Rice bran is a by-product of rice milling industries, grossly underutilised in rice-producing countries. Al- though, rice bran is rich in dietary fibre, amino acids, minerals, and other bioactive compounds, which  can have a beneficial effect on human health [1], it is com- monly used as a livestock feed as opposed to human food formulation. This is because the use of bran in food industries, especially in baked goods, poses several tech- nological problems, such as weakening of wheat dough structure and of baking quality, the decrease in bread volume and elasticity of bread crumb [2]. Bio-proces- sing technologies, such as enzymatic and fermentation processes have received more attention recently due to their potential to improve nutritional and technological

functionality of bran [3]. These processes have been em- ployed for bran sourdough production [4–6].

Sourdough is used as a starter for bread making, be- cause of its ability to improve bread quality [7, 8]. Tra- ditionally, bread volume, crumb texture, flavour, and shelf-life of bread are enhanced using sourdough fer- mentation [2, 9, 10]. This is because baked products with sourdough are better protected from bacterial and mould spoilage, and have extended shelf-life [11, 12]. However, the preservative effect of sourdough depends on the bac- teria used for the dough fermentation. Previous studies have reported an extended shelf life for sourdough bread produced with lactic acid bacteria (LAB), fermented rice bran sourdough [13, 14].

The flavour of baked foods, especially bread, is in-

fluenced by the recipe, processing condition, and vola-

Copyright © 2019, Bolarinwa et al. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.

tile organic compounds produced by a starter culture. The generation of odorants in sourdough occurs due to enzymatic and microbial processes during sourdough fermentation [15], for example, LAB is responsible for liberation of aroma precursors and acidification in sour- dough [7].

Fermentation temperature and time are important processing factors to be considered in producing a quali- ty sourdough. The extent of metabolic performances of microorganisms in sourdough depends directly on the fermentation time, while fermentation temperature plays an important role in the stability of sourdough ecosystem [16, 17]. There have been studies on the effect of pro- cessing conditions and starter cultures on the formation of volatile compounds in wheat sourdoughs [18]. Volatile

mill, Sekinchan, Selangor, Malaysia. The rice bran con- tained 7.2% of moisture, 11.7% protein, 18% fat, 8.6% fibre, 8.5% ash, and 45.9% carbohydrate, as determined in our previous study [21]. The bran was sieved using a Haver EML digital plus test sieve shaker (Harver and Boecker, 59302, OELDE,  Germany)  to  particle  size of < 300 µm.

Microbial strain. Lactobacillus plantarum. ATCC® 8014™ was purchased from the American Type Culture Collection (ATCC) (USA). SupelcoSPME fibre coated with 85-µm of Carboxen–polydimethylsilo- xane (Carboxen/PDMS) was purchased from Belle- fonte, PA (USA), SPME vials were purchased from LA-PHA-PACK (Germany). Distilled deionized  wa- ter from Sartorius, Sdn, Bhd (Malaysia) was used for

compounds produced by different species of lactobacil-

all sample preparations. N-alkanes (C –C

), and 4-me-

8        20

li in rye sourdough have also been reported [7]. There

is little information on the optimisation of fermentation conditions for rice bran sourdough production, and no information on volatile compounds in LAB fermented rice bran sourdough.

Response Surface Methodology (RSM) is a multiva- riate technique used for an analytical optimisation [19]. In recent years, RSM has been used to optimise fermen- tation conditions for cereal bran. Application of RSM for optimisation of rice bran fermentation conditions for pro- tein concentrate extraction, using baker’s yeast, has been studied [20]. RSM was also applied to study fermentation effects on the levels of phytochemicals in rye bran [5].

