Cassava starch granule structurefunction properties: inuence of time

and conditions at harvest on four cultivars of cassava starch

K. Sriroth
a,
*
, V. Santisopasri
b
, C. Petchalanuwat
c
, K. Kurotjanawong
a
, K. Piyachomkwan
d
,
C.G. Oates
e
a
Department of Biotechnology, Kasetsart University, Bangkok, Thailand.
b
Department of Biochemistry, Kasetsart University, Bangkok, Thailand.
c
Department of Agriculture, Ministry of Agriculture, Thailand.
d
Cassava Starch Technology Research Unit/National Center for Genetic Engineering and Biotechnology, Thailand.
e
Department of Biochemistry, National University of Singapore, Singapore.
Received 12 January 1998; received in revised form 11 May 1998; accepted 18 June 1998
Abstract
Impact of environmental conditions on cassava starch variability was examined by studying four commercially important cultivars,
Rayong 1, Rayong 60, Rayong 90, and Kasetsart 50 (KU 50). Age of the root and environmental conditions at harvest inuenced granule
structure and hydration properties. All cultivars were grown under identical eld conditions, and harvested at different times. Starches
extracted from cassava roots harvested at different times were characterised by unique starch granule structure and function. Apparent
amylose size of starches from all cultivars did not change signicantly during the trial period. However, apparent amylose content of starches
changed, decreasing in the older roots. Granule size distribution was affected by age of the root, gradually changing from normal to bimodal
distribution when harvested very late during the trial. The integrity and crystalline structure of starch granules also depended on the
environmental conditions, evidenced as a change in peak prole obtained by thermal analysis. This can result in the difference in water
uptake of starches, and their consequent swelling power and gelatinization. Pasting temperature of all starches increased during the dry
period, and was lowered during the wet period. Peak and nal viscosity of starch decreased from early to mid-harvest time when environ-
mental conditions became drier, and increased close to or greater than the original value when conditions became wet again. Breakdown and
setback also followed a similar trend to viscosity. This study suggests an impact of time and conditions of harvest on the structural and
functional properties of all cassava cultivars, and based on this study, it is recommended that starch should be extracted from either early or
very late harvested roots. q1999 Elsevier Science Ltd. All rights reserved.
Keywords: Cassava starch; Rayong 1; New cassava cultivars
1. Introduction
Ensuring consistent quality of cassava starch remains a
problem despite advances and upgrading of technology used
for the extraction of this material. Variation derived from
different cultivars accounts for some of the deviation
between processing operations in Thailand. However,
within a geographic area single cultivars tend to dominate
and therefore it is unlikely that this is a source of variation
for starch produced in the same factory at different times of
the year. Variation reportedly can signicantly alter the
functional properties of the starch making application of
the native material difcult in some circumstances.
In Thailand a number of cultivars of cassava are grown
for commercial purposes. One of the rst cultivars to be
developed for improved economic return was Rayong 1,
which is a selection from a local land race. Due, in part,
to its excellent agronomic traits (Rodjanaridpiched et al.,
1993) more than 1 million hectares are planted annually to
this cultivar. Since 1984, and in parallel with the growth of
the cassava industry, several new cultivars suitable for
industrial purposes have been developed, including Rayong
3, Rayong 60, Rayong 90 and Kasetsart 50 (Table 1).
Agronomic characteristics of the four cultivars summar-
ized in Table 1 have been examined in extensive trials,
designed to determine the most suitable time for harvest,
either early (9 months) or late (12 months). The largest trial,
Carbohydrate Polymers 38 (1999) 161170
0144-8617/99/$ - see front matter q 1999 Elsevier Science Ltd. All rights reserved.
PII: S0144-8617(98)00117-9
* Corresponding author. Klanarong Sriroth,Cassava and Starch Technol-
ogy Research Unit, Kasetsart Agricultural and Agro-Industrial Product
Improvement Institute (KAPI), Kasetsart University, Bangkok 10900,
Thailand. Tel.: 662 9405634; Fax: 662 9405634; e-mail: [email protected]
conducted at 9 locations in 1994, evaluated the inuence of
time of harvest on fresh root yield, dry root yield and root
dry matter (Rodjanaridpiched et al., 1993). For all cultivars
wet root yield was higher at 12 months but dry root yield
and root dry material was lower at this time. However, these
studies did not evaluate quality of starch produced at the
different harvest times, or consider the intervening months.
