UNIVERSITY OF COPENHAGEN Master’s thesis Lekka Eleni Effect of some traditional processing methods on the protein content of legumes from Ghana FACULTY OF SCIENCE DEPARTMENT OF FOOD SCIENCE Academic supervisors: Mette Holse (PhD) Birthe Møller Jespersen (Associate professor) COPENHAGEN AUGUST, 2014
78
Embed
Effect of some traditional processing methods on …curis.ku.dk/ws/files/120754000/LekkaEleniMasterthesis.pdf · III LIST OF ABBREVIATIONS ANF Anti-Nutritional Factor CHO Carbohydrates
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
UNIVERSITY OF
COPENHAGEN
Master’s thesis
Lekka Eleni
Effect of some traditional processing methods on the
protein content of legumes from Ghana
FACULTY OF SCIENCE
DEPARTMENT OF FOOD SCIENCE
Academic supervisors:
Mette Holse (PhD)
Birthe Møller Jespersen (Associate professor)
COPENHAGEN
AUGUST, 2014
I
ACKNOWLEDGEMENTS
The completion of this project was done in the implementation of the Master’s program “Gastronomy and
Health”, which was co-funded by the Act “Scholarship program of the National Scholarship Foundation of
Greece IKY”, with the resources of the Operational Program “Education and Lifelong Learning” of the
European Social Fund (ESF) and the National Strategic Reference Freamework (NSRF) 2007 – 2013.
I would like to thank my supervisors, Mette Holse (PhD) and Assosciate Professor Birthe Møller Jespersen for
trusting me with this project. I strongly appreciate their guidance, advice and comments, as well as their positive
perspective through any adversities that came up during this project. Their valuable knowledge and friendly
attitude helped me finish the present project.
A special thanks to the trainee lab technician Trine Ann–Mirll Lilla Pedersen for her insightful
recommendations and her patience in providing all the technical support.
I would also like to thank Associate Professor Thomas Skov for his assistance during the result analysis, and
Associate Professor Dennis Sandris Nielsen for his for his guidance in developing and delivering the
fermentation process.
Last but not least, I would like to dedicate this project to my parents and brother in Greece, while expressing
my utter love and gratitude to them. Without their support, love and motivation I wouldn’t have made it so far.
II
ABSTRACT
Malnutrition is a major problem in developing countries. The diets of people lack animal protein
sources and consist mainly of starchy roots, cereals and a few legumes. The present trend in
population growth indicates that the protein gap may continue to expand, unless measures are taken.
Alternative sources of protein need to be found, and attention should be directed to plant proteins.
Legumes are a good source of protein, essential amino acids, polyunsaturated fatty acids, minerals
and calories. They also contain, however, anti-nutritional factors that are known to lower the
nutritional quality of legumes. Germination and fermentation are two traditional processing methods
that are believed to reduce the anti-nutritional factors and improve the nutritional composition of the
legumes overall.
In the present study the effect of germination and fermentation on the protein content of legumes
collected in Ghana was examined. Six different legumes, namely soybean, cowpea (two cultivars),
jack bean, velvet bean and bambara groundnut were subjected to germination for 48-72h at 30oC,
and fermentation at 30oC and 37
oC for 48h. The crude protein content of the raw and processed
legumes was determined. Furthermore, the overall composition was assessed using spectroscopic
methods (FT-IR, NIR).
All six legumes showed a notable increase in their crude protein content. The increase was, in most
cases, more pronounced in the fermented samples. Regarding the increase in protein content,
germination for 72h seemed to yield better results than for 48h; however, no conclusive evidence on
the superiority of one temperature over the other was found in the fermented samples. As far as the
overall composition is concerned, both FT-IR and NIR pointed out soybean as the most
differentiated samples with regards to protein, lipid and carbohydrate content.
III
LIST OF ABBREVIATIONS
ANF Anti-Nutritional Factor
CHO Carbohydrates
CVD Cardiovascular Disease
EEA Essential Amino Acids
FAO Food an Agriculture Organization
FR-IR Fourier- Transform infrared
IR Infrared
IVPD In vitro Protein Digestibility
MDGs Millenium Development Goals
MI Myocardial Infraction
NIR Near - infrared
PC Principal Component
PCA Principal Component Analysis
PUFA Polyunsaturated Fatty Acids
RS Resistant Starch
USDA U.S. Department of Agriculture
WHO World Health Organization
IV
TABLE OF CONTENTS
Περιεχόμενα ACKNOWLEDGEMENTS ........................................................................................................................................ I
ABSTRACT ............................................................................................................................................................ II
LIST OF ABBREVIATIONS ..................................................................................................................................... III
TABLE OF CONTENTS .......................................................................................................................................... IV
TABLE OF FIGURES .............................................................................................................................................. VI
TABLE OF TABLES ............................................................................................................................................... VII
1. Nutrition trends in Africa ......................................................................................................................... 1
2.2. Health benefits ................................................................................................................................ 9
2.3. Methods of processing and cooking legumes ............................................................................... 10
3. Composition of individual legumes ....................................................................................................... 12
3.1. Soybean (Glycine max, L.) .............................................................................................................. 12
5. Experimental theory .............................................................................................................................. 29
Mid- and Near- Infrared spectroscopy ...................................................................................................... 29
AIM OF THE STUDY ............................................................................................................................................ 32
MATERIALS AND METHODS............................................................................................................................... 33
Screening of samples ..................................................................................................................................... 33
Preparation of samples for analyses ............................................................................................................. 33
Germination of samples ............................................................................................................................ 33
Fermentation of samples........................................................................................................................... 34
V
Control samples ......................................................................................................................................... 34
Defatting of soybean samples for protein determination ........................................................................ 35
Determination of moisture content .......................................................................................................... 36
Determination of fat content .................................................................................................................... 36
Determination of protein content ............................................................................................................. 37
Data analysis .................................................................................................................................................. 38
RESULTS AND DISCUSSION ................................................................................................................................ 39
Screening Test ............................................................................................................................................... 39
APPENDICES ........................................................................................................................................................... i
Appendix 1 – FT- IR spectra ............................................................................................................................... i
Appendix 2 – NIR spectra ................................................................................................................................ iv
VI
TABLE OF FIGURES
Figure 1: Overview of the main pulse producing countries in 2011 (in million tons). ........................................ 3
Figure 2: Schematic representation of wild/ underutilized legumes, their potential value, and exploitation for
development of new products. ........................................................................................................................... 5
Figure 3: FT-IR PCA scores plot of raw beans .................................................................................................... 39
Figure 4: FT-IR spectra (MSC) of the germinated, fermented, and control samples. ....................................... 41
Figure 5: Mean FT-IR spectra (MSC) of the control legume samples.. .............................................................. 42
Figure 6: PCA scores plot of the FT-IR raw (MSC) spectra (PC-1/PC-2) (a), and PCA loading plots for PC-1 (b)
and PC-2 (c)........................................................................................................................................................ 43
Figure 7: NIR raw spectra (MSC) of the germinated, fermented, and control samples .................................... 44
Figure 8: Mean FT-IR spectra (MSC) of the control legume samples.. .............................................................. 45
Figure 9: PCA scores plot of the NIR raw (MSC) spectra (PC-1/ PC-2).. ............................................................. 46
Figure 10: NIR PCA loadings plots for PC-1 (a), PC-2 (b) and PC-3 (c). .............................................................. 47
Figure 11: PCA scores plot of the NIR raw (MSC) spectra (PC-1/ PC-3).. ........................................................... 48
Figure 12: % Protein content (DM) of the germinated, fermented and control samples. ................................ 49
VII
TABLE OF TABLES
Table 1: Main non-nutrient bioactive pulse compounds and their main potential positive and beneficial
Every type of bond has a different natural frequency of vibration. Furthermore, two bonds of the
same type in two different compounds are in two slightly different environments. So, no two
molecules of different structure can have exactly the same IR and NIR absorption patterns (IR and
NIR spectra). Thus, the analysis of the molecular bonds of a sample in the IR and NIR spectrum and
creates a characteristic spectrum that acts as a ―fingerprint‖ of the sample. By comparing the spectra
of two substances, one can establish whether they are, in fact, identical. A second and more
important use of the IR and NIR spectra is to determine the structural information about a molecule.
