Research Journal of Environmental and Earth Sciences 2(1): 31-35, 2010
ISSN: 2041-0492
©M axwell Scientific Organization, 2009
Submitted Date: November 14, 2009
Accepted Date: November 26, 2009
Published Date: January 10, 2010
Extraction and Characteristics of Seed Kernel Oil from Mango
(Mangifera indica)
1,4
J.M. Nzikou , 1 A. K imbong uila, 1 L. M atos, 3 B. Loum ouamou, 1 N.P.G. Pambou -Tobi,
1
C.B . Ndangui, 2 A. A. Abena, 3 Th. Silou, 4 J. Scher and 4 S. Desobry
1
EN SP-UM NG , Laboratory of Food P hysicochem istry and B iotech nolo gy, Pole of Excellence in
Nutrition and Food, P.O. Box 69 Brazzaville-Congo
2
Faculty of Science of Health -University Marien Ngouabi P.O. Box 69 Brazzaville-Congo
3
Equipe pluridisciplinaire de recherché en alimentation et nutrition, Centre IRD.
P.O. Box 1286 Pointe-Noire Congo
4
ENSAIA-INPL , Laboratory of engineering and biomolecule, 2, avenue de la forêt de Haye,
54505 V andoeuvre-lès-Nancy (France)
Abstract: Congo mango seeds w ere collected and the kernels were separated and dried. This study was carried
out on mango seed kernels to clarify their proximate composition and the characteristics of the extracted oil
including unsaponifiable m atter and fatty acid compo sition. M ango seed kern els con tained a considerable
unsa ponifiable matter, and a low amount of crude protein. Stearic acid w as the m ain satu rated fatty acid, w hile
oleic acid w as the m ajor un saturated fatty a cid in all lipid classes. Mangifera indica seeds kernels have ash
content of 3.2% (with the presence of following minerals: Ca, K, Na, Mg and P). This oil is very rich in
unsa ponifiable m atter and can thus find its application in cosm otic industry.
Key words: Esse ntial fatty acid, Mangifera indica, minerals, nutritive values, seed kernel and u nsap onifiab le
matter
INTRODUCTION
Mangos belong to the genus Mangifera of the fam ily
Anacardiaceae. The genus Mangifera contains several
species that bear edible fruit. Most of the fruit trees that
are commonly known as mangos belong to the species
Mangifera indica.
Mango (Mangifera indica) trees are tropical fruit
bearing plants, which thrive well in Asia and Africa. As
with many fruits, the edible fleshly portion or pulp of
mango fruit is relished to the extent of commercialization.
A wide variety of processed products derived there from
include canned whole or sliced m ango pulp in brine or in
syrup, mango juice, nectar, jam, sauce, chutney and pickle
(Singh, 1960; Vandrendriessche, 1976). Over 60 varieties
of mango are identifiable (Opeke, 1982). Compositional
studies of the seed kernel of two varieties (Dhingra and
Kap oor, 1985), three varieties (Augustin and Ling, 1987),
eight varieties (Hemavathy et al., 1987 ) and 4 3 varieties
(Lakshminarayana et al., 1983) have been conducted by
Asian scholars. In a similar manner, the lipid composition
of various mango kernel varieties has drawn immense
research interest because of their potential application in
the confectionery industry as a source of a cocoa-butter
substitute (Lakshminarayana et al., 1983; Rukmini and
Vijayaraghavan, 1984; Gaydou and Bouchet, 1984; Ali
et al., 1985; Hemavathy et al., 1987). On comparison,
how ever, wide variations were found in the fatty acid
com position.
The usefulness of the wh ole mango kernel, the
defatted kernel flour and the oil, in compa rison w ith
mango juice, is yet to gain both cottage and industrial
attraction in African countries, particularly Congo. The
underutilization could be partly due to the limited
knowledge of the toxicological status of the mango
kernel, the functional properties of the kernel flour and
appropriate processing technology.
This pape r is a report on extraction and
characteristics of seed kernel oil from Mango of several
Congo kerne ls. Ow ing to its popularity in a sampled
locality, the minerals, the kernel fat and protein fractions,
were determine d and analyzed.
MATERIALS AND METHODS
This study was led to the laboratory of engineering
and biomolecule of the ENSAIA-INPL, Vandoeuvre –
lès-Nancy (France) for the period of A pr. 5, 2008 to Jun.
