-1DWQ6FL)RXQGDWLRQ6UL/DQND
DOI: http://dx.doi.org/10.4038/jnsfsr.v47i3.8699
RESEARCH ARTICLE
&RPSDULVRQRILQWHUYDULHWDOGLႇHUHQFHVLQFKHPLFDOFRPSRVLWLRQ
DQGQXWULWLRQDOSURSHUWLHVRIFRFRQXWWHVWDÀRXU
S.S.K. Marasinghe1, J.M. Nazrim Marikkar1*, C. Yalegama2, S. Wimalasiri3, G. Seneviratne1,
R. Weerasooriya1 and R. Liyanage1
1DWLRQDO,QVWLWXWHRI)XQGDPHQWDO6WXGLHV+DQWDQD5RDG.DQG\
&RFRQXW5HVHDUFK,QVWLWXWH/XQXZLOD
'HSDUWPHQWRI)RRG6FLHQFHDQG7HFKQRORJ\)DFXOW\RI$JULFXOWXUH8QLYHUVLW\RI3HUDGHQL\D3HUDGHQL\D
Submitted: 28 January 2019; Revised: 13 May 2019; Accepted: 24 May 2019
Abstract: Coconut testa is the brown colour thin outer covering
of the coconut endosperm. An attempt was made to convert
FRFRQXW WHVWD LQWR ÀRXU IRU EDNHU\ SURGXFWV ,Q WKLV VWXG\
chemical composition and nutritional properties of coconut
WHVWD ÀRXU RI IRXU ORFDO FXOWLYDUV QDPHO\ 6DQ 5DPDQ *RQ
Thambili, Ran Thambili and Tall × Tall were compared against
WKRVH RI D FRPPHUFLDO K\EULG JURZQ LQ 6UL /DQND 3DUWLDOO\
defatted coconut parings of each cultivar were oven-dried and
JURXQG LQWR FRFRQXW WHVWD ÀRXU 0RLVWXUH FUXGH IDW DVK DQG
FUXGHSURWHLQFRQWHQWVRIFRFRQXWWHVWDÀRXUZHUHGHWHUPLQHG
according to AOAC methods. The carbohydrate content was
FDOFXODWHGE\WKHGL൵HUHQFH,QWHUYDULHWDOGL൵HUHQFHVRIIDWW\
acids and micro-mineral distributions were also determined.
7KH PDLQ FRQVWLWXHQW RI FRFRQXW WHVWD ÀRXU UHJDUGOHVV RI
cultivar was carbohydrate (42.55–59.24 %) followed by protein
(23.82–32.22 %) and fat (7.93–23.49 %). Commercial hybrid
had the highest carbohydrate content (59.24 %) while the
minimum carbohydrate content was recorded for San Raman
variety (42.55 %). Highest protein content was observed in
Gon Thambili (32.22 %) variety while the least was observed
in commercial hybrid (23.82 %). The highest fat content was
noted in San Raman variety (23.49 %). Tall × Tall variety
contained the least fat content (7.93 %). Maximum ash content
was observed in Ran Thambili variety (5.30 %) while the least
ash content was for Gon Thambili variety (3.70 %). Highest
moisture content was prevalent in San Raman variety (4.27 %)
while the least was observed in commercial hybrid (2.27 %).
7KHVHUHVXOWVVXJJHVWHGWKDWFRFRQXWWHVWDÀRXULVDQXWULWLRXV
substance, which provides value addition to the under-utilised
by-product of coconut processing industry.
*
Corresponding author (nazrim.ma@nifs.ac.lk;
Keywords: &RFRQXW FRFRQXW WHVWD ÀRXU IDWW\ DFLGV PLFUR
minerals, proximate composition.
INTRODUCTION
Coconut (&RFRVQXFLIHUD L.) is a crop grown in more than
85 countries worldwide, with a total production of 54
billion nuts per annum. The island country of Sri Lanka is
WKHZRUOG¶V¿IWKODUJHVWSURGXFHURIFRFRQXWV&RFRQXWLV
grown mainly in the traditional coconut triangle, although
patches of coconut cultivation could also be seen in other
parts of the country. The endemic coconut germplasm of
Sri Lanka consists mainly of three varieties; typica (tall
palm), nana (dwarf palm), and aurantica (king coconut
palm). These varieties are generally distinguishable based
on their morphological characters as well as the breeding
habits (Liyanage, 1958; Fernando, 1999). Despite the
PDMRU FKDUDFWHULVWLF GL൵HUHQFHV WKHUH DUH RWKHU PLQRU
variations within each variety, which lead them to be
FODVVL¿HGIXUWKHULQWRGL൵HUHQWIRUPVRIFRFRQXWNQRZQ
as cultivars.
As a perennial crop, coconut is one of the most
economically important crops in the tropics, serving as
a source of food, drink, fuel, medicine, and construction
material (Lima et al., 2015). An average mature
Sri Lankan type coconut is composed of about 45 % husk,
https://orcid.org/0000-0002-0133-5852)
This article is published under the Creative Commons CC-BY-ND License (http://creativecommons.org/licenses/by-nd/4.0/).
This license permits use, distribution and reproduction, commercial and non-commercial, provided that the original work is
properly cited and is not changed in anyway.
66.0DUDVLQJKHHWDO
13 % shell, 22 % meat and 20 % water (Marikkar et al.,
2009). The most economically important part of coconut
is its endocarp; the hard dark core of the fruit. Inside
this part is a solid white endosperm of varied thickness,
depending on the maturity of the fruit (Lima et al., 2015).