The absorptions of each type of bond (N-H, C-H, O-H, C-X, C=O, C-O, C-C, C=C, C ≡C, C≡N, and
so on) are regularly found only in specific small portions of the vibrational infrared region. A small
range of absorption can be defined for each type of bond, but, outside this range, absorptions are
normally due to some other type of bond. By establishing what kinds of bonds are present in a
sample, one can conclude on its chemical structure (Bevin, Fergusson, Perry, Janik, & Cozzolino,
2006; Pavia et al., 2001). More specifically, Fourier-transform infrared (FT-IR) spectroscopy in the
mid-infrared region is widely used in determinations of the proteins‘ secondary structure. The
characteristic amide I band (stretching vibrations of C=O in the peptide bond) between 1700 and
1600 cm-1
(in the mid- infrared range) provides information on protein secondary structure due to a
sensitivity of the amide I frequency to the hydrogen bonding pattern and dipolar couplings in the
protein backbone. This sensitivity makes it possible to study not only protein folding and unfolding,
31
but also aggregation processes. Furthermore, proteins have complementary but much weaker
fingerprints in the NIR region (Bruun, Holm, Hansen, & Jacobsen, 2006; Carbonaro, Maselli, &
Nucara, 2012)
The main advantage of using spectroscopic techniques is the exploratory character of the
measurements that facilitates concurrent detection of several different and even non-anticipated
constituents (Holse, Larsen, Hansen, & Engelsen, 2011). Other reasons that render spectroscopic
techniques as the preferred analytical method in many laboratories include: minimal sample
preparation/ pretreatment, short analysis time, cost-effective to analyze a single sample or large
batches of samples, non-destructive methods for the samples, no laboratory or skilled operator
needed for routine analyses, no use of hazardous chemical reagents, and, depending on the method,
the results are usually more precise and can be more accurate than, as accurate as, or of acceptable
accuracy, when compared with the method usually employed (Batten, 1998; Blanco & Villarroya,
2002).
32
AIM OF THE STUDY
Legumes are a good and inexpensive source of protein. Therefore, they could be a good alternative to
meat protein in developing countries, where protein deficiency is high.
The aim of the project is to examine the protein content of legumes from Ghana, and evaluate how
some traditional processing methods (germination, fermentation) can alter that protein content, with
the perspective of including them in high-protein snacks.
33
MATERIALS AND METHODS
Screening of samples
The following samples were used in the screening process with FT-IR.
Cowpea: asomdwee, asetenapa and soronko, Soybean: anidaso, nangbaare, jenguma and quarshie,
Lima: koloenu, lima 104 and 204, Bambara groundnut: Nar-4, Mucuna bean and Jack bean.
One replicate of each sample was milled in a coffee mill, frozen at -20 oC and freeze-dried for at
least 24h. FT-IR spectroscopy method (see method below) was performed. A triple determination
was carried out for each sample. The data obtained where used in a Principal Component Analysis
(PCA) model with a scope to identify the most differentiated samples.
Six samples were selected for further analysis. Cowpea: asetenapa and soronko, Soybean: anidaso,
Bambara groundnut, Mucuna bean and Jack bean.
Preparation of samples for analyses
Germination of samples
20g of each sample were weighed in a beaker. Three replicates of each sample were weighed (4
replicates for Bambara groundnut and Jack bean due to mould infection). The samples were soaked
in 100 mL of 1% w/v citric acid solution for 18h at room temperature. The solution was drained off.
The samples were washed with distilled water to neutral pH, and soaked in 100 mL distilled water
for 2 hours at room temperature. The distilled water was drained off and the samples were reweighed
to determine the amount of water absorbed.
The soaked seeds were, then, placed in petri dishes containing two filter papers (Rundfilter MN 615,
9 cm ø, MACHEREY-NAGEL GmbH & Co. KG, Dűren, Germany), and 5 mL distilled water were
sprinkled on top of them. The petri dishes were then put into plastic bags, in order to keep humidity,
and kept in darkness at 30oC for either 48h or 72h. The samples were sprinkled with distilled water
daily.
34
The sprouted samples were ground coarsely (using a kitchen mini-chopper), frozen at -20oC and
freeze-dried for at least 24h. Next, they were milled in a coffee mill and stored in plastic containers
at room temperature.
Fermentation of samples
20g of each sample were weighed in a beaker. Three replicates were weighed for each sample. The
samples were soaked in 100 mL of 1% w/v citric acid solution for 18h at room temperature. The
solution was drained off. The samples were washed with distilled water to neutral pH, and soaked in
100 mL distilled water for 2 hours at room temperature. The distilled water was drained off and the
samples were reweighed to determine the amount of water absorbed. The legumes were placed in 50
mL Eppendorf tubes with excess distilled water and put in a waterbath a 90oC for 30 min (for
cowpea and soybean samples) or 60 min (for bambara groundnuts, mucuna beans and jack beans)
(boiling step). After boiling the distilled water was drained off. The boiled samples were ground
coarsely (using a kitchen mini-chopper) and stored in plastic bags at 5oC overnight.
The next day the samples were left on the lab bench until they acquired room temperature and were
then inoculated with 1,2 mL Bacillus subtilis var. natto solution (7*106 CFU/ mL) and 2 mL distilled
water. The samples were massaged in order to ensure that the bacteria were spread throughout the
sample. The inoculated samples were kept in darkness at either 30oC or 37
oC for 48h. They were
massaged twice a day to maximize the legume surface that the bacteria ferment.
The fermented samples were frozen at -20oC and freeze-dried for at least 24h. They were, then,
milled in a coffee mill and stored in plastic containers at room temperature.
Control samples
20g of each sample were weighed in a beaker. Three replicates were weighed for each sample. The
samples were soaked in 100 mL of 1% w/v citric acid solution for 18h at room temperature. The
solution was drained off. The samples were washed with distilled water to neutral pH, and soaked in
100 mL distilled water for 2 hours at room temperature. The distilled water was drained off and the
samples were reweighed to determine the amount of water absorbed. Next, the control samples were
35
ground coarsely (using a kitchen mini-chopper), frozen at -20oC and freeze-dried for at least 24h.
