Genetic diversity in Eddoe Taro (Colocasia esculenta var. antiquorum) from Indonesia based on morphological and nutritional characteristics

Abstract

Maretta D, Sobir Helianti I, Purwono, Santosa E. 2020. Genetic diversity in Eddoe Taro (Colocasia esculenta var antiquorum) from Indonesia based on morphological and nutritional characteristics. Biodiversitas 21: 3525-3533. Low yield uniformity and quality due to genetic performance become negative incentives to farmers in Eddoe Taro production. However, genetic evaluation is rarely been reported in this taro type in Indonesia. In this study, 14 eddoe genotypes collected from different regions in Indonesia were evaluated to develop a diversity map for crop improvement and future breeding activities. The genotypes were planted in the open field from September 2018 to March 2019 at the experimental station belonging to LAPTIAB-BPPT, PUSPITEK at South Tangerang District, Indonesia. Morphological and nutritional characters were accessed on the shoot and underground parts. The genotypes exhibited variation in 38 out of 48 characters in which 12 quantitative characters were distinct including oxalate level. The study revealed three findings: (i) Characters related to growth and yield had high genotypic variance coefficients, i.e., sheath length, total petiole length, plant height, number of suckers, corm and cormels weight, (ii) Genotypes clustered into two separate groups as introduced and landraces, and (iii) Landraces had high genetic variation leading to speculation of high clonal variation. Considering the findings, accession S6, S7, S18, S30, and S36 are recommended for further studies in crop improvement purposes.

Full text

B I O D I V E R S I T A S
ISSN: 1412-033X
Volume 21, Number 8, August 2020 E-ISSN: 2085-4722
Pages: 3525-3533 DOI: 10.13057/biodiv/d210814
Genetic diversity in Eddoe Taro (Colocasia esculenta var. antiquorum)
from Indonesia based on morphological and nutritional characteristics
DELVI MARETTA1,2,♥, SOBIR3,4, IS HELIANTI5, PURWONO4, EDI SANTOSA4, ♥♥
1Department Agronomy and Horticulture, Faculty of Agriculture, Graduate School, Institut Pertanian Bogor. Jl Meranti, Kampus IPB Darmaga, Bogor
16680, West Java, Indonesia. Tel.: +62-251-8629354, 8629350, ♥email: delvi.maretta@bppt.go.id
2Center for Agricultural Production Technology, Deputy of Industrial Technology for Agricultural and Biomedicine, The Agency for Assessment
and Application of Technology. Building 614, LAPTIAB-BPPT, PUSPIPTEK, South Tangerang 15314, Banten, Indonesia
3Center for Tropical Horticulture Studies, Institut Pertanian Bogor. Jl Raya Pajajaran Kampus IPB Baranangsiang, Bogor 16680 Indonesia
4Department Agronomy and Horticulture, Faculty of Agriculture, Institut Pertanian Bogor. Jl Meranti Kampus IPB Darmaga, Bogor 16680 Indonesia.
Tel.: +62-251-8629354, 8629350, ♥♥email: edisang@gmail.com
5Center for Bio-industrial Technology, Deputy of Industrial Technology for Agricultural and Biomedicine, The Agency for Assessment and Application
of Technology. Building 614, LAPTIAB-BPPT, PUSPIPTEK, South Tangerang 15314, Banten, Indonesia
Manuscript received: 1 May 2020. Revision accepted: 10 July 2020.
Abstract. Maretta D, Sobir Helianti I, Purwono, Santosa E. 2020. Genetic diversity in Eddoe Taro (Colocasia esculenta var
antiquorum) from Indonesia based on morphological and nutritional characteristics. Biodiversitas 21: 3525-3533. Low yield
uniformity and quality due to genetic performance become negative incentives to farmers in Eddoe Taro production. However, genetic
evaluation is rarely been reported in this taro type in Indonesia. In this study, 14 eddoe genotypes collected from different regions in
Indonesia were evaluated to develop a diversity map for crop improvement and future breeding activities. The genotypes were planted in
the open field from September 2018 to March 2019 at the experimental station belonging to LAPTIAB-BPPT, PUSPITEK at South
Tangerang District, Indonesia. Morphological and nutritional characters were accessed on the shoot and underground parts. The
genotypes exhibited variation in 38 out of 48 characters in which 12 quantitative characters were distinct including oxalate level. The
study revealed three findings: (i) Characters related to growth and yield had high genotypic variance coefficients, i.e., sheath length,
total petiole length, plant height, number of suckers, corm and cormels weight, (ii) Genotypes clustered into two separate groups as
introduced and landraces, and (iii) Landraces had high genetic variation leading to speculation of high clonal variation. Considering the
findings, accession S6, S7, S18, S30, and S36 are recommended for further studies in crop improvement purposes.
Keywords: Crop improvement, Eddoe Taro, genetic properties, glucomannan, oxalate
INTRODUCTION
Taro (Colocasia esculenta (L.) Schott) is an important
food in many localities in the humid tropics and subtropics
(Chaïr et al. 2016). The corm is rich of nutrients such as
carbohydrate, protein, elements (Fe, Ca, P, Mg, Na, and K)
and vitamins (A, B1, B2, B3, and C) (Ezeabara et al. 2015,
Mergedus et al. 2017); the protein level is higher than
cassava and sweet potato tubers (Temesgen and Retta
2015). Taro tubers are also important sources of
anthocyanins, cyanidin 3-glucoside, and flavonoids that act
to improve blood circulation, antioxidants, and inhibit
cancer development (Rashmi et al. 2018).
Morphologically, taro has two forms according to the
corm and cormel developments, i.e., dasheen and eddoe
types that botanically called as Colocasia esculenta var
esculenta and Colocasia esculenta var antiquorum,
respectively and var esculenta distributes widely in the
globe, while var antiquorum predominantly distributes in
China and Japan (Plucknett 1983). For this reason, var
antiquorum to some extend is called Satoimo or Japanese
taro.
In Indonesia, the production of Eddoe Taro is getting
popular to fulfill the high demand for export. Since 2013,
about 6,300 tons of frozen taro have been exported to Japan
(ITPC 2014). For such reason, intensive cultivation has
been being developed in some districts like Bantaeng
(South Sulawesi), Banggai Kepulauan (Central Sulawesi)
(Laosa et al. 2016), Kapahiang (Bengkulu) (Amelia and
Yumiati 2016) and provinces such as East and West Java
(Astuti et al. 2017) and Aceh (Rosdanelly et al. 2018). The
Indonesian Government also provides seeds and subsidies
to farmers.
