Microstructure of Flours
The scanning electron micrographs of WF and AF fractions are presented in Figure 1. As can be seen, the amaranth starch granules were round, oval, and irregular in shape.
Fourier Transforms Infrared Spectrometry Analysis of Flours
The spectra of the wheat flour and amaranth flour particle size are shown in Figure 2. The signals heights of samples spectra regarding different types of bonds stretching on the spectrums of wheat flour and amaranth fractions were interpreted according to literature data26.
Physico-chemical properties of composite flours
The physico-chemical properties of composite flour formulations are presented in Figure 3. The results revealed that the nutrient compositions were markedly influenced by the AF particle size as well as WF replacement level.
As per the graph, moisture (Fig. 3a) was decreased when PS become finest and AF replacement rose, being lower than in the control. The effect of PS and WF replacement was considerable on the protein content of the formulation (Fig. 3a). The finest AF particles size (S and M) were led to an enhancement of composite flour protein content with raising the WF replacement, whilst the larger fractions (L) decreased the protein content of these flours. The ash content of wheat-amaranth composite flour (Fig. 3b) increased proportionally with the successive replacement of the WF and with a decrease in particle size. The lipid content of formulated flours (Fig. 3b) was significantly affected when the level of replacement increased, with all samples presenting higher values than the control. Regarding the particle size influence, the lipid content in composite flours increased in the following order: S < L < M. Additionally, WF replacement with AF led to a significant decrease in the carbohydrates content of composite flours (Fig. 3c), being successively decreased with the decreased particle size. The color parameters of composite flour were measured on lightness (L*) and chroma (C*) color scale, and results are given in Figure 3d, being observed significant differences (P < 0.05). significant increase in the C* parameter with the increase of WF replacement and with a decrease of particle size was observed.
Dynamic dough rheological properties
Dynamic dough rheological properties were significantly (P ˂ 0.05) influenced by the AF particle size and WF replacement level (Table 1).
Table 1
Elastic and viscous moduli, loss tangent, maximum gelatinization temperature, and creep-recovery compliance of bread samples with different amaranth flour particle sizes and wheat flour replacement levels. FI: AF particle size; FII: AF replacement level; mean followed by the same alphabets in each column are not significantly different (P > 0.05); the first (a-c) and second (x-w) letter in each column indicates particle size and replacement level, respectively. G': elastic modulus; G'': viscous modulus; tan δ: loss tangent; Tmax: maximum gelatinization temperature; Jcmax, Jrmax: maximum creep-recovery compliance.
Sample
|
G'
(Pa)
|
G''
(Pa)
|
tan δ
(adim.)
|
Tmax
(°C)
|
Jcmax
(10‒5 Pa‒1)
|
Jrmax
(10‒5 Pa‒1)
|
Control
|
26370.00±70.15a
|
9488.00±60.00a
|
0.3598± 0.00c
|
82.74± 0.49a
|
24.46±0.04bc
|
16.62±0.00w
|
AL_5
|
33400.00±3730.00by
|
11635.50±302.40bxy
|
0.3517±0.01byz
|
81.94±0.04cz
|
20.45±0.91abxy
|
13.30±0.05az
|
AM_5
|
33010.00±2970.00by
|
11407.00±1113.00bxy
|
0.3465±.0.00ayz
|
79.77±0.91bz
|
13.54±0.07axy
|
8.63±0.04cz
|
AS_5
|
23245.00±1785.00axy
|
8066.00±805.00axy
|
0.3485±0.00byz
|
79.14±0.26cz
|
32.56±2.53cxy
|
20.00±0.05bz
|
AL_10
|
27350.00±1250.00bxy
|
9977.00±383.00bxy
|
0.3649±0.00bzw
|
79.51±1.46cy
|
16.63±2.14abx
|
9.67±0.05ax
|
AM_10
|
30610.00±830.00bxy
|
10305.00±235.00bxy
|
0.3367±0.00azw
|
77.96±0.05by
|
16.30±2.57ax
|
13.21±0.05cx
|
AS_10
|
28510.00±1480.00axy
|
9943.50±406.00axy
|
0.3490±0.00bzw
|
79.51±0.78cy
|
24.68±6.18cx
|
19.50±0.05bx
|
AL_15
|
34345.00±3005.00by
|
11925.00±1155.00by
|
0.3478±0.00by
|
78.79±0.26cy
|
22.71±4.04abxy
|
6.67±0.27ay
|
AM_15
|
34625.00±155.00by
|
11505.00±45.00by
|
0.3322±0.00ay
|
78.54±0.46by
|
16.95±3.57axy
|
13.44±0.05cy
|
AS_15
|
24450.00±1810.00ay
|
8427.00±842.00ay
|
0.3447±0.00by
|
79.34±0.17cy
|
23.93±0.30cxy
|
18.17±0.05by
|
AL_20
|
58840.00±2280.00bz
|
19010.00±850.00bz
|
0.3230±0.00bx
|
78.97±0.21cy
|
23.29±2.40abxy
|
11.24±0.05ay
|
AM_20
|
45375.00±825.00bz
|
14405.00±215.00by
|
0.3175±0.00ax
|
78.43±0.04by
|
18.15±1.40axy
|
13.54±0.05cy
|
AS_20
|
29660.00±100.00az
|
9786.00±49.00az
|
0.3299±0.00bx
|
80.61±0.48cy
|
25.50±0.35cxy
|
16.77±0.05by
|
Two-way ANOVA p value
|
FI
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P=0.0410
|
P<0.0001
|
FII
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
F IxFII
|
P<0.0001
|
P<0.0001
|
P=0.0400
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
The elastic modulus was significantly (P ˂ 0.001) higher when WF replacement increased in comparison with control, being highest when large fractions replaced WF. All dough samples, corresponding to a predominant viscoelastic nature behavior, G' > G''. Significant differences (P < 0.01) on samples' loss tangent (tan d) were recorded with the increase of WF replacement with above 10% AF, leading to a gradual decrease of this parameter, while the replacement between 5-10% did not have a significant effect on tan d. Regarding PS, significant differences were registered only between medium particle size and the other two PS (L and S). Particle size influenced Tmax due to WF replacement with AF in comparison with the control, but differences between particle sizes were observed only at the samples where was incorporated medium PS. Replacement level significantly influenced this parameter when was up to 10%. Maximum creep compliance (Jcmax) presented higher values in samples where was replaced with small PS, followed by large PS, while in samples where was incorporated medium PS, the creep compliance was lowest. The same trend was observed for all the replacement levels, which led to higher dough extensibility. Maximum creep recovery (Jrmax) was influenced significantly by both factors, PS and replacement level.
Usually, this parameter tends to decrease when the WF replacement level with AF increased, and regarding particle size, it decreases in the following order: M ˂ L ˂ S.
Bread evaluation
Physical properties
Bread technological parameters were significantly influenced by the PS and WF replacement with AF. As Figure 4a shows, the bread volume was lowest at the sample in which WF was replaced with small PS, and highest at bread with medium PS, followed by large PS. Regarding replacement level, the decrease in bread volume was more accentuated for the samples with higher levels of AF. The porosity (Fig. 4a) of the all bread-based on AF fractions, at replacement between 5-15% was higher than wheat flour bread, whilst, the 20% replacement significantly decreased bread porosity. Particle size influences crumb porosity as following trend: M, L, and S.
Bread crumb and crust colour
The lightness (L*) and chroma (C*) for the bread crust-crumb varied depending of AF particle sizes and replacement levels of WF (Fig. 4b,c). Crust lightness (L*) increased gradually with the decrease of PS and decreased with the AF replacement level. Regarding crumb lightness, followed the same trend, depending on formulated factors, PS, and replacement level. Crust chroma presented an increase with the raise of replacement level, while PS lead to an increase of bread crust chroma in the following order: L ˂ M ˂ S. Crumb C* presented an increase compared to control when replacement level of WF raise, whilst regarding particle size influence, C* tend to increase with the decrease of PS.
Textural parameters
Bread texture parameters have a direct influence on consumer perception and choice. The effect of PS and WF replacement level with AF on bread texture shows that both factors significantly affected all the textural parameters (Table 2). Crumb firmness of bread regarding PS, increased in the following order: M ˂ Control ˂ L ˂ S. Bread firmness also increased gradually with the increase of WF replacement level. Crumb springiness and cohesiveness were not affected by the AF particle size but presented significant differences between samples with 5-10% and samples with 15-20%. Crumb cohesiveness and chewiness significantly increased (P < 0.05) in all breads being higher than control, except for bread with 5 and 10% medium PS which presents lower values than control bread.
Table 2
Textural parameters of bread samples with different amaranth flour particle sizes and wheat flour replacement levels. FI: AF particle size; FII: AF replacement level; mean followed by the same alphabets in each column are not significantly different (p > 0.05), the first (a-c) and second (x-w) letter in each column indicates particle size and replacement level, respectively.
Sample
|
Firmness
(N)
|
Springiness
(adim.)
|
Cohesiveness
(adim.)