They were, then, milled in a coffee mill and stored in plastic containers at room temperature.
Defatting of soybean samples for protein determination
Prior to protein determination of the soybean samples, a defatting step was necessary in order to
reduce the fat content below 10%. The method used was modified from L‘hocine, Boye, & Arcand
(2006). 1 g of each soy sample was weighed and put in a 15 mL Eppendorf tube. 5 mL of hexane
were added in the tube and the content of the tube was mixed in a vortex mixer for at least 30 sec.
The samples were then centrifuged for 10 min at 2000 rpm and the supernatant was discarded. The
procedure was repeated two more times. The samples were air-dried in room temperature under a
fume hood for approximately 24 h.
Samples‘ assessment
FT-IR (Fourier transform infrared) spectroscopy
The absorbance measurements were performed on an Arid-Zone MB100 FT-IR instrument (ABB
Bomen, Quebec, Canada) using an Attenuated Total Reflectance (ATR) device with a triple-bounce
diamond crystal. IR spectra were recorded in the range of 4000– 530 cm−1
using a spectral resolution
of 4 cm−1
. The ground legumes were positioned on the crystal surface and squeezed towards the
diamond crystal by use of a concave needle compressor. Each spectrum represents the average of 64
scans ratioed against the background (128 scans) collected with the empty crystal and stored as
absorbance spectra. Each sample was measured in duplicate.
NIR (Near infrared) spectroscopy
The QFA flex near-infrared spectrometer (Q- Interline A/S, Tølløse, Denmark) was used to collect
the spectra. The samples were placed in a 50mm diameter sample bottle. The instrument parameters
were set as following; resolution: 4 cm-1
; Gain: C and Gain: High. Each spectrum represents the
average of 16 scans ratioed against a background (32 scans) measured on an internal white 50mm
36
diameter bottle. The spectrum measured was 4000-14000 cm-1
. Duplicate measurements were
collected from each sample.
Determination of moisture content
Moisture content was estimated by a method modified from ICC-Standard No. 110/1. Double
determination was performed in milled raw legume samples. Clean vessels were dried for 2h at
130°C and cooled down to room temperature in a desiccator. The empty closed vessels were
weighed (W1) and the weight was noted. 1 g of each sample was weighed (W2). The vessels were
placed in the drying cabinet to dry at 130°C for 2 h (180 min set – 1 hour for reaching required
temperature). After drying the vessels were placed in the desiccators (the first 5-10 min with the lid
half-open) to cool down to room temperature. After cooling down, the vessels were weighed again
(W3).
Percent moisture content was calculated as follows;
Moisture % =𝑊2 − 𝑊3 − 𝑊1 ∗ 100
𝑊2
Where,
W1 = tare weight of vessels (with cover) in grams
W2 = initial weight of sample in grams
W3 = dry weight of sample and vessel (with cover) in grams
Determination of fat content
Crude fat was determined using a modified method of AACC Method 30-25 Crude Fat in Wheat,
Corn, and Soy Flour, Feeds, and Cooked Feeds. Beakers were dried at 103°C for 30 min and cooled
down to room temperature in a desiccator. Three glass beads were added in each beaker before
weighing (W4). 3 g of milled freeze-dried sample for samples with <10% fat (cowpeas, jack bean,
bambara bean, mucuna bean) and 2g for samples with >10% fat (soybean) were weighed (W5) in the
extraction holsters. A fat-free cotton cork was added and the holsters were put in the Soxtec
apparatus (Soxtec System HT 1043 Extraction Unit, FOSS Tecator, DK) in the assigned position. 40-
37
50 mL petroleum ether HPLC grade were added in each beaker, before the beakers were also placed
in the Soxtec apparatus. The holsters with the samples were held for 15 min in boiling position,
followed by 30 min in rinsing position. The solvent was collected in the condensers; the beakers
were removed from the apparatus and left under a fume hood for 30 min for any residual solvent to
evaporate. They were then weighed again (W6).
Cotton gloves and magnet forceps were used throughout the procedure, when handling the beakers
and extraction holsters.
Percent crude fat was calculated as follows:
Crude Fat % = 𝑊6 – 𝑊4 ∗ 100
𝑊5
Where,
W4 = tare weight of beaker (with glass beads) in grams
W5 = initial weight of sample in grams
W6 = gross weight of sample fat and beaker (with glass beads) in grams
Determination of protein content
Protein content was determined by a method modified from AOAC Official Method 2001.11 Protein
(Crude) in Animal Feed, Forage (Plant Tissue), Grain, and Oilseeds. 0,5 g milled freeze-dried sample
was weighed for samples with 25-50% protein (soybean, cowpeas, mucuna bean, jack bean) and 1g
for samples with 3-25% protein (bambara groundnut), folded into N-free paper and dropped into a
Kjeldahl tube. Single determinations were carried out. Two catalyst tabs, each containing 3,5 g
K2SO4 with 0,4 g Cu catalyst (CuSO4), and 12 mL concentrated H2SO4 (98%, reagent grade) were
added into each tube. Samples were digested in 420 °C for 1 h in Kjeltec digestion block (Kjeltc
System 2020 Digestor, FOSS Tecator, DK). After digestion samples were cooled for 15-20 min. 50
mL of NaOH (40 % w/w) and 25 mL of receiver solution (H3BO3, 4 % w/v) were used for distillation
(Kjeltc System 1026 Distilling unit, FOSS Tecator, DK). Distillate was titrated with standard 0.1000
N HCl to grey endpoint. Preparing blanks, 2 catalyst tabs were mixed with 12 mL concentrated
H2SO4 and then treated identically to the samples. Blanks evaluate nitrogen from chemicals and
other extraneous sources.
38
Percent nitrogen and percent crude protein can be calculated as shown below.
% 𝑁 = 𝑇 − 𝐵 ∗ 𝑁 ∗ 14,0067 ∗ 100
𝑊
% Protein = % N * F
T – titration volume for sample, mL;
B – titration volume for blank, mL;
N – normality of HCl acid (0.1000 N);
W – sample weight, mg;
F – Conversion factor for Nitrogen to protein – 6.25
(Note: For soybean samples the sample weight was reduced to the original weight before defatting.)
Data analysis
Scatter effects in the IR and NIR raw spectra were removed by application of MSC (Multiplicative
Scatter Correction) in MatLab. The transformed spectra were then imported into LatentiX 2.12
(LatentiX Aps, Frederiksberg, Denmark). Principal Compoonent Analysis (PCA) was performed on
the mean centered spectra, with random cross validation.
Furthermore, the data on the crude protein content of the germinated, fermented, and control samples
were imported in IBM SPSS Statistics 22 (IBM Software). One-way ANOVA was performed with
protein content as the dependent variable and treatment as the independent one. The samples were
checked for their homogeneity using Levene‘s test for equality of variances. In case the assumption
for homogeneity was violated (Sig. ≤0,05) a Welch ANOVA was also performed. Post-hoc analysis
of the samples was carried out using Scheffe test. For jack bean samples Independent Samples T-test
was also performed.