However, in the field, many farmers face problems on
quality and productivity leading to the low economic
benefit of the business. In Eddoe Taro, the quality of
cormels is determined by glucomannan and oxalate
content. Glucomannan content in taro has been intensively
studied (Njintang et al. 2011; Ekowati et al. 2015). The
glucomannan is explored in concern to health and beauty
(Bateni et al. 2013; Tester and Al-ghazzewi 2016).
Glucomannan is a neutral, fermentable and viscous dietary
fiber that has been proven to reduce obese (Zalewski et al.
2015), to relieve physiological disorders especially diabetes
and cardiovascular diseases (Shah et al. 2015), to reduce
blood lipid and cholesterol (Behera and Ray 2016), and to
extend storage in the frozen form of processed meat and
fish products (Yang et al. 2017). Thus, it is desirable to
produce cormels with high glucomannan content. On the
other hand, oxalate content as anti-nutrient should be low

B I O D I V E R S I T A S
21 (8): 3525-3533, August 2020
3526
(Akalu and Geleta 2017). The mucilage of the fresh taro
causes irritation (Yu et al. 2015), stimulates kidney stones,
and impairs the absorption of minerals such as iron and
calcium in the body (Hang et al. 2013). High oxalate
content causes itching in the mouth, burning sensation, and
skin irritation (Kaushal et al. 2012; Dewi et al. 2017),
leading to low palatability.
One effort to improve quality and yield of the taro is
through genetic improvement (Banjaw 2017). Instead of
many improved skills on general taro breeding, the
diversity of eddoe type in Indonesia is still unknown. Since
genotype characterization is a fundamental step onto
selection of candidate parents for future breeding programs
(Pitoyo et al. 2018). Thus, the study aimed to evaluate
morphological and nutritional characters of Eddoe Taro in
Indonesia.
MATERIALS AND METHODS
Study site and plant materials
The experiment was conducted in rainy season from
September 2018 to January 2019 in the open field at
Experimental Station of Laboratory for Development of
Industrial Technology for Agricultural and Biomedicine
(LAPTIAB), Research Center for Science and Technology
(PUSPIPTEK) Setu subdistrict, South Tangerang, Banten,
Indonesia. The site had altitude 60 m above sea level
(6°21'26.5"S 106°39'56.2"E) with clay soil (Red Yellow
Podzolic). The soil had pH 6.1, low C/N ratio (C/N=10),
high available phosphorus (67 mg P2O5/100 g; HCl 25%
extraction), low available potassium (32 mg K2O/100 g;
HCl 25% extraction) and high cation exchangeable
capacity 14.58 me/100 g by NH4CH3CO2 extraction.
Fourteen Eddoe Taro genotypes were obtained from six
provinces (Table 1). The type had been verified in the
preliminary experiment and coded following the official
record. All genotypes are conserved in the LABTIAB
facilities with copy genotypes that are maintained at Bogor
Agriculture University, Indonesia.
Cultivation method
All genotypes were planted in two blocks, to minimize
the variation of soil fertility. The first block was located at
higher soil level, than that of the second block. In each
block, the arrangement of the genotype was randomized.
Each genotype was planted 5 plants in each block.
Before planting, the soil was plowed and harrowed
twice; and the planting site was designed using a raised bed
about 15 cm from the soil level. The width of the planting
bed was one meter, and each bed only planted a single line.
Soil liming at rate 2 t.ha-1 was applied after bedding.
Seed planting used cormlet, seized 2.5-3.5 cm in
diameter, and 30-50 g in weight depending on genotype.
Among genotypes, planting distant applied 100 cm, while
60 cm in a row within a particular genotype. In each
planting hole, a single cormlet was used. At planting, the
cormlet had no leaf was exist. Organic manure from cow
dung was applied at a week before planting, about 1 kg for
each planting hole. NPK fertilizers were applied twice. The
first application was conducted four weeks after planting
(WAP) using one-third of the total NPK dose. The second
application using the rest of the dose was conducted at 12
WAP. NPK fertilizer derived from single fertilizer, i.e., 120
kg.ha-1 Urea (46% N), 50 kg.ha1 SP36 (36% P2O5) and 150
kg.ha-1 KCl (60% K2O)/. Weeding used manual and
pesticide spraying using common chemicals according to
field conditions.
Morphological evaluation
Morphological data were obtained at maximum
vegetative growth (14 WAP) and at harvesting time (20
WAP). Morphological description followed IPGRI (1999),
and nutritional characters focused on chlorophyll, oxalate,
and glucomannan contents. Thus, a total of 48 characters
were evaluated.
The morphological evaluation focused on plant habit,
leaf, petiole, corm, cormel, and root (Figure 1). Plant habits
included plant height (Ph), plant span (Ps), and the number
of suckers. Leaf characters included leaf base, predominant
position of leaf lamina surface, leaf blade margin, leaf
blade color, leaf blade margin color, leaf lamina
appendages, leaf main vein color, vein pattern, lamina
length, lamina width, sheath length, leaf sheath edge color,
and leaf waxiness. Petiole characters included petiole
junction pattern, petiole junction color, petiole stripe,
petiole stripe color, petiole basal-ring color, the cross-
section of the lower part of petiole, and total petiole length.
Corm characters included corm manifestation, length, corm
branching, shape, weight, cortex color, flesh color of the
central part, corm flesh fiber color, skin surface, skin
thickness, degree of fibrousness of corm, and bud color.
Cormel characters included weight, number, shape, flesh
color of cormels, and root characters included color and
uniformity of color.
Figure 1. Part of the eddo taro plant. Note: Ph-plant height, Ps-
Plant span (Picture adopted from https://www.seedsofindia.com)

MARETTA et al. – Genetic diversity of Indonesian eddoe taro
3527
Nutritional analysis
Glucomannan and oxalate were analyzed from mixed
flour of corm and cormlets. The preparation followed to
Chairul and Chairul (2006). After cleaning, corm/cormlets
were peeled, thin-sliced and oven-dried at 80°C for 20
hours continuously. The dried-chips then were made into
flour using a blender (Madato type) and then shieved using
80 mesh before preparing the analysis.
