The bread making quality of Finnish spring wheats-a proposal for classification

The purpose of the present study was to provide bread making quality criteria forspring wheat (Triticum aestivum L.) grown in Finland. The bread making quality of a total of 101 wheat samples was related to flour protein concentration and dough quality determined by the farinographvalorimeter value. Nine quality classes of wheat, similar to the German system of wheat classification, were used as a basis. Because Finnish wheat breeding and cultivation primarily aim at producing grain for bread making, it was not considered necessary to define the quality classes of wheat unsuitable for bread making. Therefore, the number of quality classes were restricted to five: (class 1) medium, (class 2) medium to high, (class 3) high, (class 4) high to very high, and (class 5) very high baking quality. The influence of the environment on the quality traits should be taken into account by comparing the cultivars with a representative control cultivar.


Introduction
The baking quality of wheat is dependent on both genetic and environmental factors.By selecting wheat cultivars with an optimal high molecular weight glutenin subunit composition the bread making properties can be improved (Peltonen et  al. 1993).However, a genetically good quality wheat cultivar can be of poor industrial quality if environmental conditions and management practises are not proper (Salovaara 1986 a, Peltonen   1992).However, before the industrial baking quality of wheat cultivars grown in different environ- ments can be estimated, wheats should be divided into classes on the basis of their end-use.One approach would be to distinguish the weakness or incompleteness of the flour material and to exclude materials not meeting the required standards or specifications.Another approach would be to have a certain number of quality classes according to the type of bread and baking processes used.For ex- ample, the American Soft Red (SR) and Soft White (SW) wheat types are intended for baking cakesand biscuits.The quality of Durum wheat is suitable for semolina and pasta production, whereas Hard Red Winter (HRW) and Hard Red Spring (HRS) wheats are suitable for bread making.
In Finland there is no well defined quality classi- fication system of Finnish wheat cultivars available.Regarding industrial French breadmaking from spring wheat in Finland, the relationships be- tween analytical quality tests on a laboratory scale and performance in industrial baking processes have been studied (SALOVAARA 1986 b).In addi- tion, some requirements of good quality wheat flour such as protein and gluten content, and dough quality have been suggested (Kulhomäki and Sa- lovaara 1985, Laukkanen 1991).A classifica- tion system based on an index of 22 quality parame- ters (including grain, flour, dough and bread property parameters) has been suggested (Huttunen et   al. 1980).It is, however, complicated to determine what to emphasize in each quality parameter to form such an index.On the other hand, quality classification of Finnish wheats based on quality parameters that can describe the requirements of wheat flour to be used for the Finnish type ofbread and for the bread-making processes used in Finland will be ofadvantage both for the breeding work and for the industrial management of the wheat crop.The variation in quality due to environmental fac- tors can be taken into account by use of a well- known control cultivar (Peltonen 1992).
This study was carried out to determine bread making quality criteria for classification of wheat cultivars into quality classes.The proposed classi- fication system of wheat is based on the German wheat quality classes (Weizensorten und Backqualität 1990).In that classification system variation in quality caused by the environment is estimated by using the spring wheat cultivar 'Tur- bo' as control.The wheat cultivars are divided into nine quality classes according to their bread vol- umes in comparison to the control cultivar.The control cultivar represents medium to high baking quality.

Quality evaluation
Replicated yield samples were harvested at the caryopsis hard stage.Grain samples were pooled and mixed thoroughly for quality analyses.The following quality parameters were determined: flour protein concentration (%, FPC) by the method 46-11 (AACC 1985), wet gluten content (%, WGLUT) of flour and Zeleny sedimentation vol- ume (ml, ZEL) using standards 137 of the ICC (1982) and 116 of the ICC (1972), respectively, dough water absorption (%, FABS) by the method 54-21 of the AACC (1982), and the Brabender Farinograph valorimeter value (FVALO) by the method 54-21 of the AACC (1982).FVALO is an empirical single-figure quality score based on dough development time and tolerance to mixing derived from farinogram by means of a special template supplied by the manufacturer of the farinograph equipment.Dough stretching charac- teristics were determined with the Brabender Extensograph (standard 114, ICC 1972).Dough ex- tensibility (EXT) as resistance ratio of maximum resistance to extensibility was calculated according to FULLINGTON et al. (1987) to describe dough stretching with a single-figure.
Test baking was made from 250 g of flour ac- cording to the long fermentation process (Huttunen et al. 1980).Dough with optimum water absorption (%) was mixed for 5 minutes in a farinograph mixing bowl.The dough was allowed to pre-ferment for 90 minutes at room temperature (20°C).Remoulding of dough was done twice after 45 minutes of pre-fermentation.After pre-fermentation, the dough was allowed to rise in a baking pan for 90 minutes in a fermentation cabinet at 80% relative humidity and 32°C.The final baking was done in a laboratory oven at 230°C and the baking time was about 20 minutes.The loaf volume (ml,  LV) was measured on the baking day by the rape seed displacement method.The loaves were left overnight at room temperature before subjective crumb texture assessment.Loafresilience (LR) was measured from a piece of loaf (5.0 x 5.0 x 5.0 cm) by depressing the loaf texture 3.0 cm with a lead cube (1 kg) during 5 seconds.Five seconds after the lead cube was released the reversion of loaftexture was measured in centimeters.The reversion was determined using a scoring system of 0-10 (poor to good) as follows: Pore size Symmetry of crumb texture 0.3 1 2 3 4 0.4 0.5 0.6 5 0.7 6 0.8 7 0.9 8 1.0 All the quality analyses and baking tests were done at the Grain Research Laboratory of the State Granary, Helsinki.