The rate of decrease of dry root yield was greatest for
Rayong 60 and Rayong 1.
Plants harvested at 12 months were found to be respond-
ing to climatic change, associated with transition from dry
to wet season. The authors concluded that post dry season
the plants renew their top growth at the expense of root
growth, resulting in a reduction of root dry matter content.
This response to re-growth was variable between the
cultivars.
Cassava roots are harvested manually with a little help
from small tools. The harvest can be through the year except
in the rainy season (June/July/August). From age of
12 months, the roots are considered as ripe and are often
left in the ground until 1416 months, depending on the
economic situation. The young roots contain more water
and less starch ( , 20%) as compared with older roots
(more than 12 months), but extractability of starch from
the older roots is more difcult due to a higher ber content.
In general, starch structure-function properties alter
according to the stage of development of the plant and the
botanical source. Demonstrated by most cereals, amylose
content is lower at early stages of grain development (Inou-
chi et al., 1984; Asaoka et al., 1985). A similar trend
between amylose content ad developmental age is also
seen in potato, where increasing root size is associated
with a greater amylose content (Geddes et al., 1965). This
is, however, not the case during development of tuberous
roots such as cassava (Asaoka et al., 1992) or sweet potato
(Noda et al., 1992).
Previous work has demonstrated that both genetic consti-
tution and environmental conditions inuence functional
properties of the starch, including gelatinization properties
(Asaoka et al., 1992). However, this work did not fully
describe the inuence of rainfall, which will promote
canopy growth, on starch functional properties. These
factors will be investigated in this article which forms part
of a larger study mapping the inuence of environmental
conditions on starch structure-function properties.
The present study attempts to provide a more detailed
understanding of environmental inuences and age of the
root on quality and functional properties of the extracted
starch. This information is important for a number of
reasons, primarily it could form the basis of a model to
probe structure-function relationships of cassava starch.
Understanding the reasons responsible for variability of
this starch is required both by the processor and user, hope-
fully a better appreciation of the mechanisms responsible for
variation will lead to the production of higher quality starch.
2. Materials and methods
2.1. Materials
Cassava plants were planted in May 1995 at Rayong Field
Crops Research center, Department of Agriculture, Thai-
land using Split-plot in Randomized Complete Block
(RCB) design. During growth, fertilizer (15-15-15) was
applied at the rate of 312.5 kg/ha. No extra irrigation was
applied. Precipitation was as indicated in Fig. 1.
Soil conditions for the experiment area were; pH 5.4,
organic matter content 0.5%, phosphorus and potassium
content 24 and 20 ppm, respectively.
Roots were manually harvested each month and samples
representing all experimental cultivars collected and
analyzed for starch content by a specic gravity method
as described by Brautlecht (1953). Starch extraction was
conducted within 12 h of harvesting using water as extrac-
tion solvent and a laboratory blender for breaking the plant
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 162
Table 1
Recommended cassava cultivars in Thailand
Cultivar Year released Parents Main features
Rayong 1 1957 From local cultivars High yield, good adaptability
Rayong 60 1987 Mcol 1684 * Rayong 1 Early harvest, high yield
Rayong 90 1991 CMC 76 * V43 High root DM, high yield
Kasetsart 50 1992 Rayong 1 * Rayong 90 High root Dm, high yield,
good adaptability
Figure 1. Rainfall during the growing period. Monthly values represent the
accumulative rainfall in the previous four weeks.
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 163
T
a
b
l
e
2
E
f
f
e
c
t
o
f
c
u
l
t
i
v
a
r
s
a
n
d
h
a
r
v
e
s
t
i
n
g
d
a
t
e
o
n
a
v
e
r
a
g
e
s
t
a
r
c
h
c
o
n
t
e
n
t
(
%
)
a
n
d
y
i
e
l
d
(
t
o
n
/
h
a
)
o
f
c
a
s
s
a
v
a
r
o
o
t
s
.