Glucomannan analysis used the gravimetric method of
Widjanarko and Megawati (2015). Five-gram taro flour
was added to preheated 50 mL distilled water (75°C) plus
0.5 g Al2(SO4)3 (10% of sample weight) in Erlenmeyer
glass inside the water bath. The mixture was kept at 75°C
while stirring for 35 min, then the solution was transferred
into falcon tube 50 mL and centrifuged at 2000 rpm (25°C
for 30 min). The supernatant was transferred to a new
falcon tube and then added isopropyl alcohol 1:1 (v/v). The
supernatant was inverted until coagulated. The pellet was
then filtered using filter paper (Whatman No 1 qualitative
125 mm) equipped with a vacuum pump (Rocker 300
type). Finally, the pellet was oven-dried at 60°C for 24
hours. Percentage glucomannan content on dry basis was
estimated from ratio final weight to sample weight.
Oxalate analysis referred to Naik et al. (2014). Initially,
taro flour weighed 0.25 g was transferred in to test tube,
added 10 mL 0.25 N HCl, and then heated at 85-90°C for
15 min in the water bath. After cooling in room
temperature, the volume was adjusted into 25 mL with 0.25 N
HCl and mixed gently. The supernatant (1 mL) was then
used for oxalic acid measurement using spectrophotometric.
Before measurement, the fresh stock solution was
prepared. For this purpose, 1000 ppm standard solution was
developed from 139.6063 g oxalic acid (C2H2O4.2H2O)
diluted in 100 distilled water. The second stock was 0.02 M
KMnO4, it was prepared by dissolving 3.1606 g KMnO4
into 1 L distilled water. The third stock was 2 N H2SO4, it
was made by dissolving 27.8 mL concentrate H2SO4 with
distilled water into a final volume of 500 mL. The standard
solution was made for 0, 5, 10, 15, 20, 25, 30, 35, 40, 45
and 50 ppm of the oxalic acid standard solution. In each 1
mL solution was added 5 mL 2 N H2SO4 and 2 mL of
0.003 M KMnO4, then incubated at room temperature
(27±2oC) for at least 10 min. The measurement used UV-
1800 spectrophotometric Shimadzu at 528 nm.
Data analysis
Analysis of variance (ANOVA) was performed by F-
test to investigate the presence of statistically significant
differences among genotypes for quantitative characters.
Duncan's multiple range test p<0.05 was addressed to
estimate genotypic differences by optimizing the block as
replication.
Cladogram analysis used DARwin 6
(http://darwin.cirad.fr) by utilizing all quantitative and
qualitative data. In the clustering process, the software was
set in dissimilarity mode, unweighed neighbor-joining and
bootstrap 1000 times.
The relationships among genotypes was estimated using
STRUCTURE 2.3.4, using combining data of quantitative
and qualitative characters. The calculation used the
admixture model, length of period 100,000, number of
MCMC repeat after burned 100,000, K=3, and the number
of iteration 20. Optimum K was determined by using
Structure Harvester (Earl et al. 2012) through online
(http://taylor0.biology.ucla.edu/structureHarvester).
Genotypic and phenotypic variances estimation and the
calculation of the variance coefficient followed Syukur et
al. (2015). The formulas as follows: Genotypic variance
(2g) = MSg-MSe; Genotypic variance coefficient (GVC):
[√2g/x] x 100; Phenotypic variance (2p) = 2g + MSe;
Phenotypic variance coefficient (PVC): [√2p)/x] x 100;
Here, MSg and MSe denoted genotypic and error means
square, respectively. The 'r' represented replication (n=2)
and 'x' represented mean value. Deshmukh et al. (1986)
classified PVC and GVC as high (> 20%), medium (10 to
20%), and low (< 10%). Estimate of broad-sense
heritability (h2) was calculated based on variant value of
genetic and phenotypic follow Stansfield (1983), h2 = ; ;
the h2 was classified as high (h2 > 0.50), medium (0.20 ≤ h2
≤ 0.50), and low (h2 < 0.20).
Table 1. Site of origin of 14 Eddoe Taro accessions
Code
Local name
Collecting site (Subdistrict, District, Province)
Habitat*
(Altitude m
asl)
Origin
S6
Safira
Bontadaeng, Bantaeng, South Sulawesi
AF
750
LR
S7
Satoimo
Bontadaeng, Bantaeng, South Sulawesi
AF
750
IN
S15
Satoimo
Lembah Seulawah, Aceh Besar, Aceh
AF
150
IN
S17
Dempel
Trowulan, Mojokerto, East Java
HG
30
LR
S18
Talas Oshikawa
Kepung, Kediri, East Java
AF
60
IN
S20
Bentul
Batu, Malang, East Java
RG
445
LR
S24
Talas Oshikawa
Lamongan, Lamongan, East Java
AF
6
IN
S26
Salak
Karang Asem, Karang Asem, Bali
HG
88
LR
S28
Bentul
Pesantren, Kediri, East Java
HG
60
LR
S30
Keladi
Belitang, Belitang, South Sumatra
AF
40
LR
S33
Brentel
Sooko, Mojokerto, East Java
HG
30
LR
S34
Japanese taro
Lilirilau, Soppeng, South Sulawesi
AF
90
IN
S35
Japanese taro
Batulappa, Pinrang, South Sulawesi
AF
50
IN
S36
Ngariung Indung
Santana, Kuningan, West Java
AF
760
LR
Note: * AF-agriculture farmer field, HG-homegarden, RG-grow wild in the experimental field garden; ** LR-landrace, IN-Recently
introduced according to farmer information, but dating unknown

B I O D I V E R S I T A S
21 (8): 3525-3533, August 2020
3528
RESULTS AND DISCUSSION
Morphological and nutritional characters
Eddoe genotypes collected from Indonesia showed
variation in leaf, petiole, corm, and cormels (Figure 2).
Among 48 characters used in the genotyping, 38 characters
showed variation including qualitative and quantitative
characters. Twelve out of 15 quantitative characters
exhibited significant variation among genotypes (Table 2).
There was no significant difference for corm diameter,
number of cormlet and glucomannan content. Table 2
shows characters related to plant growth and production
like chlorophyll content, leaf size, plant size, number of the
sucker, corm size, and cormlet weight had variation among
genotypes.
