Key words: Durum wheat, bread making quality, gluten strength, physical dough properties, pasta cooking quality

Transcription

1 Relationships among bread-making quality, gluten strength, physical dough properties, and pasta cooking quality for some Canadian durum wheat genotypes B. A. Marchylo 1, J. E. Dexter 1, F. R. Clarke 2, J. M. Clarke 2, and K. R. Preston 1 1 Canadian Grain Commission, Grain Research Laboratory, Main St., Winnipeg, Manitoba, Canada R3C 3G8; 2 Agriculture and Agri-Food Canada, Semiarid Prairie Agricultural Research Centre, P.O. Box 1030, Swift Current, Saskatchewan, Canada S9H 3X2. Contribution 818 of the Canadian Grain Commission. Received 27 September 2000, accepted 1 May Marchylo, B. A., Dexter, J. E., Clarke, F. R., Clarke, J. M. and Preston, K. R Relationships among bread-making quality, gluten strength, physical dough properties, and pasta cooking quality for some Canadian durum wheat genotypes. Can. J. Plant Sci. 81: Fifty-four durum wheat (Triticum durum) genotypes entered into the 1995, 1996 and 1997 Co-operative Tests were evaluated for gluten strength characteristics using the sodium dodecyl sulphate (SDS) sedimentation test, the gluten index (GI) test, and physical dough tests including farinograph (high and low adsorption), mixograph, alveograph and extensigraph. Baking quality was evaluated for bread prepared by the Canadian short process (CSP), a short mechanical dough mixing process, and pasta quality was evaluated for spaghetti dried at both low (40 C) and high (70 C) temperatures. The effect of genotype on physical dough measurements, baking quality and spaghetti cooking quality was then determined. SDS sedimentation, GI, pasta dough farinograph (low absorption), bread dough farinograph (high absorption), extensigraph and alveograph measurements were interrelated. When baked by the CSP, the strongest genotypes exhibited mixing times and mixing energies similar to or greater than good quality bread wheat (Triticum aestivum). Although loaf volume (LV) was positively correlated to gluten strength indicators, the strongest genotypes still exhibited only about 85% of the LV expected of good-quality bread wheat of comparable protein content. Baking quality however, was not related to pasta cooking quality, and, therefore, there is potential to breed for dual-purpose durum cultivars, which combine improved baking properties and good pasta cooking quality. Key words: Durum wheat, bread making quality, gluten strength, physical dough properties, pasta cooking quality Marchylo, B. A., Dexter, J. E., Clarke, F. R., Clarke, J. M. et Preston, K. R Liens entre la qualité pour la panification, la fermeté du gluten, les propriétés physiques de la pâte et la qualité pour la production de pâtes alimentaires de certains génotypes de blé dur canadiens. Can. J. Plant Sci. 81: Les essais coopératifs de 1995, de 1996 et de 1997 portaient sur 54 variétés de blé dur (Triticum durum). À cette occasion, on a déterminé la fermeté du gluten par sédimentation SDS, selon l indice du gluten et au moyen de divers essais physiques sur la pâte, notamment avec le farinographe (absorption élevée ou faible), le mixographe, l alvéographe et l extensimètre alvéographe. La qualité boulangère a été évaluée pour le pain obtenu par le procédé canadien, qui suppose une brève période de pétrissage mécanique, tandis que la qualité des pâtes alimentaires a été évaluée avec du spaghetti séché à basse (40 C) et à haute (70 C) température. On a ensuite étudié l incidence du génotype sur les propriétés physiques de la pâte, la qualité boulangère et la qualité des spaghettis à la cuisson. La sédimentation SDS, l indice du gluten et les mesures obtenues avec le farinographe pour pâte à pâtes alimentaires (faible absorption), le farinographe pour pâte à pain (absorption élevée), l extensimètre alvéographe et l alvéographe ont des liens entre eux. Lorsqu on produit le pain selon le procédé canadien, les génotypes les plus marqués demandent un pétrissage d une durée et d une force semblables ou supérieures à celles d un blé panifiable de bonne qualité (Triticum aestivum). Même si le volume d un pain présente une corrélation positive avec les indicateurs de fermeté du gluten, les génotypes les plus marqués produisent un pain dont le volume ne correspond qu à environ 85 % du volume obtenu avec un blé panifiable de bonne qualité, aussi riche en protéines. La qualité boulangère n est toutefois pas reliée à la qualité pour la production de pâtes alimentaires, si bien qu on pourrait sélectionner des cultivars de blé dur combinant de meilleures propriétés boulangères et une bonne qualité pour la production de pâtes alimentaires. Mots clés: Blé dur, qualité pour la panification, fermeté du gluten, propriétés physiques de la pâte, qualité pour la production de pâtes alimentaires Durum wheat (Triticum durum Desf.) is the raw material of choice for making premium-quality pasta. Accordingly, cultivar improvement efforts in Canada have focused primarily on the quality factors essential to good quality pasta, i.e., bright amber colour and firmness and resilience when cooked (Matsuo 1988; Marchylo and Dexter 1997). It is generally recognised that protein content is the primary factor associated with pasta cooking quality, with gluten quality being of secondary importance (Feillet and Dexter 1996; Marchylo et al. 1998). 611 Other important end uses for durum wheat, particularly in Mediterranean countries, include couscous, various forms of baked products, bulgar and other regional products (Quaglia 1988). Interest in breeding for durum wheat suitable for both Abbreviations: CSP, Canadian Short Process; CWAD, Canadian Western Amber Durum; CWRS, Canada Western Red Spring; GI, Gluten Index; HT, high temperature; LT, low temperature; LV, loaf volume; SDS, sodium dodecyl sulfate.

