Physicochemical and sensory attributes of scones made from wheat–taro (Colocasia esculenta)–lentils (Lens culinaris) composite flour

The effect of blend proportion on the functional properties of wheat–taro–lentil composite flour and the physicochemical and sensory quality of scones was studied. The water absorption capacity (WAC), oil absorption capacity (OAC) and bulk density (BD) significantly (p < .05) increased whereas the swelling index (SWI) decreased with increase in the proportion of lentil flour. Batter viscosity significantly (p < .05) decreased from 1972 to 1968 cP with increase in the proportion of lentil flour. The crude protein content increased from 5.81% to 9.83% and the ash content decreased from 4.61% to 2.29% with increase in the lentil proportion. Specific volume, volume and height decreased with increase in the proportion of lentil. The baking loss increased from 5.36% to 13.56% with increase in taro flour proportion. The L*, a* and b* colour values and sensory attributes were also significantly affected by the blend proportions. Based on the results, lentils can be incorporated up to 30% and taro up to 10% for a protein‐enhanced and acceptable product.

Composite flours improve the nutrition, reduce the cost of product and diversify the utilization of locally available raw materials (Fenn et al., 2010;Noorfarahzilah et al., 2014;Olakanmi et al., 2022).
Legumes including lentils have been widely used to improve the protein content of baked products in general due to their relatively high protein and fibre content (Manyatsi et al., 2020;Olakanmi et al., 2022).Lentils are nutrient-dense and exhibit a good balance between protein and carbohydrates.Protein, carbohydrate and dietary fibre content of lentils is higher than that of wheat and other selected pulses (Kaale et al., 2022).Lentil is a rich source of protein (21% to 31%) with a balanced amino acid profile, plentiful low digestible carbohydrates, a range of human essential micronutrients and prebiotic carbohydrates (Johnson et al., 2020;Romano et al., 2021;Siva et al., 2017).
Taro (Colocasia esculenta) is a tuber crop widely grown in tropical and subtropical climate in the world.Taro tubers in general are low in protein and fat but are important source of carbohydrates such as resistant starch (RS) and mucilage.The high amount of RS is associated with slow digestibility which results in the slow liberation and absorption of glucose and aids in the reduction of risk of obesity, diabetes and other related diseases (Kaushal et al., 2015;Liu et al., 2006).
Taro is also rich in calcium, phosphorus, iron, potassium and magnesium besides containing thiamine, riboflavin, niacin and vitamins (Calle et al., 2020;Kaur et al., 2013) and has broader complement of vitamins and nutrients than other tubers (Kaushal et al., 2015).Taro has been used to produce several products including bread (Abera et al., 2017;Arıcı et al., 2020;Ikpeme et al., 2010), cookie and noodle (Dilek & Bilgiçli, 2021;Hajas et al., 2022;Kaushal et al., 2015;Kaushal & Sharma, 2014).Replacing wheat with protein-rich ingredient like lentil and taro which is rich in carbohydrate, mineral and RS will substantially improve the nutritional quality, health benefit and reduce the cost of bread.Tertiary blend of wheat, tubers and legumes have been used to produce baked products including cookies (Adekunle & Adeyemo, 2014;Igbabul et al., 2015), bread (Onoja et al., 2011) and cake (Olatunde et al., 2019) where the results indicated the possibility of developing nutrient-dense and in general acceptable products.Literature on baked products like cookies, bread and muffins as vehicles of novel ingredients is plenty.Relatively limited information, however, is available on scones.Scones are consumed by people of all ages and are characterized by a soft texture and sweet flavour (Rößle et al., 2011).As popular products among the consumer, scones are potential vehicles to deliver nutritional and functional/bioactive ingredients to consumers.The objectives of this work was to study the effect of blend proportion on the functional properties of the composite flours, the physicochemical and sensory attribute of wheat-taro-lentil composite flour scones.