C
u
l
t
i
v
a
r
s
A
g
e
(
m
o
n
t
h
)
6
1
0
1
2
1
4
1
6
S
t
a
r
c
h
c
o
n
t
e
n
t
(
%
)
Y
i
e
l
d
(
t
o
n
/
h
a
)
S
t
a
r
c
h
c
o
n
t
e
n
t
(
%
)
Y
i
e
l
d
(
t
o
n
/
h
a
)
S
t
a
r
c
h
c
o
n
t
e
n
t
(
%
)
Y
i
e
l
d
(
t
o
n
/
h
a
)
S
t
a
r
c
h
c
o
n
t
e
n
t
(
%
)
Y
i
e
l
d
(
t
o
n
/
h
a
)
S
t
a
r
c
h
c
o
n
t
e
n
t
(
%
)
Y
i
e
l
d
(
t
o
n
/
h
a
)
R
a
y
o
n
g
1
1
8
.
1
1
1
.
3
1
3
.
8
1
7
.
1
1
8
.
6
2
5
.
8
2
3
.
0
2
7
.
7
2
1
.
9
2
2
.
2
R
a
y
o
n
g
6
0
1
7
.
2
7
.
5
1
5
.
4
1
3
.
6
1
9
.
0
2
1
.
8
2
2
.
6
2
8
.
3
1
9
.
7
3
0
.
4
R
a
y
o
n
g
9
0
2
3
.
3
1
0
.
9
2
0
.
8
2
3
.
1
2
7
.
1
3
4
.
0
2
9
.
1
4
3
.
7
2
6
.
1
4
4
.
5
K
a
s
e
r
s
a
r
t
5
0
2
3
.
4
8
.
0
1
9
.
3
1
6
.
8
2
4
.
7
2
1
.
8
2
7
.
4
2
8
.
5
2
4
.
4
3
1
.
9
tissue to liberate the starch granules. Fiber was separated by
sieving through a 170-mesh screen (90 mm). After washing
several times, starch was oven dried at 50558C. Starch
properties were analyzed immediately on production with
suitable controls to ensure reproducibility of the methods
between analysis periods.
2.2. Proximate analysis
The content of protein, fat and ash were determined
according to AOAC methods (AOAC, 1990). The amount
of phosphorous was quantied by the method described by
FAO (Food and Agricultural Organization of the United
Nations, 1988).
2.3. Scanning electron microscopy
Dried starch samples were prepared for observation by
scanning electron microscopy by sprinkling the starch on
double-sided adhesive tape attached to a circular specimen
stub and coating with gold using a Baltzers SCD 004 sputter
coater. The samples were viewed and photographed using a
JEOL JSM-5300LV scanning electron microscope on
AGAFAPAN-APX 100 lm.
2.4. Sonication for HPSEC
An aqueous starch slurry (20 mg in 15 ml) was gelati-
nized at 1008C for 10 min, cooled to 508C and sonicated.
Samples, 15 ml, were sonicated in tubes (internal diameter,
20 mm) using a Sonicator
TM
Cell Disrupter, model XL-2020
(Heat systems Ultrasonic, Inc.) and Misonix probe (exter-
nal diameter, 4 mm) (Model 419A). Power was set at 45%
of the maximum 550 W (output frequency 20 kHz).
2.5. Amylose Size
The proportion of the major starch fractions and apparent
molecular size of amylose after sonication were determined
by high performance size exclusion chromatography
(HPSEC) according to the method of Govindasamy et al.
(1992).
2.6. Swelling and Solubility
Swelling and solubility determinations were carried out at
958C by the procedure of Schoch (1964).
2.7. Differential scanning calorimetry
Gelatinization properties of the starches were analyzed
using a Perkin Elmer Differential Scanning Calorimeter
(DSC7; Norwalk, CT) equipped with an intercooler. Alumi-
num pans (Perkin Elmer) were used for the analysis. Starch
samples were slurried with water to give a volume fraction
between 0.680.69. The starch slurry was dispensed, after
mixing, into empty aluminum pans and hermetically sealed.
Following equilibration at room temperature for 1 h the
samples were heated from 20 to 1108C at a heating rate of
108C/min. An empty pan was used as reference. Enthalpy
changes, integrated using DSC software, were calibrated on
the basis of the melting of indium metal. All pans were
reweighed after cooling. To calculate enthalpy values accu-
rately on a dry basis, each pan was carefully punctured,
dried at 1108C for 2 h and reweighed to determine moisture
content of the scanned and rescanned samples.