2 612 CANADIAN JOURNAL OF PLANT SCIENCE bread making and pasta making has increased (Liu et al 1996) because dual-purpose durum wheat can be used in place of bread wheats or in blends with high-quality baking flour. Such applications could give durum farmers alternative markets in years of high durum production (Boggini and Pogna 1989). Recently, in response to evolving market demand, the Canadian durum wheat-breeding program has developed genotypes with stronger gluten characteristics (Marchylo et al. 1998; Ames et al. 1999). Canadian durum wheat cultivars with conventional gluten strength characteristics are known to exhibit low LV and poor fermentation tolerance as compared with typical good-quality Canadian bread wheat cultivars (Dexter et al 1981a, 1994). Very strong Italian durum cultivars, which exhibit tenacious, inelastic gluten and tight inextensible dough properties, reportedly are not well suited to making high volume bread (Quaglia 1988). Recently registered medium-strength and extra strong gluten Canadian durum cultivars, however, have not been evaluated to determine their bread-making quality. Dexter and Marchylo (1997) reviewed the Canadian durum wheat cultivar development program, and how it evolved to meet market demand and technological change. They noted that there have been a number of important developments since the Canadian testing protocol for screening of late generation durum wheat breeding lines was last examined in detail (Matsuo et al. 1982). In particular, high temperature (HT) and ultra-high temperature (UHT) drying, which improve pasta texture and colour, are now used throughout the pasta manufacturing industry. Besides using higher drying temperatures to improve pasta quality, some manufacturers indicate that higher gluten strength durum wheat cultivars are preferable for the production of high-quality dried pasta. The GI test (Cubadda et al. 1992; MacDonald, 1994) has become an important indicator of durum wheat gluten strength and is now being used as a specification in durum wheat and semolina trading. Similarly, the alveograph has gained acceptance as a durum wheat gluten strength indicator (D Egidio et al. 1990) and alveograph parameters are important specifications for some processors. In response to a changing industry, and especially increasing application of high, and ultra-high temperature drying, GI and alveograph procedures are now being used as part of the quality evaluation of late generation Canadian durum wheat breeding lines (Marchylo and Dexter 1996, 1997; Marchylo et al. 1998). This study was undertaken to evaluate the bread-making quality of registered Canadian durum wheat cultivars and breeding lines, which exhibit different gluten strength characteristics, entered into the 1995, 1996 and 1997 durum wheat Co-operative tests. In addition, relationships between physical dough properties, baking quality and pasta cooking quality characteristics were investigated to assess the potential to develop dual-purpose durum cultivars. MATERIALS AND METHODS Wheat Samples Durum wheat registered cultivars and breeding lines entered in the western Canadian Durum Wheat Co-operative Tests were grown in four replicate trials at various sites in western Canada in 1995, 1996 and 1997 under the auspices of the Prairie Registration Recommending Committee for Grain. Wheat for quality testing was harvested from nine locations in 1995, seven locations in 1996, and eight locations in A composite of each line within each location was prepared first using equal quantities of grain from each replication. The final composite for quality testing was then prepared using the replicate composites from each location blended in varying proportions to give a final protein concentration near 13% and a physical condition that met the minimum standards for No. 1 or No. 2 Canadian Western Amber Durum (CWAD). Twenty-one breeding lines were included in the test each year along with four registered cultivars that were used as checks. Breeding lines that did not have the appropriate quality, disease or agronomic characteristics were dropped from the test, and new breeding lines were added the following year. In total 54 genotypes were tested for quality, 5 in all 3 yr, 11 in 2 of the years, and 38 for 1 yr. Quality Tests Sodium dodecyl sulphate sedimentation values were determined in duplicate on wheat as described by Dexter et al. (1980). Protein content (%N 5.7) was determined in duplicate on whole grain and semolina by Combustion Nitrogen Analysis (LECO Model FP-428 Dumas CNA Analyser, St. Joseph, MI) (Sweeney and Rexroad 1987). Semolina milling was performed in duplicate by the method of Dexter et al. (1990). Semolina from replicate millings of each composite sample was combined before further testing. Gluten index (American Association of Cereal Chemists AACC) was performed in duplicate. Physical dough tests were performed singly. Pasta dough farinograph curves (low water absorption similar to pasta processing) were obtained by the method of Irvine et al. (1961) and scored as described by Dexter and Matsuo (1978). Farinograph (50 g bowl), extensigraph and mixograph (35 g) curves were obtained and scored by AACC (2000) Standard Methods. ICC (1980) Standard No 121 was used to obtain alveograph curves (Chopin Model MA 82). Semolina was baked in duplicate by the CSP procedure (Dexter et al. 1994) except that bromate was excluded from the formula and replaced by ascorbic acid (150ppm) (Yamada and Preston 1994). Spaghetti was prepared by the micro-spaghetti-making process of Matsuo et al (1972), and dried by 40 C (LT) and 70 C (HT) drying cycles (Dexter et al. 1981b) for samples from the 1996 and 1997 Co-operative tests. For the 1995 crop year, pasta was dried only at LT. Processing was duplicated at each drying temperature. Spaghetti was cooked to optimum time (the time required for the white core in the centre of strands to disappear) + 15 s in duplicate. Cooked spaghetti texture was determined on the Grain Research Laboratory tenderness testing apparatus and expressed as tenderness index, an indicator of firmness (Matsuo and Irvine 1969), compressibility and recovery, measures of resilience (Matsuo and Irvine 1971), and an overall cooking score (Dexter et al. 1988). Duplicated tests are reported as means.