39
RESULTS AND DISCUSSION
Screening Test
Figure 3: FT-IR PCA scores plot of raw beans
Figure 3 shows the scores plot of the data collected in FT-IR for various raw beans, using the first
two principal components. The explained variance for the two principal components accounts for
98,4% of the total variance. In the plot two distinct clusters can be seen; one comprising of all the
soybean samples, on the right of the X-axis, and one consisting of the rest of the bean samples, close
to (0,0). Cowpea ‗Asetenapa‘ seems to be an outlier.
Selection of the sample was made, with the intention of choosing the most diverse samples, both
according to the scores plot and according to species. One sample from the soybean cluster was
selected (soybean ‗Anidaso‘). The outlier cowpea ‗Asetenapa‘ was also included as well as another
cowpea sample (‗Soronko‘) with brown seed coat. Since the rest of the samples were not sufficiently
differentiated, the rest of the samples chosen for further analysis were some underutilized legumes of
Africa (Jack bean, Velvet bean, Bambara groundnut).
The scope of the screening test was to get a vague idea of how similar or different the bean samples
are with one another; therefore, no further analysis, regarding the absorbance bands of the spectra
was performed.
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
SB-ANI2 SB-ANI SB-2ANI SB-QUA2
SB-ANI3 SB-QUA
SB-NAN3 SB-2ANI2 SB-NAN
SB-JEN3
SB-QUA3
SB-JEN2
SB-2ANI3
SB-JEN SB-NAN2
BAM3 BAM2
Scores PC#1 (95.067%)
PCA Scores [Model 1]
BAM
CP-SOR2 CP-SOR
MUC3
MUC4
CP-SOR3
JAC
JAC2
104-4
MUC5
JAC3
MUC6
104
MUC2
LIM3
CP-ASO6
104-3
KOL
CP-ASO4 CP-ASO5
104-2 KOL3 LIM LIM2
KOL2 MUC LIM4
CP-ASE2 CP-ASE CP-ASE3
Score
s P
C#2 (
3.3
39%
) 1 0 4
ΒΒΒ
C P - A S E
C P - A S O
C P - S O R
K O L
L I M
M U C
S B - 2 A N I
S B - A N I
S B - J E N
S B - N A N
S B - Q U A
J A C
40
Proximate Composition
Table 5 summarizes the proximate composition of the raw beans used in the germination and
fermentation processes. Moisture measurement was a double determination of a single raw bean
sample, whereas fat and protein measurements were based on a single determination in the triplicate
control (raw, soaked) samples.
Table 5: Proximate composition of the raw bean samples (Moisture, Fat and Protein)
Legume Moisture% Fat% Pro %
Soybean Anidaso 10,02 19,68 42,87
Cowpea Soronko 13,13 1,06 25,90
Cowpea Asetenapa 12,74 1,49 23,97
Jack bean 13,59 1,21 29,05
Bambara groundnut 10,46 6,71 18,15
Velvet bean 10,54 2,62 29,99
The moisture content falls in the range, and in some cases slightly exceeds, the standard range of 0-
13%, as reported by James (1995). Moisture content in this range is suitable for storage and
processing of legume flours without triglyceride degradation by microorganisms (James, 1995). Jack
bean is found to have the highest moisture content (13,59%), with cowpeas following, whereas the
lowest moisture content is seen in soybeans.
Regarding the fat content of the beans, soybean is, as expected, the one containing the highest
amount of lipids (19,68%). Soybean is the only legume, among the ones studied, that is classified as
an oilseed (Kayembe & Van Rensburg, 2013). The lipid content for the rest legumes examined is in
the lower range levels, or even below them, according to the literature, with cowpea soronko
containing the least fat (1,06%).
As far as protein content is concerned, soybean has, once again, the highest content (42,87%) and
bambara groundnut the lowest (18,15%). These results are in agreement with the results of Fasoyiro
et al., (2006) that found soybeans containing significantly higher protein content (p, 0,05) than other
41
minor grain legumes. They also found bambara groundnut to have the least protein, although
according to them bambara contained 22,1% protein (Fasoyiro et al., 2006).
Fourier-Transform Infrared (FT-IR) Spectroscopy
In the raw (MSC) spectra (Figure 4) the main absorbances are seen. However, it is not possible to
distinguish between the samples, due to the high number of samples and replicates. For this reason,
the mean absorbance spectrum for each sample was calculated. The mean spectra for the control
samples of the six legumes investigated are depicted in Figure 5.
Figure 4: FT-IR spectra (MSC) of the germinated, fermented, and control samples.
1000 1500 2000 2500 3000 3500 4000
-0.01
0
0.01
0.02
0.03
0.04
0.05
cm-1
42
Figure 5: Mean FT-IR spectra (MSC) of the control legume samples. Soybean is shown in blue, Cowpea Asetenapa in green, Bambara groundnut in red, Velvet bean in light blue, Cowpea Soronko in purple, and Jack bean in yellow.
The highest peak for also legumes measured appears at approximately 1080 cm-1
. Peaks at this area
of the spectrum arise from C-N and C-O bonds. Multiple peaks are also seen in the range of 1250 –
1750 cm-1
. The peak at 1550 cm-1
is related to N=O bonds. The one at 1630 cm-1
is due to the
presence of carbonyl group (C=O) and combined with the absorbance near 3300 - 3400 cm-1
(N-H
stretches) indicate the presence of amides. In this sense, it is logical that bambara groundnut, which
has the lowest protein content, exhibits lower absorption than the other samples at those
wavenumbers.
Furthermore, bambara groundnut, together with soybean, exhibit higher peaks at 1750 cm-1
, 2850
cm-1
, and 2915 cm-1
, compared to the other samples. Peaks at 1750 cm-1
arise, once again from C=O
bonds, only this time related to the presence of esters. Absorbances at 2850 – 3000 cm-1
result from
C-H stretches. Peaks arising from C-H stretches result primarily from the CH2 groups in fats,
although C-H bonds from carbohydrates and proteins also contribute (Holse et al., 2011). Since
soybean and bambara groundnut have the highest lipid content, the highest absorbances at these
wavenumbers were not a surprise.
Mean spectra of the germinated and fermented samples for each legume were also created, in order
to examine how the change in the legumes composition affects their spectra (Data shown in
Appendix 1). For all the different legumes, the differences in the composition are mainly translated
1000 1500 2000 2500 3000 3500 4000
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
cm-1
43
as a change in the peak at 3300 – 3400 cm-1
, arising from N-H stretches and related to the primary
and secondary amines and amides. However, the trends for the protein content according to the FT-
IR spectra do not correspond to the trends according to the protein content measured with Kjeldahl.
A PCA model was calculated, using the raw (MSC) spectra. Table 6a shows the scores plot of data
using the first two principal components (PCs), for which the accumulated variance contribution rate
was up to 78, 2%. In the plot the data set was split in three groups; one containing the soybean
samples (colored in the plot), one the bambara groundnut ones (green), and one comprising of the
rest of the samples.
Figure 6: PCA scores plot of the FT-IR raw (MSC) spectra (PC-1/PC-2) (a), and PCA loading plots for PC-1 (b) and PC-2 (c). In the scores plot Soybean samples are colored red, Bambara groundnut green, Velvet bean pink, Cowpea Asetenapa blue, Cowpea Soronko light blue and Jack bean black.