2.8. Granule size distribution
Granule size distribution was determined by image analy-
sis (Carl Zeiss, KS400 v2) recorded directly from a Carl
Zeiss Axiophol 2 microscope. Starch samples were
suspended in 80% sucrose solution to minimize the effects
of refractive index and provide a sharper image of the gran-
ules (Baldwin, 1994).
2.9. Pasting proles
Pasting proles were recorded on a RVA 4 (Newport
Scientic, Australia) using standard program Number 1.
The starch sample was 3.00 g (on 14% moisture basis).
The starch suspension was held at 508C for 1 min and subse-
quently heated to 958C at 12.28C/min. Holding time at 958C
was 2.5 min. Subsequently, the sample was cooled to 508C
at 12.28C/min, where it was kept for 2.1 min. A rotation
speed of the paddle of 160 rpm was used and cooling was
performed with water circulated through cooler.
3. Results and discussion
During the study two rainy seasons were recorded, the
rst coinciding with planting and the second in the later
months of the trial. Months 7, 8, 9 and 11 were compara-
tively dry months interspersed with one wet month (month
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 164
Table 3
Proximate analysis of cassava starches.
Cultivars Protein (% w/w) Lipid (% w/w) Ash (% w/w) Phosphorus (mg/kg)
Rayong 1 0.17 ^ 0.04 nil 0.10 ^ 0.02 2.45 ^ 0.08
Rayong 60 0.15 ^ 0.02 0.01 0.15 ^ 0.04 2.20 ^ 0.14
Rayong 90 0.28 ^ 0.06 nil 0.08 ^ 0.01 2.04 ^ 0.05
KU 50 0.30 ^ 0.04 0.01 0.15 ^ 0.02 2.04 ^ 0.05
10). The roots were harvested at 6, 10, 12, 14 and
16 months.
Previous work (unreported) suggested that rainfall had a
strong inuence if high in the month immediately prior to
harvest. However, time of harvest was not a major determi-
nant during the growing period up to 16 months. For this
reason and in order to investigate the inuence of cumula-
tive conditions, samples were harvested during only ``wet''
months, and harvesting pre and post dry period was during
months that had experienced a similar level of rainfall.
For all cultivars, starch content was lowest at 10th month
(Table 2). Plants harvested in the 10th month were subjected
to heavy rainfall in the previous four weeks. Prior to which,
drought conditions had prevailed for about 2 months. Starch
content increased again and reached the maximum at 14th
month. In general, root yield increased with a greater age of
the cassava plant.
The proximate composition of the starches is presented in
Table 3. Starches from different cultivars were low in lipid
(,0.01%), protein (0.15%0.30%), ash (0.08%0.15%)
and phosphorous (2.042.45 mg/kg). Values were in the
range expected for cassava starches (Rickard et al., 1991).
3.1. Starch structure
3.1.1. Macromolecule components
Starch structure was different if extracted from cassava
harvested prior to the onset of the rainy season. Apparent
amylose size did not appear to be inuenced by either age of
the root or growing conditions immediately prior to harvest
(Table 4).
The proportion of amylose changed within a narrow band
during the period of the trial. Starch extracted from roots
harvested late in the trial, (14th and 16th month), was gener-
ally lower in amylose content (Table 4). Previous studies
(Inouchi et al., 1984; Asaoka et al., 1985) have reported a
variability of amylose content as inuenced by time of
harvest. In cereals and potato amylose increases in later
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 165
Table 4
Apparent size and content of amylose in cassava starch at different harvest time
a
.
Time of harvest / months
6 10 14 16
Dp Content
b
(%) DP Content
b
(%) DP Content
b
(%) DP Content
b
(%)
Rayong 1 1175 24.1 1188 23.5 1148 20.6 1161 20.6
Rayong 60 1122 22.5 1096 20.5 1035 20.7 1148 20.8
Rayong 90 1202 23.1 1216 21.3 1122 22.3 1135 22.5
KU 50 1175 21.4 1161 21.5 1135 19.5 1121 19.6
a
Values are the mean of ve determinations, the range of each values was , ^ 0.2%
b
Amylose content, expressed as a percentage of the total granule carbohydrate
Figure 2. Rayong 1 size distribution prole. A representative example for cassava starch granules obtained from Rayong cultivars as inuenced by time of
harvest. X axis represents the size of starch granule in micron whereas Y axis represents the percentage of distribution.
stages of development (Inouchi et al., 1984; Asaoka et al.,
1985; Gedds et al., 1965). In comparison for cassava and
sweet potato there was little difference in amylose content
(Asaoka et al., 1992; Noda et al., 1992). Results of the
present study are not in agreement with those of Asaoka
et al. (1992), amylose content was highest (for all cultivars)
at the early harvest time, but dropped to a constant level
during the dry period. In two samples, Rayong 60 & 90,
amylose content decreased during the extended dry period.