3 MARCHYLO ET AL. BREAD-MAKING QUALITY OF CANADIAN DURUM WHEAT 613 Table 1. Strength indicators for three categories of durum cultivars as illustrated by the cultivars Hercules, Kyle, AC Morse and AC Navigator Statistical Analysis For each quality trait, the following analyses were performed with SAS (Version 6.12; SAS Institute, Inc. 1992) on the response of the composited genotypes: average response, range of response and the standard deviation of the breeding lines were calculated within each year. The entries common to each year combination differed, so mixed analysis was performed with PROC MIXED (Littell et al. 1996) over 1995, 1996 and Mixed models make use of all the information available in an unbalanced data set (Stroup 1989), and can increase efficiency (Patterson 1997). The MODEL statement contains any fixed factors and the RANDOM statement contains any random factors. Genotypes were first considered fixed and years random to determine if any breeding lines differed significantly, and how they compared with the registered cultivars Hercules, Kyle, AC Navigator and AC Morse. Mixed analysis was performed with genotypes and years random to estimate variances. With composites of selected locations within each year, our estimate of genetic variance is inflated with genotype by year and genotype by location variance, and the error (sampling) variance is inflated with genotype by year by location variance. Correlations were calculated among the quality characteristics for each year and were pooled over 1995, 1996 and 1997 if homogeneous (Steele and Torrie 1997). Cooking quality for pasta dried at HT or LT was compared using a paired difference t-test. RESULTS AND DISCUSSION Gluten Strength Characterisation The efforts to strengthen gluten properties of Canadian durum wheat have produced cultivars that can be broadly classified into three categories of gluten strength as shown in Table 1. Those with strength characteristics consistent with Kyle, which currently is the predominant durum cultivar grown in western Canada, fall into one category. These breeding lines have gluten with good extensibility but lower resistance to extension and they have been modelled after Hercules, which has been the regulatory standard for the CWAD class for many years (Marchylo and Dexter 1996; SDS z P/L z W z sedimentation GI z MWI z (height (joules (ml) (%) (cm 2 ) 1.1/length) 10 4 ) Conventional Strength Hercules y Kyle y Medium Strength AC Morse y Extra Strong AC Navigator x SED w (excluding AC Navigator) SED (including AC Navigator) z SDS = sodium dodecyl sulfate, GI = gluten index, MWI = mixograph work input; P/L = ratio of alveograph P and L, W = Alveograph W. y Average over 3 yr, 1995, 1996 and x Average over 2 yr, 1996 and w Standard error of a difference. Dexter and Marchylo 1997). A new category includes genotypes with gluten that can be characterised as extra strong. This gluten is much more elastic and shows considerable resistance to extension. AC Navigator (Clarke et al. 2000), a new cultivar that was granted interim registration in 1998, is typical of the extra strong gluten types. The third category exhibits gluten strength characteristics intermediate to those of the conventional and extra strong types. AC Morse, which was registered in 1996, is typical of the intermediate gluten strength category and exhibits the strength that has been targeted for future conventional CWAD cultivars. SDS sedimentation, GI, mixograph work input (MWI) and Alveograph P/L ( P = maximum height of the curve and is a measure of resistance and, L = the length of the curve and measures extensibility) and W (area under the curve 1.1 and measure of overall strength) values all ranked AC Navigator > AC Morse > Kyle and Hercules as significantly different (P < 0.05) in strength (Table 1). The relative strength of genotypes within these three categories is compared using these commonly used strength indicators as shown in Table 2. The majority of breeding lines were stronger than the most widely grown commercial cultivar Kyle and similar to the new intermediate strength cultivar AC Morse. The increase in strength is especially evident for SDS, GI and W strength parameters but less so for MWI and P/L. The high number of breeding lines with P/L values close to that of Kyle is probably due to the higher extensibility associated with Canadian durum cultivars and breeding lines. A small number of breeding lines with strength in the extra strong category is to be expected because breeding for high strength has only recently been initiated. Mean, range and standard deviation of composited samples of the durum genotypes in 1995, 1996 and 1997 for each quality measurement are summarised in Table 3. Gluten strength exhibited a wide range each year as evident from SDS sedimentation, GI, and physical dough parameters. AC Navigator and other extra strong genotypes, which were entered into the co-operative testing program beginning in 1996, contributed to the higher mean and maximum values for GI, SDS sedimentation and physical dough strength indicators in 1996 and 1997 as compared with 1995.