In an earlier report, Nathanael (1966) pointed out that
the coconut endosperm had some unique features such
that the layer closest to the water cavity was least rich
in oil (~56.3%) while the layer nearest to the brown
testa was richest in oil (~75.4 %). Later investigations
into this aspect showed that the oil characteristics of the
EURZQWHVWDZHUHVOLJKWO\GL൵HUHQWIURPWKRVHRIWKHRLO
characteristics of inner layers of the kernel (Marikkar &
Nasyrah, 2012).
of each cultivar were subjected to cold press oil extraction
using a micro oil expeller (Komet DD85 machine,
*HUPDQ\ 3DUWLDOO\GHIDWWHGFRFRQXWWHVWD OHVVWKDQ
RLOFRQWHQW ZHUHJURXQGLQWR¿QHFRFRQXWWHVWDÀRXUXVLQJ
a grinder. The entire process was carried out according to
WKHVHTXHQFHLOOXVWUDWHGLQ)LJXUH7KHJURXQGHGÀRXU
samples were then stored at refrigerated (4 °C) condition
for further analysis. All chemicals used in this study were
RIDQDO\WLFDOJUDGHXQOHVVRWKHUZLVHVSHFL¿HG
6HDVRQHGQXWV
'HKXVNLQJ
Coconut testa is the brown coloured thin outer
covering of the coconut endosperm. It is an underutilised
by-product of the desiccated coconut industry; it is often
XVHGDVDQLPDOIHHG3UHYLRXVLQYHVWLJDWLRQVE\$SSDLDK
et al. (2014) showed that it is a rich source of bioactive
FRPSRXQGV VXFK DV SKHQROLFV DQG ÀDYDQRLGV ,W KDV
potential to be used as a functional ingredient by the food
SURFHVVLQJLQGXVWU\$OWKRXJKGHYHORSPHQWRIÀRXURXW
of coconut kernel has been the interest of researchers for
preparation of snack foods (Yalegama et al., 2013), the
VWXGLHVRQXWLOLVDWLRQRIFRFRQXWWHVWDDVDVRXUFHRIÀRXU
is scanty. Hence, the aim of this study was to evaluate
LQWHUFXOWLYDU GL൵HUHQFHV LQ SUR[LPDWH FRPSRVLWLRQ
mineral content and fatty acid distribution of coconut
WHVWDÀRXURIIRXULQGLJHQRXVFRFRQXWFXOWLYDUVQDPHO\
Gon Thambili (GT), Ran Thambili (RT), San Raman
(SR), Tall × Tall (T×T) and commercial hybrid coconut
(COM). It is believed that this information would be
YLWDOIRUKHOSLQJWRGHYHORSFRFRQXWWHVWDÀRXUDVDUDZ
material for nutritional improvement of the Sri Lankan
society.
'HVKHOOLQJ
3DULQJ
&RFRQXWWHVWD
'U\LQJ
2LOH[WUDFWLRQ
METHODOLOGY
Coconuts of twelve-month maturity were collected from
¿YHGL൵HUHQWORFDOFXOWLYDUV LH*757657î7DQG
COM) maintained at the varietal blocks of the Coconut
Research Institute, Lunuwila, Sri Lanka during the period
August 2018 to October 2018. Fifty nuts of each cultivar
were sampled for seasoning followed by de-husking.
Shells of the nuts were removed manually using
hammers while de-pairing was done using manually
operated knives. The fresh testa of individual cultivars
were disintegrated separately to medium size particles
using a disintegrator (Unitex Engineers, Sri Lanka). The
disintegrated parings were then dried at 70 °C using a
cabinet-type dehydrator (Wessberg, Martin, Germany)
for 8 hrs. Two kilogram samples of dried coconut testa
2LO
'HIDWWHGPHDO
*ULQGLQJ
7HVWDIORXU
Figure 1: 3URFHVVÀRZGLDJUDPIRUSURGXFWLRQRIFRFRQXWWHVWDÀRXU
)LJXUH3URFHVVIORZGLDJUDPIRUSURGXFWLRQRIFRFRQXWWHVWDIORXU
September 2019
Journal of the National Science Foundation of Sri Lanka 47(3)
$QDO\VLVRIFRFRQXWWHVWDÀRXU
Analysis of proximate composition
Moisture, crude fat, crude protein and ash content of
FRFRQXWWHVWDÀRXUZHUHGHWHUPLQHGDFFRUGLQJWRPHWKRGV
described in AOAC (2005) manual. The carbohydrate
FRQWHQW RI WKH ÀRXU ZDV FDOFXODWHG E\ WKH GL൵HUHQFH
[100- (crude protein + crude fat + ash + moisture+ crude
¿EUH @
Analysis of micro-minerals
'LJHVWLRQRIÀRXUVDPSOHVZDVFDUULHGRXWLQDPLFURZDYH
digester (CEM MARS 6, USA) with the addition of 3 mL
RI QLWULF DFLG WR J RI ÀRXU 7KH GLJHVW ZDV
¿OWHUHGLQWRDP/YROXPHWULFÀDVNDQGPDGHXSWR
mark with distilled water. This solution was used for
DQDO\VLVRIPLFURPLQHUDOVXVLQJ,&3VSHFWURSKRWRPHWHU
7KHUPRVFLHQWL¿FL&$3VHULHV86$
$QDO\VLVRIIDWW\DFLGSUR¿OH
A sample portion of oil (0.4 g) was weighed into a screw
capped glass tube and 4.0 mL portion of methanol and 0.1
mL portion of methanolic KOH were added. The mixture
was heated to 60 °C in a water bath for 10 min and
allowed to cool. Into this, 2 mL portion of hexane and 4
mL portion of distilled water were added. Contents were
agitated at 2500 rpm for 10 min in a vortex. After allowing
the contents to separate into layers, the upper layer was
injected into a gas chromatograph (GC-2010 Shimadzu
&RUSRUDWLRQ-DSDQ ¿WWHGZLWKDÀDPHLRQLVDWLRQGHWHFWRU
(FID). The temperature of the oven was programmed
as follows: the initial temperature was 130 °C (1 min
hold), then increased to 170 °C (6.5 °C min-1), 170 °C
to 215 °C (2.75 °C min-1) and maintained at 215 °C for
12 min. Thereafter, the temperature was increased from
215 °C to 230 °C (4 °C min-1) and maintained at 230 °C
for 3 min. Temperatures of the injector and detector
were maintained at 270 °C and 280 °C, respectively.
+\GURJHQ ZDV XVHG DV WKH FDUULHU JDV DW D ÀRZ UDWH RI
43 cm/sec. Split ratio of the injector was 50:1. Retention
time of each peak was compared with that of standard
fatty acid methyl esters to identify individual fatty acids.
The percentage of each fatty acid was calculated by
dividing the peak area of the individual fatty acid by the
total of the peak areas gained for all fatty acids.
Statistical analysis
All the results from analyses were expressed as the
mean value ± standard deviation. Data were statistically
analysed by one-way analysis of variance (ANOVA)
using Tukey’s test of MINITAB (version 14) statistical
package at 0.05 probability level.
Journal of the National Science Foundation of Sri Lanka 47(3)
RESULTS AND DISCUSSION
Moisture
Moisture is the most abundant component of most plant
foods and is also a crucial factor to determine the shelflife stability of processed products (Coultate, 2009).
