation parameter B value was the only fundamentalrheological, chemical, or functional parameter that appeared tohave even the slightest association with baking quality. Thesimple linear regression of the baking tests on B value (rate ofstress relaxation) were significant (p<0.05).Examination of the fundamental rheological parameters andthe different groups of cultivars showed that greater rates ofstress relaxation were found with the doughs made from the poorbreadmaking cultivars than with those made from the goodbreadmaking cultivars. Also, doughs made from hard strong wheatshad greater resistance to deformation (deformability modulus)than those made from the soft weak wheats.The gluten of cultivars with poor breadmaking ability werefound to have a greater sulfhydryl content than the gluten ofcultivars with good breadmaking ability. No differences werefound in disulfide content and solubility of gluten from good andpoor breadmaking cultivars.103Some considerations for future research are to test thepredictability of some of the prediction equations with anotherdata set. There are many more combinations of parameters thatcan be regressed and perhaps obtain a model with betterpredictability of baking tests.The fundamental rheological parameters studied here did notappear to be important to breadmaking. Perhaps rheology is notreally that essential in breadmaking. Extensibilty of the doughto form and retain gas bubbles appears to be important butbesides this, maybe other rheological properties are not criticalfor good breadmaking. The study of some other fundamentalrheological principles in relation to breadmaking might confirmor deny this.Although difficult to obtain because of the insolubility ofgluten, the measurements of other types of bonding such ashydrophobicity, charge content, and hydrogen bonding might proveuseful in describing good breadmaking cultivars.104REFERENCES AACC. 1962. Approved Methods of the American Association ofCereal Chemists, 7th ed. Method 38-10, Approved 1961. TheAssociation, St. Paul, MN.Baker, R.J., and Campbell, A.B. 1971. Evaluation of screeningtests for quality of bread wheat. Can. J. Plant Sci.51:449.Bass, E.J. 1988. Wheat flour milling. Ch. 1 In: "Wheat:Chemistry and Technology." Vol. II, 3rd Ed.^Y.Pomeranz (Ed.)^American Association of CerealChemists, St. Paul, MN.Beveridge, T., Toma, S.J., and Nakai, S. 1974. Determination ofS-H and S-S groups in some food proteins using Ellman'sreagent. J. Food Sci. 39:49.Bloksma, A.H. 1990. Rheology of the breadmaking process. CerealFoods World 35:228.Bloksma, A.H., and Bushuk, W. 1988. Rheology and chemistry ofdough. Ch. 4 In: "Wheat: Chemistry and Technology." Vol.II, 3rd Ed. Y. Pomeranz (Ed.) American Association ofCereal Chemists, St. Paul, MN.Buchanan, A.M., and Nicholas, E.M.^1980.^Sprouting, alpha-amylase and breadmaking quality. Cereal Res. Commun. 8:23.Buckley, D.J., Lukow. 0.M., Lalonde, J.L.M., and Woods, S.M.1990.^A multilevel computer system for dough qualityanalysis and report distribution.^Cereal Foods World35:472.Bushuk, W. 1985. Flour proteins: Structure and functionality indough and bread. Cereal Foods World 30:447.Concon, J.M. and Soltess, D. 1973.^Rapid micro-Kjeldahldigestion of cereal grains and other biological materials.Anal. Biochem. 53:55.Cullen-Refai, A., Faubion, J.M., and Hoseney, R.C. 1988.Lubricated uniaxial compression of fermenting doughs. CerealChem. 