4 614 CANADIAN JOURNAL OF PLANT SCIENCE Table 2. Number of genotypes with values lower, similar or higher (P < 0.05) for commonly used durum wheat strength indicators compared to Hercules, Kyle, AC Morse amd AC Navigator z SDS z P/L z W z sedimentation GI z MWI z (height (joules (ml) (%) (cm 2 ) 1.1/length) 10 4 ) Compared to Hercules Lower Similar Higher Compared to Kyle Lower Similar Higher Compared to AC Morse Lower Similar Higher Compared to AC Navigator Lower Similar Higher z SDS = sodium dodecyl sulfate, GI = gluten index, MWI = mixograph work input; P/L = ration of alveograph P and L, W = Alveograph W. The high contribution of genotype to the variance in gluten strength and physical dough measurement in this study is consistent with previous reports for durum wheat grown in western Canada (Matsuo et al. 1982; Ames et al 1999) and Europe (Autran et al. 1986; Mariani et al. 1995). In the current study, SDS sedimentation and GI exhibited comparable correlations to physical dough measurements (Table 4 and 5). The SDS sedimentation test has been used effectively in durum wheat gluten strength screening in Canada for many years. More recently, however, GI has become an important quality index internationally and, therefore, it was introduced into the Canadian durum quality screening program (Marchylo and Dexter 1996; Marchylo et al. 1998). While the SDS sedimentation is ideal for early generation tests (Dexter et al. 1980; Dick and Quick 1983), the increased use of GI in international trade makes it useful for assessment of commercial acceptance of potential cultivars. The pasta dough farinograph (Matsuo 1988) and the mixograph (Bendelow 1967) procedures were, for many years, the primary durum wheat physical dough tests used to evaluate later generation Canadian durum wheat breeding lines. The mixograph is a primary physical dough testing procedure in the US durum wheat breeding program (Dick and Youngs 1988), while the alveograph is the physical dough test of choice in Italy (D Egidio et al. 1990) and other parts of Europe. Because of its importance in commercial trade it also was introduced into the late generation Canadian testing protocol (Marchylo and Dexter 1996; Clarke et al. 1998; Marchylo et al. 1998). Interrelationships to gluten strength indicators and the other physical dough procedures did not indicate any clear advantage of one technique over the other (Tables 4 and 5). A disadvantage of both the alveograph and the mixograph procedures is that they are performed at constant water absorption. Durum wheat is very hard, and milling technique can have a profound effect on water absorption and rheological parameters (Dexter et al. 1994). In this study, milling technique was carefully controlled, and water absorption differences, as determined by the farinograph, were only 4 to 5% each year (Table 3). A water absorption range of that magnitude will have some influence on physical dough properties, particularly alveograph values, but should not be enough to mask intrinsic differences among genotypes (Dexter et al. 1994). Farinograph and extensigraph testing, which are normally not part of the durum wheat Co-operative Test evaluation, were included in this study because these tests are commonly used to evaluate baking performance. The extensigraph, unlike the alveograph, is performed at variable water absorption based on farinograph absorption. Extensigraph area and alveograph W (work, related to area under the curve) were strongly correlated, confirming that variations in water absorption were small enough not to affect dough strength rankings by the alveograph (Table 5). Other correlations between extensigraph area, alveograph W, ratio of extensigraph maximum resistance to extension and length and ratio of alveograph P (related to maximum height) and length were heterogenous, but in all cases were highly significant (P < 0.01) for at least 2 of the 3 yr. These and other heterogeneous relationships were probably due to environmental effects. Relationships among Quality Measurements Most of the quality measurements considered in this study, for either bread or pasta, identified significant (P < 0.05) differences among genotypes (Table 4). Protein content, which is strongly related to LV of Canadian durum (Dexter et al. 1994) for sample sets exhibiting a wide range of protein, was not significantly correlated to LV in this study. Breeding lines with low protein content are screened out during early stages of the breeding program. Consequently, protein content differences among Cooperative Test breeding lines were small in all 3 yr, showing a protein range of only about 1% (Table 3), which moderated its effect on baking quality. All CSP baking quality measurements were affected by genotype (Table 4). SDS sedimentation, GI and physical dough measurements were consistently correlated (P < 0.05) with CSP mixing time and mixing energy (Table 5). The extra strong durum breeding lines exhibited CSP mixing times and energies (results not shown) equal to or higher than those reported by Preston et al. (1997) for hard red spring wheat cultivars registered in the Canada Western Red Spring (CWRS) wheat class. Durum wheat LV, however, was only weakly correlated to gluten strength. Although LV was significantly correlated to SDS sedimentation volume, in agreement with the results of Peña et al. (1994), and GI, correlations to physical dough properties were fewer and weaker than for CSP mixing time and mixing energy. The highest LV obtained, around 1800 cm 3, was much lower than for CWRS cultivars, which exhibit LV in