According to the data presented in Table 1, the mean
PRLVWXUH FRQWHQW RI WKH WHVWD ÀRXU RI FRFRQXW FXOWLYDUV
was found to range from 1.8 to 4.6 %. The moisture
FRQWHQW RI 65 ZDV VLJQL¿FDQWO\ KLJKHU WKDQ WKRVH RI
*7 7î7 DQG 57 7KHUH ZDV QR VLJQL¿FDQW GL൵HUHQFH
among the mean moisture contents of GT, T×T and
RT. However, the mean moisture content of COM
ZDVVLJQL¿FDQWO\ORZHUWKDQWKRVHRI*77î7DQG57
(p < 0.003). When compared with previous reports, the
PRLVWXUHFRQWHQWVRIFRFRQXWWHVWDÀRXUXVHGLQWKLVVWXG\
ZHUHORZHUWKDQWKRVHRIFRPPHUFLDOZKHDWÀRXU 1DVLU
et al., 2UJDQLVPV QDWXUDOO\ SUHVHQW LQ WKH ÀRXU
VWDUW WR JURZ DW KLJK PRLVWXUHV SURGXFLQJ R൵ RGRXUV
DQG ÀDYRXUV +HQFH 1DVLU et al. (2003) suggested that
ZKHDW ÀRXU KDYLQJ OHVV WKDQ PRLVWXUH ZRXOG EH
appropriate for extended shelf life. Further, more mold
growth and insect infestation has been noticed in wheat
ÀRXUKDYLQJKLJKHUPRLVWXUHGXULQJVWRUDJH
Protein
3URWHLQV DUH WKH WKLUG PRVW DEXQGDQW FODVV RI
macromolecules in food systems; they perform
numerous biological functions in living systems (Chang,
1998). According to the data presented in Table 1,
WKH PHDQ SURWHLQ FRQWHQW RI WKH WHVWD ÀRXU RI FRFRQXW
cultivars was found to range from 23.8 to 32.2 %. The
lowest protein content was found with COM while the
highest was recorded for GT. However, there was no
VLJQL¿FDQW GL൵HUHQFH EHWZHHQ WKH SURWHLQ FRQWHQWV RI
RT and SR cultivars. The mean protein content of T×T
ZDVVLJQL¿FDQWO\KLJKHUWKDQWKRVHRI5765DQG&20
3UHYLRXVUHVHDUFKHUVKDYHH[DPLQHGWKHSURWHLQFRQWHQW
RI GHIDWWHG FRFRQXW ÀRXU REWDLQHG IURP WKH ZKROH
endosperm, but not the protein content of the testa.
In an early report, Ediriweera and Kashizumi (1991)
pointed out that the whole endosperm of fresh coconut
has about 4 % protein and the value might increase in
defatted meals after extraction of milk. In a study on
mixed coconut types, Yalegama and Chavan (2006)
IRXQG WKDW FRFRQXW ÀRXU REWDLQHG DIWHU RLO H[WUDFWLRQ
of the whole kernel has around 18 to 20 % of protein.
According to another report, Beansch et al. (2004)
UHSRUWHG WKDW GHIDWWHG FRFRQXW ÀRXU REWDLQHG DIWHU WKH
extraction of virgin coconut oil contained about 20 %
protein; the value was higher than those reported for
September 2019
66.0DUDVLQJKHHWDO
FRPPHUFLDOO\PLOOHGZKHDWÀRXUZKLFKFRQWDLQHGDERXW
10.33 % protein. The protein content is an important
SDUDPHWHU IRU EUHDG PDNLQJ DV ÀRXU FRQWDLQLQJ KLJKHU
SURWHLQ FRQWHQWV ZRXOG EH PRUH H[SHQVLYH WKDQ ÀRXUV
Table 1:
of lower protein content. Another important feature of
FRFRQXWWHVWDÀRXULVWKDWLWVSURWHLQLVJOXWHQIUHHZKLFK
is advantageous for people with celiac disease or gluten
intolerance.
,QWHUYDULHWDOGL൵HUHQFHVLQSUR[LPDWHFRPSRVLWLRQRIFRFRQXWWHVWDÀRXURIGL൵HUHQWFRFRQXWFXOWLYDUV
3DUDPHWHU
Moisture (%)
Ash (%)
Crude protein (%)
Crude fat (%)
Total carbohydrates
E\GL൵HUHQFH
SR
GT
&XOWLYDU
RT
T×T
COM
4.27 ± 0.31c
5.00 ± 0.57c,d
24.69 ± 0.74a
23.49 ± 4.91d
42.55
3.40 ± 0.53b
3.70 ± 0.14a
32.22 ± 2.48c
13.41 ± 4.56c
47.27
3.07 ± 0.11b
5.30 ± 0.14d
25.39 ± 0.25a
13.28 ± 0.06c
52.96
2.80 ± 0.40b
4.20 ± 0.00b
28.37 ± 0.00b
7.93 ± 2.22a
56.7
2.27 ± 0.42a
4.50 ± 0.14c
23.82 ± 0.99a
10.17 ± 1.84b
59.24
(DFKYDOXHLQWKHWDEOHUHSUHVHQWVWKHPHDQRIWKUHHUHSOLFDWHV0HDQVZLWKLQHDFKURZEHDULQJGL൵HUHQWVXSHUVFULSWVDUH
VLJQL¿FDQWO\ S GL൵HUHQW
SR - San Raman; GT - Gon Thambili; RT - Ran Thambili; T×T - Tall×Tall; COM - commercial hybrid
Fat
Fatty acid distribution
Dietary fat or lipid is one of the most important
macronutrients that provides energy and essential fatty
acids to various functions of the human body (Raihana
et al., 2015). Fat contents of food usually vary from
very low to high depending on the source of origin,
variety, geographical location, etc. (De Man, 1999). The
data presented in Table 1 compared the inter-varietal
GL൵HUHQFHV RI IDW FRQWHQW DPRQJ WKH ORFDOO\ DYDLODEOH
coconut cultivars. The mean fat content of the testa
ÀRXURIFRFRQXWFXOWLYDUVZDVIRXQGWRYDU\IURP
to 23.49 %. According to literature, previous researchers
KDYHH[DPLQHGWKHIDWFRQWHQWRIFRFRQXWÀRXUREWDLQHG