Reversion cm
Loaf resilience (LR) 2.0 0 2.1 -2.3 0.51.0 2.4 -2.6 1.5 -2.0 Classification of wheat cultivars 2.7 -2.9 2.5 -3.0 3.0 -3.2 3.5 -4.0The German classification system of wheat quality (Weizensorten und Backqualität 1990) was applied to classify the quality ofFinnish spring wheat culti- vars.The German system separates wheat varieties into nine different classes according to a control cultivar.The quality of the control cultivar was based on loaf volume and it was assigned to class 6 (Table 1).In the present study, cultivar 'Ruso' was used as a control, because it was the only cultivar which was grown during the whole study period.The cultivar 'Ruso' was used as a control also in the state official experiments during the study period (Mustonen et al. 1986).Each season the quality properties were compared to the control cultivar 'Ruso' (Qi/Qc x 100, where Qi is the actual quality property of a cultivar or breeding line and Qc is the quality property of the control cultivar).One way analysis of variance (ANOVA) and the MSTAT 4.0 (MSTAT 1989)program were used to calculate the mean values of the quality properties for each class.The values obtained were coded with numbers from 1 to 9 to indicate which Finnish quality class corre- sponds to the German quality class (Table 1).
Simple correlations between quality parameters were computed using linear regression analysis.
recorded, but the correlation coefficients were weak.Therefore, for further analysis, we concen- trated on FPC, ZEL, FVALO, LV, and LAPP values in order to provide a classification system.The FPC, ZEL, FVALO, LV and LAPP values were classified in comparison to thecontrol cultivar 'Ruso' (Table 4) using the ranges (from <75.6 to >110.5) of quality classes (1-9) described in Table 1.ANOVA was used to form the corresponding range for each class from the means of the quality parameters.The classification of wheats according to FPC indicated that in class 6 (medium to high baking quality), the corresponding FPC values varied from 11.6 to 12.0% (Table 4).The limit for class 9 (very high quality) was determined by the FPC value 13.3%.Laukkanen (1991) pointed out that a protein concentration of 13% would be optimal for the industrial uses in Finland.All wheats represented at least medium baking quality (class 5) when compared with the control cultivar 'Ruso'.
The classification based on ZEL showed that the range of ZEL values in class 6 is 40.9 -44.9 ml.According to Laukkanen (1991) a ZEL value of 30 would be optimal for the industrial uses in Fin-land.The lowest ZEL values in this study were above 38 ml, indicating medium baking quality (class 5).More frequently the ZEL values fell into class 9.
The classification based on FVALO showed that the limit of very low baking quality (class 1) was 38.6.The range of FVALO values was 48.8 -51.3 in class 6 (medium to high quality).An FVALO value above 56.4indicated a class 9 category.
According to LV values, the baking quality varied from class 4 to class 9. On average, the baking quality classified according to LV indicated classes 6 and 7.In class 6, LV values varied from 1438 to 1513 ml.A LV value above 1663 ml indicated very high quality (class 9).
The LAPP values divided the quality into all nine classes, although a great deal ofthe LAPP belonged to classes 8 and 9. LAPP over 607.8 was an indicator of very high baking quality, a LAPP score less than 415.8 being a characteristic of class 1 (very low baking quality).
demonstrate their dependence on each other (Table 5) The strongest correlation was found between FPC and FVALO (r=0.55).The other correlations were poor (r<0.42).Therefore, classifications based on FPC and FVALO were tested against 13 independent wheat cultivars grown during 1989-90, and their mean values were calculated (Table 6) Our results showed that cultivars 'Benito' and 'Katepwa' -both FIRS wheat types -were categorised into class 9 (very high baking quality) as compared with control cultivar 'Ruso' (class 6).HRS wheat types are commonly imported to com- pensate for the poor quality of domestic grain.Cul- tivars 'Heta', 'Satu' and 'Tähti' showed high to very high baking quality (class 8), and 'Polkka' high quality (class 7).Medium to high baking quality (Class 6) included control cultivar 'Ruso' and cultivars 'Apu', 'Reno', and 'Runar', while culti- vars 'Kadett', 'Tapio' and 'Luja' belonged clearly to the lowest class (class 5 = medium quality).These results confirmed the earlier reports on the baking quality of these wheat cultivars in industrial baking (Salovaara 1986 a), in breeding pro- grammes (Juuti 1988), and in commercial wheat production (Huttunen and Lilja 1990).

