Industrial Crops and Products 16 (2002) 11 – 21

www.elsevier.com/locate/indcrop

Edible canna (Canna edulis) as a complementary starch

source to cassava for the starch industry
Kuakoon Piyachomkwan a, Sunee Chotineeranat a, Chukiet Kijkhunasatian a,
Ratchata Tonwitowat b, Siripatr Prammanee c, Christopher G. Oates d,
Klanarong Sriroth e,*
a
Cassa6a and Starch Technology Research Unit, National Center for Genetic Engineering and Biotechnology (BIOTEC),
Bangkok, Thailand
b
Plant Science Di6ision, Rajamangala Institute of Technology, Pathumtani, Thailand
c
Faculty of Liberal Arts and Science, Kasetsart Un6iersity, Bangkok, Thailand
d
Agro Food Resources (Thailand) Co., Ltd., Bangkok, Thailand
e
Department of Biotechnology, Faculty of Agro-Industry, Kasetsart Uni6ersity, Jatujak, Bangkok 10900, Thailand
Accepted 19 December 2001

Abstract

Edible canna (Canna edulis Ker) as an alternative starch source was evaluated on the basis of genetic characteristics,
agronomic traits and starch properties. Four canna varieties indigenous to Thailand were examined including
Thai-green, Japanese-green, Thai-purple and Chinese-purple and compared with cassava (Manihot esculenta Crantz).
Using the Random Amplified Polymorphic DNA (RAPD) technique employing ten 10-base primers, four primers
implied that at least three types of canna including Thai-green, Japanese-green and Thai/Chinese-purple existed and
corresponded to plant characteristics as identified by flower, stem, leaf and rhizome colors. Despite genetic
diversification, starch properties were not variable. All four varieties produced 30.4– 38.4 tonne/ha of rhizomes with
starch content about 13% (wet basis). Starch yields of canna (4.1– 4.9 tonnes/ha) were comparatively lower than
cassava (6.5 tonnes/ha). The starches were characterized by giant granules (10– 80 mm), and compared with cassava
starch pastes had a higher peak viscosity (930–1060 BU for canna starches and 815 BU for cassava starch), occurring
at a higher temperature. Pastes of canna starch were more stable and when cooled, viscosity increased to 1800 BU.
Gelatinized pastes of canna starches also rapidly formed good gels on cooling. It is evident that edible canna provides
starches with very attractive properties and totally different from cassava and is the greatest promise for the new base
starch to be employed complementarily with cassava starch. © 2002 Published by Elsevier Science B.V.

Keywords: Canna; Cassava; Variety; Starch

1. Introduction
* Corresponding author. Tel.: + 662-9405634; fax: + 662-
9405634. Starches have a wide range of food and non-
E-mail address: aapkrs@ku.ac.th (K. Sriroth). food applications. The inherent functional diver-

0926-6690/02/$ - see front matter © 2002 Published by Elsevier Science B.V.