65:401.D'Egidio, M.G., Mariani, B.M., Nardi, S., Novaro, P., andCubadda, R. 1990. Chemical and technological variables andtheir relationships: A predictive equation for pasta cookingquality. Cereal Chem. 67:275.105Dreese, P.C., Faubion, J.M., and Hoseney, R.C. 1988.^Dynamicrheological properties of flour, gluten, and gluten-starchdoughs II. Effect of various processing and ingredientschanges. Cereal Chem. 65:354.Ewart, J.A.D. 1968. A hypothesis for the structure and rheologyof glutenin. J. Sci. Food Agr. 19:617.Ewart, J.A.D. 1977.^Re-examination of the linear gluteninhypothesis. J. Sci. Food Agr. 28:191.Ewart, J.A.D. 1979.^Glutenin structure.^J. Sci. Food Agr.30:482.Finney, K.F., and Barmore, M.A. 1948. Loaf volume and proteincontent of hard winter and spring wheats. Cereal Chem.25:291.Finney, K.F., Yamazaki, W.T., Youngs, V.L., and Rubenthaler, G.L.1987. Quality of hard, soft, and durum wheats. In: "Wheatand wheat improvement." Agronomy monograph 13, ASA,Madison, WI.Fowler, D.B., and DeLaRoche, I.A., 1975.^Wheat qualityevaluation. 2. Relationships among prediction tests. Can.J. Plant Sci. 55:251.Greenaway, W.T., Hurst, N.S., Neustadt, M.H., and Zeleny, L.1966. Micro sedimentation test for wheat. Cereal Sci.Today 11:197.Hoseney, R.C., Hsu, K.H., and Junge, R.C. 1979. A simple spreadtest to measure the rheological properties of fermentingdough. Cereal Chem. 56:141.Johnson, E.A., Segars, R.A., Kapsalis, J.G., Normand, M.D., andPeleg, M. 1980. Evaluation of the compressive deformabilitymodulus of fresh and cooked fish flesh. J. Food Sci. 45:1318.Kaufman, S.P., Hoseney, R.C., and Fennema, 0. 1986.^Doughrheology - A review of structural models and the role ofdisulfide interchange reactions. Cereal Foods World 31:820.Kilborn, R.H., and Tipples, K.H. 1981.^Canadian test bakingprocedures I. GRL remix method and variations. CerealFoods World 26:624.Kolster, P., and Vereijken, J.M. 1993. Evaluating HMW gluteninsubunits to improve bread-making quality of wheat. CerealFoods World. 38:76.106Kosmolak, F.G. 1978. Grinding time - screening test for kernelhardness in wheat. Can. J. Plant Sci. 58:415.Kruger, J.E., and Tkachuk, R. 1969. Wheat alpha-amylases. I.Isolation. Cereal Chem. 46:219.Kuchumova, L.P., and Strelnikova, M.M. 1968. Sulfhydryl groupsand disulfide bonds of the gluten of hard and soft wheats.Prikl. Biokhim. Mikrobiol. 3:592. (Chem. Abstr. 68:38259h).Kulp, K., Roewe-Smith, P., and Lorenz, K.^1983.^Preharvestsprouting of winter wheat. I. Rheological properties offlours and physicochemical characteristics of starches.Cereal Chem. 60:355.Launay, B., and Bure, J. 1974. Stress relaxation in wheat flourdoughs following a finite period of shearing. I.Qualitative Study. Cereal Chem. 51:151.Law, C.N., Payne, P.I., Worland, A.J., Miller, T.E., Harris,P.A.,Snape, J.W., and Reader, S.M. 1984. Studies of geneticalvariation affecting grain protein type and amount in wheat.In: "Cereal Grain Protein Improvement. Proceedings of aMeeting." FAO/IAEA/GSF/SIDA. IAEA, Vienna.Lukow, O.M., Zhang, H. and Czarnecki, E. 1990.^Milling,rheological, and end-use quality of Chinese and Canadianspring wheat cultivars. Cereal Chem. 67:170.Lukow, O.M., and McVetty, P.B.E. 1991. Effect of cultivar andenvironment on quality of characteristics of spring wheat.Cereal Chem. 68:597.MacRitchie, F. 1973. Conversion of a weak flour to a strong oneby increasing the proportion of its high molecular weightgluten protein. J. Sci. Food Agr. 24:1325.MacRitchie, F. 1975. Mechanical degradation of gluten proteinsduring high-speed mixing of doughs. J. Polymer Sci.Symposium No. 49:85.MacRitchie, F. 1978. Difference in baking quality between wheatflours. J. Food Technol. 