from the whole endosperm, but not the fat content of the
WHVWD WR FRPSDUH WKH LQWHUYDULHWDO GL൵HUHQFHV RI ORFDO
cultivars. For instance, Yalegama and Chavan (2006)
UHSRUWHGWKDWFRFRQXWÀRXUREWDLQHGDIWHURLOH[WUDFWLRQ
of the whole kernel had about 10 to 13 % fat. In a
separate communication, Beansch et al. (2004) stated
WKDWGHIDWWHGFRFRQXWÀRXUREWDLQHGDIWHUWKHH[WUDFWLRQ
of virgin coconut oil contained about 12.0 % fat (w/w,
dry basis). According to Najwa et al. (2017), the fat
content of defatted coconut residue left after extraction
of coconut milk was found to be 17.26 %. All these
indicated that the method of preparation or the nature of
VDPSOLQJ LQ GL൵HUHQW VWXGLHV FRXOG KDYH FRQWULEXWHG WR
WKHREVHUYHGYDULDWLRQLQIDWFRQWHQWRIGL൵HUHQWFRFRQXW
WHVWDÀRXUVDPSOHV
The fatty acid distributions of oils extracted from testa
ÀRXURIGL൵HUHQWFXOWLYDUVZHUHFRPSDUHGDVVKRZQLQ
Table 2. The oil samples consisted of 88.75–91.23 %
saturated fatty acids (SFA) and 8.76–11.19 % unsaturated
IDWW\DFLG 86)$ $PRQJWKHGL൵HUHQWFXOWLYDUVODXULF
acid was the dominant fatty acid (42.65–45.97 %),
followed by myristic acid (19.69–21.46 %) and palmitic
acid (9.42–10.24 %). In a previous study reporting the
composition of coconut testa oil, lauric acid (42.28 %)
was found to be the predominant fatty acid, followed
by myristic acid (18.99 %) and palmitic (11.57 %) acid
(Zhang et al., 2015). In another study to compare the
composition of coconut testa oil and ordinary coconut oil,
Marikkar and Nasyrah (2012) observed the proportion
of fatty acids in the order of lauric > myristic > palmitic
DFLGV+RZHYHUVRPHGL൵HUHQFHVZHUHREVHUYHGLQWKH
proportional distribution of fatty acids in the oils of
copra testa and wet-coconut testa; they took the order
of lauric > myristic > oleic acids (Appaiah et al., 2014).
With reference to the report of Appaiah et al. (2014), the
mean proportion of lauric acid observed in the present
study (44.92 %) was slightly higher than those of copra
testa (40.9 %) and wet-coconut testa (32.4 %). When
FRPSDUHGWRRLOVIURPFRFRQXWWHVWDÀRXURIWKHSUHVHQW
study (7.02 %), copra testa (12.2 %) and wet-coconut
testa (17.8 5%) had considerably higher oleic acid
contents. However, the mean percentage of myristic
September 2019
Journal of the National Science Foundation of Sri Lanka 47(3)
$QDO\VLVRIFRFRQXWWHVWDÀRXU
Table 2: ,QWHUYDULHWDOGL൵HUHQFHVLQIDWW\DFLGFRPSRVLWLRQVRIFRFRQXWWHVWDÀRXURIGL൵HUHQWFRFRQXWFXOWLYDUV
Fatty acid
C8:0
C10:0
C12:0
C14:0
C16:0
C18:0
C18:1
C18:2
SFA
USFA
SR
RT
Cultivar
GT
T×T
COM
7.62 ± 0.01b
4.96 ± 0.01c
45.48 ± 0.03b,c
20.25 ± 0b
9.84 ± 0.03b
2.23 ± 0.69a
7.22 ± 0.65a,b
2.38 ± 0b
90.39 ± 0.66b
9.6 ± 0.65a
8.05 ± 0.08c
5.32 ± 0.03d
45.97 ± 0.23c
19.69 ± 0.06a
9.42 ± 0.11a
2.55 ± 0.35a
6.63 ± 0.20a
2.34 ± 0.04b
91.01 ± 0.16b
8.98 ± 0.17a
7.60 ±0.01b
4.72 ± 0.11b
45.23 ± 0.1b
21.46 ± 0.03c
9.55 ± 0.06a
2.66 ± 0.09a
6.27 ± 0.05a
2.49 ± 0.07b
91.23 ± 0.02b
8.76 ± 0.02a
8.14 ± 0.03c
5.21 ± 0.04d
45.29 ± 0.11b
20.2 ± 0.04b
9.43 ± 0.04a
2.52 ± 0.11a
7.09 ± 0.05a,b
2.09 ± 0.04a
90.80 ± 0.08b
9.19 ± 0.08a
7.33 ± 0.01a
4.45 ± 0.03a
42.65 ± 0.21a
21.15 ± 0.21c
10.24 ± 0.04c
2.92 ± 0.02a
7.91 ± 0.01b
3.28 ± 0.02c
88.75 ± 0.06a
11.19 ± 0.01b
(DFKYDOXHLQWKHWDEOHUHSUHVHQWVWKHPHDQRIWKUHHUHSOLFDWHV0HDQVZLWKLQHDFKURZEHDULQJGL൵HUHQWVXSHUVFULSWVDUH
VLJQL¿FDQWO\ S GL൵HUHQW
SR - San Raman; GT - Gon Thambili; RT - Ran Thambili; T×T - Tall×Tall; COM - commercial hybrid
C:8 - caprylic; C:10 - caproic; C12:0 - lauric, C14:0 - myristic, C16:0 - palmitic, C18:0 - stearic; C18:1 - oleic; C18:2
- linoleic; SFA - saturated fatty acid; USFA - unsaturated fatty acid
acid (20.55 %) in the present study was comparatively
similar to those of copra testa (20.9 %) and wet-coconut
testa (20.2 %).
7KH LQWHUYDULHWDO GL൵HUHQFHV RI WKH GLVWULEXWLRQ
of individual fatty acids among cultivars RT, GT, SR,
T×T and COM are of considerable interest in nutrition.
Generally, in this study there is no particular pattern of
change among the distribution of various fatty acids.