| Processing method of taro, lentil and composite flour formulations
Raw taro corms were washed with water, peeled to remove skins and sliced.The slices were blanched with distilled water at 90 C for about 2 min.After blanching, taro was dried in a dehydrator (KM-1500 MRC, Holon, Israel) set at 50 C for 4 h, ground into fine powder using a small scale grinder (Excalibur 4926T, Sacramento, USA), then flour was sifted using a number 70 (190 μm) mesh sieve (Kaur et al., 2013;Kaushal & Sharma, 2014).Whole lentil flour was processed according to a modified method illustrated by Morales et al. (2015).About 500 g of black lentils were ground using a food processor (KM-1500 MRC, Laboratories, Holon, Israel) into very fine flour and then sifted using a 350 μm mesh sieve mesh before use.The composite flours were formulated from wheat (W), lentil (L) and taro (T).The three components were blended in different per cent proportions (70W:0T:30L, 70W:10T:20L, 70W:15T:15L, 70W:20T:10L and 70W:30T:0L) and a control 100%W.
2.2 | Functional properties of the flours 2.2.1 | Water absorption capacity (WAC) WAC of samples was determined using the method outlined by Chandra et al. (2015) with some modifications.About 0.5 g of each sample was dissolved in 10 mL of distilled water in centrifuge tubes and vortexed for 30 s.The dispersion was allowed to stand at room temperature for 30 min, centrifuged at 3000 rpm for 25 min.The supernatant was then filtered with Whatman No. 1 filter paper, and the volume retrieved was accurately measured.The difference between the initial volumes of distilled water added to the sample and the volume obtained after filtration was determined.The results were reported in mL of water absorbed per gram of sample.

WAC mL=g
ð Þ¼ amount of water absorbed weight of sample :

| Oil absorption capacity (OAC)
OAC was determined using the method described by Chandra et al. (2015) where about 1 g of sample (W 0 ) was weighed into a preweighed 15 mL centrifuge tube and thoroughly mixed with 10 mL (V 1 ) of refined palm oil using vortex mixer.Samples were allowed to stand for 30 min.The sample oil mixture was then centrifuged at 3000 rpm for 20 min.Immediately after centrifugation, the supernatant was carefully poured into a 10 mL graduated cylinder, and the volume recorded (V 2 ).OAC (millimetre of oil per gram of sample) was calculated as The SWI was measured by the method outlined by Chandra et al. (2015).About 25 g of each sample was placed into a 210 mL measuring cylinder and 150 mL of distilled water was added.The solution was allowed to stand for 4 h before observing the level of swelling.
The SWI was calculated as follows:

| Production of scones
All ingredients (360 g flour, 100 g margarine, 125 g sugar, 125 mL milk, 2.5 g salt and one egg) were measured according to the recipe.
The dough was brought together by whisking eggs and milk and pouring over the dry ingredients.The mix was quickly and gently folded into a cohesive mass using a spatula.Scones were baked in a preheated oven at 230 C for 18 min.After baking, scones were removed from the oven and cooled for further analysis.

| Determination of proximate composition of scones
Moisture and ash were determined according to American Association of Cereal Chemist (AACC) methods (AACC, 2000).The crude protein was determined by Association of Official Analytical Chemists (AOAC, 2000).Soxhlet method was employed to determine fat content.

| Evaluation of physical properties of scones
Physical properties were assessed by measuring the weight, volume, height, baking loss and specific volume.Rapeseed displacement method was used for loaf volume determination (AACC, 2000).The weight of the scones was determined using a weighing balance (E,I,M series N17250, Milton Keynes, UK).Specific volume was computed as the ratio between volume and weight.Scones were cut at the centre from the highest point to measure the height using a Vernier calliper (Nasir et al., 2020).The baking loss was computed as the difference between fresh dough weight before placing in the oven and the final scone weight after baking (Hajas et al., 2022).The colour measurement was done using a digital colorimeter (CLRM-310, MRC, Israel), which determined colour on the basis of three determinants L* for lightness and a* and b* for chromaticity coordinates (a*-red to green, b*-blue to yellow).

| Data analysis
One-way ANOVA was carried out to compare means using IBM SPSS version 20.Post hoc analysis using Duncan's test was done to measure significant differences at p < .05,and principal component analysis was done to determine the variability among formulations and associations between measured variables using SAS-JMP Pro 16.