The first PC (PC-1) is thought to be related to the fat content, since the samples with high fat content
(soybean, bambara groundnut) are placed on the left side of the vertical axis, taking negative score
values for PC-1, whereas the ones with lower lipid content have positive score values and are
positioned on the right side of the vertical axis. Furthermore, in the second PC (PC-2) there is a
44
tendency that the samples with the highest protein content (soybean, velvet bean) are placed at the
bottom of the plot, while the ones containing the least protein (bambara groundnut, cowpea
asetenapa) are on the top. The loading plots for PC-1 and PC-2 are depicted in Figure 6b and Figure
6c respectively.
Near infrared (NIR) spectroscopy
NIR spectroscopy measures overtones and combinations of the fundamental molecular vibrations.
The signals arise from anharmonic molecular bonds, namely bonds containing at least one hydrogen
atom. Thus, NIR is excellent at detecting the four bulk components in foods: water (O-H), fat (C-H),
protein (N-H) and carbohydrate (C-H and O-H) (Holse et al., 2011).
From the raw NIR spectra (Figure 7) a difference in the peaks of some samples can be observed.
However, it is not possible to tell which the deviating samples are. Therefore, the mean spectrum for
each sample was calculated, and the mean spectra (MSC) of the control samples are presented in
Figure 8.
Figure 7: NIR raw spectra (MSC) of the germinated, fermented, and control samples
Figure 8: Mean FT-IR spectra (MSC) of the control legume samples. Soybean is shown in blue, Cowpea Asetenapa in green, Bambara groundnut in red, Velvet bean in light blue, Cowpea Soronko in purple, and Jack bean in yellow.
From the mean spectra of the control samples in Figure 8 it seems that the deviating samples in Figure
7 are the soybean samples.
Peaks arising from C-H stretches are visible as combination bands at around 4300 – 4350 cm-1
and as
1st and 2nd overtones at 5650 – 5800 cm-1
and 8200-8400 cm-1
, respectively. As stated earlier the C-
H bonds are primarily related to the CH2 groups in fat, so it is only relevant that soybean exhibits the
highest peaks among the samples at these wavenumbers.
The broad peak at approx. 6300 – 7000 cm-1
is related to the 1st overtones of O-H and N-H stretches,
related to carbohydrates and proteins, respectively. Soybean and velvet bean, the samples with the
highest protein content, have the lowest absorptions in this area of the spectra. So, it could be
assumed that the peaks arise primarily from the O-H stretches, rather than the N-H ones.
Another interesting peak can be seen at 4750 cm-1
, for all the samples but soybean. The peak arises
from a combination band (O-H deformation/ C-O stretch) that is related to starch. Instead, soybean,
together with velvet bean, have a peak at 4600 cm-1
that arises from a combination band (2 * amide I
+ amide III) relevant to protein content.
The peak at 5170 cm-1
is an O-H stretch/ O-H deformation combination band primarily from the
carbohydrates (Holse et al., 2011).
Additional plots were made comparing the mean spectra of all treatments for each legume
individually (plots in Appendix 2). Little variations were observed between different treatments. The
main deviation is seen at 5170 cm-1
related to carbohydrates. It is well established that during
germination carbohydrates are broken down to provide energy for the development of sprouts
(Donangelo et al., 1995; Mugendi & Njagi, 2010). Furthermore, during fermentation microorganisms
hydrolyze non-digestible carbohydrates (poly- and oligo-saccharides) into sugars (Parkouda et al.,
2009).
A PCA model was calculated using the raw (MSC) spectra of all the samples. Figure 9 shows the
scores plot for the first two PCs, accounting for 58,3% and 33,2% of the explained variance,
respectively. The samples are separated in three distinct clusters in the scores plot; soybeans (pink)
are the first cluster, velvet beans (blue) the second and the rest of the samples comprise the third
cluster. It seems as PC-1 (PC-1 loadings plot shown in Figure 10a) is highly correlated to the protein
content, since the beans with the highest protein content are placed on the left of the X- axis, whereas
the rest on the right.
Figure 9: PCA scores plot of the NIR raw (MSC) spectra (PC-1/ PC-2). Soybean samples are colored pink, Bambara groundnut red, Velvet bean blue, Cowpea Asetenapa light blue, Cowpea Soronko green and Jack bean black.
It is not clear how the samples are plotted according to PC-2, but a possible explanation could be that
the samples with the lowest carbohydrate content are placed in the bottom of the plot; especially
since the PC-2 loadings plot (Figure 10b) has a distinct peak at around 4750 cm-1
Figure 11: PCA scores plot of the NIR raw (MSC) spectra (PC-1/ PC-3). Soybean samples are colored pink, Bambara groundnut red, Velvet bean blue, Cowpea Asetenapa light blue, Cowpea Soronko green and Jack bean black.
In the scores plot two well defined groups can be distinguished. One on the left side of the X-axis,
consisting of the soybean and velvet bean samples, and the other on the right side of the X-axis
comprising of the rest of the samples. The grouping of the samples is related to their protein content,
since in the former group there are the legumes with the highest protein content, while in the latter
the ones containing lower amounts of protein.
In general, from both the FT-IR and NIR plots it can be concluded that soybean is the most
differentiated legume. This is logical, since, according to the literature, soybeans contain higher
amounts of protein and lipids and lower amounts of carbohydrates compared to the rest of the
samples (See Table 2).
Effect of fermentation and germination on protein content
The protein content for each sample was determined using the Kjeldahl method. The results are
presented in Table 6.
Table 6: % Protein content (DM) of the germinated, fermented and control legume samples (dry weight basis)
% Protein (DM)
Control Germination
48h
Germination
72h
Fermentation
30°C
Fermentation
37°C
Soybean
Anidaso 1 42,87 ± 1,05
a 47,79 ± 2,17 b
(11,5)
46,95 ± 0,37 ab
(9,5)
50,29 ± 1,78 b
(17,3)
49,99 ± 1,68 b
(16,6)
Cowpea
Soronko 1
25,90 ± 0,00 a
26,49 ± 0,44 ab
(2,3)
28,12 ± 0,20 b
(8,6)
27,95 ± 0,41 ab
(7,9)
27,77 ± 1,48 ab
(7,2)
Cowpea
Asetenapa 1
23,97 ± 0,18 a
24,45 ± 0,36 a
(2,0)
25,79 ± 0,10 b
(7,6)
25,85 ± 0,10 b
(7,8)
26,72 ± 1,20 c
(11,5)
Jack bean 2 29,05 ± 0,17
a
29,14 ± 1,57 ab
(0,3)
31,20 ± 1,13 b
(7,4)
31,07 ± 0,57 b
(7,0)
29,12 ± 1,17 b
(0,2)
Bambara
groundnut 1
18,15 ± 0,31 a
19,76 ± 0,11 b
(8,9)
19,65 ± 0,37 b
(8,3)
20,28 ± 0,41 b
(11,7)
20,86 ± 1,10 b
(14,9)
Velvet bean 1 29,99 ± 0,73
a
31,74 ± 0,37 b
(5,8)
31,86 ± 0,63 b
(6,2)
33,14 ± 0,39 b
(10,5)
33,03 ± 0,53 b
(10,1)
Mean values ± St. Deviation
The numbers in the parentheses refer to the %increase of the protein content compared to the protein content of the control group.