PII: S 0 9 2 6 - 6 6 9 0 ( 0 2 ) 0 0 0 0 3 - 1
12 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21
sity of materials extracted from different biologi-
cal sources adds to the range of applications.
However, commercial starch sources are limited
to a few crops, which are further restricted within
geographic regions to one or two crops. Impor-
tant starch sources are cereals such as corn (Zea
mays L.) and wheat (Triticum spp.) in North
America; rice (Oryza sati6a L.) in Asia; roots and
tubers such as potato (Solanum spp.) in Europe
and cassava (Manihot esculenta Crantz) in Asia,
Latin America and Africa. The portfolio of com-
mercial sources does not represent the entire spec-
trum of potential starch functionality. Novel
sources include roots and tubers such as ahipa
(Pachyrhizus ahipa), arracacha (Arracacia xanth-
orrhiza), yam (Dioscorea spp.), arrowroot
(Maranta arundinacea), yacon (Smallanthus
sonchifolius), sweet potato (Ipomoea batatas) and
edible canna (Canna edulis Ker). Opportunities
for improving the potential value, and in turn
production, of these crops are possible if their
utilization can be expanded, such as a source of
industrial raw material of starch production.
Studies supporting utilization should consider im-
proving production, and for down-stream pro-
cesses characterizing this raw material in terms of
starch properties. Expanding utility of these crops
is dependent on expanding or developing their
utilization in either fresh or primary processed
forms.
Edible canna (C. edulis Ker), one of the Andean
root and tuber crops known as Achira is a native
crop of South America (Hermann and Heller,
1997). This crop is distributed across many re-
gions of the world, and is known by a variety of
local names such as ‘Chisgua’ in Columbia, ‘Ca-
pacho’ in Venezuela, ‘Imbirg’ in Brazil, ‘Tous les
mois’ in West Indies, ‘Queensland arrowroot’ in
Australia, ‘Zembu’ in Philippines, ‘Lotus tuber’ in
Taiwan and ‘Sagu’ in Thailand (Tonwitowat,
1994). This plant has high starch rhizomes that
are traditionally used as a staple food for native
people. This crop is typically cultivated in sub-
tropical highland, at an altitude of 1000– 2500 m,
mainly in China and Vietnam, with the variable
rhizome production (150 000–258 4000 tonnes of
rhizomes per year). Interest in canna in parts of
Asia suggests potential for this crop as a raw
material supply to industry especially after a suc-
cessful field trial of canna in low land plantation
(Tonwitowat, 1994). The rhizomes can be pro-
cessed into starch, which is used for making trans-
parent noodles being regarded as higher quality
than mungbean. Preliminary studies of canna
starch suggest some interesting properties, such as
paste properties. Two types of edible canna
namely the ‘green’ and ‘purple’ types are docu-
mented (Kay, 1973; Tawornmas, 1987). However,
there are at least four types of edible canna,
including Thai-green, Japanese-green, Thai-purple
and Chinese-purple, found locally in Asia (Ton-
witowat, 1994).
In this study, genetic characteristics, agronomic
traits and starch properties of these four canna
varieties were investigated and compared with
cassava, the most important commercial root crop
in Asia and Latin America, to evaluate the useful-
ness of canna starch as a complementary product
to cassava starch. Well-characterized properties of
these four canna varieties will be also advanta-
geous for plant breeders, farmers, starch proces-
sors and users to develop this crop industrially.
2. Materials and methods
2.1. Field study
Four canna varieties, Thai-green, Japanese-
green, Thai-purple and Chinese-purple were field
planted at Department of Agronomy, Rajaman-
gala Institute (Pathumthani campus) in July 1999.
A Randomized Complete Block Design (RCBD)
of four replicates was employed and planting area
extended over 24 m2 (planting space of 1× 1 m,
four rows). Three month old sprouts germinated
from underground rhizomes were used as planting
materials. After 8 months, the rhizomes were har-
vested for starch extraction and leaves were col-
lected and kept at − 70 °C prior to use for
genetic analysis. Agronomic characteristics and