Moreover, oxalate content in the corm and cormlets
significantly varied among genotypes (Table 2). Low
oxalate level is an important indicator for taro palatability
affected by genotype (Nurilmala and Mardiana 2019); the
level of the oxalate content could be reduced by cooking
(Hang et al. 2013). We confirmed here that genotype had a
different level of oxalate. This fact could be an important
consideration in crop improvement to address farmer's
problems in cormel quality.
From 33 qualitative characters observed, 28 characters
showed variation among genotypes. The variation was
found in leaf blade margin, leaf lamina appendages, leaf
blade margin color, leaf sheath color, and vein pattern.
Here, seven characters were presented; these characters
could be easily distinguished for genotyping including by
farmers (Table 3). Among leaf morphological characters,
sheath color and vein patterns were easily recognized.
Three accessions had red-purple sheath, i.e., S26, S28, and
S33 genotypes, while the others were light green. The
common vein pattern was V-pattern, followed by Y-pattern
extending to secondary vein and Y-pattern. Y-pattern was
found in single accession S20 from Malang District (Table
1). Petiole varied in whole color, junction, stripe, and
basal-ring color. Here, the whole petiole color was the most
prominent character compared to other petiole characters.
The whole petiole mostly was light green, but brown
existed for S26 and S30, and purple was found in S28 and
S33 genotypes (Table 3). Rudyatmi and Rahayu (2014)
reported that taro with purple petiole was known as black
taro.
Corm exhibited variation in shape, cortex color, flesh
color of the central part, flesh fiber color, skin thickness,
fibrousness degree of corm and bud color; with corm shape
and bud color were the most discernible characters. S28
was the only genotype with cylindrical corms (Table 3).
Corm with dumb-bell shape existed in two genotypes,
conical shape in four and round shape in seven genotypes.
Corm dumb-bell and round shape was also found in C.
esculenta characterized by Sinaga et al. (2017). Bud color
was yellow-green in S20, S28 and S33 genotypes, and
pink-red in the other 11 accessions. Table 3 shows that all
genotypes with round corm had pink-red corm bud. Corm
commonly had a big size and located at the central such as
in S6, S20, and S36 genotypes. However, for some
genotypes like S18, S34 and S35 had the corm size almost
the same size as its cormlets (Figure 2C).
Table 2. ANOVA for quantitative characters of 14 Eddoe Taro genotypes from Indonesia
Characters
MeanSD
SS
MS
F Value
R-Square
CV
Sig
Leaf lamina
Length (cm)
36.18.6
1607.76
123.67
5.81
0.86
12.79
**
Width (cm)
12.72.8
158.99
12.23
4.89
0.84
12.49
**
Sheath length (cm)
22.15.5
688.66
52.97
6.81
0.88
12.64
**
Total petiole length (cm)
62.123.1
12285.81
945.06
8.24
0.90
17.24
**
Plant width (cm)
93.723.7
10208.65
785.28
3.25
0.79
16.58
*
Plant height (cm)
81.827.8
17775.32
1367.33
8.28
0.90
15.71
**
Number of sucker
5.73.4
214.69
16.51
3.34
0.79
39.27
*
Corm
Length (cm)
61.013.5
4232.96
325.61
7.08
0.88
11.12
**
Diameter (cm)
56.57.2
885.72
68.13
1.87
0.66
10.69
ns
Weight (g)
95.536.9
1375410.69
105800.82
5.97
0.81
24.04
**
Cormlet
Number
24.310.1
1705.90
131.22
1.91
0.67
34.17
ns
Weight (g)
425.847.3
1375410.69
105800.82
5.97
0.87
31.27
**
Chlorophyll (mg/g)
57.65.2
520.89
40.07
2.58
0.72
6.84
*
Oxalate (ppm)
94.832.8
15738.99
1210.69
3.47
0.85
19.69
*
Glucomannan (%)
5.71.5
39.40
3.03
2.23
0.69
20.43
ns
Note: SD-Standard Deviation; SS-Sum of Squares; MS-Mean Square; CV-coefficient of variant; ** : significant (p<0.01); * :
significant (p<0.05), ns : not significant

MARETTA et al. – Genetic diversity of Indonesian eddoe taro
3529
Figure 2. The appearance of leaf lamina (A), petiol (B), corm (C), and cormlets (D) of 14 Eddoe Taro genotypes from Indonesia
Table 3. Selected morphological data of 14 Eddoe Taro genotypes from Indonesia
Characters
Genotype code
S6
S7
S15
S17
S 18
S20
S24
S 26
S28
S30
S 33
S34
S35
S 36
Sheath color
1
1
1
1
1
1
1
2
2
1
2
1
1
1
Vein pattern at leaf base
3
3
3
3
3
1
3
2
2
2
2
3
3
3
Whole petiole color
3
3
3
3
3
3
3
1
2
1
2
3
3
3
Corm shape
4
4
4
4
4
3
4
3
1
3
3
2
4
2
Bud color
2
2
2
2
2
1
2
2
1
2
1
2
2
2
Cormlet shape
4
4
4
4
4
3
4
2
3
1
3
4
4
4
Root color
2
2
2
2
2
1
2
2
1
2
1
2
2
2
Note: Sheath color (1-light green, 2-red purplish), Vein pattern at leaf base (1-Y pattern, 2- Y-pattern extending to the secondary vein, 3-
V pattern), Whole petiole color (1-brown, 2-purple, 3-light green), Corm shape (1-cylindrical, 2-dumb-bell, 3-conical, 4-round), Bud
color (1-yellow green, 2-pink-red), Cormlet shape (1-conical, 2-elongated, 3-elongated-curved, 4-elliptical), Root color (1-pinkish
white, 2-white)
Genetic properties
GVC and PVC for important quantitative characters
significantly varied from low to high (Table 4). Six
characters had high criteria GVC, i.e, sheath length, total
petiole length, plant height, number of sucker, corm and
cormlets weight. Chlorophyll content contributed to low
GVC, and it was the lowest among characters. On the other
side, all PVC was high, except for the character of
chlorophyll content. In the present study, all PVC values
were higher than the GVC. This finding is a common case
in the genotypic evaluation such as in other crops (Effendy
et al. 2018).