The same superscript in the same row means no significant difference (p≤0,05)
1. The statistical significance was calculated with One-way ANOVA and Scheffe post-hoc test for multiple comparisons.
2. The statistical significance was calculated with independent-samples T-test for each pair of samples.
49
For each legume, an increase in the protein content is seen in the germinated and fermented samples
compared to the control ones. In many cases the increase in the protein content was found to be
statistically significant, according to the analysis of variance (ANOVA). However, the small amount
of samples (3-4) for each group deteriorates the importance of the results. For example, although
Jack bean samples varied significantly according to ANOVA, the post-hoc test failed to reveal where
the significant difference lies. For this reason, maybe a look at the tendencies, by visualization of the
results, could be more appropriate (Figure 12).
Figure 12: % Protein content (DM) of the germinated, fermented and control samples of (a) Soybean Anidaso, (b) Cowpea Soronko, (c) Cowpea Asetenapa, (d) Jack bean, (e) Bambara groundnut and (f) Velvet bean. Coloumn order from left to right: Raw, Germinated 48 hours, Germinated 72 hours, Fermented at 30°C and Fermented at 37°C.
50
The protein content in the germinated samples was found to be higher than the control ones, for all
legumes and this can -in most cases- already be seen from the column charts (Figure 12). This came
to no surprise; most researchers have observed an increase in the protein content of legumes, and
most of the times this increase is statistically significant (El-adawy, 2002; Ghavidel & Prakash,
2007; V. a Obatolu, 2002). Only a few found no changes (Trugo et al., 1999), or even slight decrease
(not significant) (Martínez-Villaluenga et al., 2007; Torres et al., 2007). In the present study the
highest increase in germinated samples was that of soybeans germinated for 48h (11,5%) and the
lowest at jack beans (0,3%), also at 48h. The %increase in protein content for the 48h was highly
deviating, but for 72h it was in the range of 6,2% (velvet bean) to 9,5% (soybean). The big difference
in %increase (6,3 – 7,1%) in the protein content between 48h and 72h, for jack beans and both
varieties of cowpeas, might mean that 48h are not enough for the germination of these legumes. In
general, most of the legumes exhibited higher protein contents after 72h of fermentation, rather than
48h. However, soybean and bambara groundnut had a highest protein content at 48h of germination,
compared to 72h, although the difference was not statistically significant. This is in accordance with
the results of Akpapunam et al. (1996); in their study they germinated soybeans and bambara
groundnuts for 1 – 5 days and they found an increase in the protein content up to 48h and a decrease
thereafter. Nevertheless, they removed the vegetative parts prior to the analysis, so that could be the
reason for their results.
For the fermentation, the increase observed was even higher (Figure 12). Once again, the highest
increase was observed in soybean samples fermented at 30oC (17,3%) and the lowest at jack beans
fermented at 37oC (0,2%). Mugendi & Njagi (2010) examined the effect of fermentation and
germination on the protein content of velvet beans and found that fermentation resulted in 17,87%
versus 3,13% for germination. Other than that, the results are controversial. Many researchers have
found a significant increase in protein in fermented legumes (Azeke et al., 2005; Baik & Han, 2012;
Reyes-Moreno et al., 2004); but there is also a great deal of them that observed a decreased protein
content after fermentation (Allagheny et al., 1996; Granito et al., 2005; Marisela Granito & Alvarez,
2006). The fermentation conditions and the microorganisms used in the studies varied a lot, and that
could be the reason for the contradicting results. The bacteria used in the present study was
B.subtilis, a bacteria naturally found in many traditional fermented products in Africa (Parkouda et
al., 2009).
In this study, there is no doubt that fermentation lead to elevated levels of protein. What is
inconclusive is the effect of temperature on the outcome. Some of the legumes exhibited better
results, regarding the protein content, at 30oC and other at 37
oC. An error during the experimental
work, resulted in one Cowpea Asetenapa sample being fermented for 24h at 30oC followed by 24h at
51
37oC, instead of 48h at 37
oC; nevertheless, the protein content of this sample was comparable to the
other Asetenapa samples fermented at 37oC.
Among the samples, soybeans and bambara groundnuts were the ones with the highest %protein
increase, both for germination and for fermentation, distinctively higher than that of the other
legumes. Both these legumes had the highest fat content. Fat is one of the storage compounds in
seeds that is degraded during germination (Mugendi & Njagi, 2010), and it known that Bacillus spp.
possess lipolytic activity, to a different extend depending on the strain (Ouoba et al., 2003). So,
perhaps the high protein is a result of altered protein proportion on a dry weigh basis rather than
actual increase (Akpapunam et al., 1996).
All in all, the germination and fermentation processes were deemed adequate, although
improvements could be made in the experimental procedure.
52
CONCLUSIONS
The study shows that the proximate composition (protein, fat, moisture) of the legumes examined
falls within the range according to the literature available. Soybean is the most differentiated legume,
of the ones studied, in regards to both fat and protein content, and overall according to spectroscopic
techniques (FT-IR, NIR).
Both germination and fermentation increased the protein content of the legumes considerably.
Germination for 72h resulted in higher protein levels compared to 48h of germination. In fermented
samples, the evidence on the effect of temperature in yielding higher protein contents was
inconclusive. In general fermentation brought about higher protein levels compared to germination.
Spectroscopic methods, showed a change in the protein (FT-IR) and carbohydrate (NIR) content of
the germinated and fermented samples, compared to the control ones.
PERSPECTIVES
Further investigation of the effect of temperature and time on the protein content of germinated and
fermented samples could lead to improved processing techniques.
Additionally, the determination of true protein, instead of crude protein, could be of interest since
alkaline fermentation results in elevated levels of ammonia (Allagheny et al., 1996), that could
tamper with the protein levels as determined by Kjeldahl.
Last but not least, the effect of germination and fermentation on the amino acid profile and protein
digestibility of the legumes could be of utmost importance in designing nutritious and acceptable
food products based on processed legumes.
53
BIBLIOGRAPHY
Adebowale, Y. A., Adeyemi, I. A., & Oshodi, A. A. (2005). Variability in the physicochemical,
nutritional and antinutritional attributes of six Mucuna species. Food Chemistry, 89(1), 37–48.
Adebowale, Y. A., Schwarzenbolz, U., & Henle, T. (2011). Protein Isolates from Bambara
Groundnut ( Voandzeia Subterranean L.): Chemical Characterization and Functional Properties.
International Journal of Food Properties, 14(4), 758–775.
Agbede, J. O., & Aletor, V. A. (2005). Studies of the chemical composition and protein quality
evaluation of differently processed Canavalia ensiformis and Mucuna pruriens seed flours.
Journal of Food Composition and Analysis, 18(1), 89–103.
Ahenkora, K., Dadzie, M., & Osei-Bonsu, P. (1999). Composition and functional properties of raw
and heat processed velvet bean ( Mucuna pruriens (L.) DC. var utilis) flours. International
Journal of Food Science & Technology, 34(2), 131–135.