13:187.MacRitchie, F. 1984. Baking quality of wheat flours. Adv. FoodRes.^29:201.MacRitchie, F. 1987.^Evaluation of contributions from wheatprotein fractions to dough mixing and breadmaking.^J.Cereal Sci. 6:259.107MacRitchie, F. 1989. Identifying the baking quality relatedcomponents of wheat flours. Cereal Foods World 34:548.Matsuo, R.R., Dexter, J.E., Kosmolak, F.G., and Leslie, D. 1982.Statistical evaluation of tests for assessing spaghetti-making quality of durum wheat. Cereal Chem. 59:222.Mok, C., and Dick, J.W. 1991. Response of starch of differentwheat classes to ball milling. Cereal Chem. 68:409.Nakai, S., and Li-Chan, E. 1993. Recent advances in structureand function of food proteins: QSAR approach. Crit. Rev.Food Sci. Nut. 33:477.Orth, R.A., Baker, R.J., and Bushuk, W. 1972.^Statisticalevaluation of techniques for predicting baking quality ofwheat cultivars. Can. J. Plant Sci. 52:139.Orth, R.A., and Bushuk, W. 1972. A comparative study ofthe proteins of wheats of diverse baking qualities.Cereal Chem. 49:268.Peleg, M. 1979. Characteristics of the stress relaxation curvesof solid foods. J. Food Sci. 44:277.Preston, K.R., Lukow, 0.M., and Morgan, B. 1992. Analysis ofrelationships between flour quality properties and proteinfractions in a world wheat collection. Cereal Chem. 69:560.Pomeranz, Y. 1987.^Grain quality.^Ch. 7 In: "Modern CerealScience and Technology." VCH, New York, NY.Pomeranz, Y. 1988. Composition and functionality of wheat flourcomponents. Ch. 5. In: "Wheat: Chemistry and Technology."Vol. II, 3rd Ed. Y. Pomeranz (Ed.) American Association ofCereal Chemists, St. Paul, MN.Ponte, J.G. Jr., DeStefanis, V.A., Titcomb, S.T., and Cotton,R.H. 1967. Study of gluten properties as influenced bycertain organic solvents. Cereal Chem. 44:211.SAS. 1985.^SAS users guide:Statistics.^SAS Institute, Inc.,Cary, Inc.Schofield, J.D., and Booth, M.R. 1983. Wheat Proteins and theirtechnological significance. In: "Developments in FoodProteins." Vol. 2. B.J.F. Hudson (Ed.) Applied SciencePublishers, London.108Shuey, W.C., and Schmitz, A.O. 1982. Instrument construction.Ch. II. In: "The Amylograph Handbook." W.C. Shuey, andK.H. Tipples (Eds.) American Association of CerealChemists, St. Paul, MN.Singh, N.K., Donovan, R., and MacRitchie, F. 1990.^Use ofsonication and size-exclusion high-performance liquidchromatography in the study of wheat flour proteins. II.Relative quantity of glutenin as a measure of breadmakingquality. Cereal Chem. 69:560.SYSTAT\SYGRAPH. 1991. "The System of Statistics and Graphics Forthe PC." SYSTAT, Inc. Evanston, IL.Tipples, K.H. 1982. Uses and applications. Ch. III. In: "TheAmylograph Handbook." W.C. Shuey, and K.H. Tipples (Eds.)American Association of Cereal Chemists, St. Paul, MN.Tsen, C.C., and Bushuk, W. 1968. Reactive and total sulfhydryland disulfide contents of flours of different mixingproperties. Cereal Chem. 45:58.Williams, P.C., Jaby El-Haramein, F., Ortiz-Fereira, G., andSrivastava, J.P. 1988.^Preliminary observations on thedetermination of wheat^strength by near-infraredreflectance. Cereal Chem. 65:109.Wrigley, C.W. 1993.^A molecular picture of wheat quality:Finding and fitting the jigsaw pieces. Cereal Foods World38:68.Zeleny, L. 1947. A simple sedimentation test for estimating thebreadbaking and gluten qualities of wheat flour. CerealChem. 24:465.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086260/manifest

Comment

Related Items