In the breeding process, genotype selection based on
characters with high heritability and high genetic variation
is desirable. Heritability value would determine the
selection method of plant characters because it gives a
portion of genetic and phenotypic variation that is inherited
(Sleper and Poehlman 2006). According to Jalata et al.
(2011), broad genetic diversity would speed the success of
the selection and breeding progress. Effendy et al. (2018)
pointed out that high genetic variation of particular
characters within a population reflected the span of genetic
control and the high genetic control means expected
product of the breeding process could be estimated more
precisely. In the present experiment, the estimate of broad-
sense heritability and PVC of all characters showed high,
except the chlorophyll content that showed medium and
low. High GVC found in characters sheath length, total
petiole length, plant height, number of suckers, weight of
corm, cormlets weight and oxalate content (Table 4). The
high genetic diversity would enlarge chances of success
selection in plant breeding due to the higher frequency of

B I O D I V E R S I T A S
21 (8): 3525-3533, August 2020
3530
genes (Hapsari 2014). Confirmable to Eze and Nwofia
(2016) that plant height, number of suckers, length, and
weight of corm and cormlet weight has high heritability.
This research result reveals additional information that
oxalate content also has a high value of heritability, PVC
and GVC. It indicates that the quantitative characters as
listed in Table 4 could be used in the selection program.
On the contrary, Mulualem and Michael (2013)
reported that the quantitative characters of Colocasia
esculenta had low or medium PVC, GVC, and heritability.
Low heritability value indicated a character inherited
complicatedly and influenced by environment factors
(Meydina et al. 2015). Refers to Sleper and Poehlman
(2006) the effectiveness of selection depends on the
variability of the individual in the population and the
variability among plants due to the environment. For that
reason, the selection of these all Eddoe Taro genotypes
highly recommended to be examined in multi-location or
different season conditions to find out which genotypes
perform consistently in a wide range of environments.
Genotype grouping
Genotypes separated into two groups with seven-
member in each group, Group-1 (S18, S20, S26, S28, S30,
S33, and S36) and Group-2 (S6, S7, S15, S17, S24, S34,
and S35) (Figure 3). Group-2 represented the 'satoimo
group' judge from the local name, except Safira (S6) and
Dempel (S17). Cladogram showed S6 separated to other
members including S7. Unexpectedly, S6 and S7 genotypes
that geographically close (Table 1), were separated
distantly in the cluster Group-2.
According to farmers that cultivated the S7 accession,
they received the seed from the local government of South
Sulawesi province. The local government imported seeds
from Japan through a trading company. The seeds were
propagated by the trading company and then distributed to
some Districts within South Sulawesi, including other
provinces such as Aceh and East Java, so farmers
maintained the accessions name as 'satoimo' of Japanese.
Therefore, it is presumable that Satoimo group shared
similar ancestor based on morphological characters; and
the ancestor was likely derived from S7 in Bantaeng
District, South Sulawesi. In Aceh, Eddoe Taro was
reintroduced around 2014 almost the same time
reintroduction to Buleleng District in Bali (Taufiq 2015);
thus it supported that Aceh accession S15 clustered in
Satoimo group.
High genetic variation within satoimo group as shown
in Figure 3, indicated that clonal variation could be high in
eddoe type. There probably was genetic distant between S6
and S7 although geographically close due to clonal
variation. On the other side, Group-1 represented the local
landrace. It is still unclear, S18 that had a similar name
with S24 'Talas Oshikawa' clustered to a different group. In
general, morphological variation within Group-1 was larger
than the variation within Group-2 (Figure 3). It needs
further evaluation using molecular markers the reliability in
this grouping.
Genotype grouping using a cladogram was consistent
with grouping based on population assessment using
STRUCTURE (Figure 4). The population grouped into two
types, Type-1 and Type-2. The Type-1 was indicated by
light grey color consisted of eight genotypes (S6, S7, S15,
S17, S18, S24, S34, S35, S36) and shows high uniformity.
The type-1 population included all accessions in the Group-
2 cladogram (Figure 3), and the clustering was exactly
matched with a local name especially S18 that unable to be
resolved using the cladogram. Here, almost all accessions
under Type-1 were taken from the commercial field at
which based on farmer interview the parent seeds were
imported except for S6, S17, and S36 that obtained from
local farmers. Considering these genotypes shared the same
ancestor, this finding envisages the hypothesis that clonal
variation could be high in Eddoe Taro.
Type-2 had five members, i.e., S20, S26, S28, S30, and
S33 indicated by a mixture of black and dark grey colors
(Figure 4). Type-2 accessions corresponding to Group-1 in
the cladogram. This type, most probably as original
landraces in Indonesia.
Table 4. Genotypic and phenotypic variances and it’s coefficient for 12 quantitative characters of Eddoe Taro in Indonesia
Character
2g
2p
h2
GVC (%)
PVC (%)
Leaf lamina
Length
51.20
72.48
0.71
***
19.84
**
23.60
***
Width
4.86
7.37
0.66
***
17.42
**
21.44
***
Sheath length
22.59
30.38
0.74
***
21.54
***
24.97
***
Total petiole length
415.18
529.88
0.78
***
32.80
***
37.06
***
Plant size
Plant span
271.96
513.32
0.53
***
17.60
**
24.18
***
Plant height
601.09
766.24
0.78
***
29.96
***
33.83
***
Number of suckers
5.79
10.73
0.54
***
42.50
***
57.87
***
Corm
Length
139.82
185.80
0.75
***
19.40
**
22.36
***
Weight
878.63
1,406.01
0.62
***
31.03
***
39.25
***
Cormlets weight
44,036.55
61,764.27
0.71
***
49.28
***
58.36
***
Chlorophyll content
12.26
27.80
0.44
**
6.08
*
9.15
*
Oxalate content
431.12
779.57
0.55
***
21.90
**
29.45
***
Note: GVC-Genotypic variance coefficient; PVC-phenotypic variance coefficient; ***high, **medium, *low

MARETTA et al. – Genetic diversity of Indonesian eddoe taro
3531
Figure 3. Cladogram of 14 Eddoe Taro genotypes collected from Indonesia. Bar indicates dissimilarity distant. The genotype name is
presented in Table 1
Figure 4. The genetic population structure of Eddoe Taro in Indonesia is drawn by STRUCTURE
Figure 4 shows that genetically, S36 seemed as an
intermediary type between Type-1 and Type-2 clusters.