rhizome yield of plants were also recorded. Cas-
sava variety Kasetsart 50 (KU50), the most popu-
lar variety grown by farmers, was also planted in
a similar way and the roots were harvested at 8
months for starch extraction.
 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21
2.2. Genetic analysis
Genomic DNA was extracted from leaf tissues
frozen in liquid nitrogen following a modified
method of Dellarporta and Wood (1983) with
RNase treatment, including phenol– chloroform
extraction. The concentration and purity of DNA
samples were analyzed spectrophotometrically
(Pharmacia Biotech) and DNA purity was further
confirmed by 0.7% agarose gel electrophoresis
with ethidium bromide staining. Genetic mapping
of extracted DNA was accomplished by Random
Amplified Polymorphic DNA (RAPD) technique.
Ten primers containing ten base pairs were used
to amplify the genomic DNAs. Six of them were
the universal random primers obtained from
Pharmacia Biotech (Sweden), including primer 1
(5%-GGTGCGGGAA-3%), primer 2 (5%-GTTTCG-
CTCC-3%), primer 3 (5%-GTAGACCCGT-3%), pri-
mer 4 (5%-AAAAGCCCGT-3%), primer 5 (5%-GTT-
TCGCTCC-3%), primer 6 (-5%CCCGTCAGCA-3%).
Two primers including primer 7 (5%-TGGT-
CACTGA-3%) and primer 8 (-5%TGGACACTGA-
3%) were used following the method of William et
al. (1990). The last two primers were Operon no.
06 (primer 9; 5%-AAGGCGGCAG-3%) and Operon
no. 18 (primer 10; 5%-TGCCCAGCCT-3%). In 25
ml of reaction mixture; there were 1×PCR buffer
(3 mM MgCl2, 30 mM KCl, 10 mM Tris – HCl,
pH 8.0), 0.4 mM each of dNTPs (Promega), 25
pmol of the primer and 0.6 u Taq DNA poly-
merase (Promega) and 20 ng of genome DNA.
The thermal program was 45 cycles of 95 °C for
1 min, 36 °C for 1 min and 72 °C for 2 min
(Gene Amp PCR System 2400, Perkin– Elmer).
DNA amplified products were analyzed on 2%
agarose (SeaKem® LE Agarose) in TBE buffer
(0.89 M Tris–base pH 8.3 containing 0.89 mM
boric acid and 24 mM EDTA) at 90 V in 5.0 h.
The gel was stained with ethidium bromide (0.5
mg/ml) for 30 min.
2.3. Rhizome and starch properties
Physical and chemical properties of harvested
rhizomes were investigated for protein, fat, fiber
and ash contents according to the AOAC meth-
ods (AOAC, 1990). Starch was extracted by
13
grinding peeled rhizomes with water and filtering
through a 88 mm sieve. Starch slurry was then
collected and settled. After starch sedimentation,
water was decanted and starch cake was rewashed
twice before drying at 50 °C. Starch purity was
determined by a polarimetric method (AOAC,
1995). Properties of dried starch were character-
ized and compared with cassava starch extracted
in the same manner from 8 month old roots of
Kasetsart 50 (KU50).
2.3.1. Granule morphology and size
Starch granules were viewed under normal and
polarized light microscope (Meiji Techno., Japan)
at 4× magnification. Images were taken with a
Meiji CK 3800 (Meiji Techno., Japan) color video
camera and the granule sizes were then analyzed
by image analysis software (Image Pro Plus 3.0,
Media Cybernatic, LP, USA).
2.3.2. Whiteness
Starch whiteness was estimated using a KETT
Digital Whiteness Meter Model C-100 (KETT
Electric Laboratory, Germany) with a blue filter
(u 440 nm) as a calibration plate.
2.3.3. Paste clarity
Starch suspension (1% w/v) was gelatinized in a
boiling bath for 30 min. The clarity of paste, after
cooling, at room temperature for 60 min was
measured as the percentage light transmittance
(%T) at 650 nm, using water as a blank, accord-
ing to the method of Craig et al. (1989).
2.3.4. Swelling and solubility
Swelling and solubility determinations were car-
ried out at 55, 65, 75 and 85 °C by the procedure
of Schoch (1964).
2.3.5. Paste properties
Pasting profiles of starch slurries were investi-
gated by a Visco Amylograph (Barbender, Ger-
many) using 30.00 g (dry basis) starch suspended
in 470 ml distilled water (6.00%starch). The starch
suspension was heated from 50 to 95 °C within 40
min and the holding time at 95 °C was 30 min.
After which, starch paste was cooled down to
50 °C at the same rate and this temperature was
14 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21