| Functional properties of composite flour blends and dough
The functional properties of the composite flours are presented in Table 1.The results indicated that the blend proportions significantly influenced ( p < .05) the WAC.The WAC increased with increase in the proportion of lentil flour whereas it decreased with increase in the proportion of the taro flour.Water absorption of the flour is a very good indicator of the quality of the final baked product.Lentil is a good source of protein.The increase in the WAC with increase in lentil flour proportion could be attributed to the increase in the proportion of protein and fibre contributed by lentils (Fouad & Ali Rehab, 2015;Hajas et al., 2022;Hasmadi et al., 2020;Kaale et al., 2022) which may have higher ability to compete and bind water (Mashayekh et al., 2008;Mohammed et al., 2012).Similar trend has been observed with composite flours where protein sources were incorporated (Aider et al., 2012;Ammar et al., 2009;Katunzi-Kilewela et al., 2023;Mohammed et al., 2012;Ribotta et al., 2005;Shongwe et al., 2022).Decrease in water absorption with increase in the proportion of taro flour in wheat-taro composite flour has been reported in an earlier study (Ikpeme et al., 2010).
The result indicated that the blend proportion significantly ( p < .05)influenced the OAC where there was a general increase in OAC with increase in the proportion of lentil flour.Proteins offer higher OAC due to their hydrophilic and hydrophobic parts (Katunzi-Kilewela et al., 2023).OAC indicates the flour's emulsifying capacity and the amount of oil that can be adsorbed during cooking.Protein is a major chemical component that affects OAC (Chandra et al., 2015).
Increase in OAC with increase in protein-rich components of composite flours has been reported in previous studies.Hydrophobic proteins play the main role in oil absorption (Manyatsi et al., 2020).Moreover, high OAC with increase in protein could be attributed also to the protein present which physically entrap the oil by capillary attraction (Katunzi-Kilewela et al., 2023;Kaushal et al., 2012).The BD was significantly ( p < .05)influenced by the blend proportions (Table 1).
There was an increase in the BD with increase in the proportion lentil flour.BD is mainly influenced by particle size distribution, and the results found in this study are within the ranges of other composite flours made from starch and protein sources (Katunzi-Kilewela et al., 2023;Manyatsi et al., 2020).The high BD of flours suggests their suitability for use in food preparations (Chandra et al., 2015;Kaushal et al., 2012).
The proportion of the component flours brought about a significant difference in the SWI.The results indicated that SWI increased with increase in taro flour proportion (Table 1).The SWI could be influenced by particle size, the flour chemical composition and processing methods.High SWI with increase in taro flour proportion could be attributed to swelling capacity of taro starch due to hydration (Chandra et al., 2015).Reduction in SWI with increase in lentil flour proportion could be attributed to the formation of a starch-protein complex which reduces swelling capacity (Menon et al., 2015).

| Batter viscosity
The batter viscosity was significantly ( p < .05)influenced by the blend proportion (Table 1).The batter viscosity increased with increase in the proportion of taro flour whereas increase in the proportion lentil flour resulted in a decrease in the batter viscosity.The viscosity of the control sample was not significantly different from 70W:15T:15L and 70W:20T:10L samples.The reduction in batter viscosity with increase in lentil flour is in line with previous studies where the hardness of composite flour made from rice and black lentil was 1.5 mJ as compared with 2.8 mJ for dough made of rice (Hajas et al., 2022).Increase in degree of softness or decrease in dough strength of dough containing high protein components including lentil has also been reported indicating that lentil flour has a negative impact on the formation of gluten network (Ammar et al., 2016;Bourré et al., 2019;Katunzi-Kilewela et al., 2023;Mohammed et al., 2012;Previtali et al., 2014;Romano et al., 2021).

| Proximate composition of scones
The proximate composition of scone from the different treatments are presented in Table 2.The minimum protein content was observed with the formulation where there was no lentil flour.The protein content decreased with increase in the proportion of taro flour and increased with increase in the proportion of lentil flour.
Lentil flour is reported to be a good source of protein and has a protein content in the ranges of 20.6% to 31% (Dhull et al., 2023;Joshi et al., 2017;Kaale et al., 2022); thus, increase in the protein content with increase in lentil flour was as expected.Increase in crude protein content from 11.15% to 17.9% (Man & P aucean, 2013) and from 13.8% to 19.02% (Portman et al., 2018) due to increase in lentil flour from 0% to 40% has been reported for wheat-lentil composite flour bread.The crude fat content in scones made with the composite samples was significantly lower than the control.However, there was no significant difference ( p > .05) among the samples made from composite flours.The low fat content of taro and lentil resulted in the low crude fat content of the composite flour scones.
The ash content increased with increase in the proportion of taro flour and was significantly influenced ( p < .05)by the blend proportion (Table 2).Taro flour has in general high ash content (Arıcı et al., 2020;Dilek & Bilgiçli, 2021).The moisture content increased with increase in the proportion of lentil flour.The moisture content of the control was not significantly different from the samples with 70W:30T:0L and 70W:20T:10L.The samples with relatively high proportion of lentil flour have high moisture content.An increase in water solvent retention capacity was reported with increase in the proportion of lentil flour in wheat-lentil cotyledon composite flour (Portman et al., 2018).
T A B L E 1 The effect of wheat, taro and lentil flour blend proportions on the functional properties of composite flour and batter viscosity.