5 MARCHYLO ET AL. BREAD-MAKING QUALITY OF CANADIAN DURUM WHEAT 615 Table 3. Mean, range and standard deviation of quality factors for durum wheat genotypes hravested in 1995, 1996, Quality measurement Min. Max. Mean STD Min. Max. Mean STD Min. Max. Mean STD Wheat Protein (%) SDS sedimentation (ml) Semolina Protein (%) Gluten index (%) Pasta farinograph Mix time (min) Max. consistency (BU) Bandwidth (BU) Bread farinograph Absorption (%) Dev. time (min) Stability (min) Mixograph Mix time (min) Peak resistance (mm) Peak bandwidth (mm) Stability Work input (cm 2 ) Extensigraph (135 min) Area (cm 2 ) R MAX /length (BU/mm) Alveograph W (Joules 10 4 ) P/L (height 1.1/length) CSP Bread (200g) Absorption (%) Mixing time (min) Mixing energy (Wh 1 kg 1 ) Loaf volume (cm 3 ) C Spaghetti Cooking score (units) C Spaghetti Cooking score (units) NA z NA NA NA NA z NA = not available. excess of 2000 cm 3 at protein contents comparable to durum wheat breeding lines in this study (Preston et al. 1997). As durum wheat genotypes became stronger, they exhibited increasingly inextensible dough properties as seen by significant relationships of GI and SDS sedimentation to both extensigraph R max /L and alveograph P/L (Table 5). Quaglia (1988) identified inextensible dough characteristics as a factor limiting LV potential of strong Italian durum wheat cultivars. Inextensible dough may reduce LV by limiting oven spring. In addition, inextensible dough is not desirable for baking because dough handling (sheeting) properties are inferior. More recently, Ammar et al. (2000) also suggested that inadequate dough extensibility as shown by lower alveograph L and greater P/L prevents durum wheat from achieving LV equivalent to those of bread wheat. In this study, however, there was no significant correlation between dough extensibility (L) and LV (results not shown) but alveograph W did show a significant, although weak relationship (Table 5). Some genotypes did have extensibilities (alveograph L) that were equal to or greater than those of bread wheat but, typically, the alveograph P values were much lower as were W values. Consequently, to develop durum wheat cultivars with LV equivalent to those of bread wheat it may be necessary to achieve an appropriate balance of resistance to extension (i.e., alveograph P or tenacity) and extensibility (alveograph L) in conjunction with increased alveograph W values or overall strength. Short process baking was chosen for this project because Dexter et al. (1994) determined that registered CWAD cultivars, with traditional Hercules strength, exhibited better baking performance in the absence of bulk fermentation. More recently, Dexter et al. (1998) reported that extra strong

6 616 CANADIAN JOURNAL OF PLANT SCIENCE Table 4. Source of variance for durum wheat quality factors Source of variance, % Quality measurement (code) F-test for breeding lines Line (L) Year (Y) Error Wheat Protein (%) (WPR) 5.28** SDS sedimentation (ml) (SDS) 22.9** Semolina Protein (%) (SPR) 6.49** Gluten index (%) (GI) 12.92** Pasta farinograph Mix time (min) (PMT) 4.21** Max. consistency (BU) (PMC) 5.92** Bandwidth (BU) (PBW) 5.57** Bread farinograph Absorption (%) (FAB) 6.72* Dev. time (min) (FDT) 25.4** Stability (min) (FST) 5.28** Mixograph Mix time (min) (MMT) 7.55** Peak resistance (mm) (MPR) 5.49** Peak bandwidth (mm) (MBW) 2.60* Mixing stability (mm) (MMS) 5.02** Work input (cm 2 ) (MWI) 6.91** Extensigraph (135 min) Area (cm 2 ) (EXA) 37.0** R MAX /length (BU/mm) (ERL) 57.1** Alveograph W (Joules 10 4 ) (AVW) 11.8** P/L (height 1.1/length) (APL) 18.6** CSP Bread (200g) Absorption (%) (BAB) 6.63** Mixing time (min) (BMT) 5.11** Mixing energy (Wh 1 kg 1 ) (BME) 2.35* Loaf volume (cm 3 ) (LV) 2.51* C Spaghetti Cooking score (units) (LCS) z 5.08** C Spaghetti Cooking score (units) (HCS) z 4.94** z LCS and HCS data are for 1996 and 1997 only. **, *** P < 0.05 and P 0.01, respectively. durum wheat genotypes from Italy and Canada exhibited a direct relationship between gluten strength and LV for a long fermentation baking process. The benefit of extra strength on LV may have been more evident if a longer fermentation baking process had been included in the current study. However, the strongest genotypes examined by Dexter et al. (1998), which were equal to or stronger than the extra strong genotypes examined in this study, still achieved only about 80% of the LV obtained from bread wheat using a long fermentation process. There were no significant correlations between bread quality measurements and either LT or HT pasta cooking quality (Table 5). This result was not surprising considering that the cooking score was so weakly influenced by genotype. If baking performance and cooking quality are not related, then development of durum wheat cultivars combining improved baking performance and good cooking quality is achievable. Protein content, which has been clearly identified as the primary factor associated with pasta cooking quality (Feillet and Dexter 1996) was not significantly related (P > 0.05) to either LT or HT spaghetti cooking quality in the current study (Table 5) due to the narrow protein range of the sample set. As would be expected from the low genotypic variance in cooking quality, few significant correlations with other test measurements were observed (P < 0.05) (Table 5). Neither GI nor SDS sedimentation was related to cooking quality. Some significant correlations between physical dough prop-