27, 20 08.
Materials: Mango (Ma ngifera indica ) fruits were plucked
directly from several trees and studied during the 2007
and 2008 fruiting seasons. The Kibangou variety, which
ripens from late October until late December, was
Corresponding Author: J.M . Nzik ou, ENS P-U MNG , Laboratory of Food Physicochemistry and Biotechnology, Pole
of Excellence in Nutrition and Food, P.0. box 69 Brazzaville-Congo
31
Res. J. Environ. Earth Sci., 2(1): 31-35, 2010
obtained from about 447± 10 km south of the Brazzaville,
capital of Congo. The kernels w ere remo ved from their
tenacious leathery coat, dried an d finely groun d into flour.
Chem ical analysis: Determinations for peroxide, iodine,
and saponification values, unsaponifiable matter and free
fatty acid contents were carried out using Pena et al.,
(1992) standard analytical me thods. The fatty acid
composition was determined by conversion of oil to fatty
acid methyl esters prepared by adding 950 :l of n-hexane
50 mg of oil followed by 50 :l of sodium methoxide
using the method of Cocks et al., (1966). The mixtures
were vortex for 5s and allowed to settle for 5 min. The top
layer (1 :l) was injected into a gas chromatograph (Model
GC- 14A, Shimadzu Corporation, Kyoto, Japan) equipped
with a flame-ionisation detector and a polar capillary
column (BPX70 0.25), 0.32 mm internal diameter, 60 m
length and 0.25 :m film thickness (SGE Incorporated,
USA) to obtain individual peaks of fatty acid methyl
esters. The detector temperature was 240 ºC and column
temperature was 110 ºC held for one minute and increased
at the rate o f 8 ºC.minG 1 to 220 ºC and held for one
minu te. The run time was 32 min. comp aring their
retention time w ith those of standards identified the fatty
acid methyl esters peaks. Percent relative fatty acid was
calculated based on the peak area of a fatty acid species to
the total peak area of all the fatty acids in the oil sample.
The minerals were determined by atomicabsorption
spectrophotometry. One gram sam ples, in triplicate, were
dry ashed in a muffle furnace at 550 ºC for 8 h un til a
white residue of constant weight was obtained. The
minerals were extracted from ash by adding 20.0 ml of
2.5% HCl, heated in a steam bath to reduce the volume to
about 7.0 ml, and this was transferred quantitatively to a
50 ml volumetric flask. It was diluted to volume (50 ml)
with deionised water, stored in clean polyethylene bottles
and mineral contents determine d using an atomic
absorption spectrophotometer (Perkin-Elmer, Model
2380, USA). T hese bottles and flasks w ere rinsed in dilute
hydrochloric acid (0.10 M HC l) to arrest microbial action
which may affect the concentrations of the anions and
cations in the samples. The instrum ent w as calibrated with
standard solutions.
M ethods: Proxima te analysis of Mangifera indica seed
Moisture, crude protein (micro-Kjeldahl), crude fiber and
oil (Soxhlet) contents were determined using the methods
described by Pearson (1976), whereas the ash content was
determined using the method of Pomera nz an d M eloan
(1994) and total carbohydrate was determined by
difference. All determinations were done in triplicate.
Oil extraction: For solvent extraction (soxlhet method),
50g of mango kernel flour were placed into a cellulose
paper cone and extracted using light petroleum ether
(b.p 40–60 ºC) in a 5-l Soxhlet extractor for 8 h (Pena
et al., 1992). The oil was then recovered by evaporating
off the solvent using rotary evapo rator Mo del N-1 (Eyela,
Tokyo Rikakikal Co., Ltd., Japan) and residual solvent
was removed by drying in an oven at 60 ºC for 1 h and
flushing with 99.9% nitrogen. For m ethanol/chloroform
extraction (Bligh and Dyer, 1959), 100g of the mango
seeds kernels w ere hom ogenised with a chloroform
mixture methanol (1:1) and water. Two phases was
obtained, aqueous layer (methanol-water) and organ ic
layer (chloroform). Oil was recovered by evaporating off
the solvent (chloroform) using rotary evaporator Model
N-1 (Eyela, Tokyo Rikakikal Co., Ltd., Japan) and
residual solvent w as removed by drying in an oven at
60 ºC for 1 h and flushing w ith 99.9% nitrogen All
experiments were done in triplicates and the mean and
standard deviations were calculated.