Discussion
The evaluation of Finnish spring wheat cultivars indicated that the bread making quality could be 9.0 Katepwa (AC) 9 9 9.0 Heta (Hja)  9 7 so Satu (WW) 8 8 8.0 Tähti (Jo)   7 9 8.0 Polkka (Sv)   7 8 7.5 Apu (Jo) 8 5 6.5 Reno (NLH)  6 7 6.5 Runar (NLH)  6 6 Our results are in agreement with Baker et al. (1971) and FOWLER and de la ROCHE (1975) who concluded that FPC in addition to farinograph dough quality were effective indicators of baking quality, thus replacing baking tests in quality breed- ing.Varis and Juuti (1975) showed that both farinograph dough quality and bread volume were improved by N fertilization strategies, raising the protein concentration in winter wheats.SA- LOVAARA (1986b) reported that farinograph dough quality and wet gluten content were the best indic- ators of industrial French bread making quality of spring wheat.However, our results indicated that wet gluten content (WGLUT) correlated poorly (Table 3) with FPC, dough quality (FABS,   FVALO, EXT) and loaf quality (LV, LR.LAPP)   parameters.In the present situation in Finland, FPC would be more a suitable indicator of quality than WGLUT.The pricing system of wheat, too, is based on protein concentration, in addition to starch quality and test weight (Viljaliite 1992).It is recom- mended to combine both FPC and FVALO in the pricing system in order to give a better under- standing of the quality of flour.This is mainly due to high protein concentration not necessarily being linked to high dough quality.Peltonen et al.  (1993) have indicated that farinographic data de- scribes well the quality of gluten of the cultivars which have different combinations ofhigh molecular weight glutenin subunits.Thus, to optimize the system, FVALO should be included to contribute information also about genetic factors such as gluten composition.
The correlation between loaf volume and the other quality traits in the classes proposed was low (Table 5).This may indicate that the test baking procedure used in Finland (Huttunen et al. 1980)  is not suitable for predicting the bread-making quality of flour for the particular type of bread and the baking processes used by the Finnish bakeries.In a Swedish wheat breeding programme, the bread making potential ofcultivars is tested at two differ- ent dough mixing intensities before release, thus indicating the suitability of a cultivar for different bread making processes (Svensson 1987).In fu- ture, it would be important to focus on the different test baking procedures in relation to industrial bak- ing performance.In Finland, optimizing the mixing time instead of using constant mixing for 5 minutes may give improvements.It could then be possible to provide quality classes for wheat based on loaf volume only, as does the German classification system (Weizensorten und Backqualität 1990).
Because wheat breeding and cultivation in Finland primarily aim at producing grain for bread making (Kivi 1969), quality classes lower than me- dium baking quality (classes 1-4, Table I) are not necessarily required.Therefore, we suggest that the classification system of wheat in Finland should contain five classes.This idea is also supported by the results in Table 6, whereby the commercial Finnish wheat cultivars were divided into five classes.Quality classes and the corresponding range in FPC and FVALO values of the recom- mended classification system are given in Table 7.
Wheat of the highest quality classes (classes 4 and 5) with strong gluten could be used to compensate for the baking quality of wheats in class 1.In turn, wheats in classes 2 and 3 could be used directly for bread making.The class ranges presented in Table 7 are depended on the control cultivar used.Obvi- ously, the cultivar 'Ruso' is now too old (released in 1967) to be used as a control.Later released cultivars are of better bread-making quality than 'Ruso' (Table 6).Therefore, another control culti- var than 'Ruso' should be chosen in the future.

Table 1 .
German quality classes for wheat based on loaf volume.Comparison has been made to a control cultivar(Weizensorten und Backqualität 1990).
*Range of quality classes taken from the German system.

Table 3 .
Significant phenotypic correlations between quality characteristics.

Table 5 .
Significant correlations between qualities proposed in Table4.

Table 6 .
Division of spring wheat cultivars grown in 1989-90 at Hankkija Plant Breeding Institute into quality classes according to flour protein concentration(FPC)and farinograph valorimetric value (FVALO).

Table 7 .
Proposal for quaky classes in Finland based on flour protein concentration (FPC) and farinograph valorimetric value(FVALO).