Fig. 1. DNA polymorphisms of four canna varities from RAPD using (a) primer 1 (5%-GGTGCGGGAA-3%) (b) primer 7
(5%-TGGTCACTGA-3%) (c) primer 9 (5%-AAGGCGGCAG-3%) and (d) primer 10 (5%-TGCCCAGCCT-3%). Lane 1: Thai-green, Lane
2: Japanese-green, Lane 3: Thai-purple, Lane 4: Chinese-purple and Lane M: marker 100 bp DNA ladder of 1500, 1200, 1000, 900,
800, 700, 600, 517, 500, 400, 300, 200 and 100 bp.

Fig. 2. Rhizomes of edible canna (C. edulis) of different varieties: (a) Thai-green (b) Japanese-green (c) Thai-purple and (d)
Chinese-purple.
 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21 15

Table 1
Phenotypes and agronomic traits of four edible canna (C. edulis) varieties

Characteristics Variety

Thai-green Japanese-green Thai-purple Chinese-purple

Stem color Green Green Bronze Bronze

Leaf color Green Green Green/purple Green/purple
Plant height (cm) 235 248 243 243
50% Inflorescence (days after planting) 86 79 80 93
Flower color Orange Red Red Red
Harvest time (months) 8 8 8 8
Number of shoots per plant at harvest 40 49 43 33
Number of rhizomes per plant at harvest 46 50 48 44
Rhizome color Yellow Yellow Yellow/purple Yellow/purple
Rhizome yield at harvest (tonnes/ha) 33.5 38.4 33.7 30.4
Total plant yield at harvest (tonnes/ha) 22.2 25.8 22.9 21.9
Harvest index 0.64 0.62 0.64 0.58
Starch content of harvest rhizomes (% wet basis) 13.52 12.74 13.23 13.63
Starch yield (tonnes/ha) 4.53 4.89 4.46 4.14