6LJQL¿FDQWGL൵HUHQFHVZHUHQRWLFHGDPRQJWKHFXOWLYDUV
with regard to the distribution of fatty acids such as
FDSU\OLFDFLGDQGFDSURLFDFLGDOWKRXJKWKHGL൵HUHQFHV
were minute. The proportion of lauric acid was highest
for RT while the same for COM was lowest. Likewise,
the proportion of myristic acid was highest for GT while
WKHVDPHIRU57ZDVORZHVW$PRQJDOO¿YHFXOWLYDUVWKH
proportions of stearic acid and unsaturated fatty acids such
as oleic and linoleic acids were low. As a result, coconut
WHVWDÀRXURIWKHVHFXOWLYDUVPLJKWGLVSOD\EHWWHUVKHOI
OLIHVWDELOLW\WKDQZKHDWÀRXU,WLVEHFDXVHFRFRQXWWHVWD
ÀRXU ZLWK OHVV DPRXQWV RI XQVDWXUDWHG IDWW\ DFLGV WKDQ
ZKHDWÀRXUZRXOGEHFRPHOHVVSURQHWRDXWRR[LGDWLRQ
In a previous study, Nikolic et al. (2015) reported that the
SUHGRPLQDQWIDWW\DFLGRIZKHDWÀRXUZDVOLQROHLFDFLG
(66.57 %), followed by palmitic acid (15.36 %) and oleic
acid (13.34 %). According to another report by Nikolic
et al WKH PDMRU IDWW\ DFLG RI ZKHDW ÀRXU ZDV
found to be linoleic acid (57.67 %), followed by oleic
acid (20.28 %) and palmitic acid (19.56 %).
Journal of the National Science Foundation of Sri Lanka 47(3)
Ash
Ash is the composite material of minerals present in
ÀRXU 'HWHUPLQDWLRQ RI WKH DVK DQG PLQHUDO FRQWHQW RI
foods is important for a number of reasons. For instance,
the quality of many foods depends on the concentration
and type of minerals they contain, including the taste,
DSSHDUDQFHWH[WXUHDQGVWDELOLW\3UHYLRXVVWXGLHVKDYH
VKRZQWKDWDVKFRQWHQWRIZKHDWÀRXUYDULHVIURPDERXW
1.50 to 2.00 % (NDSU, 2018). It is generally accepted
WKDW WKH DVK FRQWHQW RI ÀRXU GRHV QRW D൵HFW WKH EDNLQJ
performance in majority of the cases (Borla et al., 2004).
The data presented in Table 1 shows that ash contents of
the samples ranged from 3.6 to 5.4 %. Mean ash content
RI*7ZDVVLJQL¿FDQWO\ S ORZHUWKDQWKRVHRI
65 57 7î7 DQG &20 +RZHYHU QR VLJQL¿FDQW S !
GL൵HUHQFHVZHUHQRWLFHGDPRQJPHDQDVKFRQWHQWV
of SR, RT, T×T and COM. In a previous study, Yalegama
et al. (2013) observed the changing pattern of ash content
DPRQJFRFRQXWÀRXURUUHVLGXHVDPSOHVREWDLQHGIURP
GL൵HUHQW PHWKRGV RI SURFHVVLQJ )RU LQVWDQFH WKH DVK
contents of coconut residue samples obtained after milk
H[WUDFWLRQ E\ WZR GL൵HUHQW PDFKLQHV ZHUH IRXQG WR EH
1.5 % (Yalegama et al., 2013) and 0.54 % (Najwa et al.,
7KHVHGL൵HUHQFHVFRXOGEHGXHWRWKHGL൵HUHQFHLQ
H[WUDFWLRQH൶FLHQFLHVRIWKHWZRPLONH[WUDFWRUPDFKLQHV
used by these two groups. However, higher ash content
YDOXHVZHUHQRWLFHGIRUFRFRQXWÀRXUREWDLQHGWKURXJK
virgin coconut oil extraction. This could probably
September 2019
66.0DUDVLQJKHHWDO
be because more minerals are washed away during
the aqueous exaction of coconut milk, while they are
retained with the defatted residue coming from coconut
oil extraction.
Micro-mineral distribution
&RPSDULQJ PLFURPLQHUDO GLVWULEXWLRQ LQ IRRG VWX൵V LV
generally important for the assessment of nutritional
values. Although they are required in minute quantities,
micro minerals are essential to catalyse enzymatic
biochemical reactions of various metabolisms. The
data presented in Table 3 compares the distribution of
PLFURPLQHUDOVSUHVHQWLQWKHFRFRQXWWHVWDÀRXURIWKH
cultivars. Mn was the most prevalent mineral (73.71–
94.1 mg/kg), followed by Zn (29.65–57.34 mg/kg) and
Cu (29.94–45.14 mg/kg). According to previous studies,
whole coconut kernel was known to possess minerals
such as Fe, Cu, Mn and Zn (Yalegama et al., 2013). Mn
is an essential micro-mineral that acts as a cofactor to
many enzymes involved in bone formation and various
other metabolic processes. It is said to be present in trace
amounts in a variety of food items such as nuts, whole
grains, and some vegetables.