For the other two samples amylose decreased during the dry
period and continued the trend into the wet period. This
suggests that inuences on amylose content are more
complicated than length of time in the ground.
3.2. Starch granule
The size and morphological appearance of the starch was
as expected for cassava starch (Rickard et al., 1991).
Size distribution of the granules followed a trend similar
to swelling with respect to harvest time. Slight differences
are evident between cultivars (Fig. 2), though with the
exception of KU 50, all were similarly effected by condition
at harvest.
The range of granule sizes approached a normal distribu-
tion in samples harvested up to 14th month. For all cultivars
the mean size was centered at about 15 mm and the distribu-
tion was from about 822 mm. There was no signicant
shift in the mean granule size with harvest time, but for
all Rayong cultivars there was a slight shoulder at about
12 mm, which developed to such an extent that by the
16th month granule size was bimodally distributed. The
two distinct populations of granule were centered at about
15 and 12 mm. In contrast, size distribution for granules of
Ku 50 appeared to be similar for all times of harvest and
normally distributed.
Electron micrographs suggest that granules are consider-
ably irregular in shape with oval, round and truncated gran-
ules present within the same sample. There were no
apparent differences, between species with respect to gran-
ule morphology (Fig. 3 and data not shown).
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 166
Figure 3. Scanning electron photomicrographs of cassava starch from Rayong 60 at different harvest times (a) 6th, (b) 10th, (c) 14th and (d) 16th month.
Comparative information on the integrity and crystal
structure of starches can be obtained by thermal analysis.
For starches in this study, intercultivar comparisons are only
possible for starches collected in the early months. There
were only slight difference between the cultivars in respect
to gelatinization temperature and enthalpy. Peak tempera-
tures (T
p
) were similar for all cultivars and in the range 69
718C (Kasetsart 50 and Rayong 90/Rayong 1 respectively)
(Fig. 4). Starches extracted from roots harvested in later
months of the trial exhibited unusual thermal properties
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 167
Figure 4. Representative thermoanalysis proles for cassava starch harvested at different times. (a) Rayong 90; (b) KU 50. Dry sample weight: (a) Rayong 90
(1) 4.54, (2) 4.22, (3) 4.34 and (4) 4.50 mg for 6, 10, 14 and 16 months respectively; (b) KU 50 (1) 4.56, (2) 4.48, (3) 4.45 and (4) 4.51 mg for 6, 10, 14 and
16 months respectively. The bar corresponds to 1 mW.
that made comparison between the starches difcult. The
single endothermic peak evident in samples harvested
early in the study developed into a biphasic peak with
concomitant increase in temperature of the original peak
(Fig. 4). In this study starch samples were hydrated to a
point close to that known to inuence the gelatinization
mechanism. For starch water mixtures containing more
than 60% water (WSB) a single symmetrical endothermic
peak is usually observed. At moisture contents lower than
the critical point, size of the transition progressively
decreases with concomitant development of a second transi-
tion. The later peak moves to higher temperature as the
amount of water is further reduced.
Differences in thermal behavior as inuenced by moisture
content are assumed indicative of non-equilibrium melting
and support the notion that structure reorganization during
heating is water dependent. At lower moisture, melting
temperature and transition can not be accurately determined
because of the resulting broad fusion range.
Temperature of gelatinization (T
P
, T
o
, T
f
) for cassava
starch is variably reported in the literature, values (T
p
)
range from as low as 588C (Asaoka et al., 1992) to as
high as 658C (Yamada et al., 1987) for cassava starch
from Thailand. Such differences suggest that genetic varia-
tion and environmental conditions impact on structure-func-
tion of the starch. Comparisons with the current study would
be difcult, as the moisture content is lower than that custo-
marily used for starch thermoanalysis. Though the range
recorded in this study for T
P
(69718C) is comparable,
considering the lower moisture content. Studies at high
moisture conrm that the gelatinization temperature (T
P
)
is as expected at 668C (data not presented).