The use of rice bran sourdough is gaining populari- ty in bread making [13, 14, 21]. It is therefore important to establish optimum fermentation temperature and time for quality rice bran sourdough preparation. Further- more, bread quality mostly depends on taste and aroma. Apart from aroma development during baking, aroma precursors are also produced during fermentation pro- cess. Volatile organic compounds (VOCs) in fermented sourdough contribute to bread flavour [22] and are pro- duced from enzymatic, or autoxidation of flour lipids, microbial, yeast metabolism, and Maillard reaction [15]. However, publications about VOCs in sourdough are mostly about wheat, wheat germ, buckwheat, rye, rice, maize, Kefir or chestnut-flour sourdough using either GC-MS, or HRGC-MS [15].

thyl-2-pentanol were purchased from  Sigma-Aldrich (St Louis, USA). All other chemicals and reagents were of analytical grade.

Optimisation of rice bran fermentation condi- tions. We used RSM to investigate the combined effects of significant variables (fermentation temperature and time) in the production of rice bran sourdough. The ado- pted experimental design was a central composite design (CCD). The minimum and maximum independent vari- able levels were selected on the basis of previous sour- dough studies [18, 5]. The CCD was divided into three blocks to calculate the reproducibility of the method. The CCD’s use required 13 runs for each block and the total of 39 runs for the whole design. Table 1 shows the experimental runs  in  the  order  they  were  performed. The dependent/response variables for the rice bran sour- dough quality were pH and total titratable acidity (TTA). The independent variable range was 20–35°C of fermen- tation temperature and 6–20 hours of fermentation time. For reproducibility, the  experiments  were  carried  out in triplicate. For each of the response variable, a model summary, and a lack of fit test were constructed for li- near, square, interaction and quadratic models. We in- vestigated the behaviour of the response surface for the

response function (Yp, predicted response) using a re- gression polynomial equation. The generalised polyno- mial mode to predict the response variable was:

Although, VOCs of rice bran oil have been charac-

Y = β + β T + β T + β T 2 + β

T 2 + β

T T ,          (1)

terised  by  GC-MS  [23],  to  our  knowledge,  there  has

p         0         1   e

2    i         11   e

22    i

12   e   i

been no study on the characterisation of VOCs in rice

where β

was intercept, β and β

were coefficients. The

0                                              1                 2

bran  sourdough  using  HS-SPME  method.  The  obje-

ctives of this study were to optimise important fer- mentation factors (temperature and time) for rice bran sourdough production using RSM and to determine the effect of fermentation conditions on pH, titratable acidi- ty (TTA), and total phenolics in L. plantarum fermented rice bran sourdough. For the first time, VOCs profile of rice bran sourdough samples fermented at different tem- perature and time was also identified and quantified by SPME-GC/MS analysis.

STUDY OBJECTS AND METHODS

Raw materials. Rice variety MR bran 219 was ob- tained from Padiberas Nasional Berhad (BERNAS) rice

most accurate model was chosen based on a number of

significant terms (p < 0.05), percentage of regression, the lack of fit test, and P-value for a regression model. A three-dimensional response surface plot was generated for each response variable. The reliability of the model was evaluated from the regression of the model (R2) and predictive power of the model (Q2). Generally, for a good fi model R2 should be ≥ 0.8 [24], however, the models with R2 and Q2 values of  ≥ 0.9 are considered excellent.

Optimal processing conditions to determine rice bran sourdough production were calculated by per- forming a multiple response method called desirabili- ty. The desires and priorities for each of the variables

were incorporated into the optimisation method. pH and

Table 1. Composition of various runs of central compo- site design (CCD)

was incubated on a plate count agar, and the colonies were counted. The bacteria suspension (109  CFU/ml) in

                                                                                                                peptone water (0.1%) was used for bran fermentation.

pH and total titratable acidity (TTA) of rice bran sourdough. pH and TTA were measured as described by [25]. Frozen rice bran sourdough samples were thawed overnight in a refrigerator at 4°C. 10 g of the thawed sourdough were mixed with 100 ml of distilled water to obtain a suspension. The pH value of the sus- pension was measured by using a pH meter. TTA was determined by titrating the suspension against 0.1M NaOH to a final pH value of 8.5. TTA was expressed as the amount of NaOH (ml) used for titration and as a mean value of three replicates.