7 MARCHYLO ET AL. BREAD-MAKING QUALITY OF CANADIAN DURUM WHEAT 617 Table 5. Pooled (1995, 1996 and 1997) simple correlations z ( 10 2 ) among quality measurements WPR SDS SPR GI PMT PMC PBW FAB FDT FST MMT MPR MBW MMS MWI EXA ERL AVW APL LCS y WPR 1 SDS 1 SPR 85 1 GI 80 1 PMT PMC PBW 62 H H 1 FAB +H FDT FST 71 H MMT MPR +H 58 +H MBW H 56 H H 1 MMS H 66 H +H 63 H MWI 67 H +H EXA H 84 +H H 82 +H H +H 1 ERL H +H AVW H H +H 53 +H 86 +H H 1 APL H +H H +H +H 1 LCS y HCS y 56 BAB +H BMT BME LV HCS y BAB BMT BME LV z Non-significant (P < 0.05) correlations not reported; +H = heterogeneous correlations with the significant ones poistive; H = heterogeneous correlations with the significant ones negative; H = heterogeneous correlations with significant ones a mixture of positive or negative; See Table 4 for measurement codes. y LCS and HCS data are for 1996 and 1997 only H 1 +H 1 1

8 618 CANADIAN JOURNAL OF PLANT SCIENCE erties such as pasta dough farinograph bandwidth, mixograph bandwidth at peak, mixograph work input and alveograph P/L to LT cooking measurements were evident, but they were weak (r 0.59 or less). HT cooking quality was not correlated significantly to gluten strength measurements or physical dough properties. These results substantiate the report of D Egidio et al. (1990) that gluten strength is a moderate factor in determining LT pasta texture for Italian durum wheat, but of little importance in determining HT pasta texture. Kovacs et al. (1997) found much stronger correlations between strength measurements such as SDS sedimentation, mixograph and alveograph parameters and pasta cooking quality for durum wheat genotypes grown for 3 yr in western Canada. That study, however, included weak gluten genotypes expressing gamma-gliadin-42 and their associated type 1 low molecular weight glutenin subunits, which are well known to be associated with inferior cooking quality (Damidaux et al. 1978; Kosmolak et al. 1980; Masci et al. 1995). In Canada, screening for gamma-gliadin 45 genotypes is performed early in the breeding program (Clarke et al. 1993, 1998). All genotypes in this study, regardless of strength, were gamma-45 types, which have the associated low molecular weight glutenins related to superior cooking quality (Pogna et al. 1988). The small genotype variance seen in cooking quality in this study suggests that selection for gamma-45 genotypes, and for gluten strength by SDS sedimentation, GI and alveograph W or P/L produces genotypes with similar cooking quality potential. CONCLUSIONS SDS sedimentation, GI and physical dough test measurements using the mixograph, alveograph, farinograph and extensigraph were interrelated. Significant correlations among these quality measurements and both durum wheat baking quality and pasta cooking quality indicate that they can be used to predict both characteristics. GI and alveograph tests are of particular importance as primary tests for selecting for gluten strength because they are becoming internationally accepted as durum wheat strength indicators and as marketing specifications. In this study, although some correlations between physical dough properties and end product quality were significant, they were relatively weak, which indicates that considerable variation between quality measurements and baking and pasta making quality remains unexplained. It appears, then, that further work is necessary to determine other sources of variation that influence durum wheat baking and pasta making quality. Genotype contributed to LT cooking quality variance, but because LT drying is no longer widely used, inclusion in breeding line evaluations is of limited value. Statistical evaluation also showed that genotype contributed to HT cooking quality variance. Because previous work (De Stefanis and Sgrulletta 1990) has suggested that the magnitude of improvement of pasta texture with HT and ultra-ht drying differs among genotypes, it may be appropriate to assess breeding lines at more than one HT drying temperature. In addition, pasta processing at HT is an important tool to evaluate the interaction between drying temperature and pasta colour of breeding lines. Breeding for increased strength in Canadian durum wheat cultivars will be an asset in baking. The strongest genotypes in this study exhibited mixing times and mixing energies similar to or greater than high-quality bread wheat cultivars, which will impart more mixing tolerance and fermentation tolerance. LV was positively correlated with gluten strength indicators, but the strongest genotypes still exhibited only about 85% of the LV of good quality bread wheat. An