Root and cereal starches are assumed to have similar
melting mechanisms. The melting of starch granules is
governed by the speed in which water can enter the starch
granule. The event is thought to comprise of two stages,
involving plasticizing of the amorphous amylose followed
by hydration and melting of the crystallites. In excess water
rapid plasticizing can occur and the two endotherms merge
as one, in the range of the single endotherm, as the rate of
water penetration will be the same as the rate of crystallite
disruption. At intermediate water conditions the endother-
mic peak splits forming a biphasic endotherm, as the amor-
phous regions are slowly plasticized causing a delay in
hydration which forces gelatinization temperatures towards
higher regions.
This study suggests that starch granules may differ in
their response to water uptake and consequent gelatiniza-
tion, depending on environmental conditions. This phenom-
enon is not apparent when samples are examined at higher
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 168
Table 5
Swelling power of cassava starch at 958C as inuenced by time of harvest.
Time of harvest (months)
6 10 12 14 16
Rayong 1 55 ^ 1 58 ^ 7 59 ^ 7 53 ^ 1 51 ^ 2
Rayong 60 58 ^ 7 61 ^ 4 54 ^ 7 48 ^ 1 45 ^ 6
Rayong 90 60 ^ 6 59 ^ 2 55 ^ 2 59 ^ 7 66 ^ 1
KU 50 54 ^ 4 77 ^ 12 64 ^ 3 51 ^ 1 51 ^ 2
Table 6
Pasting properties of cassava starch as inuenced by time of harvest.
Pasting temperature (8C) Peak viscosity (RVU) Trough (RVU) Final viscosity (RVU) Breakdown* Setback*
Rayong 1
6 71.35 ^ 0.35 419 ^ 21
b
169 ^ 17 284 ^ 10
a
250 ^ 6
b
115 ^ 12
a
10 71.38 ^ 0.04 363 ^ 10
c
154 ^ 17 221 ^ 10
c
209 ^ 7
c
67 ^ 8
c
12 72.60 ^ 0.76 357 ^ 2
c
158 ^ 8 205 ^ 10
c
200 ^ 10
c
47 ^ 4
d
14 72.00 ^ 0.64 436 ^ 33
b
160 ^ 20 245 ^ 16
b
265 ^ 29
b
74 ^ 12
bc
16 71.08 ^ 0.35 502 ^ 4
a
182 ^ 15 268 ^ 10
ab
324 ^ 14
a
85 ^ 5
b
Rayong 60
6 70.98 ^ 0.16 402 ^ 12
ab
156 ^ 3
ab
266 ^ 10
a
246 ^ 15
ab
111 ^ 11
a
10 71.30 ^ 0.99 382 ^ 25
bc
156 ^ 18
ab
238 ^ 13
b
226 ^ 24
b
82 ^ 6
b
12 70.80 ^ 0.57 336 ^ 11
c
133 ^ 8
c
183 ^ 8
c
203 ^ 12
b
50 ^ 4
c
14 70.30 ^ 0.64 351 ^ 37
c
147 ^ 7
bc
196 ^ 3
c
205 ^ 47
b
50 ^ 7
c
16 69.50 ^ 0.78 459 ^ 45
a
169 ^ 4
a
265 ^ 14
a
288 ^ 41
a
96 ^ 13
ab
Rayong 90
6 71.10 ^ 0.82 396 ^ 3
b
145 ^ 21 271 ^ 10
a
251 ^ 20
b
125 ^ 13
a
10 72.90 ^ 0.79 395 ^ 27
b
157 ^ 31 243 ^ 19
a
239 ^ 15
b
71 ^ 4
b
12 72.10 ^ 0.49 304 ^ 2
c
158 ^ 14 177 ^ 4
b
150 ^ 18
d
23 ^ 20
c
14 72.07 ^ 0.40 351 ^ 53
bc
163 ^ 23 244 ^ 31
a
198 ^ 30
c
69 ^ 22
b
16 70.68 ^ 0.66 480 ^ 12
a
177 ^ 6 262 ^ 6
a
302 ^ 15
a
84 ^ 8
b
KU 50
6 72.57 ^ 0.70
bc
439 ^ 4
b
174 ^ 13 278 ^ 17 265 ^ 9
b
98 ^ 6
a
10 74.32 ^ 0.60
a
369 ^ 4
c
145 ^ 19 213 ^ 23 214 ^ 7
c
69 ^ 4
ab
12 73.22 ^ 0. 435 ^ 1
b
170 ^ 27 224 ^ 26 254 ^ 27
ab
54 ^ 8
b
14 71.25 ^ 0.42
cd
404 ^ 26
c
160 ^ 4 253 ^ 29 243 ^ 28
ab
92 ^ 31
a
16 70.38 ^ 0.53
d
531 ^ 8
a
164 ^ 4 261 ^ 13 369 ^ 6
a
92 ^ 17
a
or lower moisture contents. The precise mechanism is not
known, but would appear to be subtle and assumed to repre-
sent structural differences that cause an apparent variation in
the rate of hydration for late harvested samples. For starches
extracted from other botanical sources physical properties
of starch granules, such as gelatinization temperature, are
affected by environmental temperature during growth. This
is probably not a key issue in tropical crops where tempera-
ture varies very little during the year. However, variation in
rainfall could lead to differences in soil temperature in the
immediate environment of the growing roots. Tempering,
mechanical damage and thermal history of the starch during