| Physical properties of scones
The weight of scones increased with increase in lentil flour and decreased with increase in the taro flour proportion.This could be attributed to the high water holding capacity of lentil flour which is associated with its high protein content resulting in heavy dough.Composite flours with protein-rich components have been reported to result in increased weight of baked products (Manyatsi et al., 2020;Mohammed et al., 2012;Nasir et al., 2020;Ribotta et al., 2005;Shongwe et al., 2022).This increase in weight might be attributed to the high water holding capacity of the proteins.
There was an increase in the height and volume with increase in the proportion of taro flour (Table 3).Increase in lentil flour proportion resulted in decreased height and volume of scones.This could be due to decreased gluten proportion in the blends.Similar trends have been observed as a result of increase in protein-rich components of baked products made from composite flours (Manyatsi et al., 2020;Mohammed et al., 2012;Nasir et al., 2020).The specific volume also exhibited similar trend as height and volume where the maximum and the minimum were 3.63 ± 0.04 and 1.37 ± 0.04 cm 3 /g for the control and samples with 70W:0T:30L, respectively.A significant reduction in the specific volume was reported with increase in lentil flour from 5% to 40% which was attributed to low hydration of wheat gluten (Hernandez-Aguilar et al., 2020;Portman et al., 2018).Similar trends were reported by previous studies for baked products made from composite flours containing protein sources (Manyatsi et al., 2020;Mohammed et al., 2012;Nasir et al., 2020).The reduction of specific volume with increase in protein content could be attributed to the relative reduction in the starch proportion which contributes to expansion upon hydration (Blazek & Copeland, 2008).
Baking loss decreased with increase in the proportion of lentil flour and increased with the proportion of taro flour (Table 3).The baking loss values and the trends observed in this study are comparable with baking loss values reported in literature (Ammar et al., 2016;Hajas et al., 2022;Mohammed et al., 2012).The water retention capacity was improved by increasing the lentil flour composition hence have high water content (reference).

| Colour assessment (L*, a* and b* values)
Colour was significantly ( p < .05)influenced by the blend proportions (Table 4).The results indicated that the lightness (L*) value decreased with increase in lentil flour proportion and increased with increase in The effect of wheat, taro an lentil flour blend proportions on proximate composition of scones.T A B L E 3 The effect of wheat, taro and lentil flour blend proportions on physical properties of scones.the proportion of taro flour.Lower values of L* with increase in lentil flour proportion has been reported for bread made with wheat-lentil flour composite bread (Aider et al., 2012;Atudorei et al., 2022;Hernandez-Aguilar et al., 2020;Portman et al., 2018).The negative a* values indicated that the product was greenish though it increased significantly with increase in the lentil flour and decrease in taro flour proportion.Similar trend was reported for wheat-sprouted lentil composite flour with increase in lentil proportion (Atudorei et al., 2022;Hernandez-Aguilar et al., 2020).The b* value ranged from 28.27 ± 4.10 to 17.89 ± 1.12 for the control and 70W:30T:0L, respectively.
The positive b* value indicates a yellowish colour.There was no significant difference among samples containing taro and lentil flours.
Wheat and wheat-lentil composite flour bread with 10% lentil showed no significant difference in b* value (Hernandez-Aguilar et al., 2020).The results indicate that the control was more yellow than the other blends.

| Sensory attributes
The sensory scores for the degree of liking rated on a 9-point hedonic scale of the different attributes are presented in Table 5. Regarding the colour, there was no significant difference ( p > .05) between all the treatments though the sample with 70W:0T:30L had a score of 6.06 ± 1.87 which has intensity of like slightly and the control exhibited a an average score of 6.84 ± 2.13 (like moderately).The colour of bread made with 5% and 7.5% germinated lentil flour was appreciated more than 100% wheat flour bread (Atudorei et al., 2022).High proportion of taro resulted in taste score comparable with the control.
Samples with 7.5% and 5% germinated lentil flour were found to have high scores (Atudorei et al., 2022).Except for 70W:0T:30L, there was no significant difference ( p > .05) in taste between the control and the rest of the treatments.Cookies made from black lentil flour showed a better taste compared with cookies made from rice flour indicating the potential of lentil flour to produce baked products of acceptable taste (Hajas et al., 2022).The taste, smell, texture, flavour and global acceptability were reported to have the highest scores for the samples with 7.5% and 5% germinated lentil flour addition (Atudorei et al., 2022).The aroma and texture were not significantly ( p > .05)influenced by the proportion of lentil and taro flour.In terms of the overall acceptability, there was a significant difference between the control and the rest of the samples indicating that lentil can be incorporated up to 30%.Earlier studies also reported where 10% and 20% showed the sensory scores similar to the 100% durum wheat with overall quality of 7.38 and 6.8, respectively, in durum wheatlentil flour composite sample bread (Previtali et al., 2014).showing their differences (Figure 1).The loading plot (Figure 2) showed the association between the flour functional properties, the scones physical properties and the sensory attributes.The physical properties of scones including specific volume, height, volume, baking loss and the L* (lightness) value were positively associated.These physical properties of scones were also positively correlated with sensory taste, aroma, texture and colour.High specific volume, volume, height and L* values are linked with high taro flour proportion.The functional properties of the composite flour like BD, WAC and OAC were associated with the weight, protein and moisture content of scones.Composite flours with high BD, WAC and OAC resulted in heavy, high protein and moisture content scones.These associations are linked with high protein and fibre contents of lentil flour (Dhull et al., 2023;Kaale et al., 2022).