appropriate balance between resistance to extension and extensibility in conjunction with increased alveograph W values is probably needed in order to breed for durum wheat cultivars with LV equivalent to those of bread wheat. Because baking quality was not related to pasta cooking quality, it is possible that dual-purpose durum cultivars with both good baking properties and pasta cooking quality can be developed. To achieve this goal, more research is required to determine if the cause of inferior LV in durum wheat is due to protein composition or to other grain components. ACKNOWLEDGEMENTS We thank the many dedicated staff of the Grain Research Laboratory and Semiarid Prairie Agriculture Research Centre for their excellent technical contributions that led to the completion of this work. We thank the Prairie Registration Recommending Committee for Grains for allowing the use of data recorded in the minutes of their annual meetings. American Association of Cereal Chemists Approved methods of the AACC. 10th ed. Method 38-12, Wet gluten and gluten index; Method 54-10, Extensigraph method, general; Method 54-12, Farinograph method for flour; Method 54-40A. Mixograph method. AACC, St Paul, MN. Ames, N. P., Clarke, J. M., Marchylo, B. A., Dexter, J. E. and Woods, S. M Effect of environment and genotype on durum wheat gluten strength and pasta viscoelasticity. Cereal Chem. 76: Ammar, K., Kronstad, W. E. and Morris, C. F Breadmaking quality of selected durum wheat genotypes and its relationship with high molecular weight glutenin subunits allelic variation and gluten protein composition. Cereal Chem. 77: Autran, J. C., Abecassis, J. and Feillet, P Statistical evaluation of different technological and bicohemical tests for quality assessment in durum wheats. Cereal Chem. 63: Bendelow, V. M Early generation screening for gluten strength in durum wheat breeding. Cereal Sci. Today 12: Boggini, G. and Pogna, N. E The breadmaking quality and storage protein composition of Italian durum wheat. J. Cereal Sci. 9: Clarke, J. M., Howes, N. K., McLeod, J. G. and DePauw, R. M Selection for gluten strength in three durum wheat crosses. Crop Sci. 33: Clarke, J. M. Marchylo, B. A., Kovacs, M. I. P., Noll, J. S., McCaig, T. N. and Howes, N. K Breeding durum wheat for pasta quality in Canada. Euphytica 100: Clarke, J. M., Mcleod, J. G., Depauw, R. M., Marchylo, B. A., McCaig, T. N., Knox, R. E., Fernandez, M. R. and Ames, N AC Navigator durum wheat. Can. J. Plant Sci. 80: Cubadda, R., Carcea, M. and Pasqui, L. A Suitability of the gluten index test for assessing gluten strength in durum wheat and semolina. Cereal Foods World 37:

9 MARCHYLO ET AL. BREAD-MAKING QUALITY OF CANADIAN DURUM WHEAT 619 Damidaux, R., Autran, J. C., Grignac, P. and Feillet, P Mise en évidence de relations applicables en sélection entre l électrophorégramme des gliadines et les propriétés viscoélastiques du gluten de Triticum durum Desf. C. R. Acad. Sci. Sér. D. 287: D Egidio, M. G., Mariani, B. M., Nardi, S., Novaro, P. and Cubadda, R Chemical and technological variables and their relationships: A predictive equation for pasta cooking quality. Cereal Chem. 67: De Stefanis, E. and Sgrulletta, D Effects of high-temperature drying on technological properties of pasta. J. Cereal Sci. 12: Dexter, J. E. and Matsuo, R. R Relationship between durum wheat protein properties and pasta dough rheology and spaghetti cooking quality. J. Agric. Food Chem. 28: Dexter, J. E. and Marchylo, B. A Risposte canadesi alle esigenze di qualità di grano duro da parte dei pastifici: Tendeze passate, presenti e future. Tecnica Molitoria 47: [Available in English on Canadian Grain Commission website at Dexter, J. E., Marchylo, B. A., Preston, K. R., Clarke, J. M. and Carcea, M Comparison of the quality characteristics of some Italian and North American durum wheat cultivars. Pages in D. B. Fowler, W. E. Geddes, A. M. Johnson, and K. R. Preston, eds. Wheat protein production and marketing. Proc. Symp. on Wheat Protein, University Extension Press, University of Saskatchewan, Saskatoon, SK. Dexter, J. E., Martin, D. G. and Matsuo, R. R Effect of roll flute orientation on semolina experimental milling performance and semolina quality. Can. Inst. Food Sci. Technol. J. 21: Dexter, J. E., Matsuo, R. R., Kosmolak, F. G., Leisle, D. and Marchylo, B. A The suitability of the SDS-sedimentation test for assessing gluten strength in durum wheat. Can. J. Plant Sci. 60: Dexter, J. E., Matsuo, R. R. and Kruger, J. E The spaghetti-making quality of commercial durum wheat samples with variable α-amylase activity. Cereal Chem. 67: Dexter, J. E., Matsuo, R. R. and Morgan, B. C. 1981b. High temperature drying: Effect on spaghetti properties. J. Cereal Sci. 46: Dexter, J. E., Matsuo, R. R., Preston, K. R. and Kilborn, R. H. 1981a. Comparison of gluten strength, mixing properties, baking quality and spaghetti quality of some Canadian durum and common wheats. Can. Inst. Food Sci. Technol. J. 14: Dexter, J. E., Preston, K. R., Martin, D. G. and Gander, E. J The effects of protein content and starch damage on the physical dough properties and bread-making quality of Canadian durum wheat. J. Cereal Sci. 20: Dick, J. W. and Quick, J. S A modified screening test