Table 2
Chemical composition of edible canna rhizomes of different varieties and cassava roots

Sample Content (% dry basis)

Protein Fat Fiber Ash

Thai-green canna 3.269 0.12 0.25 9 0.01 3.00 9 0.11 5.56 90.01
Japanese-green canna 2.899 0.04 1.63 90.22 2.90 9 0.06 5.26 90.01
Thai-purple canna 2.399 0.03 0.29 90.09 3.18 9 0.24 5.70 90.08
Chinese-purple canna 3.16 9 0.36 0.34 9 0.06 2.99 9 0.00 5.11 90.06
Cassava 1.159 0.04 0.22 90.03 1.31 9 0.06 1.91 90.01

held for 30 min. The viscosity of starch paste,
including peak viscosity (P), final viscosity at
95 °C (H), viscosity at 50 °C (C) and gelatiniza-
tion temperature was recorded. The breakdown
(P–H), setback (C– P) and consistency (C– H)
were also reported.
2.3.6. Enzyme susceptibility
Enzyme susceptibility of starch was measured
by using a-amylase and amyloglucosidase (Novo
Nordisk Co., Bagsvaerd, Denmark), similar to the
method of Wang et al. (1995). The 10 ml reaction
mixture contained a 2.0% starch suspension in 0.1
M sodium acetate buffer (pH 5.0) and 40 ppm
Ca2 + and the enzyme solution containing 3 KNU
of a-amylase (Termamyl® 120L from Bacillus
licheniformis with 120 Kilo Novo a-amylase Unit
(KNU)/g; 1 KNU is the amount of enzyme that
hydrolyzes 5.26 g of starch per h under the fol-
lowing standard conditions, i.e. at 37 °C, pH 5.6
for 7–20 min) and 6 AGU of amyloglucosidase
(AMG 300L from Aspergillus niger with 300
Amyloglucosidase Unit (AGU)/ml; 1 AGU is the
amount of enzyme which hydrolyzes 1 mmole
maltose per min under the following standard
conditions, i.e. at 25 °C, pH 4.3 for 30 min).
After incubating at 37 °C for 48 h, the content of
reducing sugar was analyzed using the Somogyi–
Nelson method (Somogyi, 1952) and total sugars
by the method of Dubois et al. (1956).
16 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21

2.3.7. Acid susceptibility Table 3

Starch samples (1% w/v) were suspended in Whiteness and paste clarity of canna starches of different
varieties and cassava starch
aqueous sulfuric solution (1.35 N H2SO4) and
incubated at 37 °C with intermittent shaking. After Sample Whiteness Paste clarity
24 and 72 h, samples were collected and centrifuged (%transmittance
at 10 000 rpm 5 min. The supernatant was analyzed at 650 nm)
for total sugar contents by the phenol– sulfuric
Thai-green canna 91.5 90.07 43.55 9 0.35
method (Dubois et al., 1956). Japanese-green canna 90.6 90.00 49.70 9 0.42
Thai-purple canna 89.9 9 0.00 43.15 9 0.07
2.3.8. Starch gel texture Chinese-purple canna 90.8 9 0.07 51.30 9 1.13
To prepare gels, starches (9.2% w/w) were Cassava 93.4 90.07 38.00 9 0.28
cooked in the Rapid Visco Analyzer according to
Sriroth et al. (1998). The paste was covered with a
parafilm sheet and allowed to cool and set at the

Fig. 4. Swelling power and % solubility of canna starches of

different varieties and cassava starch.

room temperature. The set gel was then stored at

4 °C for 15 h and equilibrated at the room temper-
Fig. 3. Granular appearance of canna starches under a (a) ature for 4 h before texture analysis using a Texture
normal light and (b) polarized light microscope at 40 × mag- Analyzer (LLOYD TA 500, IntroEnterprise Co.
nification. LTD, USA; Collado and Corke, 1999). The gel of
Table 4
Paste viscosity properties by a Brabender Visco Amylograph of canna starches obtained from different varieties and cassava starch (6% w/w)

Sample Gelatinization- Peak viscosity Final viscosity at Viscosity at 50 °C Breakdown Setback Consistency
temperature (°C) (P, BU) 95 °C (H, BU) (C, BU) (P–H, BU) (C–P, BU) (C–H, BU)