| Cluster and principal component analysis
The bi-plot (Figure 3) showed the association between the composition of the composite flour and the physicochemical and sensory attributes.High percentage of taro is associated positively with taste, aroma, texture, specific volume, volume, baking loss, height, L* and batter viscosity.On the other hand, high proportion of lentil flour is associated with the functional properties of the flour (WAC, OAC and BD) protein content, moisture content and weight of scones.The biplot revealed that ash content is associated with the taro and lentil components.The ash content of lentils and taro has been reported to be higher than wheat and other cereals (Dilek & Bilgiçli, 2021;Kaale et al., 2022;Njintang et al., 2007).The a* value is associated with the proportion of lentil flour whereas the b* value is strongly associated with the wheat evident from its closeness with the control sample.
The hierarchical dendrogram (Figure 4) clustered the scones into four clusters which are 100%W, the wheat-taro-lentil, wheat-taro and wheat-lentil composite flour scones.This suggests that a range of possibilities exist to blend taro and lentil to come up with an optimal product.This calls for a further study to optimize the blend proportions.
Principal component analysis was conducted to visualize the variation in the properties among the formulations and the association between measured variables.The analysis indicated that about 93.4% of the variation was explained by three principal components.The first principal component (PC1) explained 62.5% of the variations whereas the second (PC2) and the third (PC3) components explained 21.9% and 9%, respectively.The score plot (Figure1) showed the similarities and differences between formulations.The distance between the locations of any two formulations on the score plot is directly proportional to the degree of difference or similarity between them.The 100%W and 70W:0T:30L were on the extreme right and left showing T A B L E 5 The effect of wheat, taro and lentil flour blend proportions on sensory attributes of scones.
Score plot showing the variability of the wheat, taro and lentil blend formulations.F I G U R E 2 PCA loading plot showing the association between variables of wheat, taro and lentil flour blend formulations.F I G U R E 3 Bi-plot showing the association between flour blends and the response variables.the differences in these two flours in PC1 whereas the 70W:0T:30L and 70W:30T:0L appeared on either side of the score plot in PC2 4 | CONCLUSIONThe functional properties of wheat-taro-lentil composite flour were significantly influenced by the blend proportion.WAC, OAC and BD density increased with increase in the proportion of lentil flour.Increase in the proportion of lentil flour also increased the protein content whereas increase in the proportion of taro increased the ash content indicating the mineral contribution of taro.The physical properties of scones including specific volume, height, volume, baking loss and the L* (lightness) value were positively associated.These physical properties of scones were also positively correlated with sensory taste, aroma, texture and colour.High percentage of taro is associated positively with taste, aroma, texture, specific volume, volume, baking loss, height, L* and batter viscosity.On the other hand, high proportion of lentil flour is associated with the functional properties of the flour (WAC, OAC and BD) protein content, moisture content and weight of scones.The aroma and texture were not significantly (p > .05)influenced by the proportion of lentil and taro flour.In terms of the overall acceptability, there was no significant difference (p > .05) between the control and the rest of the samples indicating that lentil can be incorporated up to 30%.
Dendrogram showing the four clusters of scones formulations prepared from wheat, taro and lentil flour.
Note: Means followed by the same letter in a column are not significantly different ( p > .05).
Note: Means followed by the same letter in a column are not significantly different ( p > .05).
Note: Means followed by the same letter in a column are not significantly different ( p > .05).The effect of wheat, taro and lentil flour blend proportions on colour values.
T A B L E 4Note: Means followed by the same letter in a column are not significantly different (p > .05).
Note: Means followed by the same letter in a column are not significantly different ( p > .05).