In this study, Mn content of the samples ranged
from 73.71 to 94.1 mg/kg; there was no statistically
VLJQL¿FDQW GL൵HUHQFH DPRQJ WKH PHDQ 0Q FRQWHQW RI
RT, GT and COM. However, the mean Mn content of
65 ZDV VLJQL¿FDQWO\ S KLJKHU WKDQ WKRVH RI
RT, GT and COM. Meanwhile the mean Mn content
RI 7î7 ZDV VLJQL¿FDQWO\ S ORZHU WKDQ WKRVH
of RT, GT and COM. Zn was the next most abundant
PLFURPLQHUDOGHWHFWHGLQFRFRQXWWHVWDÀRXURIWKHORFDO
FRFRQXWFXOWLYDUV$FFRUGLQJWRVFLHQWL¿FVWXGLHV=QLV
WKH FRIDFWRU IRU PDQ\ HQ]\PHV D൵HFWLQJ JURZWK DQG
GLJHVWLRQLWVGH¿FLHQF\FDQOHDGWRJURZWKUHWDUGDWLRQ
sexual immaturity and impaired immune response
(Coultate, 2009). Generally, protein containing foods
are a good source of zinc. In this study, Zn contents
of the samples ranged from 29.65–57.34 mg/kg. There
ZDV QR VWDWLVWLFDOO\ VLJQL¿FDQW S ! GL൵HUHQFH
in the mean Zn contents among SR, RT, GT, T×T and
COM. The next important micro-mineral is Cu, which
plays an important role in several enzymatic reactions
(Coultate, 2009). It is a constituent of enzymes such
as tyrosinase, cytochrome oxidase, ascorbic acid
oxidase, uricase, monoamine oxidase, etc. Legumes,
ZKROHJUDLQVQXWVVKHOO¿VKDQGVHHGVDUHVRPHRWKHU
sources that provide Cu in human nutrition. Since Cu
is a transition metal, citrus fruit juices might help Cu
DEVRUSWLRQWKURXJKPHWDOFKHODWLQJH൵HFW$FFRUGLQJWR
7DEOH&XFRQWHQWVRIFRFRQXWWHVWDÀRXURIFXOWLYDUV
were found to range from 29.94–45.14 mg/kg. The
mean Cu content was highest for RT while the mean
Cu content decreased in order of RT, SR and GT in a
VWDWLVWLFDOO\ VLJQL¿FDQW S PDQQHU +RZHYHU
WKHUH ZDV QR VLJQL¿FDQW GL൵HUHQFH EHWZHHQ WKH PHDQ
Cu content of T×T and COM. The data presented in
Table 3 shows that Fe contents of the samples were in
the range of 0.48–2.6 mg/kg. Fe is an essential mineral
for hemoglobin and myoglobin, which are part of
the oxygen transport system of the human body. Iron
EDODQFHLVDOVRFULWLFDOIRUEUDLQIXQFWLRQWKHGH¿FLHQF\
might lead to tiredness, fatigue and anemia (Coultate,
5HGPHDWV¿VKSRXOWU\HJJVDQGOHJXPHVDUH
Table 3: ,QWHUYDULHWDOGL൵HUHQFHVLQPLQHUDOFRPSRVLWLRQRIFRFRQXWWHVWDÀRXURIGL൵HUHQWFXOWLYDUV
Mineral element
Ni (mg/kg)
Zn (mg/kg)
Mn (mg/kg)
Cr (mg/kg)
Co (mg/kg)
Cu (mg/kg)
Fe (mg/kg)
Ba (mg/kg)
Mo (mg/kg)
SR
RT
Cultivar
GT
T×T
COM
6.54 ± 0.67c
46.7 ± 6.50b
93.82 ± 0.38d
6.78 ± 0.65
0.15 ± 0.57
38.7 ± 0.70
2.60 ± 0.16c
1.1 ± 0.02c
0.26 ± 0.01b
6.69 ± 1.5c
54.5 ± 4.00c
89.20 ± 0.41c
7.56 ± 0.32
0.21 ± 0.03
44.7 ± 0.70
1.38 ± 0.41b
1.19 ± 0.11c
0.31 ± 0.08b
3.17 ± 78.4b
44.5 ± 8.50b
85.62 ± 3.02b
1.97 ± 0.02
0.11 ± 0.00
34.8 ± 0.90
0.91 ± 0.64a
0.52 ± 0.09b
0.12 ± 0.00a
2.12 ± 0.10a
33.4 ± 5.30a
75.30 ± 2.25a
0.68 ± 0.02
0.07 ± 0.01a
30.5 ± 0.80a
0.48 ± 0.26a
0.36 ± 0.10a
0.12 ± 0.01a
3.89 ± 0.04b
36.2 ± 0.70a
83.60 ± 1.43b
0.35 ± 0.02a
0.16 ± 0.00
31.0 ± 0.70a
0.61 ± 0.00a
0.51 ± 0.03b
0.09 ± 0.03a
(DFK YDOXH LQ WKH WDEOH UHSUHVHQWV WKH PHDQ RI WKUHH UHSOLFDWHV 0HDQV ZLWKLQ HDFK URZ EHDULQJ GL൵HUHQW VXSHUVFULSWV DUH
VLJQL¿FDQWO\ S GL൵HUHQW
SR - San Raman; GT - Gon Thambili; RT - Ran Thambili; T×T - Tall×Tall; COM - commercial hybrid
September 2019
Journal of the National Science Foundation of Sri Lanka 47(3)
$QDO\VLVRIFRFRQXWWHVWDÀRXU
usually good sources of Fe. The mean Fe contents of
RT (1.38 ± 0.41 mg/kg), GT (0.91 ± 0.64 mg/kg), T×T
(0.48 ± 0.26 mg/kg) and COM (0.61 ± 0.00 mg/kg) were
PRUH RU OHVV HTXDO EXW VLJQL¿FDQWO\ ORZHU S
than that of SR (2.60 ± 0.16 mg/kg).
Total carbohydrate content
Total carbohydrates consist of multiple nutrients, which
LQFOXGH GLHWDU\ ¿EUH VXJDUV DQG VWDUFKHV 7KH GDWD
presented in Table 1 compares the total carbohydrate
FRQWHQWVRIFRFRQXWWHVWDÀRXUREWDLQHGIURP¿YHORFDO
coconut cultivars. The total carbohydrate contents of
the samples were found to range from 59.24–42.55 %;
the mean total carbohydrate content was lowest for SR
variety, while the highest value of the same was recorded
IRU &20 7KH LQWHUYDULHWDO GL൵HUHQFHV RI WRWDO
FDUERK\GUDWHV DPRQJ WKH FXOWLYDUV ZHUH VLJQL¿FDQWO\
S GL൵HUHQW7KHFDUERK\GUDWHFRQWHQWRIFRFRQXW
WHVWDÀRXULVJHQHUDOO\ORZHUWKDQWKDWRIWUDGLWLRQDOJUDLQ
ÀRXUVVXFKDVZKHDWÀRXU.DVVHJQ UHSRUWHGWKDW
J RI ZKHDW ÀRXU PLJKW FRQWDLQ DURXQG ± J RI
carbohydrates while David et al. (2015) found that the
WRWDOFDUERK\GUDWHFRQWHQWRIVRIWZKHDWÀRXUZDVDURXQG
83 %. According to previous reports of Yalegama and
Chavan (2006), the total carbohydrate content of coconut
ÀRXUZDVDURXQG,QDVHSDUDWHVWXG\%HDQVFKet al.