Akpapunam, M. A., & Achinewhu, S. C. (1985). Effects of cooking, germination and fermentation
on the chemical composition of Nigerian cowpea (Vigna unguiculata). Plant Foods for Human
Nutrition, 35, 353–358.
Akpapunam, M. A., Igbedioh, S. O., & Aremo, I. (1996). Effect of malting time on chemical
composition and functional properties of soybean and bambara groundnut flours. International
Journal of Food Sciences and Nutrition, 47(1), 27–33.
Alexandratos, N. (1995). World Agriculture: Towards 2010. A FAO Study. New York: Food and
Agriculture Organisation of the United Nations and John Wiley and Sons.
Ali, M., & Kumar, S. (2000). Problems and Prospects of Pulses Research in India. Indian Farming,
50, 4–13.
Allagheny, N., Obanu, Z. A., Campbell-Platt, G., & Owens, J. D. (1996). Control of ammonia
formation during Bacillus subtilis fermentation of legumes. International Journal of Food
Microbiology, 29(2-3), 321–333.
Amarowicz, R., & Pegg, R. B. (2008). Legumes as a source of natural antioxidants. European
Journal of Lipid Science and Technology, 110(10), 865–878.
Amarteifio, J. O., Tibe, O., & Njogu, R. M. (2006). The mineral composition of bambara groundnut
(Vigna subterranea (L) Verdc) grown in Southern Africa. African Journal of Biotechnology,
5(23), 2408–2411.
Antova, G. A., Stoilova, T. D., & Ivanova, M. M. (2014). Proximate and lipid composition of
cowpea (Vigna unguiculata L.) cultivated in Bulgaria. Journal of Food Composition and
Analysis, 33(2), 146–152.
Aranda, P., Dostalova, J., Frias, J., Lopez-Jurado, M., Kozlowska, H., Pokorny, J., … Zdyunczyk, Z.
(2001). Nutrition. In C. L. Hedley (Ed.), Carbohydrates in Grain Legume Seeds. Improving
54
Nutritional Quality and Agronomic Characteristics (pp. 61–87). Wallingford, UK: CAB
International.
Arun, A., Sridhar, K., Raviraja, N., Schmidt, E., & Jung, K. (2003). Nutritional and antinutritional
components of Canavalia spp. seeds from the west coast sand dunes of India. Plant Foods for
Human Nutrition, 58(3), 1–13.
Atiim, P. (2011). Assessment of Postharvest Losses in Soybeans Production in the Builsa District in
the Upper East Region and Savelugu District in the Northern Region of Ghana. Kwame
Nkrumah University of Science and Technology.
Aykroyd, W. R., & Doughty, J. (1982). Legumes in human nutrition. Food and Agriculture
Organization of the United Nations. FAO Food Nutrition Papers, 20, 1–152.
Azeke, M. a, Fretzdorff, B., Buening-Pfaue, H., Holzapfel, W., & Betsche, T. (2005). Nutritional
value of African yambean (Sphenostylis stenocarpa L): improvement by lactic acid
fermentation. Journal of the Science of Food and Agriculture, 85(6), 963–970.
Baik, B., & Han, I. (2012). Cooking, Roasting, and Fermentation of Chickpeas, Lentils, Peas, and
Soybeans for Fortification of Leavened Bread. Cereal Chemistry, 89(6), 269–275.
Bamshaiye, O. M., Adegbola, J. A., & Bamishaiye, E. I. (2011). Bambara groundnut : an Under-
Utilized Nut in Africa. Advances in Agricultural Biotechnology, 1, 60–72.
Barampama, Z., & Simard, R. E. (1995). Effects of soaking, cooking and fermentation on
composition, in-vitro starch digestibility and nutritive value of common beans. Plant Foods for
Human Nutrition (Dordrecht, Netherlands), 48(4), 349–365.
Batten, G. D. (1998). Plant analysis using near infrared reflectance spectroscopy: the potential and
the limitations. Australian Journal of Experimental Agriculture, 38, 697–706.
Bau, H., Villaume, C., Nicolas, J.-P., & Mejean, L. (1997). Effect of germination on chemical
composition, biochemical constituents and antinutritional factors of soya bean (Glycine max)
seeds. Journal of the Science of Food and Agriculture, 73, 1–9.
Bevin, C. J., Fergusson, A. J., Perry, W. B., Janik, L. J., & Cozzolino, D. (2006). Development of a
rapid ―fingerprinting‖ system for wine authenticity by mid-infrared spectroscopy. Journal of
Agricultural and Food Chemistry, 54(26), 9713–8.
Bhat, R., & Karim, A. A. (2009). Exploring the Nutritional Potential of Wild and Underutilized
Legumes. Comprehensive Reviews in Food Science and Food Safety, 8(4), 305–331.
Blanco, M., & Villarroya, I. (2002). NIR spectroscopy: a rapid-response analytical tool. Trends in
Analytical Chemistry, 21(4), 240–250.
Bojňanská, T., Frančáková, H., Líšková, M., & Tokár, M. (2012). Legumes – The Alternative Raw
Materials For Bread Production. Journal of Microbiology, Biotechnology and Food Sciences,
1(February special issue), 876–886.
55
Bourne, L., Langenhoven, M., Steyn, K., Jooste, P., Laubscher, J., & Van der Vyver, E. (1993).
Nutrient intake in the urban African population of the Cape Peninsula, South Africa. The Brisk
study. The Central African Journal of Medicine, 39(12), 238–247.
Bressani, R., & Elfas, L. G. (1980). Nutritional values of legume crops for humans and animals. In
R. J. Summerfield & A. H. Bunting (Eds.), Advances in legume science (pp. 57–66). Londom:
FAO.
Brough, S. H., & Azam-Ali, S. N. (1992). The effect of soil moisture on the proximate composition
of bambara groundnut (Vigna subterranea (L) Verdc). Journal of the Science of Food and
Agriculture, 60(2), 197–203.
Bruun, S. W., Holm, J., Hansen, S. I., & Jacobsen, S. (2006). Application of near-infrared and
Fourier transform infrared spectroscopy in the characterization of ligand-induced conformation
changes in folate binding protein purified from bovine milk: influence of buffer type and pH.
Applied Spectroscopy, 60(7), 737–46.
Carbonaro, M., Maselli, P., & Nucara, A. (2012). Relationship between digestibility and secondary
structure of raw and thermally treated legume proteins: a Fourier transform infrared (FT-IR)
spectroscopic study. Amino Acids, 43(2), 911–921.
Champ, M. M.-J. (2002). Non-nutrient bioactive substances of pulses. The British Journal of
Nutrition, 88 Suppl 3, 307–19.
Czarnecka, M., Czarnecki, Z., Nowak, J., & Roszyk, H. (1998). Effect of lactic fermentation and
extrusion of bean and pea seeds on nutritional and functional properties. Nahrung / Food,
42(01), 7–11.
Dalgetty, D. D., & Baik, B.-K. (2003). Isolation and Characterization of Cotyledon Fibers from Peas,
Lentils, and Chickpeas. Cereal Chemistry, 80(3), 310–315.