Total phenolic compound (TPC) analysis. A modified method of [28] was used for a total pheno- lic compound determination. 200 mg of the rice bran sourdough  fermented  with  L.  plantarum  at   diffe- rent temperatures and time were extracted with 4 ml of 70% aqueous acetone by shaking in a rotary shaker (260 rpm, 2 hours) at room temperature. The su- pernatant  liquid  was  collected   after   centrifu- gation (Hereaus Instruments D867 bench top centrifuge, Germany) at 3,000 g for 10 min. 200 µl of the supernatant were added to 1.5 ml of freshly pre- pared Folin Ciocalteu reagent (1:10 v/v with water). After 5 min equilibration, 1.5 ml of the sodium carbo- nate (60 g/l) solution was added to the mixture. The mixture   was   incubated   at   room   temperature   for

90  min.  The  absorbance  of  the  mixture  was  read  at 725 nm using 70% aqueous acetone as blank. The gal-

Note: a Denotes centre point runs

titratable acidity (TTA) were chosen as the two most important quality parameters in this study, because sourdough quality depended mainly on them [25]. Since a rice bran falls within a high extraction rate category with 95% extraction rate, the desired pH target was in the range of 3.5 to 4.5 [26], while the desired TTA was in the range of 16 to 22 ml NaOH [26, 27].

serial dilution (10–2  to 10–10). A serial suspension (0.1 ml)

lic acid calibration curve (0 to 1,000 µg/ml) was plotted, and TPC was expressed as micrograms (µg) of gallic acid equivalents (GAE) per gram of rice bran sourdough. Characterisation of volatile compounds (VOCs) in rice bran sourdough. Volatile acids are generally pro- duced, when cereals are fermented by lactic acid bacte- ria, such as L. plantarum. Volatile acids of cereal dough, such as rye dough and baked product (bread), are repor- ted to be acetic and lactic acids. These acids are respon-

sible for an acidic bread flavor [29].

Headspace solid phase microextraction (HS-SPME) analysis. The volatile fraction of rice bran sourdough samples was analysed by headspace sam- pling using solid phase microextraction (SPME) method described by [30], with some modifications. The VOCs of rice bran sourdoughs fermented with L. plantarum at different temperature and time (Table 2) were identified and quantified. 2 g of a sourdough sample fermented at

different temperature and time (Table 2, sample A–J) were placed in a 20 ml headspace vial. 2 ml of a dis- tilled deionised water and 5 µl of 4-methyl-2-pentanol (internal standard, 100 mg/l) were added to the head- space vial. The vial was placed in a thermostatic block on a stirrer at 60°C, then the fibre was inserted and maintained in the sample head space for 30 min, after which it was removed and immediately inserted into the GC/MS injector port for the desorption of compounds. A silica fibre coated with 85-µm of Carboxen–polydimet- hylsiloxane (Carboxen/PDMS) was used for the analy- ses. The fibre was conditioned prior to its first use. The internal standard (4-methyl-2-pentanol) was used for all analyses to monitor the SPME extraction and fibre per- formance during the analysis, as well as for semi-quanti- tative analysis of VOCs.

Gas chromatography–mass spectrometry (GC/MS) analysis and VOCs identification. The method described by [31] and [30] was adopted for the GC/MS analysis. The VOCs of the  sourdough  sam- ples  were  determined  with  a  gas  chromatograph  and a mass spectrometer equipped with a 30 m, 0.25 mm ID, film thickness of 0.25 µm, and capillary column (TG-WAXMS, Thermo Scientific, USA). The gas car- rier was helium with a flow of 1.5 ml/min. SPME injec- tions were splitless (straight glass line, 0.75 mm ID) at 240°C for 20 min (during which time thermal desorp- tion of analytes from the fibre occurred). The oven para- meters were as follows: the initial temperature was 40°C held for 3 min, followed by an increase to 240°C at a rate of 5°C /min, then held for 10 min. A mass spectrometer was operated in a scan mode from m/z 33–300 (2 s/scan) at an ionisation potential of 70 eV.