Phy sical an d chem ical an alysis o f crud e oil:
Thermal behaviour: The thermal property of the oil
samples was investigated by differential scanning
calorimetry using a Perkin–Elmer Diamond DSC
(Norwalk, USA). T he instrume nt was calibrated using
indium and zinc. The purge gas used was 99.99% nitrogen
with a flow rate of 100 m l.minG 1 and a pressu re of 20 psi.
Sam ple weights ranged from 5–7 mg and were subjected
to the following temperature program: Frozen oil sa mple
was heated at 50 ºC in an oven until completely melted.
Oil sample was placed in an aluminium volatile pan and
was cooled to -50 ºC and held for 2 min, it was then
heated from - 50 to 50 ºC at the rate of 5 ºC .minG 1 (normal
rate) (Che M an and S we, 19 95), and he ld - 50 ºC
isothermally for 2 min and cooled from -50 to 50 ºC at the
rate of 5 ºC per minute. The heating and cooling thermo
grams for the normal and the fast (hyper DSC) scan rates
were recorded and the onset, peak, and offset
temperatures were tabu lated. These v alues provide
information on the temperature at which the melting
process starts, the temperature at which most of the TAG
have melted, and the complete melting temperature of the
oil, respectively.
Statistical analysis: Values represented are the means
and standard de viations for three replicates. Statistical
analy sis was carried out by Excel Version 8.0 software.
Significance was defined at P < 0.05.
RESULTS AND DISCUSSION
Proximate analysis of Mango seed kern el: Results
obtained showed that the seeds contained 45.2% moisture,
13% crude oil, 6.36% crude proteins, 32.24%
carbo hydrate (by difference), 2.02% crude fiber and 3.2%
ash (Table 1). The light percentage of oil makes this seed
kernel a not distinct potential for the oil industry.
According to Dhingra and Kapoor ( 1985). Variation in
oil yield may be due to the differences in variety of plant,
cultivation climate, ripening stage, the harvesting time of
the seeds kernels and the extraction method used.
32
Res. J. Environ. Earth Sci., 2(1): 31-35, 2010
Table 1: Proximate analysis of Mango seed kernel
Characteristic
Obtained values a Reported values b
(M ± S .D.)
-----------------------------1
2
M o is tu re co nte nt (% )
45.2 ± 0.17
38.55
50.98
Cru de p rotein c (% )
6.36 ± 1.07
5.34
5.25
F ats /o ils (% )
13.0 ± 1.28
7.82
6.98
C ru de fib er (% )
2.02 ± 0.80
1.75
1.65
A sh co nte nt (% )
3.2 ± 0.30
2.75
2.47
To tal carb ohy drate d (% ) 32.24
nd
nd
nd, not determined.
a
M ± S.D. mean ± standard deviation.
b
(1) D hin gra an d K ap oo r (19 85 ), va riety Ch au sa.
b
(2) D hing ra an d K apo or (1 985 ), varie ty D ush eri.
c
Crude protein = N (%) x 6.25
d
Carbohydrate obtained by difference
significant role in photosynthesis, carbohydrate
metabolism, nucleic acids a nd binding agen ts of cell walls
(Russel, 1973). Calcium assists in teeth development
(Brody, 1994). M agnesium is essential mineral for
enzyme activity, like calcium and chloride; magnesium
also plays a role in regulating the acid-alkaline balance in
the body. Phosphorus is needed for bone growth, kidney
function and cell grow th. It also plays a ro le in
maintaining
the
body’s
acid-alkaline
balance
(Fallon, 2001 ).
Oil extraction: Characteristics of the oil were compared
with Mangifera indica varieties others country, described
by Dhingra and Kapoor (1985). The extracted oils were
solid at room temperature. The oil conten t of Mang ifera
indica “Congo-Brazzav ille” seeds kernels and the level at
which the differences are significant are sh own in
Table 3. The oil extraction with the Soxlhet method had
the highest yield, due to the increased ability of the
solvent to overcome forces that bind lipids within the
sample matrix (Lumley et al., 1991). The Blye and Dyer
method, showed the low yield due to losses during the
separation of the two phases, aqu eous layer (methano lwater) and organ ic layer (chloroform). The results of the
above au thors agree w ith those of the present work.