extraction and drying may also inuence thermal properties.
This is unlikely to be the case in this study, granules look to
be morphologically similar and care was exercised during
the extraction, ensuring experimental uniformity through-
out.
3.3. Hydration properties
Swelling power of starches followed less of a general
trend in some samples, for example Rayong 60 and KU
50 (Table 5). For Rayong 60 starch, swelling power was
higher when extracted from root harvested at 6, 10 and
12 months. Swelling power of KU 50 peaked in the 10th
month, and then decreased through the extended growing
period. For Rayong 1 and Rayong 90, swelling power did
not decrease with a similar trend, suggesting that their swel-
ling power might be affected by other factors such as gran-
ule size, structure and composition.
3.4. Pasting properties
To detect pasting differences, Rapid Visco Analyser
(RVA) was used (Table 6). Pasting temperatures were little
effected by harvest time irrespective of the cassava cultivar.
Viscosity was high in samples harvested at 6th month,
however at the 10th month peak viscosity dropped, reaching
a low value by the 12th month for all Rayong cultivars and
the 10th month for KU 50. The level of decrease ranged
from 15%23% for Rayong cultivars and 16% for KU 50.
With further growing time in the wet season, roots produced
starch with a peak viscosity higher than that of the 6th
month harvested sample; the percentage increased ranged
from 14%21%, when comparing the increase in peak visc-
osity between 6th and 16th month samples. Final viscosity
followed a similar trend to that of peak viscosity; being low
at 10th12th month, and becoming high again in the wet
season at 14th16th month. At age of 6th month, all culti-
vars showed the most susceptibility to breakdown and
setback, then the viscosity became more stable until the
age of 14th16thmonth. The change in breakdown and
setback was compatible with the change in peak and nal
viscosity of starch solutions, and also the amylose content,
implying a relation between pasting properties and amylose
content. The root at 6th month contained high amylose
content (21%24%), and produced starch with high nal
viscosity and setback. Amylose content of starch decreased
later during the dry period, and was also lower in nal
viscosity and setback. Starch content of the roots was also
lowest at 10 months, suggesting utilization of starch by the
plant with the onset of rain, immediately post dry period
(Table 2). It is reasonable to assume that the internal archi-
tecture of the granules will be different, as they will contain
a lower proportion of amorphous material. These structural
changes are assumed to impact on the functional properties
of the starch and are presently being studied.
4. Conclusions
Previous work has shown differences between starch
harvested at two extreme times, but the current work
suggests that starch properties can change in mid-cycle.
Results of this study also show that starch granules differ
in their response to water uptake and consequent gelatiniza-
tion depending on the environmental conditions at the time
of harvest, expressed as a change in peak prole obtained by
thermal analysis for samples harvested at 14th and 16th
month. This change is similar to that seen when moisture
content is reduced. The precise mechanism is not known,
and is not evident at higher or lower moisture content. This
effect is therefore subtle, but real, and can be assumed to
arise from structural differences, which give rise to apparent
variation in the rate of hydration for late harvest samples.