Identification of volatile compounds was achieved by comparing their mass spectra with the reference mass spectra library (NIST, version 2.0) and by matching the retention indices (RI) calculated according to the equa-

tion of [32], based on a series of alkanes, ranging from

sponse factor. Blank of the fibre and blank of the empty vial were conducted after every 10 analyses. All the analyses were performed in replicates.

Statistical analysis. The experimental design and optimisation procedure were performed using Minitab® version 16.0 software (Pennsylvania, USA). The data obtained were subjected to one way analysis of variance (ANOVA) to determine regression coefficients and a statistical significance of model terms to fit the mathe- matical models to the experimental data. Significant dif- ferences in the total phenolic compounds were compared by means of Duncan’s multiple comparison tests at 95% confidence level (p ≤ 0.05). The data were reported as mean values of replicate analyses.

RESULTS AND DISCUSSION

Influence of fermentation conditions on pH, and titratable acidity (TTA). The pH and TTA of the rice bran before fermentation (control) was 6.11 and 11.89, respectively (the results are not shown). The acidity le- vels (pH and TTA) of the rice bran sourdough fermented at different times and temperatures  are  presented  in Fig. 1. The pH of the sourdough decreases with increa- sing fermentation time and temperature, whereas the TTA values increase.

The  pH  of  the  rice  bran  sourdough  ranged  from

4.0 to 4.6, while the TTA ranged from 13.9 to 23.4 ml of 0.1M NaOH/10 g sourdough (Fig. 1). Fermentation time was reported to be the most important parameter of wheat sourdough acidity, fermented with a single or combined starter of lactobacilli or yeast [18]. In con- trary, the findings of this study showed that fermentation temperature and time  were the most  influential para- meters of L. plantarum fermented rice bran sourdough acidity. The most acidic rice bran sourdough was the sample, fermented at 35°C for 13 hours, while the least acidic sourdough was the sample, fermented at 20°C for

used for compounds eluting after C

). A semi-quanti-

strong interaction between fermentation time and tem-

perature during sourdough fermentation. The maximum

TTA and the minimum pH values were obtained in the

Sample code   Fermentation tempera-

ture, °C

Fermentation time, h

A                     22.2                                8.05

B                      22.2                                17.95

C                      20.0                                13.00

D                     27.5                                6.00

E                      27.5                                13.00

F                      27.5                                20.00

G                     32.8                                17.95

H                     32.8                                8.05

I                       35.0                                13.00

J*                                35.0                                13.00

Note: * Control sample (sourdough sample fermented without the ad- dition of L. Plantarum)

Fig. 1. pH and total titratable acidity (TTA) profiles of rice bran sourdoughs fermented with L. plantarum at different fermentation time and temperature.

sourdoughs, fermented at 35°C for 13 hours. This is probably because at the optimum proliferation tempera- ture (37°C, which is close to 35°C) of L. Plantarum, and the optimum time for LAB activity (> 12 hours), L. plan- tarum would produce more lactic acid, which will lead to a reduction in pH, increased TTA, and thus, higher acidity in the sourdough. Similar observations were also reported in previous studies [18, 22, 15].

Optimisation and validation process. RSM was used  to  develop  a  prediction  model  for  optimising

L. plantarum fermentation conditions for rice bran sour- dough production. The independent and dependent vari- ables were analysed to obtain a regression equation, that could predict response (pH and TTA) within the given range. The developed mathematical models for the rice bran sourdough are presented in Table 3. High values of R2 (the fit of the model to the experimental data) and Q2 (predictive power of the model) were obtained in the acidity model. The obtained R2 values  (0.88)  indicate that 88% of the variation can be explained by the fit- ted model (Table 3). Large Q2 values (> 0.9) imply that the model has a good predictive ability with a small prediction error. A significant linear effect of time and temperature was observed in pH values of the rice bran sourdoughs. The latter  indicates that the  influence of fermentation time on pH values was dependent on fer- mentation temperature. And both independent  factors had quadratic effect on TTA values (Table 3). p-values for both models significantly affected the acidity lev- els with the p-value lower than 0.05, however, the p-va- lue (0.000) for the reduced quadratic model of TTA was more significant than that of the pH linear model (p-va- lue = 0.017). These results agreed with the results of the previous studies of L. plantarum fermented wheat sour- doughs [18, 33].