Thai-green canna 68.9 1058 989 1661 69 603 672

Japanese-green canna 69.1 1044 1011 1801 33 757 790
Thai-purple canna 69.4 932 921 1567 11 635 646
Chinese-purple canna 68.8 1047 1000 1668 47 621 668
Cassava 63.4 815 247 461 568 −354 214
K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21
17
18 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21

20 mm in height was compressed by 40% of the 3. Results and discussion

original height at a speed of 20 mm/min using a
cylindrical probe with a diameter of 50 mm. The
analysis was performed into two cycles. The peak
height at 30% compression of the first cycle was
termed as hardness. Cohesiveness is the ratio of
the area of work during the second compression
to the first compression. Springiness is the ratio of
the compressed time of the second cycle to the
first cycle. Gumminess is the product of hardness
and cohesiveness.
3.1. Genetic characteristics
Genetic diversity of four canna varieties includ-
ing Thai-green, Japanese-green, Thai-purple and
Chinese-purple were probed by the RAPD tech-
nique using ten 10-base primers. Except primer 4,
nine sets of primers were able to generate DNA
polymorphism patterns in all canna varieties.
DNA patterns which were amplified by primer 2,

Fig. 5. Paste viscosity profiles of canna starches of different varieties and cassava starch when determined by a Brabender Visco
Amylograph using 30.00 g (dry basis) starch in 470 ml distilled water.

Table 5
Texture Profile Analysis (TPA) of canna and cassava starch gels

Hardness (N) Cohesiveness Springiness (mm) Gumminess (N)

Thai-green canna 34.989 2.95 0.40 90.03 7.44 90.13 13.96 91.54
Japanese-green canna 40.98 9 2.64 0.46 90.05 7.67 90.28 18.58 90.98
Thai-purple canna 38.209 2.74 0.36 90.02 7.69 90.08 13.88 91.51
Chinese-purple canna 30.449 1.25 0.449 0.04 7.76 90.19 13.29 90.72
Cassava n.a. n.a. n.a. n.a.

n.a.= not applicable.

 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21 19

Table 6
Acida and enzymeb susceptibility of canna starches of different varieties and cassava starch

Acid susceptibility (%) Enzyme susceptibility (%)

At 24 h at 72 h

Thai-green canna 43.55 9 0.35 91.5 90.07 36.56 9 3.73

Japanese-green canna 49.709 0.42 90.6 9 0.00 31.32 9 3.51
Thai-purple canna 43.159 0.07 89.9 90.00 44.55 9 4.37
Chinese-purple canna 51.309 1.13 90.8 90.07 39.42 9 5.22
Cassava 38.009 0.28 93.4 9 0.07 54.56 9 5.72

a
By 1.35 N sulfuric acid at 24 and 72 h.
b
By 3 KNU a-amylase (Termamyl® 120L from B. licheniformis with 120 Kilo Novo a-amylase Unit (KNU)/g; 1 KNU is the
amount of enzyme that hydrolyzes 5.26g of starch/hr under the following standard conditions, i.e. at 37 °C, pH 5.6 for 7–20 min)
and 6 AGU glucoamylase (AMG 300L from A. niger with 300 Amyloglucosidase Unit (AGU)/ml; 1 AGU is the amount of enzyme
which hydrolyzes 1 mmole maltose per minute under the following standard conditions, i.e. at 25 °C, pH 4.3 for 30 min).

3 and 8 exhibited genetic similarity in all four
varieties. When using primer 5 and 6, there were
slightly different genetic patterns among the four
varieties. Genetic differences between Thai-green
and Japanese-green were observed by the amplifica-
tion with primer 1, 7, 9 and 10, they were different
from the other two varieties (Thai-purple and
Chinese-purple) (Fig. 1). Only two varieties of
canna are reported in the literature which are the
‘green’ and ‘purple’ types. However, this study
suggests that at least three types of canna exist
including Thai-green, Japanese-green and Thai/
Chinese-purple.
those achieved in field trials in the highlands of
Ecuador of 56 tonnes/ha (Hermann et al., 1996).
Harvest index, the ratio of the rhizome to total
plant weight (: 0.62) and starch content of har-
vested rhizomes (: 13% wet basis) of the four
varieties corresponded to a previous study; harvest
index of 0.56 and rhizome starch content of 14%
wet basis (Hermann et al., 1996). In this trial, starch
yields of four canna varieties were 4.14 to 4.89
tonnes/ha, which is lower than that attained in field
trials in Ecuador (7.8 tonnes/ha; Hermann et al.,
1996). Starch yields of canna were also lower than
those of cassava (6.5 tonnes/ha).