(2004) also reported that the total carbohydrate content
RI GHIDWWHG FRFRQXW ÀRXU REWDLQHG DIWHU WKH H[WUDFWLRQ
of virgin coconut oil was about 52.0 % (w/w, dry
basis). The occurrence of higher proportions of fat and
SURWHLQLQFRFRQXWÀRXUZRXOGEHDUHDVRQIRUWKHORZHU
proportion of total carbohydrates. In addition, the quality
RIFDUERK\GUDWHLVDOVRGHSHQGHQWRQLWV¿EUHFRQWHQWDQG
JO\FHPLF LQGH[ $V LW FRQWDLQV HQRXJK ¿EURXV PDWWHU
it can also be useful as a thickening agent in sauces or
soups. In a previous study, Leelavathi and Rao (1993)
UHSRUWHGWKDWZKHDWÀRXUFRQWDLQHGDORZDPRXQWRIWRWDO
GLHWDU\ ¿EUH DQG WKHUHIRUH VXSSOHPHQWDWLRQ RI ZKHDW
ÀRXU ZLWK GHIDWWHG FRFRQXW ÀRXU FRXOG LQFUHDVH WKH
GLHWDU\¿EUHFRQWHQWRIIRRGIRUPXODWLRQV
CONCLUSIONS
,Q WKLV VWXG\ LQWHUYDULHWDO GL൵HUHQFHV RI FKHPLFDO
composition and nutritional properties of coconut testa
ÀRXU RI GL൵HUHQW LQGLJHQRXV FXOWLYDUV ZHUH FRPSDUHG
,QJHQHUDOFRFRQXWWHVWDÀRXURIDOOFXOWLYDUVGLVSOD\HG
KLJKHU FRQWHQWV RI SURWHLQ DQG IDW WKDQ ZKHDW ÀRXU
KHQFH SDUWLDO VXEVWLWXWLRQ RI ZKHDW ÀRXU ZLWK FRFRQXW
WHVWD ÀRXU ZRXOG LPSURYH WKH QXWULWLRQDO TXDOLW\ RI
ÀRXUEDVHGSURGXFWV$PRQJWKHGL൵HUHQWFXOWLYDUVWKH
Journal of the National Science Foundation of Sri Lanka 47(3)
highest protein content was observed in GT while the
lowest protein value was observed for COM variety.
The highest fat content was noted in SR while least fat
content was found in T×T. The maximum ash content
was found in RT while the lowest ash content was in
GT variety. COM hybrid had the highest carbohydrate
content while the lowest carbohydrate content was
UHFRUGHGIRU65YDULHW\,QJHQHUDOFRFRQXWWHVWDÀRXU
of all cultivars contained micro-minerals such as Mn,
&X DQG =Q 7KHUH ZHUH QRWLFHDEOH FXOWLYDU GL൵HUHQFHV
with regard to mineral composition. Highest amounts
of Fe and Mn were present in SR while the highest
content of Cu and Zn were found with RT. All these
¿QGLQJV VXJJHVW WKDW FRFRQXW WHVWD ÀRXU FDQ EHFRPH D
potential source for value addition purposes and reduce
the wastage of under-utilised coconut testa generated by
the coconut processing sector.
Acknowledgements
Authors gratefully acknowledge provision of coconut
samples and analytical services by the Coconut Research
Institute of Sri Lanka.
REFERENCES
AOAC International (2005). 2৽FLDO 0HWKRGV RI $QDO\VLV RI
$2$& ,QWHUQDWLRQDO, 18th edition. AOAC International,
Rockville, Maryland, USA.
$SSDLDK36XQLO/.XPDU3.3 .ULVKQD$**
Composition of coconut testa, coconut kernel and its oil.
-RXUQDO RI WKH $PHULFDQ 2LO &KHPLVWV¶ 6RFLHW\ 91: 917–
924.
DOI: https://doi.org/10.1007/s11746-014-2447-9
Baensch W., Yalegama C. & Jayasundera J.M.M.A. (2004).
New technologies of coconut processing-part I: process for
production of virgin coconut oil and low fat / high protein
FRFRQXWÀRXUIURPFRFRQXWNHUQHOSS±3URFHHGLQJ
,, RI WKH ,QWHUQDWLRQDO &RFRQXW &RQIHUHQFH (eds.
76* 3HLULV &6 5DQDVLQJKH &RFRQXW 5HVHDUFK
Institute of Sri Lanka, Lunuwila.
%RUOD 23., Motta E.L., Saiz A.I. & Fritz R. (2004). Quality
parameters and baking performance of commercial gluten
ÀRXUV/:7)RRG6FLHQFHDQG7HFKQRORJ\ 37: 723–729.
DOI: https://doi.org/10.1016/j.lwt.2004.02.013
Chang S.K.C. (1998). )RRG$QDO\VLV (ed. S.S. Nielson). Aspens
3XEOLVKHU,QF86$
&RXOWDWH73 )RRG The Chemistry of its Components.
The Royal Society of Chemistry, Cambridge, UK.
'DYLG2$UWKXU(6DPXHO2.%DGX( 3DWULFN6
3UR[LPDWH FRPSRVLWLRQ DQG VRPH IXQFWLRQDO SURSHUWLHV
RI VRIW ZKHDW ÀRXU ,QWHUQDWLRQDO -RXUQDO RI ,QQRYDWLYH
5HVHDUFK LQ 6FLHQFH (QJLQHHULQJ DQG 7HFKQROogy
4: 753–758.
September 2019
De Man J.M. (1999). 3ULQFLSOHV RI )RRG &KHPLVWU\. Van
Nostrand Reinhold, New York, USA.
Ediriweera N. & Hashizumi K. (1991). Extractability of
coconut protein and changes in its gel forming properties
on heating. -RXUQDORIWKH1DWLRQDO6FLHQFH&RXQFLORI6UL
/DQND19: 115–128.
DOI: https://doi.org/10.4038/jnsfsr.v19i2.8159
)HUQDQGR :08 $3&&%8527523&2*(17
3URMHFW RQ WKH DVVHVVPHQW RI WKH SHUIRUPDQFH RI KLJK
yielding coconut varieties (HYCVs)/hybrids and the
varietal preferences of coconut farmers: a country report
for Sri Lanka. Asian and 3DFL¿F &RFRQXW &RPPXQLW\,
Yogyakarta, Indonesia.
Kasseng H.H. (2018). Determination of proximate composition
and bioactive compounds of the Abyssinian purple wheat.
&RJHQW)RRGDQG$JULFXOWXUH4(1): 1421415.
DOI: https://doi.org/10.1080/23311932.2017.1421415
/HHODYDWKL. 5DR3+ 'HYHORSPHQWRIKLJK¿EHU
biscuits using wheat bran. -RXUQDO RI )RRG 6FLHQFH DQG
7HFKQRORJ\ 30: 187–190.