Fig. 2. Surface plot of effects of fermentation temperature, and fermentation time, on pH levels in L. Plantarum fer- mented rice bran sourdough.

The levels of the independent variables generated a targeted range of pH (4.0 to 4.6) and TTA (14–23 ml of 0.1M NaOH/10g) for the rice bran sourdough (Fig. 1). The optimum pH for the sourdough was obtained in 19.5 hours at 34.8°C while, the optimum TTA value was obtained in

10.65 hours at 32.6°C. Based on the multiple optimisation of the fermentation of rice bran sourdough, the optimum pH value of 4.3 and TTA value of 19.5 ml of 0.1M NaOH were obtained, when the sourdough was fermented  at 33°C for 12.5 hours.

A two-sample t-test was conducted to test the ad- equacy of the final reduced models by comparing the experimental values with predicted values. The data ob- tained showed that there were no significant differences (p > 0.05) between the experimental and predicted va- lues (Table 4). This observation showed the agreement between the two values, and thus, the adequacy of the response surface equations fitted by RSM was verified.

The influence of fermentation parameters on to- tal phenolic compounds. Since sourdough can im- prove phytochemicals in baked goods, we investigated the effect of fermentation conditions on the level of to- tal phenolic compounds (TPC) in rice bran sourdough. The result is presented in Table 5. The highest amount of TPC (484.9 µg GAE/g) was detected in sourdough fermented at the high temperature (32.8°C) for a long time (17.95 hours). The least TPC (410.2 µg GAE/g) was found in rice bran sourdough that was fermented at a low temperature (22.2°C) for 8.05 hours. Although, the TPC values of the sourdoughs increased with the rise

Table 3. Effect of fermentation factors, expressed as their corresponding coefficients obtained in the model of rice bran sourdough acidity fermented with L. Plantarum

Fig. 3. Surface plot of effects of fermentation temperature and fermentation time on titratable acidity (TTA) levels in

L. Plantarum fermented rice bran sourdough.

Table 4. Predicted and experimental values of responses at the optimum pH and titratable acidity (TTA)

fermented rice bran sourdough (A-I) was more than the

VOCs from naturally fermented rice bran sourdough (J).

                                                                                                               The most abundant alcohol among the VOCs was etha-

Response variables      Predicted valuea       Experimental valuea pH   4.22         4.03 ± 0.14

 TTA                                              19.10                     23.32 ± 1.14              

Note: a no significant difference (p > 0.05) between predicted and ex- perimental values

of fermentation temperature and time, there were not significant differences (p > 0.05) in the TPC of all the sourdoughs. This indicates that a high TPC value will be obtained at the optimum fermentation temperature (33°C) and time (12.5 hours). The trend of the TPC con- tents of the rice bran sourdoughs is in congruent with the

nol, a primary product of fermentation, followed by 2-Methoxy-4-vinylphenol. 2-Methoxy-4-vinylphenol (spicy) is an aromatic substance used as a flavouring agent [15, 36]. Ethanol production increased with the rise of fermentation temperature and time (Table  6). This could be because the longer fermentation time is enough for the microbial strain to degrade simple su- gars in the sourdough  into  carbon-dioxide  and  etha- nol [37]. Heterofermentative LAB has been reported to produce a high amount of ethanol and carbon-dioxide during rye sourdough fermentation [7], thus, L. planta- rum (a facultatively heterofermentative LAB) [38] was

previous findings on rye bran [5]. A higher TPC value

probably responsible for increased ethanol and CO2

pro-

obtained at a higher fermentation temperature and time

could be due to the fermentation in liberating and syn- thesising bioactive compounds, such as phenolic com- pounds in cereal brans [5, 34, 35].