Since S36 contained many share characters from
population Type-1 lead to join the Type-1 in the present
analysis. Considering that Type-1 just recently introduced,
it is difficult to conclude that S36 was a result of breeding
Type-1 and Type-2 genotypes. Moreover, based on genetic
composition it is likely that Type-2 genotypes were clonal
variant from S36 in Kuningan District.
Source of genetic variation
We speculated that morphological and nutritional
variation in Eddoe Taro from Indonesia arose from a
combination of multiple introduction and clonal variation.
The first introduction could be around the 1940s or before.
This assumption based on information from Prana (2007)
where the eddoe type is only found at an isolated site like
'Talas bithek' in Tana Toraja District and 'Talas salak' in
Buleleng District, Bali, in which these sites had intense
interaction with Japanese in the past time. Fortunately, we
incorporated 'Talas salak' from Karangasem District, Bali
(code S26) that geographically less than 100 km; it was
involved in Group-1 and Type-2. Here, Group-1 or Type-2
was represented as the first introduction. According to
information from senior farmers (> 80-year old) in Sooko
village, Mojokerto, he claimed that the seeds of 'Brentel'
(S33) were introduced by the Japanese army in around
1940s to support their logistic in East Java. His information
was confirmed in the cladogram dan dendrogram (Figures
3 and 4). It is confident to note that S20, S26, S28, S30,
and S33 arisen from the first introduction.
The second introduction was represented by genotypes
belongs to Group-2 cladogram or Type-1 dendrogram such
as S6, S7, S15, S17, S24, S34 and S35 (Figures 3 and 4).
The introduction could be around the 2000s through South
Sulawesi after the establishment of the 'Satoimo
Consortium Project in 2004' involving Japanese and
Indonesian companies (Kallo et al. 2019). According to
Das et al. (2015), Eddoe Taro has diversity in chromosome
numbers. Thus, it is probable that the Japanese introduced
different cultivars for the first and second periods.
Presence of genetic variation within-cluster group in
Figure 3 probably due to clonal variation after generations.
According to Vandenbroucke et al. (2016), clonal variation
in taro was about 3%. After the first phase of the
introduction, the development of Eddoe Taro could be

B I O D I V E R S I T A S
21 (8): 3525-3533, August 2020
3532
restricted by unknown reasons. The success of the Green
Revolution in Indonesia leading to the high availability of
rice (Poerwanto et al. 2012, Yunus et al. 2016), could be
one explanation Eddoe Taro became less utilized. During
the survey, many farmers in East Java stated that Eddoe
Taro was abandoned ones, even sometimes called a weed.
The eddoe underwent dormant during the dry season; the
dormancy might limit continuous availability, unlike
dasheen type that available throughout the year. The case
nearly similar to clonally propagated Amorphophallus
paeoniifolius as a neglected crop in Java after the green
revolution (Santosa et al. 2017).
As a result, eddoe genotypes such as S36 were
cultivated in a limited area at highland in Kuningan
District, West Java along the edge of a vegetable field close
to Ceremai Mt. Locally, S36 called Ngariung indung or
lahun indung means 'mother carrying the son'. Sometimes
the boiling cormlets were available in the local market.
According to Kuningan people, particular cormlets have
been known available in the market since the 1970s.
S36 was an exceptional genotype. According to the
clade diagram, it is grouped with genotypes into the first
phase of introduction (Figure 3) but genetic composition
indicated a high proportion of the second phase of
introduction (Figure 4). Judgment from the timeline, the
S36 was most likely as descendent from the first phase of
introduction but underwent genetic manipulation. The
genetic manipulation is less likely from natural mating
because flowering on eddoe type in the field was a rare
case. According to Susepah (2018), Sundanese in
Kuningan is a famous retailer that travels across Indonesian
cities. The Kuningan genotype probably originated from
clonal variation, like in clonally propagated A.
paeoniifolius (Santosa et al. 2010). Nevertheless, the
hypothesis needs further evaluation because in the present
experiment the original variety of satoimo from Japan was
not incorporated.
To develop better cormlets quality and production, it is
important to consider this genetic diversity and historical
data. After years, first introduced-genotype probably has
adapted with Indonesian agro climate, while second
introduced-genotype has superiority in the palatability and
global acceptance. Therefore, it is recommended to use
genotype S6, S7, S18, S30 and S36 for further breeding
purposes. In conclusion, genetic diversity based on
morphological characters was considered high in eddo taro
in Indonesia as representing by major characters such as
sheath and petiole color, vein pattern at the leaf base, leaf
size, plant size, cormels size and oxalate content. On the
other hand, the cormlets number and glucomannan content
were statistically similar. The cause of such high diversity
is presumably caused by multiple introductions from Japan
and variation in clonal propagation. However, it needs
further clarification using a robust genotyping method such
as a molecular marker. Based on distinct grouping, five
genotypes, i.e., S6, S7, S18, S30, and S36 could be
considered as parents in the future breeding program and a
multi-location examination is recommended to determine
the consistency perform of genotypes.
ACKNOWLEDGEMENTS
The authors' thanks to The Ministry of Research and
Technology, the Republic of Indonesia for financial
support through Saintek Scholarship and The Agency of
Application and Assessment of Technology Indonesia
(BPPT) for providing research facilities.
REFERENCES
Akalu ZK, Geleta SH. 2017. Antinutritional levels of tubers of Colocasia
esculenta L. Schott (Taro) and Dioscorea alata (Yam) cultivated in
Ethiopia. J Nutr Food Sci 7: 2. DOI: 10.4172/2155-9600.1000585.
Amelia DD, Yumiati Y. 2016. Analysis of satoimo taro farming
(Colocasia esculenta var. antiqourum) (Case study in Sukasari
Village, Kabawetan subdistrict, Kepahiang District. J Agritepa 2 (2):
188-198. [Indonesian]
Astuti SD, Andarwulan N, Fardiaz D, Purnomo EH. 2017. The
characterization of satoimo taro flour produced by controlled
fermentation using L. plantarum and S. cerevisiae. Proc. Semin. Nas.