All the cultivars examined show this phenomenon. In a
similar manner the size distribution prole of the granules
changed from a normal distribution, gradually developing
into a bimodal distribution. Again the mechanism is not
known, and we are unable to speculate whether the two
phenomena are related.
This study also suggests that starch extracted from roots
harvested late will have good pasting and swelling proper-
ties. This may be related to the large amylose size in these
starches or greater proportion of amylopectin. The dry
season has an impact on starch properties, as does the
immediate onset of the rainy season. We would therefore
recommend based on these ndings that starch be extracted
from either early or very late harvested roots.
The work reported in this article forms part of a more
detailed project to probe the structure-function relationships
of cassava starch, future efforts will involve the production
of cassava under controlled environmental conditions with
subsequent isolation and characterization of structural or
functional features of the starch unique to a particular set
of environmental conditions.
Acknowledgements
This work was kindly supported by Thai Tapioca Devel-
opment Institute (TTDI) and National Center for Genetic
Engineering and Biotechnology (BIOTEC), Ministry of
Science, Technology and Environment, Thailand. Technical
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 169
assistance of Miss Lee Woan Ping, National University of
Singapore is gratefully acknowledged.
References
Asaoka, M., Blanshard, J. M. V., & Rickard, J. E. (1992). Effects of cultivar
and growth season on the gelatinization properties of casava starch.
Journal Science Food & Agriculture, 59, 5358.
Asaoka, M., Okumo, K., Sugimoto, Y., & Fuwa, H. (1985). Developmental
changes in the structure of endosperm starch of rice (Oryza sativa L.).
Agric. Biol. Chem., 49, 19731978.
Association of Ofcial Analytical Chemists (AOAC) (1990). Ofcial
method of analysis (15th ed.). Virginia: The Association of Ofcial
Agricultural Chemists.
Baldwin, P. M. (1994). Holes in starch granules: confocal. SEM and light
microscopy studies of starch granule structure. Starch/Starke, 46, 341
346.
Brautlecht, C.A. (1953). Starch: its sources, production and uses. New
York: Reinhold.
Food and Agricultural Organization of the United Nations. 1988 Specica-
tions for Identity and Purity. FAO Food and Nutrition Paper.
Geddes, R., Greenwood, C. T., & Mackenzie, S. (1965). Studies on the
biosynthesis of starch granules. Carbohydrate Research, 1, 7182.
Govindasamy, S., Oates, C. G., & Wong, H. W. (1992). Characterisation of
changes of sago starch components during hydrolysis by thermostable
alpha-amylase. Carbohydrate Polymers, 18, 89100.
Inouchi, N., Glover, D. V., Sugimoto, Y., & Fuwa, H. (1984). Develop-
mental changes in starch properties of several endosperm mutants of
maize. Starch/Starke, 36, 812.
Noda, T., Takahatana, Y., & Nagata, T. (1992). Developmental changes in
properties of sweet potato starches. Starch/Starke, 44, 405409.
Rickard, J. R., Asoaka, M. A., & Blanashard, J. M. V. (1991). The physi-
cochemical properties of cassava starch. Tropical Science, 31, 189
207.
Rodjanaridpiched, C., Limsila, A., Suprahan, D., Boonsenf, O., Poolsan-
guan, P., Tiraporn, C., & Kawano, K. (1993). Recent progress in
cassava varietal improvement in Thailand. In R.H. Howelr (Ed.),
Cassava breeding, agronomy research and technology transfer in
Asia, Proc. 4th Regional workshop held in Trivandrum, India (pp.
125134).
Schoch, T.J. (1964) Swelling power and solubility of granular starches. In
R.L. Whistler, R.J. Smith, & J.N. Be Miller (Eds.) Methods in
Carbohydrate Chemistry, Vol. 4. (pp. 106108). New York: Academic
Press.
Yamada, S., Nagashima, N., & Kawabata, A. (1987). Thermal properties of
some tropical starches by differential scanning calorimetry. Journal
Home Econ. Japan., 38, 647.
K. Sriroth et al. / Carbohydrate Polymers 38 (1999) 161170 170