Volatile organic compounds (VOCs) of rice bran sourdough. The list of the volatile compounds detected in rice bran sourdough fermented at different tempera- ture and time is presented in Table 6. It includes only those detected in more than one sample. Volatile analy- sis of the ten (10) rice bran sourdough samples (Table 2, A-J) showed the presence of 47 compounds. They were grouped into different chemical classes as esters (15), alcohols (7), acids (6), alkanes (5), pyridine (1), siloxane

derivatives (3), aldehydes (2), phenol (1), furan (1), aro-

matic hydrocarbon (1), ketose (1), ketone (1), nitriles (1), and others (2). The VOCs from the rice bran sourdough could be generated from enzymatic and microbial pro- cesses during sourdough fermentation (e.g., acids, alco- hol, esters, ketones, and aldehydes), lipid oxidation (e.g., aldehydes, and ketones), and Maillard and caramelisation reactions (e.g., furans, aldehydes, ketones, and pyridines), or from other compounds present in the flour [15].

Rice bran sourdough fermented at a higher tempera- ture (> 27°C) for a longer time (> 8 hours) produced the largest number of VOCs (Table 6). Furthermore, the amount of volatile compounds released from L. plantarum

Table 5. Total phenolic compounds (TPC) in rice bran sourdoughs fermented with L. plantarum at different temperature and time

duction during rice bran sourdough fermentation. Etha-

nol has been cited as a specific volatile compound in a wheat sourdough bread [15]. The highest amount of eth- anol and 2-Methoxy-4-vinylphenol was detected in sour- dough fermented at 35°C for 13h (sample I). This is an indication that rice bran sourdough can be a good alter- native to wheat sourdough for bread production.

Estra-1,3,5(10)-trien-17á-ol, 2,3-Butanediol, 3-Eth- yl-2-pentanol, 2-methyl-1-Hexadecanol, and 1-Ni- tro-2-acetamido-1,2-dideoxy-d-glucitol were other alcohols detected in the rice bran sourdoughs. Es- tra-1,3,5(10)-trien-17á-ol    was    present    in    all    the

L. plantarum fermented sourdough, but not found in the naturally fermented rice-bran sourdough (Table 6). This indicates that Estra-1,3,5(10)-trien-17á-ol is one of the main volatile compounds specific in L. plantarum fer- mented rice bran sourdoughs. This compound was also reported to be one of the main volatile compounds in moringa leave extract [40].

The presence of esters in sourdough and bread is an indicator of pleasant fruity aromas [39]. Esters in rice bran sourdough may be derived from products of free fatty acids reaction with some alcohols [41]. Hexadeca- noic acid, 1-(hydroxymethyl)-1,2-ethanediyl ester, was the most abundant ester in rice bran sourdough, and its highest quantity was detected in the sample fermented at 35°C for 13 hours (I). Higher quantities of this es- ter were also found in sample F and G, which were fer- mented for more than 17 hours, indicating that the longer fermentation time favours the formation of the esters. According to the literature, the LAB fermentation re-

Fermentation temperature, °C

Fermentation time, h

TPC, µg GAE/g

quires more than 12 hours to produce a sufficient amount

of volatiles [15]. The higher concentration of the esters

22.2                               8.05                    410.15a ± 19.75

32.8                               8.05                    472.73a ± 79.54

22.2                               17.95                  434.00a ± 9.04

32.8                               17.95                  484.86a ± 34.96

20.0                               13.00                  419.31a ± 37.50

35.0                               13.00                  473.92a ± 57.22

27.5                               6.00                    440.98a ± 59.70

27.5                               20.00                  454.20a ± 32.80

27.5                               13                       445.01a ± 25.61

Note: Values are mean ± standard deviations of triplicate analyses. Values in the same column with the same superscript letter are not significantly different (p>0.05)