Pangan, Gizi Kesehatan 796-809. [Indonesian]
Banjaw DT. 2017. Review of taro (Colocasia esculenta) genetics and
breeding. J Hortic 4: 1. DOI: 10.4172/2376-0354.1000196
Bateni E, Tester R, Al-Ghazzewi F, Bateni S, Alvani K, Piggott J. 2013.
The use of konjac glucomannan hydrolysates (GMH) to improve the
health of the skin and reduce acne vulgaris. Am J Dermatol Venereol.
2 (2): 10-14. DOI: 10.5923/j.ajdv.20130202.02.
Behera SS, Ray RC. 2016. Konjac glucomannan, a promising
polysaccharide of Amorphophallus konjac K. Koch in health care. Int
J Biol Macromol 92: 942-956. DOI: 10.1016/j.ijbiomac.2016.07.098
Chairul C, Chairul S. 2006. Isolation of glucomannan from two species of
Araceae: Taro (Colocasia esculenta (L .) Schott) and elephant foot
yam (Amorphophallus campanulatus Blumei). Berita Biol 8 (3): 171-
178. [Indonesian]
Chaïr H, Traore RE, Duval MF, Rivallan R, Mukherjee A, Aboagye LM,
Rensburg WJ, Andrianavalona V, Carvalho P, Saborio F, Prana MS,
Komolong B, Lawac F, Lebot V. 2016. Genetic diversification and
dispersal of taro (Colocasia esculenta (L.) Schott). PLoS ONE. DOI:
10.1371/journal.pone.0157712
Das AB, Das A, Pradhan C, Naskar SK. 2015. Genotypic variations of ten
Indian cultivars of Colocasia esculenta var. antiquorum Schott.
evident by chromosomal and RAPD markers. Caryologia 68(1): 44-
54. DOI: 10.1080/00087114.2015.1013335
Deshmukh SN, Basu MS, Reddy PS. 1986. Genetic variability, character
association and path coefficient analysis of quantitative traits in
Virginia bunch varieties of groundnut. Indian J Agro Sci 56: 515-518.
Dewi SK, Dwiloka B, Setiani BE. 2017. Reduction of oxalate levels in
taro tubers by adding activated charcoal to the steaming method. J.
Apl. Teknol. Pangan 6 (2): 2. [Indonesia]
Earl, Dent A, vonHoldt, M. Bridgett. 2012. Structure Harvester: A website
and program for visualizing structure output and implementing
Evanno method. Conserv. Genet. Resour 4 (2): 359-361. DOI:
10.1007/s12686-011-9548-7
Effendy, Respatijarti, Waluyo B. 2018. Genetic variability and heritability
characters of yield component and yield of physalis (Physalis sp.). J
Agro 5 (1): 30-38. DOI: 10.15575/1864
Ekowati G, Yanuwiadi B, Azrianingsih R. 2015. A source of
glucomannan from edible Araceae in East Java. J-PAL 6 (1): 32-41.
[Indonesian]
Ezeabara CA, Okeke C, Amadi J. 2015. Phytochemical, proximate,
mineral and vitamin investigations of cormels of five varieties of
Colocasia esculenta (L.) Schott found in Anambra State, Southeastern
Nigeria. Am J Life Sci Resour 3 (4): 273-281.
Eze CE, Nwofia GE. 2016. Variability and inter-relationships between
yield and associated traits in taro (Colocasia esculenta (L.) Schott). J
Exp Agric Intl 14 (2): 1-13. DOI: 10.9734/JEAI/2016/27053.
Hang DT, Vanhanen L, Savage G. 2013. Effect of simple processing
methods on oxalate content of taro petioles and leaves grown in
central Viet Nam. Food Sci Technol 50: 259-263.
Hapsari RT. 2014. Estimation of genetic diversity and correlations
between yield components of early maturing green beans. Bull
Plasma Nutfah 20 (2): 51-58. [Indonesian]

MARETTA et al. – Genetic diversity of Indonesian eddoe taro
3533
IPGRI. 1999. Descriptor for Taro (Colocasia esculenta). International
Plant Genetic Resources Institute, Rome
ITPC. 2014. Market Brief Satoimo (Taro). ITPC, Osaka.
Jalata Z, Ayana A, Zeleke H. 2011. Variability, heritability and genetic
advance for some yield and yield-related traits in Ethiopian barley
(Hordeum vulgare L.) landraces and crosses. Intl J Plant Breeding
Genet 5 (1): 44-52. DOI: 10.3923/ijpbg.2011.44.52
Kallo R, Satna A, Nappu MB. 2019. Prospects for the development of
Japanese satoimo taro in South Sulawesi. Bull Diseminasi Teknol
Pertanian 1 (1): 1-5. [Indonesian]
Kaushal P, Kumar V, Sharma HK. 2012. Comparative study of
physicochemical, functional, antinutritional and pasting properties of
taro (Colocasia esculenta), rice (Oryza sativa) flour, pigeon pea
(Cajanus cajan) flour and their blends. Food Sci Technol 48: 59-68
Laosa A, Darman S, Alam MN. 2016. Analysis of production and income
of Japanese taro farming system in Tinangkung Village South
Tinangkung Sub District of Banggai Kepulauan District. J Agrol 23
(3): 174-181. [Indonesian]
Mergedus A, Kristl J, Ivancic A, Sober A, Sustar V, Krizan T, Lebot V.
2015. Variation of mineral composition in different parts of taro
(Colocasia esculenta) corms. Food Chem 170: 37-46. DOI:
10.1016/j.foodchem.2014.08.025.
Meydina A, Barmawi M, Sa’diyah N. 2015. Genetic variability and
heritability of agronomy characters of soybean (Glycine max [L.]
Merrill) F5 generation as the results of crosses WILIS X B3570. J
Penelitian Pertanian Terapan 15 (3): 200-207
Mulualem T, Michael G. 2013. Study on genotypic variability estimates
and interrelationship of agronomic traits for selection of taro
(Colocasia esculenta (L.) Schott) in Ethiopia. Sky J Agric Resour 2
(11): 154-158.
Naik VV, Patil N, Aparadh VT, Karadge BA. 2014. Methodology in
determination of oxalic acid in plant tissue: A comparative approach.
J Global Trends Pharm Sci 5 (2): 1662-1672.
Njintang NY, Boudjeko T, Tatsadjieu LN, Nguema-ona E, Scher J,
Mbofung CMF. 2011. Compositional, spectroscopic and rheological
analyses of mucilage isolated from taro (Colocasia esculenta L.