Table 2: Mineral elemental Composition of Mango seed flour
Mineral Elem ents
Comp osition (mg/100g) of Seed
Calcium, Ca
10.21±0.05
Magnesium, Mg
22.34±0.01
Po tass ium , K
158.0±0.12
Sodium, Na
2.70±0.02
Ph osp ho rus , P
20.0±0.42
Values are mean ± S.D of triplicate determination
Tab le 3:
Physical and c hem ical properties of Mango seed kernel oil extracted
using solvent process
Properties
Obtained values
Reported values a
-----------------------------------------------------------Bligh & D yer
Soxlhet
Solvent extract
b
B
A
Oil ( %)
12.0±1.32
14.0 ± 2.14
7.4
A
A
PV
0.32±0.12
0.20 ± 0. 48
nd
A
B
FFA (as % o leic acid)
4.48±0.44
3.93 ± 0.18
5.35
A
A
IV (w ijs)
32.0±0.31
43.0 ± 1.24
39.5
A
A
Saponification value
210.0 ± 2.54
206.0 ± 2. 81
207 .5
A
B
Unsaponifiable matter
4.80 ± 0.12
4.35 ± 0.18
nd
C on te nt ( %)
nd, not determined.
Means for the d eterm ined v alues in the same row followed by the same superscript
letter are not significantly different (P < 0.05).
a
Dhingra and Kapoor (1985)
b
Oil = weight of extracted oil x 100/weight of seed.
Abbreviations: PV: Peroxide Value, FFA : Free Fatty Acid, IV: Iodine Value.
Phy sical an d chem ical properties of oil:
Physical properties:
Differential Scanning Calorimetry (DSC):DS C is
suitable to determine these physical properties. The
results of thermal analysis of oils are presented in
Table 4. The obtained peaks were asymmetries and may
indicate the presence of three components in oil extracted
from the two me thods. The first peak s at low melting
points appe ar at – 22.58 ºC ( H f = -14.43 J.gG 1 ) for Blye
and Dyer method and – 22.53 ºC ( H f = -15.37 J.gG 1 ) for
Soxlhet method. These peaks correspond to triglycerides
formed by poly unsaturated acids (PUFA ). The second
melting points are at – 1.25 ºC ( H f = +5 4.36 J.gG 1 ) for
Blye and Dyer method and – 1.33 ºC ( H f = +5 9.86 J.gG 1 )
for Soxlhet method. This is a characteristic of mono
unsaturated acids (MUFA ). The last peaks appear
to +7.51 ºC ( H f = +9.17J.gG 1 ) for Blye and Dyer method
and +6.30 ºC ( H f = +6 .49 J.gG 1 ) for Soxlhet method,
suggest the presence of mixed triglycerides groups with
different melting points.
Tab le 4: Melting beha viour o f Mango seed kernel oil using different
scan rates. Experimental conditions: temperature program set
at -50 °C f or 1 0 m in, risin g to 5 0 °C at rate o f 5 °C .min G 1 .
Thermogram
5 °C /min
-------------------------------------------------Bligh and Dyer
Soxlhet
P e ak 1 [° C ]
-22.58
-22.53
)H [J/g]
-14.43
-15.37
P e ak 2 [° C ]
)H [J/g]
-1.25
+54.36
-1.33
+59.86
P e ak 3 [° C ]
)H [J/g]
+7.51
+9.17
+6.30
+6.49
M inerals: It is of interest to note that the most prevalent
mineral element in Mangifera indica seeds kernels is
potassium which is a high as 158.0 ± 0.12 m g/100g d ry
mater (Table 2), followed in descending order by
Magnesium (22.34±0.01 mg/100g dry mater), Phosphorus
(20.0±0.42 mg/100g dry mater), Calcium (10.21±0.05
mg/100g dry mater) and Sodium (2.70 ±0.42 m g/100g d ry
mater). Potassium is an essential nutrient and has an
important role is the synthesis of amino acids and proteins
(Malik, 1982). Calcium and M agnesium plays a
Chem ical properties: The chemical properties of oil are
amongst the most important properties that determines the
present condition of the oil. Free fatty acid and peroxide
values are valuable measures of oil quality. The iodine
value is the measure of the degree of unsaturation of the
oil. The free fatty acid and the unsaponifiable matter
content of the Soxlhet method w ere significantly higher
(P < 0.05) than those of the Blye and dyer method
(Table 3). There was no significant difference in the
33
Res. J. Environ. Earth Sci., 2(1): 31-35, 2010
Tab le 5: R elative perc ent c om pos ition o f fatty a cid in M ang o se ed o il
Fatty acid
Determined values
--------------------------------------------------------Bligh and Dyer
Soxlhet
C1 4 :0
–
–
A
C1 6 :0
6.38 ± 0.82
6.48 ± 0.33 A
C1 6 :1
C1 8 :0
37.51± 1.20 A
37.94± 2.10 AB
C1 8 : 1
46. 67± 0. 32 B
45.76± 0. 50
A
C1 8 :2
7.21 ± 0.88
7.45 ± 2.18 A
A
C1 8 :3
2.22±0.53
2.37±0.42 A
Saturated
43.89
44.42
Unsaturated
56 .1
55.58
M eans fo r the determ ined v alues in th e sam e row follow ed by the sam e sup erscript letter
a
(1) D hin gra an d K ap oo r (19 85 ), va riety Ch au sa.