in sample I, compared to sample F and G, could be, be-

cause sample I was fermented at 35°C for 13 hours, that was close to the optimum fermentation conditions (33°C for 12.5 hours) for rice bran sourdough production. Other esters found in the sourdough were 9-Octadecenoic acid, (2-phenyl-1,3-ioxolan-4-y l) methyl ester, Methoxyacetic acid, 3-tridecyl ester, and Pentanoic acid, 4-methyl- ethyl ester. These esters were also detected at the rela- tively high amount in the sourdoughs. 9-Octadecenoic acid, (2-phenyl-1,3-ioxolan-4-y l) methyl ester, was also reported in buckwheat sprout [42].

Seven (7) esters were detected in two (2) samples

(H and I) fermented at a high temperature for a short

time (32.8°C, 8 hours; sample H) and a longer time (35°C, 13 hours; sample I). Although, rice bran sour- dough fermented at the lowest temperature for a short time (20°C, 8 h; sample A) had the largest number of esters (8); the quantity of Hexadecanoic acid, 1-(hy- droxymethyl)-1,2-ethanediyl ester, was the least in this sample. This indicates that high fermentation tempera- ture and a long time favour the production of a large amount of esters in sourdough. However, there was vari- ation in the type and quantity of esters present in all the rice bran sourdough samples.

A long fermentation time and a  high  tempera- ture also favour the production of acids  in  the  rice bran sourdoughs (Table 6). This observation is in conformity with the result  of  the  sourdough  acid- ity (Fig. 1), where sourdough sample fermented  at 35°C for 13 hours had the highest acidity level. Sour- dough acids have been reported to serve as a bio-pre- servative [39]. This is probably why L. plantarum fermented rice bran sourdough taftoon bread had a higher anti-mould activity than the doughs fermented by L. Casei, and L. acidophilus [13]. Rice bran sour- dough fermented with L. plantarum at a higher tem- perature (> 32°C) for a longer time (> 12 h) (G and I) had the  largest  number  of  volatile  acids.  Thus,  these sourdoughs can extend a bread shelf-life and improve its flavour [43, 44].

Lactic and acetic acids are important for the flavour and taste of sourdough bread [45]. Acetic acid is the most common volatile acids reported to be present in sourdoughs [46, 47]. This acid was detected in a large quantity (143 µg/kg) in rice bran sourdough fermented at 35°C for 13 hours (I), but not detected in the sourdough samples fermented at < 27°C for a short or long time. Even though LAB was used for sourdough fermentation, lactic acid was not detected in any of the rice bran sour- dough samples.

Low levels of ketone were detected in the nat- urally    fermented    sourdough,    compared    to    the

L. plantarum fermented sourdough. The presence of ke- tone (acetoin) in the sourdough could be due to a long fermentation period. As shown in Table 6, the longer the fermentation time, the higher the amount of acetoin in the sourdoughs. Mild heating of flour for a long period under oxidising conditions has been reported to induce lipid peroxidation, consequently resulting in a lipid deg- radation to aldehydes and ketones [31]. The sourdough sample fermented at 32.8°C for 17:95 hours (G) had the higher acetoin value (146 µg/kg), than the sourdough fermented at 35°C for 13 hours (I). Acetoin was also reported to be present in a whole wheat sourdough [39] and chestnut-flour-based-sourdoughs [30]. Since aceto- in is characterised with a butter, or popcorn like aroma [46, 47], the presence of acetoin in rice bran sourdough produced by fermentation of rice bran at a high tempera- ture for a long time is an indicator, that the sourdough will add a special good flavour to bread, and retain bread freshness for a long time. The obtained results confirm the facts already known in the scientific literature [39].

Aldehydes are related to oxidative status of foods,

characterised as off-odours in bread [46, 47]. The pres-