Donangelo, C. M., Trugo, L. C., Trugo, N. M. F., & Eggum, B. O. (1995). Effect of germination of
legume seeds on chemical composition and on protein and energy utilization in rats. Food
Chemistry, 53, 23–27.
Doyle, J. J. (1994). Phylogeny of the legume family: an approach to understanding the origins of
nodulation. Annual Review of Ecology and Systematics, 25, 325–349.
El-adawy, T. A. (2002). Nutritional composition and antinutritional factors of chickpeas ( Cicer
arietinum L .) undergoing different. Plant Foods for Human Nutrition, 57, 83–97.
Ezeagu, I. E., & Gowda, L. R. (2006). Protein extractability, fractionation and amono acid
composition of some leguminous seeds found n Nigeria. Journal of Food Biochemistry, 30(1),
1–11.
Famurewa, J. A. V., & Raji, A. O. (2005). Parameters Affecting Milling Qualities of Undefatted
Soybeans ( Glycine max , l . Merill ) ( 1 ): Selected Thermal Treatment. International Journal
of Food Engineering, 1(4), 1–9.
56
FAO. (2014). FAO Statistical Yearbook 2014 - Africa Food and Agriculture. Accra: Food and
Angriculture Organization of the United Nations, Regional Office for Africa.
FAO/WHO. (1990). Protein quality evaluation. Report of a joint FAO/WHO expert consultation.
Bethesda, MD, 4 - 8 December 1989.
Farzana, W., & Khalil, I. A. (1999). Protein quality of tropical food legumes. Journal of Science and
Technology, 23, 13–19.
Fasoyiro, S. B., Ajibade, S. R., Omole, a. J., Adeniyan, O. N., & Farinde, E. O. (2006). Proximate,
minerals and antinutritional factors of some underutilized grain legumes in south-western
Nigeria. Nutrition & Food Science, 36(1), 18–23.
Ghavidel, R. A., & Prakash, J. (2007). The impact of germination and dehulling on nutrients,
antinutrients, in vitro iron and calcium bioavailability and in vitro starch and protein
digestibility of some legume seeds. LWT - Food Science and Technology, 40(7), 1292–1299.
Gómez, C. (2004). Cowpea: Post-harvest Operations. Rome, Italy: Food and Agriculture
Organisation of United Nations.
Granito, M., & Alvarez, G. (2006). Lactic acid fermentation of black beans (Phaseolus vulgaris) :
microbiological and chemical characterization. Journal of the Science of Food and Agriculture,
86, 1164–1171.
Granito, M., Torres, A., Frias, J., Guerra, M., & Vidal-Valverde, C. (2005). Influence of
fermentation on the nutritional value of two varieties of Vigna sinensis. European Food
Research and Technology, 220(2), 176–181.
Grusak, M. A. (2002). Enhancing Mineral Content in Plant Food Products. Journal of the American
College of Nutrition, 21(sup3), 178S–183S.
Guillon, F., & Champ, M. (1996). Grain Legumes and Transit in Humans. Grain Legumes AEP, 11,
18.
Gulewicz, P., Martínez-Villaluenga, C., Frias, J., Ciesiołka, D., Gulewicz, K., & Vidal-Valverde, C.
(2008). Effect of germination on the protein fraction composition of different lupin seeds. Food
Chemistry, 107(2), 830–844.
Hansen, E. B. (2004). Microorganisms. In Y. H. Hui (Ed.), Handbook of food and beverage
fermentation technology (pp. 10–26). New York: Marcel Dekker.
Hillocks, Bennett, C., & Mponda, O. M. (2012). Bambara nut: A review of utilization, market
potential and crop improvement. African Crop Science Journal, 20(1), 1–16.
Holse, M., Larsen, F. H., Hansen, Å., & Engelsen, S. B. (2011). Characterization of marama bean
(Tylosema esculentum) by comparative spectroscopy: NMR, FT-Raman, FT-IR and NIR. Food
Research International, 44(1), 373–384.
57
Honfoga, B. G., & van den Boom, G. J. (2003). Food-consumption patterns in central West Africa,
1961 to 2000, and challenges to combating malnutrition. Food and Nutrition Bulletin, 24(2),
167–82.
Hui, Y. H. (2012). Fermented Plant Products and Their Manufacturing. In Y. H. Hui, E. Ö. Evranuz,
& F. N. Arroyo-López (Eds.), Handbook of Plant-Based Fermented Food and Beverage
Technology (2nd ed., pp. 3–21). Boca Raton, Fla.: CRC Press.
Iqbal, A., Khalil, I. a., Ateeq, N., & Sayyar Khan, M. (2006). Nutritional quality of important food
legumes. Food Chemistry, 97(2), 331–335.
James, C. S. (1995). Analytical chemistry of foods (pp. 39–80). New York, USA: Chapman and Hall
Publishers.
Josephine, M. R., & Janardhanan, K. (1992). Studies on chemical composition and anti-nutritional
factors in three germplasm seed materials of the tribe pulse, Mucuna pruriens (L.) DC. Journal
of Food Chemistry, 43, 13–18.
Josephsen, J., & Jespersen, L. (2004). Starter Cultures and Fermented Products. In Y. H. Hui (Ed.),
Handbook of food and beverage fermentation technology (pp. 27–59). New York: Marcel
Dekker.
Kabagambe, E. K., Baylin, A., Ruiz-Narvarez, E., Siles, X., & Campos, H. (2005). Decreased
consumption of dried mature beans is positively associated with urbanization and nonfatal acute
myocardial infarction. The Journal of Nutrition, 135(7), 1770–1775.
Kayembe, N. C., & Van Rensburg, C. J. (2013). Germination as a processing technique for soybeans
in small-scale farming. South African Journal of Animal Science, 43(2), 167–173.
Khattab, R. Y., Arntfield, S. D., & Nyachoti, C. M. (2009). Nutritional quality of legume seeds as
affected by some physical treatments, Part 1: Protein quality evaluation. LWT - Food Science
and Technology, 42(6), 1107–1112.
Kiers, J. L., Nout, R. M. J., & Rombouts, F. M. (2000). In vitro digestibility of processed and
fermented soya bean , cowpea and maize. Journal of the Science of Food and Agriculture,
1331(January), 1325–1331.
Kiers, J. L., Van Laeken, a E., Rombouts, F. M., & Nout, M. J. (2000). In vitro digestibility of
Bacillus fermented soya bean. International Journal of Food Microbiology, 60(2-3), 163–9.
Kolapo, A. L. (2011). Soybean : Africa ‘ s Potential Cinderella Food Crop. In Prof. Tzi-Bun Ng
(Ed.), Soybean - Biochemistry, Chemistry and Physiology. InTech.
Kolapo, A. L., & Oladimeji, G. R. (2008). Production and quality evaluation of Soy-corn milk.
Journal of Applied Biosciences, 1(2), 40–45.
Kolapo, A. L., & Sanni, M. O. (2009). A comparative evaluation of the macronutrient and
micronutrient profiles of soybean-fortified gari and tapioca. Food and Nutrition Bulletin, 30(1),
90–94.
58
Kuo, Y.-H., Rozan, P., Lambein, F., Frias, J., & Vidal-Valverde, C. (2004). Effects of different
germination conditions on the contents of free protein and non-protein amino acids of