Schott) corms. J Food Sci Technol. DOI: 10.1007/s13197-011-0580-
0.
Nurilmala F, Mardiana D. 2019. Nutrients and anti-nutrients content
analysis of Bogor taro mutant clone (Colocasia esculenta). IOP Conf.
Series: Earth Environ Sci 334: 012070. DOI: 10.1088/1755-
1315/334/1/012070
Pitoyo A, Prameta AA, Marsusi, Suratman, Suranto. 2018. Morphological,
anatomical and isozyme variability among taro (Colocasia esculenta)
accessions from southeastern part of Central Java, Indonesia.
Biodiversitas 19 (5): 1811-1819. DOI: 10.13057/biodiv/d190531.
Plucknett DL. 1983. Taxonomy of the genus colocasia. In: Wang JK (eds)
A Review of Colocasia esculenta and its potentials. Univ Hawaii,
Honolulu, USA.
Prana MS. 2007. Study on flowering biology of taro (Colocasia esculenta
(L.) Schott.). Biodiversitas 8 (1): 63-66.
Poerwanto R, Sulaeman A, Wattimena GA. 2012. The history and
development of the green revolution, the biotechnology revolution
and the sustainable green revolution. In: Poerwanto R, Siregar IZ,
Suryani A (eds) Merevolusi Revolusi Hijau. IPB Press, Bogor.
[Indonesian]
Rashmi DR, Raghu N, Gopenath TS, Palanisamy P, Bakthavatchalam P,
Karthikeyan M, Gnanasekaran A, Ranjith MS, Chandrashekrappa
GK, Basalingappa KM. 2018. Taro (Colocasia esculenta): An
overview. J Med Plants Stud 6 (4): 156-161.
Rosdanelly CNR, Agussabti, Azhar. 2018. The prospect of Japanese taro
cultivation (Colocasia esculenta var antiquorum) at Agricultural
Training Centre Saree, Aceh Besar. J Pertanian Unsyiah 3 (1): 380-
390. http://www.jim.unsyiah.ac.id/JFP. [Indonesian]
Rudyatmi E, Rahayu ES. 2014. Characterization of Middle Java local taro
(Germplasm identification as part of conservation effort). Sainteknol
12 (1): 1-8 [Indonesian]
Santosa E, Mine Y, Nakata M, Lian C, Sugiyama N. 2010. Genetic
diversity of cultivated elephant foot yam (Amorphophallus
paeoniifolius) in Kuningan, West Java as revealed by microsatellite
markers. J Appl Hortic 12 (2): 125-128.
Santosa E, Lian CL, Sugiyama N, Misra RS, Boonkorkaew P, Thanomchit
K. 2017. Population structure of elephant foot yams (Amorphophallus
paeoniifolius (Dennst.) Nicolson) in Asia. PLoS ONE 12: 6. DOI:
10.1371/journal.pone.0180000.
Shah RB, Li B, Wang L, Liu S, Li Y, Wei X, Weiping J, Zhenshun L.
2015. Health benefits of konjac glucomannan with special focus on
diabetes. Bioact Carbohyd Diet Fibre 5 (2): 179-187. DOI:
10.1016/j.bcdf.2015.03.007.
Sinaga KA, Murningsih, Jumari. 2017. Identification of edible Araceae in
Semarang, Middle Java. Bioma 19 (1): 18-21. [Indonesian]
Sleper DA, Poehlman JM (eds.). 2006. Breeding Field Crop. 5th ed.
Blackwell Publishing, Oxford.
Stansfield WD. 1983. Theory and Problems of Genetic. 2nd ed. McGraw-
Hill, New York.
Susepah I. 2018. Profile and performance of mindring enterprise in the
informal sector, especially the Kuningan migrants in Padukuhan
Pandean VII Sidoluhur, Godean, Sleman, Yogyakarta. Jurnal
Pemberdayaan Masyarakat 2 (1): 87-108. DOI:
10.14421/jpm.2018.021-05. [Indonesian]
Syukur M, Sujiprihati S, Yunianti R. 2015. Technic of Plant Breeding. 2nd
ed. Penebar Swadaya, Jakarta. [Indonesian]
Taufiq FM. 2015. Japanese Taro, Export Value Food Commodity.
https://www.kompasiana.com/masfathan66. [Indonesian]
Temesgen M, Retta N. 2015. Nutritional potential, health and food
security benefits of taro Colocasia esculenta (L.): A Review. Food
Sci Qual Manag 36: 23-30.
Tester RF, Al-Ghazzewi FH. 2016. Beneficial health characteristics of
native and hydrolysed konjac (Amorphophallus konjac) glucomannan.
J Sci Food Agric 96: 3283-3291. DOI: 10.1002/jsfa.7571.
Vandenbroucke H, Mournet P, Vignes H, Chaïr H, Malapa R, Duval MF,
Lebot V. 2016. Somaclonal variants of taro (Colocasia esculenta
Schott) and yam (Dioscorea alata L.) are incorporated into farmer's
varietal portfolios in Vanuatu. Genet Resour Crop Evol 63: 495-511.
Widjanarko SB, Megawati J. 2015. Comparation between colorimetric
and gravimetric methods of glucomannan analysis to konjac. J Pangan
Agroindustri 3 (4): 1584-1588. [Indonesian]
Yang D, Yuan Y, Wang L, Wang X, Mu R, Pang J, Xiao J, Zheng Y.
2017. A review on konjac glucomannan gels: Microstructure and
application. Intl J Mol Sci 18 (11): 2250. DOI:
10.3390/ijms18112250.
Yu JG, Liu P, Duan JA, Tang ZX, Yang Y. 2015. Itches-stimulating
compounds from Colocasia esculenta (taro): Bioactive-guided
screening and LC-MS/MS identification. Bioorg Med Chem Lett 25:
4382-4386
Yunus A, Salman D, Demmallino EB, Viantika NM. 2016. Sociotechnical
change and institutional adjustment in paddy rice farming during post
green revolution in Indonesia. Intl J Agric Syst 4 (2): 218-227.
Zalewski BM, Chmielewska A, Szajewska H. 2015. The effect of
glucomannan on body weight in overweight or obese children and
adults: A systematic review of randomized controlled trials. Nutr 31:
437-444.