a
(2) D hing ra an d K apo or (1 985 ), varie ty D ush eri.
a
(1) S un da y S Ar og ba (19 97 ).
Table 6. Co mparison of the p rofile in fatty vegetable oil acids
Oils
C1 4 :0
C1 4 :1
C1 6 :0
Palm
1.0
44 .5
Safou
45 .5
Maize
10 .5
Groundnut
10 .0
Cotton
0.9
23 .0
Hazel nut
7.0
Soybean
11 .0
Jatropha
15 .6
Mango
6.43
C1 6 :1
0.2
0.1
1.0
-
iodine and saponification values, in the two extraction
methods (P > 0.05). The slightly higher value of
unsaponifiable matter in the Soxlhet method may be due
to the ability of the solvent to extract other lipid
associated substances like, sterols, fat soluble vitamins,
hydrocarbons and p igme nts (Bastic et al., 1978; Salunke
et al., 1992).
C1 8 :0
4.6
2.8
2.5
2.0
2.2
2.0
4.0
5.8
37.73
Reported values a
-------------------------------------------------------------1
2
3
–
–
7.5
10 .5
7.18
38 .9
38 .7
27 .0
42 .6
43 .2
48 .3
5.7
6.2
14 .0
5.3
2.9
–
46.08
46 .2
37 .5
53 .6
52 .3
62 .3
are no t significantly d ifferent (P < 0.05).
C1 8 :1
38 .7
28
28
46 .0
17 .7
74 .5
22 .0
40 .1
46.22
C1 8 :2
10 .5
24 .9
58 .5
31 .0
55 .8
16 .5
54 .3
37 .6
7.33
C1 8 :3
0.3
1.24
1.0
7.5
2.30
C2 0 : 0
0.3
0.5
-
greater than the saturated fatty acids. High unsaponifiable
matters content (4.58%) guarantees the use the o ils in
cosmetics industry.
REFERENCES
Ali, M.A., M.A. Gafur, M.S. Rahman and G.M. Ahmed,
1985. Variations in fat content and lipid class
composition in ten different varieties. J. Am. Oil.
Chem. S oc., 62(3): 520 -523.
Augustin, M.A. and E.T. Ling, 1987. Composition of
mango seed kernel. Pertanika., 10(1): 53-59.
Bastic, M., L. Ba stic, J.A Jabano vic and G . Spiteller,
1 9 7 8 . H y d r o c a r b o n s a n d o t h e r w e a k ly
unsa ponifiables in som e veg etable oils. J. Am. O il
Chem., Soc., 55: 886-8 92.
Bligh, E.G. and W .J. Dyer, 1959. A rapid method of total
lipid extraction and purification. Can. J. Biochem.
Physiol., 37: 911-917.
Brody, T., 1994. N utritional Biochem istry, San Diego,
CA: Academic Press. 2nd Edn., pp: 761-794.
Che Man, Y. B. and P. Z. Swe, 1995. Thermal analysis of
failed-batch Palm oil by differential scanning
calorimetry. J. Am . Oil Chem . Soc., 72(12):
1529-1532.
Cocks, L.V . and C .Van Rede, 1966. Laboratory handbook
for oil and fats analysts. London: Academic Press.
pp: 88
Dh ingra, S. and A. C. Kapoor, 1985. Nutritive value of
mango seed kernel. J. Sci. Food. Agr., 6: 752-756.