for rapid estimation of gluten strength in early-generation durum wheat breeding lines. Cereal Chem. 60: Dick, J. W. and Youngs, V. L Evaluation of durum wheat, semolina, and pasta in the United States. Pages in G. Fabriani and C. Lintas, eds. Durum: Chemistry and Technology. American Association of Cereal Chemists, St. Paul, MN. Feillet, P. and Dexter, J. E Quality requirements of durum wheat for semolina milling and pasta production. Pages in J. E. Kruger, R. R. Matsuo, and J. W. Dick, eds. Pasta and Noodle Technology. American Association of Cereal Chemists, St. Paul, MN. ICC Standard Methods of the ICC. Standard No 121. International Association for Cereal Science and Technology, Vienna, Austria. Irvine, G. N., Bradley, J. W. and Martin, G. C A Farinograph technique for macaroni doughs. Cereal Chem. 38: Kosmolak, F. G., Dexter, J. E., Matsuo, R. R., Leisle, D. and Marchylo, B. A A relationship between durum wheat quality and gliadin electrophoregrams. Can. J. Plant Sci. 60: Kovacs, M. I. P., Poste, L. M., Butler, G., Woods, S. M., Leisle, D., Noll, J. S. and Dahlke, G Durum wheat quality: Comparison of chemical and rheological screening tests with sensory analysis. J. Cereal Sci. 25: Littell, R. C., Milliken, G. A., Stroup, W. W. and Wolfinger, R. D SAS(R) system for Mixed Models. SAS Institute, Inc., Cary, NC. p Liu, C. Y., Shepherd, K. W. and Rathjen, A. J Improvement of durum wheat pasta making and bread making qualities. Cereal Chem. 73: MacDonald, C. E Collaborative study on wet gluten and gluten index determinations of wheat flour or meal (AACC method 38-12). Cereal Foods World 39: Marchylo, B. A. and Dexter, J. E Durum wheat and pasta quality now and into the 21st century. Pages in C. W. Wrigley, ed. Conference Proceedings of the 46th Australian Cereal Chemistry Conference Cereals 96, Royal Australian Chemical Institute, Sydney, Australia. Marchylo, B. A. and Dexter, J. E Pasta. National Research Council of Canada s Plant Biotechnology Institute PBI Bull. pp Marchylo, B. A., Dexter, J. E., Clarke, J. M. and Ames, N Effects of protein content on CWAD quality. Pages in D. B. Fowler, W. E. Geddes, A. M. Johnston, and K. R. Preston, eds. Proc. Symp. on Wheat Protein and Marketing, University Extension Press, University of Saskatchewan, Saskatoon, SK. Mariani, B. M., D Egidio, M. G. and Novaro, P Durum wheat quality evaluation: Influence of genotype and environment. Cereal Chem. 72: Masci, S., Lew, E. J.-L., Lafiandra, D., Porceddu, E. and Kasarda, D. D Characterisation of low molecular weight glutenin subunits in durum wheat by reversed-phase high-performance liquid chromatography and N-terminal sequencing. Cereal Chem. 72: Matsuo, R. R Evaluation of durum wheat, semolina, and pasta in Canada. Pages in G. Fabriani and C. Lintas, eds. Durum: Chemistry and technology, American Association of Cereal Chemists, St Paul, MN. Matsuo, R. R. and Irvine, G. N Spaghetti tenderness testing apparatus. Cereal Chem. 46: 1 6. Matsuo, R. R. and Irvine, G. N Note on an improved apparatus for testing spaghetti tenderness. Cereal Chem. 48: Matsuo, R. R., Bradley, J. W. and Irvine, G. N Effect of protein content on the cooking quality of spaghetti. Cereal Chem. 49: Matsuo, R. R., Dexter, J. E., Kosmolak, F. G. and Leisle, D Statistical evaluation of tests for assessing spaghetti-making quality of durum wheat. Cereal Chem. 59: Patterson, H. D Analysis of series of variety trials. Pages in R. A. Kempton and P. N. Fox, eds. Statistical methods for plant variety evaluation. Chapman & Hall, London. UK. Peña, R. J., Zarco-Hernandez, J., Amaya-Celis, A. and Mujeeb-Kazi, A Relationships between chromosome 1Bencoded glutenin subunit compositions and bread-making quality characteristics of some durum wheat (Triticum turgidum) cultivars. J. Cereal Sci. 19: Pogna, N., Lafiandra, D., Feillet, P. and Autran, J. C Evidence for a direct causal effect of low molecular weight subunits of glutenins on gluten viscoelasticity in durum wheats. J. Cereal Sci. 7:

10 620 CANADIAN JOURNAL OF PLANT SCIENCE Preston, K. R., Quail, K. J., Zounnis, S. and Gras, P. W No-time dough baking performance and mixing properties of Canadian Red Spring wheat cultivars using Canadian and Australian test procedures. Aust. J. Agric. Res. 48: Quaglia, G. B Other durum wheat products. Pages in G, Fabriani and C. Lintas eds. Durum: Chemistry and technology. American Association of Cereal Chemists. St. Paul, MN. SAS Institute, Inc SAS technical report P-229, SAS/STAT Software: Changes and enhancements. Release SAS Institute, Inc., Cary, NC. Steel, R. G. D. and Torrie, J. H Principles and procedures of statistics. A biometrical approach. 2nd ed. McGraw-Hill, New York, NY. 666 pp. Stroup, W. W Why mixed models? Applications of mixed models in agriculture and related disciplines. Louisiana Agriculture Experimental Station, Southern Cooperative Series Bull. No. 343, Baxton Rouge. pp Sweeny, R. A. and Rexroad, P. R Comparison of LECO FP-228 Nitrogen Determinator with AOAC copper catalysed Kjeldhal method for crude protein. J. Assoc. Off. Anal. Chem. 70: Yamada, Y. and Preston, K. R Sponge-and-dough bread: Effects of oxidants on bread and oven rise properties of a Canadian red spring wheat patent flour. Cereal Chem. 71: