Physical and textural properties and sensory acceptability of wheat bread partially incorporated with unripe non-commercial banana cultivars

2.1 Materials

Five bunches of Luvhele (Musa ABB) and Mabonde (Musa AAA) bananas at unripe stages were purchased from rural households in Thohoyandou, Limpopo province, South Africa. Wheat flour (Sasko^®) had a protein content of 11.4%, total fat (2.3%), dietary fibre (4.1%), and energy (1,456 kJ). Sugar (Selati^®), salt (Cerebos^®), margarine (Siqalo^®), water, and dry yeast (Anchor yeast^®) were purchased from a local supermarket.

2.2 Preparation of UBFs

The unripe banana fruits were detached from the bunch, washed, peeled manually, sliced into 5 mm width using a sharp knife, dipped in citric acid solution for 10 min to prevent enzymatic browning, spread evenly on a stainless steel tray, and dried in a conventional oven dryer (Model 278, digital oven, Labotec, Midrand, South Africa) at 60°C for 12 h. The dried slices were milled using a laboratory hammer mill (Retsch KG 5657, Retsch GmbH, Haan, Germany); the flour was passed through a 0.25 mm mesh sieve, packaged in a plastic bag, wrapped, and kept at room temperature (26 ± 2°C) for further use. Luvhele and Mabonde flours had a moisture content of 10.03 and 11.62%, and ash content of 1.57%, as determined by the AOAC method. The total soluble solids were 0.93 and 0.62% for Luvhele and Mabonde flours, respectively.

2.3 Preparation of breads

The dough was baked using the straight dough method [23], using different ratios of UBF (0–10%) in substituting part of the wheat flour. These included breads from Luvhele banana flour (LBF) and Mabonde banana flour (MBF). Prototype tests were performed to obtain the maximum amount of UBF that could be used in bread production. More than 20% of UBF in wheat flour produced bread with poor technological and sensory characteristics, limiting the utilisation of UBF by 20%. The dough was formulated with a mixture of wheat flour and four substitution levels of LBF and MBF (2.5, 5, 7.5, and 10% w/w). The ingredients were mixed manually for 3 to 4 min in a bowl. The dough was fermented at room temperature for 60 min and then punched back and scaled. The dough was manually moulded, placed in lightly greased bread pans, and proofed for 90 min at room temperature (26 ± 2°C) (Table 1).

The breads were baked in a preheated oven (Defy, Model DSS700, Midrand, South Africa) for 15 min at 230°C. The 100% wheat flour bread served as a control. The physical characteristics and sensory attributes of the bread samples were determined after cooling overnight at room temperature. Three batches of bread were produced, and the experiment was repeated three times for the reliability of the results.

2.4 Physical properties of breads

The loaf volume was measured using the rapeseed displacement method [24]. The bread was placed in a container of known volume, and the basin was filled to the brim with rapeseed. The bread was removed, and a measuring cylinder was used to measure the volume of the loaf. The ratio between the apparent volume of the baked breads and their mass was used to determine the breads’ specific volume. The weight of breads was measured in a semi-weighing balance (BPS 51 plus, Boeco, Germany). The loaves were weighed in triplicate. The density of bread was evaluated using the following equation: density (g/cm³) = weight of bread/loaf volume of bread.

Texture profile analysis values (hardness, chewiness, cohesiveness, and springiness) of breads were determined using a TA-XTplus texture analyser (Stable Micro System, Surrey, UK) fitted with a 25 kg load cell and an aluminium cylindrical probe with a 50 mm diameter. The texture profile of the bread was determined within 24 h of baking. The test parameters were at a constant rate: pre-test speed: 5 mm/s, test speed: 5 mm/s, and post-test speed: 5 mm/s, compression distance: 60%, and slice thickness: 20 mm. Hardness was explained as the peak force needed for the first compression. The springiness was explained as the distance between the end of the first compression and the beginning of the second compression of the slice back to its original height. Cohesiveness was measured as the ratio of the area under the second compression curve to the area under the first compression curve. Chewiness: Hardness × cohesiveness × springiness [25].

The crust and crumb colour was measured using Hunterlab LabScan XE Spectrophotometer CIELAB. The standardisation was done in compliance with the manufacturer’s instructions. The colour was then denoted using CIE-L* a* b* coordinates, where L* indicates 100 = white; 0 = black, a* (green-red component, where negative values correspond to green and positive values to red), and b* (blue-yellow component, where negative values correspond to blue and positive values to yellow). Chroma was calculated by Chroma = ( a * ) 2 + ( b * ) 2 , and the Hue angle was calculated as Hue = tan⁻¹ b a . The colour difference was calculated as ∆E = [ ( L − Lc ) 2 + ( a − ac ) 2 + ( b − bc ) 2 ] [26]. Coordinates (Lc, ac, and bc) refer to the control sample.

2.5 Sensory evaluation of breads

Panelists were verbally approached and recruited 72 h before sensory evaluation. Seventy semi-trained panelists (40 females and 30 males) in the age range of 19–50 collaborated in the sensory evaluation of bread samples. Panelists were selected from students and staff in the Faculty of Science, Engineering and Agriculture, University of Venda, South Africa, and were regular consumers of bread. Panelists were screened for critical conditions such as wheat allergy and they were given consent form prior to bread evaluation. The sensory evaluation was conducted in a sensory laboratory in individual panel booths using a hedonic scale sheet. The bread samples (with dimensions measuring 10 cm × 3 cm, including crumb and crust) were served in polystyrene plate at room temperature and marked with randomly selected three-digit numbers. The instructions were given to the panelists to evaluate the bread samples for the following attributes: colour, texture, taste, and overall acceptability. A sensory evaluation questionnaire with the nine-point hedonic scale (9 = extreme like, 5 = neither like or dislike, and 1 = extreme dislike), developed in English was distributed to panelists. Panelists were requested to clean their mouths with water in between bread tasting. Participation in the study was on a voluntary basis, and panelists were not given information about the bread samples. No monetary incentive was given to panelists for participating in the study. Permission to conduct a sensory evaluation was obtained from the University’s Ethics Committee, and a certificate of approval was issued.

2.6 Statistical analysis

Statistical analyses were carried out using SPSS 26.0 (Chicago, IL, USA). The data were analysed using one-way analysis of variance (ANOVA), and the mean values were compared using the Fisher least significant difference test (P ≤ 0.05). The mean and standard deviation of three replicates were reported for each analysis. The independent variable was the ratio of mixing the wheat flour with LBF and MBF. Dependent variables were the measured values such as physical properties, colour profile, texture, and sensory properties. The primary sensory evaluation data were subjected to statistical calculations, and, on their basis, a regression analysis was performed for each of the tested bread samples.

3.1 Physical properties of breads incorporated with unripe non-commercial banana flour

The physical (volume, weight, density, and specific volume) properties of different breads are provided in Table 2. Breads showed a significant difference (P ≤ 0.05) regarding all physical properties. The loaf volume of composite breads significantly decreased as the proportion of UBF increased, ranging from 156.40 to 123.40 cm³ (LBF) and 270.00 to 249.33 cm³ (MBF), respectively. Consumers are mostly attracted to bread because of its volume, especially higher volume, and the loaf size reduction that occurs during baking is not desirable [27]. The reduced loaf volume of UBF-added breads could be attributed to less gluten in both banana cultivars, which diluted the gluten of wheat flour [28,29]. Moreover, Nwanekezi [30] reported that adding non-wheat flour to wheat flour during breadmaking reduces the number of gluten proteins, resulting in low volume. The coagulation of gluten during baking due to the impact of heat serves as a foundation for the loaf to become fairly rigid and not collapse [31]. Similar results were reported by Almoraie [32], where the inclusion of walnut flour reduced the volume of bread, ranging from 944.2 to 645.9 cm³.

There was a significant increase in weight in all composite breads compared with the control, with values varying from 431.32 to 448.37 g (LBF) and 449.66 to 463.28 g (MBF), respectively. The addition of water during dough development and the capacity of UBF to hold water could be attributed to the increase in the weight of composite breads [28]. The weight of the bread was also attributed to factors such as moisture content and characteristics of the dough [33]. Comparably, Mudau et al. [26] found an increase in the weight of bread incorporated with finger millet flour, with values varying from 141.77 to 148.52 g. On the other hand, no significant difference (P ≥ 0.05) was observed between the weight of control and bread samples incorporated with prickly pear peel flour [34].

The composite breads had increased density compared to the control with values ranging from 0.28 to 0.45 g/cm³ (LBF) and from 0.17 to 0.19 g/cm³ (MBF), respectively. The increased density in composite bread samples could be due to factors such as the dilution of the gluten network, which might have resulted in a compact structure with less visible air pocket [24]. Gómez and Oliete [35] reported that the addition of fibre ingredients like UBF in bread production results in increased cell density. Amini-Khoozani et al. [25] reported similar results wherein the incorporation of UBF (Cavendish of Musa sp.) increased the density of wheat bread.

Specific volume is directly associated with the ability of the dough mass to retain gas production during proofing or fermentation, which in turn is due to the gluten that is available in the wheat flour. The composite breads had low specific volumes ranging from 5.88 to 4.90 cm³/g (LBF) and from 5.91 to 5.37 cm³/g (MBF), respectively. The low specific volume of composite breads might be attributed to the high amount of UBF in both banana cultivars, which competed with the dough for moisture, affecting the secondary structure of the dough’s gluten. Furthermore, the decrease in the specific volume of composite breads could be attributed to the reduced holding of carbon dioxide gas in the mixed dough [36]. Blandino et al. [37] argued that a high amount of dietary fibre could reduce bread volume because it interferes with the optional generation of gluten matrix during fermentation and baking, making bread less extensible. The fibre content of UBF varies from 6.77 to 65.76 g/100 g [38]. Similar results of decrease in a specific volume of wheat bread were observed by incorporation of other non-gluten flours such as Chinese yam at 25% [31] with values ranging from 4.47 to 3.80 mL/g and cassava at 30% [39] with values ranging from 2.5 to 1.5 cm³/g. However, the inclusion of prickly pear peel flour at 10% increased the specific volume of wheat bread [34].

3.2 Textural properties of breads incorporated with unripe non-commercial banana flours

Table 3 shows the textural properties of the breads incorporated with UBF. The hardness values of bread samples increased significantly as the UBF ratio increased, ranging from 8.73 to 9.92 N (LBF) and 8.68 to 9.96 N (MBF), respectively. When the amount of UBF increased in the flour, the resultant bread became harder, as shown by the higher hardness values in breads incorporated with both non-commercial banana cultivars. Modification in the texture of bread might be attributed to the low amounts of protein in UBF, which did not generate a network related to the gluten network, leading to a hard and more solid dough. The increased hardness of composite breads could have resulted from starch gel within inter-granular spaces that provided rigidity and resulted in the hardness of the breads [40]. Interconnection between gluten and fibrous materials might have increased the hardness of the composite breads. The hardness of composite breads might be due to the high amount of dietary fibre in UBF which interfered with the gluten matrix [41]. Moreover, the high density of bread correlates with a more compact and harder texture. The hardness of the breads is predominantly due to the amylopectin and amylose matrix which contribute to the texture of bread. Results are similar to a report by Amini-Khoozani et al. [25], whereby the inclusion of whole green banana flour improved the hardness of the wheat bread.

The chewiness of breads was associated with increased UBF, with the control sample being the chewiest (15.22 N), and breads with 10% UBF being the least chewy with a value of 11.97 for LBF and 12.02 (MBF), respectively. The weak starch–gluten network contributed to the low chewiness of composite breads. This weakening is linked to the expansion of starch granules and the gluten network’s water absorption during dough production [42]. The results align with the report of Li et al. [31], wherein the inclusion of Chinese yam flour (5–15%) decreased the chewiness of bread ranging from 170.90 to 155.38 g.

The cohesiveness of breads significantly decreased (P ≤ 0.05) as UBF increased in the formulation, with values ranging from 0.80 to 0.71 (LBF) and 0.84 to 0.71 (MBF), respectively. The reduction in cohesiveness of the composite breads might be associated with high integration of crumb structure, as reported by Salinas et al. [43]. The decreased cohesiveness values of composite breads show that the crumb is more prone to rupture. Similar results were obtained by Thakaeng et al. [44], wherein the incorporation of UBF (Cavendish spp.) decreased the cohesiveness of the bread with values ranging from 0.55 to 0.27 g s.

The more wheat flour was replaced with UBF, the less springiness the bread crumb, with values ranging from 0.88 to 0.77 cm (LBF) and 0.86 to 0.77 cm (MBF); Feili et al. [45] mentioned that the association between gelatinised starch and gluten dough, which contributes to the springiness of the dough, can make a constant sponge structure of the bread upon heat treatment or baking. A low level of gluten in UBF of both banana cultivars contributed to the dough’s lower capacity to hold gas, resulting in the low springiness of composite breads. It is also possible that the interaction of protein and polysaccharides in the gluten network might have expedited starch entrapment, thereby building up the dough’s structure and reducing the bread’s texture [46]. It was expected that bread (control) with higher gluten content would be more elastic. The interactions between gluten dough and gelatinised starch, which can create a sponge structure due to high temperature, are thought to cause high springiness values [34]. The hydration of non-gluten flours, like UBF, has weak foaming characteristics and might have resulted in the formation of fewer pockets during the fermentation process, leading to a low springiness in the composite breads [47]. Ho et al. [48] found similar results wherein the cohesiveness and springiness of breads incorporated with banana pseudostem flour decreased.

3.3 Colour profile of breads incorporated with unripe non-commercial banana flour

The crust and crumb colour properties of control and composite breads are depicted in Table 4, and Figure 1 shows the pictures of the formulated bread samples. The L* values of the crumb breads incorporated with UBF were significantly lower compared with the control (74.70), ranging from 61.00 to 66. 78 (LBF) and 62.62 to 65.55 (MBF), respectively. The addition of UBF significantly affected crumb lightness, as observed visually that the crumb of bread samples containing UBF was darker while the control sample had a white colour. Zuwariah and Azizh [49] reported that the crumb of bread composited with banana flour turns darker than wheat bread. Crumb lightness can be modified by the nature and particle size of the milled flour [50]. Bananas are very rich in polyphenols; nonetheless, the darkening influence might be due to the phenolase activity in the banana while milled into flour [36]. Furthermore, the decreased crust lightness of composite breads could be due to chemical reactions which occur in bread crust, such as Maillard browning caused by a reaction between wheat protein, added sugars, and caramelisation, which are influenced by the distribution of water and reaction of added sugars [51].

Figure 1

White bread formulated with (a) Luvhele flour and (b) Mabonde flour.

Similar to crumb, the crust lightness was reduced as the percentage of UBF increased, with values varying from 58.32 to 52.89 (LBF) and 58.26 to 53.21 (MBF), respectively. No significant difference was observed between the crust of control bread and the crust of breads containing 2.5% of UBF. The decrease in the a* and b* values of composite breads characterised the colour shift from crumb to the crust. The availability of phytochemicals such as carotenoids, polyphenols, and xantophylls that are related to the colour characteristics of banana fruit might be the cause of the decrease in crust lightness of composite breads [52]. On the other hand, the majority of bread samples, except for breads incorporated with 10% UBF, were within the acceptable range of L* value for the bread crust, which is 54–62 [39]. Parafati et al. [34] also reported low L* values in bread enriched with 20% prickly pear peel flour, and values ranged from 66.6 to 47.9 for crust and 65.5 to 57.9 for crumb.

The redness (a*) and yellowness (b*) of the crumb and crust of breads decreased with the incorporation of UBF, as shown in Table 3. The low b* values of composite breads might be due to oxidation of carotenoid, which is naturally available in wheat and UBF flours. The loss of carotenoids during the production steps of bread compared to other bakery products has been reported [53]. Liu et al. [54] observed similar results wherein the inclusion of yam flour decreased the b ^* value of breads. Chroma (C*) and hue angle (h°) were determined to have a complete characterisation of the colour of breads. Chroma is mainly associated with colour purity. Control bread showed a significantly higher purity of yellowness in crumb and crust than breads containing UBF. It was noted that the inclusion of UBF reduced the hue value in the red quadrant. However, the obtained hue was positive, which reflected a yellow-orange hue. The decrease in hue values in the red quadrant with an increase in the ratio of UBF was notable in both crust and crumb.

The colour difference (ΔE) of the crust and crumb of composite breads increased with the increasing levels of LBF and UBF. The increase in the ΔE value of composite breads might be attributed to the composition of ingredients and generation of red pigments because of the Maillard reaction during baking [55]. The ΔE values of all composite breads were higher than three, suggesting an obvious colour difference for the human eye [56].

3.4 Sensory properties of breads incorporated with unripe non-commercial banana flours

The mean scores for sensory parameters, colour, aroma, taste, texture, and overall acceptability, were obtained from a nine-point hedonic scale and are shown in Table 5. There was no significant difference (P ≥ 0.05) between control bread and composite breads up to 7.5% UBF regarding aroma, taste, and overall acceptability. On the other hand, a significant difference was observed among breads with regard to colour and texture characteristics. Moreover, the consumer acceptance of wheat breads was affected by an increased ratio of 10% UBF in all sensory attributes. Ajala et al. [57] reported similar results wherein the addition of plantain flour at 50% resulted in bread with dark brownish colour, which was not well accepted by consumers. A high ratio of UBF resulted in a negative effect on consumer colour acceptance, as consumers showed neither like nor dislike for breads. The overall acceptability in control bread was 7.69, while in 10% UBF breads it was 7.31 (LBF) and 7.27 (MBF), respectively. This decrease in overall acceptance was due to the low score of mouth feel of breads made from a different ratio of UBF. A similar observation of consumers’ lower overall acceptance of breads was reported in breads incorporated with plantain and soybean flour [49].

Texture attributes contribute strongly to consumer freshness perceptions. Control samples and breads with 2.5% UBF were rated best in terms of texture, and this might be due to gluten in wheat flour, which contributed to the development of the elastic dough. It is well known that wheat flour’s unique baking properties enable it to produce bread with a desirable texture [50]. The texture of bread is associated with the water-holding ability of functional components like starch, protein, and other hydrocolloids. An increase in the UBF ratio increased the hardness of breads (Table 3), which could be affected by the crumb cell’s size or structure and is also greatly influenced by the amount and characteristics of dough’s gluten protein [51]. The results suggest that overall, UBF of Luvhele and Mabonde banana cultivars at a level of 2.5 to 7.5% (w/w) might be incorporated in bread production without affecting sensory parameters such as aroma, taste, and overall acceptability.