Lima E.B.C., Sousa C.N.S., Meneses L.N., Ximenes
N.C., Santos Júnior M.A., Vasconcelos G.S., Lima
1%& 3DWURFtQLR 0&$ 0DFHGR ' 9DVFRQFHORV
S.M.M. (2015). &RFRV QXFLIHUD (L.) (Arecaceae): a
phytochemical and pharmacological review. Brazilian
-RXUQDORI0HGLFDODQG%LRORJLFDO5HVHDUFK48: 953–964.
DOI: https://doi.org/10.1590/1414-431x20154773
Liyanage D.V. (1958). Varieties and forms of the coconut palm
grown in Ceylon. &H\ORQ&RFRQXW4XDUWHUO\ 9: 1–10.
Marikkar J.M.N., Banu M.K.I. & Yalegama C. (2009).
(YDOXDWLRQ RI WKH PRGL¿HG&H\ORQ FRSUD NLOQ IRU
accelerated production of ball copra. ,QWHUQDWLRQDO )RRG
5HVHDUFK-RXUQDO 16: 175–181.
Marikkar J.M.N. & Madurapperuma W.S. (2012). Coconut. ,Q
7URSLFDODQG6XEWURSLFDO)UXLWV3RVWKDUYHVW3K\VLRORJ\
3URFHVVLQJDQG3DFNDJLQJ (ed. Muhammed Siddiq). John:LOH\3XEOLVKLQJ&R$PHV,RZD86$
DOI: https://doi.org/10.1002/9781118324097.ch9
Marikkar J.M.N. & Nasyrah A.R. (2012). Distinguishing
coconut oil from coconut pairing oil using principle
component analysis of fatty acid data. ,QWHUQDWLRQDO
-RXUQDORI&RFRQXW5HVHDUFKDQG'HYHORSPHQW 28: 9–13.
Nasir M., Masood S.B., Faquir M.A., Sharief K. &
Minhas R. (൵HFWRIPRLVWXUHRQVKHOIOLIHRIZKHDW
ÀRXU,QWHUQDWLRQDO-RXUQDORI$JULFXOWXUHDQG%LRORJ\ 5:
458–459.
Nathanael W.R.N. (1966). Moisture and other quality factors of
copra. &H\ORQ&RFRQXW4XDUWHUO\ 17: 1–41.
NDSU (2018). Available at KWWSVZZZQGVXHGXIDFXOW\
September 2019
66.0DUDVLQJKHHWDO
VLPVHNZKHDWÀRXUKWPO. Accessed 4 December 2018.
1LNROLF15DGXORYLü10RPFLORYLF%1LNROLF*6/D]LF
M.L. & Todorovic Z. (2008). Fatty acids composition
and rheology properties of wheat and wheat and white or
EURZQ ULFH ÀRXU PL[WXUH (XURSHDQ )RRG 5HVHDUFK DQG
7HFKQRORJ\ 227: 1543–1548.
DOI: https://doi.org/10.1007/s00217-008-0877-z
1LNROLü16WRMDQRYLü-0DVWLORYLü-/D]Lü0.DUDEHJRYLü
, 6WRMDQRYLü * 5KHRORJ\ SURSHUWLHV DQG
acylglycerols and fatty acid composition of the wheat
ÀRXU VXSSOHPHQWHG ZLWK %ROHWXV HGXOLV ÀRXU $GYDQFHG
7HFKQRORJLHV 4: 79–85.
DOI: https://doi.org/10.1007/s13197-016-2448-9
Nur Ain Najwa N.M., Abbasiliasi S., Marikkar J.M.N.,
Lamasudin D.U., Shuhaimi M., Arbakariya A., Mehnoursh
A. & Yazid A.M.M. (2017). Defatted coconut residue crude
polysaccharides as potential prebiotics on proliferation and
acidifying activity of probiotics LQYLWUR-RXUQDORI)RRG
6FLHQFHDQG7HFKQRORJ\ 54: 164–173.
DOI: https://doi.org/10.1007/s13197-016-2448-9
Raihana A.R., Marikkar J.M.N., Ismail A. & Musthafa S.
(2015). A review on food values of selected fruits’ seeds.
,QWHUQDWLRQDO-RXUQDORI)RRG3URSHUWLHV 18: 2380–2392.
DOI: https://doi.org/10.1080/10942912.2014.980946
5DPDVZDP\/ 3UHSDUDWLRQRIFRFRQXWÀRXULWVNHHSLQJ
quality and acceptability in recipes. ,QGLDQ&RFRQXW-RXUQDO
36: 13–16.
5DQDVLQJKH &6 :LPDODVHNDUD 5 GH 6DUDP 36$
)HUQDQGR :3.) 3UHVHUYDWLRQ RI \RXQJ NLQJ
coconuts (&RFRVQXFLIHUD var. DXUDQWLDFD) during simulated
sea shipment. $VHDQ)RRG-RXUQDO 12: 175–181.
Ranasinghe C.S., Wimalasekara R. & Jayasekara C. (1999).
(൵HFW RI VWRUDJH WHPSHUDWXUH DQG ZUDSSLQJ WUHDWPHQWV
on the keeping quality of tender king coconut. &RFRV 13:
16–22.
DOI: https://doi.org/10.4038/cocos.v13i0.2171
Yalegama L.L.W.C., Karunaratne D.N., Sivakanesan R. &
Jayasekara C. (2013). Chemical and functional properties
RI¿EUHFRQFHQWUDWHVREWDLQHGIURPE\SURGXFWVRIFRFRQXW
kernel. )RRG&KHPLVWU\ 141: 124–130.
DOI: https://doi.org/10.1016/j.foodchem.2013.02.118
Yalegama L.L.W.C. & Chavan J.K. (2006). Studies on utilization
RIFRFRQXWÀRXUDVDVRXUFHRIFHOOZDOOSRO\VDFFKDULGHV
7URSLFDO$JULFXOWXUDO5HVHDUFK 18: 126–134.
=KDQJ< =KHQJ< 'XDQ . *XL 4 3UHSDUDWLRQ
DQWLR[LGDQWDFWLYLW\DQGSURWHFWLYHH൵HFWRIFRFRQXWWHVWDRLO
extraction on oxidative damage to human serum albumin.
,QWHUQDWLRQDO-RXUQDORI)RRG6FLHQFHDQG7HFKQRORJ\ 51:
946–953. DOI: https://doi.org/10.1111/ijfs.12945
Journal of the National Science Foundation of Sri Lanka 47(3)