Fallon, S. and M.G. Enig, 2001. Nourishing Traditions.
The Cookbook that Challenges P olicitally Correct
Nutrition and the Diet Dictocrats, pp: 40-45.
Fatty acid composition: The major saturated fatty acids
in Mangifera indica seed kernel oil were stearic (37.73%)
and palmitic (6.43%) acids and the main un saturated fatty
acids are oleic (46.22%), linoleic (7.33% ) and linolenic
(2.30%) acids (T able 5). There was no significant
difference (P > 0.05) in the a moun ts of the major fatty
acids in the two oil samples. In the two oil samples of
Mang ifera indica contained saturated and unsaturated
acids (44.16 and 55.84%) respectively. The proportion of
unsaturated fatty acids w as greater than the saturated fatty
acids. Mangifera indica seed kernel oil is pred omin antly
made up of stearic and oleic acids (37.73 and 46.22%)
respectively. One notes 2.22 and 2.37% of linolenic acid
C18:3 respectively in the two methods Blye and Dyer and
Soxlhet (Table 5). The results obtained are in agreement
with those of the literature Dhingra a nd K apoor (1985).
The comparison of the composition in fatty acids of
Mang ifera indica seed kerne l oil with that of vege table
oils indicates that this plant is rich in acids stearic (C18:0)
and oleic (C18:1), Table 6.
CONCLUSION
Mang ifera indica seed kernel oil has a low content of
protein . Stearic and oleic acid s we re the principal fatty
acids and the proportion of unsaturated fatty acids was
34
Res. J. Environ. Earth Sci., 2(1): 31-35, 2010
Gaydou, E.M. and P. Bouchet, 1984. Sterols, methyl
sterols, triterpene alcohols and fatty acids of the
kernel fat of di†erent Malagasy mango (M angifera
indica) varieties. J. Am. O il. Chem. So c., 61(10):
1589-1593.
Hemavathy, J., J.V. Prabhakar, D.P. Sen, 1987.
Composition of polar lipids of alphonso mango
(Ma ngifera indica) kernel. J. Food . Sci., 52(3):
833-834.
Lakshminarayana, G., T.C. Rao, P.A. Ramalingaswamy
and T. Chandrasekhara Rao, 1983.Varietal variations
in content, Characteristics and composition of mango
seeds and fat. J. Am. Oil. Chem. Soc., 60(1): 88-89.
Lumley, I.D. an d R.K. Colwell, 1991. Fats from fatty
foods and d etermination of fat content. In Analysis
of Fats and Fatty Foods. J.B. Rossell and J.L.R.
Pritcha rd, (Eds.), pp: 23 8-247.
Malik, C.P. and A.K. Srivastava, 1982. Text book of plant
physiology. New Delhi: Ludhiana.
Opeke, L.K ., 1982. T ropical Tree Crops. John Wiley,
New D elhi, India.
Pearson, D., 19 76. The C hem ical Analysis of Foods. 7th
Edn. R.K . Chu rchillLivingstone, Edinbu rgh, U .K.,
pp:488-496.
Pena, D.G., R.G.L. Anguiano and J.J.M A rredondo, 1992.
Modification of the method 1 AOAC (CB-method)
for the detection of aflatoxins, Bull. Environ.
Contam. Toxicol., 49 : 485-489.
Pomeranz, Y. and C. M eloan, 1994. Food analysis:
Theory and practice, 3rd Edn., New Y ork: Chapman
and Hall, pp: 778.
Rukmin i, C. and M. Vijayaraghavan, 1984. Nutritional
and toxicological evaluation of mango kernel oil. J.
Am. Oil. Chem. Soc., 61(4): 780-792.
Russel, E.W., 1973. Soil conditions and plant growth.
Superg ene Zo ne, M . Nedra, 19 (in Russian).
Salunke, D.K., J.K. Chavan, R.N. Adsule and S.S.
Kadam, 1992. W orld oil seeds: Chemistry,
techn ology and u tilization. pp: 170 -173.
Singh, L.B., 1960 . The Mango (Botany , Cultivation and
Utilization). Leonard Hill, London, UK.
Vandrendriessche, H., 1976. Tropical fruit Products. In:
Tropical fruit processing Industry. Development
Centre of the Organization, Paris, France, pp: 15-16.
35