|
Gumminess
(N)
|
Chewiness
(J)
|
Control
|
7.71±0.04a
|
1.3458±0.19c
|
0.7664±0.02c
|
602.30±13.92a
|
602.30±13.92a
|
AL_5
|
8.19±0.11bx
|
1.2475±0.00ay
|
0.8578±0.01bw
|
623.74±8.77cx
|
623.74±8.77cx
|
AM_5
|
6.06±0.02ax
|
1.1544±0.00ay
|
0.7411±0.00aw
|
458.35±4.44ax
|
458.35±4.44ax
|
AS_5
|
9.19±0.73cx
|
1.2073±0.00ay
|
0.8850±0.00cw
|
678.22±5.85bx
|
678.22±5.85bx
|
AL_10
|
12.10±0.02by
|
1.1319±0.01ay
|
0.7305±0.00bz
|
894.00±7.14cy
|
894.00±7.14cy
|
AM_10
|
6.19±0.02ay
|
1.0527±0.05ay
|
0.7197±0.00az
|
454.64±3.81ay
|
454.64±3.81ay
|
AS_10
|
12.10±0.02cy
|
1.1453±0.03ay
|
0.8650±0.00cz
|
743.70±43.98by
|
743.70±43.98by
|
AL_15
|
21.12±0.29bz
|
1.0000±0.00az
|
0.6930±0.00by
|
1445.94±5.56cz
|
1445.94±5.56cz
|
AM_15
|
12.64±0.69az
|
1.0000±0.00az
|
0.6930±0.00ay
|
764.03±4.84az
|
764.03±4.84az
|
AS_15
|
21.57±0.40cz
|
1.0015±0.00az
|
0.6930±0.01cy
|
1070.65±0.65bz
|
1070.65±0.65bz
|
AL_20
|
28.43±0.67bw
|
0.9985±0.00ay
|
0.6732±0.00bx
|
1950.93±18.68cw
|
1950.93±18.68cw
|
AM_20
|
28.43±0.67aw
|
0.9980±0.00az
|
0.6732±0.00ax
|
938.21±2.77aw
|
938.21±2.77aw
|
AS_20
|
32.89±0.02cw
|
0.9988±0.00az
|
0.6400±0.02cz
|
1112.50±7.50bw
|
1112.50±7.50bw
|
Two-way ANOVA p value
|
FI
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
FII
|
P<0.0001
|
P=0.2600
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
FIxF II
|
P<0.0001
|
P=0.4460
|
P<0.0001
|
P<0.0001
|
P<0.0001
|
Sensory evaluation
Sensory evaluation results revealed some improvements regarding overall acceptance, crust surface, taste, crumb structure, and smell for bread which contains medium and large particle sizes up to 10%, compared to control (Fig. 5). For bread with a small AF particle size, for all replacement levels were observed a decrease in sensorial acceptance in comparison with bread control.
Relations between assessed characteristics
By applying Pearson's correlation analysis between assessed characteristics, a series of siginificant (P ˂ 0.05) correlation coefficients (0.56 ˃ r ˂ 0.98) was found. Flours humidity was strongly positive correlated with loss tangent (r = 0.67), bread springiness (r = 0.78), gumminess (r = 0.61), bread volume (r = 0.95) and bread overall aceptability (r = 0.65), bread taste (r = 0.78) and crumb structure (r = 0.81). Instead this physical parameter of flour was nevatively associated with bread gumminess and chewiness (r = ‒ 0.58). In this way, it seems that flour humidity is a good indicator for flour quality which has direct correlation with dough and bread properties. High positive correlation were found between flour lipids and bread firmness (r = 0.66) and gumminess (r = 0.67), and elastic modulus (G') (r = 0.58), while lipids are negatively associated with loss tangent (r = ‒ 0.70), bread volume (r = ‒ 0.85), bread springiness (r = ‒ 0.82), bread cohesiveniss (r = ‒ 0.72), and with all sensorial characteristics: structure (r = ‒ 0.80), smell (r = ‒ 0.73), overall aceptability (r = ‒ 0.53). Regarding the bread texture, it was found significant (P ˂ 0.05) correlation with dough rheology and consumer accpetance of final product. Bread firmness is positive correlated with elastic modulus (r = 0.67) and viscous modulus (r = 0.69), while with bread volume (r = ‒ 0.58), bread porosity (r = ‒ 0.70), and bread oberall aceptability (r = ‒ 0.56) is negatively associated. Bread springiness is positive correlated with loss tangent (r = 0.56), bread volume (r = 0.67), and bread structure (r = 0.74). Bread gumminess and chewiness are positve associated with elastic modulus (r = 0.59), viscous modulus (r = 0.60), while in a negative way is associated with bread volume (r = ‒ 0.65), bread porosity (r = ‒ 0.74), and sensory characteristics of bread: overall aceptability (r = ‒ 0.64).
The principal component analysis (PCA) was used to highlight the similarities or dissimilarities between the determined characteristics (Fig. 6). The loadings of the studied characteristics on the first principal component, PC1 (49.26%), and the second principal component, PC2 (19.06%) described 68.32% of the total variance.
The dough maximum gelatinization temperature (Tmax) and creep-recovery compliance (Jcmax - Jrmax) have a small contribution to the data variation, as is suggested by their position on the graphic, close to the center. Instead closeness of single parameters for example flour humidity, loss tangent (tan δ), bread volume, and sensory characteristics confirms a tight pair correlation, as well as the association between elastic and viscous moduli (G', G''), bread firmness, gumminess, and chewiness. The PC1 was associated with flour humidity and lipids, tan δ, bread volume and porosity, bread textural parameters, and bread sensory characteristics while PC2 was associated with flour protein, ash, and carbohydrates, elastic, and viscous moduli (G', G''). It can be remarked a high opposition between protein-ash and carbohydrates, bread firmness and porosity, viscoelastic moduli and bread springiness. A strong correlation was observed between wheat flour bread and bread with medium and large PS when WF was replaced at a 5% level (AL_5 and AM_5) and 10%, respectively (AL_10). Composite flour with a 20% AF of medium and small fractions were associated with lipids, ash, and protein, whilst samples with large fractions (AL_20) was associated with viscoelastic moduli, bread firmness, gumminess, and chewiness.