3.2. Phenotypes and agronomic traits 3.3. Rhizome and starch properties

Similarly, on the basis of plant characteristics,
three types of canna could be categorized including
Thai-green, Japanese-green and Thai/Chinese-pur-
ple. The first two could be readily differentiated
from the last by color of stems, leaves and rhi-
zomes. They were in green for the former and
green/purple for the latter (Fig. 2). Thai- and
Japanese-green types were distinguished by flower
color. Other phenotypes and agronomic traits of
these four varieties were similar as summarized in
Table 1. Rhizome yields for the four canna varieties
varied from 30.4 to 38.4 tonnes/ha with the
Japanese-green type providing the highest yield
(Table 1) and slightly greater than typical root
yields of cassava (25 tonnes/ha). However, average
canna yields reported in this study are lower than
Potential use of canna for industrial purposes
was evaluated by characterizing rhizome and starch
properties. Similar to other root and tuber crops,
canna rhizomes are high in starch content with
small amounts of other impurities (protein : 2.39–
3.26, fat : 0.25–1.63, fiber :2.90–3.18 and ash
: 5.11–5.70% dry basis; Table 2). Canna rhizomes
contain a higher ash content than cassava roots
(Table 2). Canna starch can be readily extracted
from the rhizomes by simple techniques including
grinding, sedimentation and drying which are effec-
tive enough to produce pure canna starch (starch
: 94.0–96.6, protein : 0.05–0.20, fat : 0.01–
0.15, fiber : 0.40–0.90 and ash :0.70–0.90% dry
basis). All varieties of canna starch had rounded
and oval-shaped granules with wide size distribu-
20 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21
tion ranging from 5 to 100 mm (Fig. 3). Granular
sizes were similar to potato starch, the large gran-
ules (:70%) were 30– 70 mm and much larger
than cassava starch (granule size ranging from 5
to 30 mm with the average size of 12– 15 mm).
In general, genetic factors did not seem to
contribute to the variation in starch functional
properties. All four canna varieties provided
starch with high whiteness (89.9– 91.5), close to
cassava starch (93.4) (Table 3). When gelatinized,
the cooked starch paste was very clear and clearer
than cassava starch (Table 3). Independent of
variety, all canna starches exhibited lower swelling
power than cassava starch at all temperatures
tested, while starch solubility was lower only at
low temperatures (Fig. 4). Among four varieties,
Thai-green canna yielded starch with the highest
swelling power and solubility, which were evident
at high temperatures. Significant changes in
swelling power of all four canna starches occurred
at higher temperatures (\65 °C) compared with
cassava starch (\ 55 °C), as ascertained by
Brabender Visco amylograph (Table 4). Gela-
tinization of the four canna starches began at
higher temperatures than cassava starch (starch
gelatinization temperatures were 69 and 63 °C,
respectively). Paste profiles of all four canna
starches were similar, Japanese-green canna gave
the highest cold paste viscosity (Fig. 5). All were
considerably different to cassava starches. Viscos-
ity after holding at 95 °C for canna paste (final
viscosity at 95 °C, H, was 921 – 1011 BU) de-
creased slightly from the maximum viscosity
(peak viscosity, P, was 932– 1058 BU) indicating
very good paste stability (breakdown, P– H was
11– 69 BU). After cooling to 50 °C, paste viscos-
ity of all four canna markedly increased (Viscosity
at 50 °C, C, was 1567 –1801 BU), resulting in
high setback (603–757 BU). In contrast, cassava
starch provided hot paste with low stability
(breakdown, P –H was 568 BU), after cooling,
paste viscosity also increased, yet it was still lower
than the peak viscosity (815 BU). Consequently,
setback of cassava starch was negative (− 354
BU). The results are comparable to previous re-
ports (Perez et al., 1997, 1998). Gels formed from
the gelatinized canna starches had superior tex-
tural properties than cassava starch (Table 5).
Granule susceptibility of canna starches to acid
and enzymatic hydrolysis was also evaluated in
comparison with cassava starch. The initial acid
hydrolysis of canna starches was more rapid than
cassava starch but all were degraded to about the
same extent after prolonged hydrolysis (Table 6).
Starch granules were more susceptible to acid
than enzymatic hydrolysis. Degree of enzymatic
hydrolysis of canna starches was slightly lower
than that of cassava starch. Among four canna
varieties, Thai-purple canna had the highest de-
gree of enzymatic hydrolysis (Table 6).
It is conclusive that edible canna provides
starches with distinct paste and gel characteristics
and totally different from cassava. The starches
are potential candidates for industrial utilization
as a new base starch. Interesting is to use canna
starch as a complementary starch to cassava
starch, allowing the production of tailor-made
composite starches with improved and desired
properties.
Acknowledgements
This work was partially supported by the Min-
istry of Industry, Thailand.
References
Association of Official Analytical Chemists (AOAC), 1990.
Official Methods of Analysis. 15th ed. The Association of
Official Analytical Chemists, Virginia.
Association of Official Analytical Chemists (AOAC), 1995.
Official Methods of Analysis. 16th ed. The Association of
Official Analytical Chemists, Virginia.
Collado, L.S., Corke, H., 1999. Heat-moisture treatment ef-
fects on sweetpotato starches differing in amylose content.
Food Chem. 65, 339 – 346.
Craig, S.A.S., Maningat, C.C., Seib, P.A., Hoseney, R.C.,
1989. Starch paste clarity. Cereal Chem. 66, 173 – 182.
Dellarporta, S.L., Wood, J.B., 1983. A plant DNA miniprepa-
ration version II. Plant Mol. Bio. Rep. 1 (4), 19 – 21.
Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A.,
Smith, F., 1956. Colorimetric method for determination of
sugars and related substances. Anal. Chem. 28, 350 – 356.
Hermann, M., Heller, J., 1997. Andrean Roots and Tubers:
Ahipa, Arracacha, Maca and Yacon. Promoting the Con-
servation and Use of Underutilized and Neglected Crops,
vol. 21. Institute of Plant Genetics and Crop Plant Re-
 K. Piyachomkwan et al. / Industrial Crops and Products 16 (2002) 11–21
search, Gatersleben/International Plant Genetic Resources
Institute, Rome, Italy.
Hermann, M., Uptmoor, R., Freire, I., Montalvo, J. L., 1996.
Crop growth and starch productivity of edible canna.
International Potato Center. http://www.cipotato.org/
market/pgmrprts/pr95%2D96/program6/prog611.htm.
Kay, D.E., 1973. TPI Crop and Product Digest No. 2 Root
Crops. Prod. Institute, London, p. 245p.
Perez, E., Lares, M., Gonzalez, Z., 1997. Some characteristics
of sagu (Canna edulis Kerr) and zulu (Maranta sp.) rhi-
zomes. J. Agric. Food Chem. 45, 2511 –2514.
Perez, E.E., Breene, W.M., Bahnassey, Y.A., 1998. Variations
in the gelatinization profiles of cassava, sagu and arrow-
root native starches as measured with different thermal and
mechanical methods. Starch/Starke 50, 70 –72.
Somogyi, M., 1952. Notes on sugar determination. J. Biol.
Chem. 195, 19 – 23.
Schoch, T.J., 1964. Swelling power and solubility of granular
starches. In: Whistler, R.L., Smith, R.J., Bemiller, J.N.
21
(Eds.), Method in Carbohydrate Chemistry, vol. 4. Aca-
demic Press, New York, pp. 106 – 108.
Sriroth, K., Wanlapatit, S., Piyachomkwan, K., Oates, C.G.,
1998. Improved cassava starch granule stability in the
presence of sulphur dioxide. Starch/Starke 50, 466 – 473.
Tawornmas, D., 1987. Sagu. Department of Agriculture, Min-
istry of Agriculture and Cooperatives, Bangkok, p. 4.
Tonwitowat, R., 1994. Yield and some agronomic characteris-
tics of edible canna (Canna edulis Ker) as affected by
varieties, planting dates, fertilizer rates and plant densities.
Ph.D. Thesis. Kasetsart university, Bangkok.
Wang, W.J., Powell, A.D., Oates, C.G., 1995. Pattern of
enzyme hydrolysis in raw sago starch: effects of processing
history. Carbohydrate Polymers 25, 91 – 97.
William, J.G.K., Kubelik, A.R., Rafalsk, J.A., Tingey, S.V.,
1990. DNA polymorphisms amplified by arbitrary primers
are useful as genetic markers. Nucl. Acid Res. 18, 6531 –
6535.