cOO

Sebil r
cssc
- J

P r o c e e d in g s off the 16th

A n n u a l C o n fe re n c e off th e
C ro p S cien ce Society
o f E th io p ia n

N o v e m b e r , 2017
A d d is A b a b a , E th io p ia
Crop Science Society of Ethiopia (CSSE)

The Crop Science Society of Ethiopia (CSSE) is a non-profit making professional association
established in 1987. It was inaugurated as a society in December 1991 with the objective of
contributing towards the development of Ethiopian agriculture in general and the solving of crop
production-related problems in particular through promoting effective research, documenting
and disseminating scientific information, encouraging professional growth, and fostering inter­
disciplinary interactions and dialogues among crop scientists, policy makers and other
governmental and non-governmental partners and stakeholders involved in the sector.

So far, the society has organized sixteen conferences and is pleased to produce the Sixteen
SEBIL issue of the proceedings

Members of the Executive Committee of the Crop Science Society of Ethiopia (CSSE)

Kindie Tesfaye President, International Maize and Wheat Improvement

Centre/CIMMYT, Addis Ababa, Ethiopia
Edossa Etissa Secretary, Ethiopian Institute of Agricultural Research, Melkasa
Research Center
Belaynesh Hailu Treasurer, Ethiopian institute of Agricultural Research, Holetta
Research Center
Girma Mamo Accountant, Ethiopian Institute of Agricultural Research, Melkasa
Research Center
Fetein Abay Internal Audit, Mekele University, Mekele
Tesfaye Balemi Editor-in-chief, International Maize and Wheat Improvement
Center/CIMMYT, Addis Ababa, Ethiopia
Dagnachew Lule Associate Editor, Oromia Agricultural Research Institute
Shimelis Admasu Public Relations, Addis Ababa University

SEBIL is published biannually by the executive committee of CSSE. For subscription,

contributions and information regarding the proceeding and membership please write to the:
Secretary, Crop Science Society of Ethiopia, P.O.Box 2003
Addis Ababa
Ethiopia

The opinions expressed in this volume are those of the authors and do not necessarily reflect
the views of the society. Trade names mentioned in the bulletin are not intended to endorse
those products.

Correct citations: Crop Science Society of Ethiopia (CSSE). 2016. Sebil Vol. 16. Proceedings of
the Biennial Conference, 1-2 October 2015, Addis Ababa, Ethiopia

Published November 2017

Sebil
Proceedings o f the 16th
Annual Conference o f the
Crop Science Society
o f Ethiopia

November, 2 0 i r
Addis Ababa, Ethiopia

Edited by
Tesfaye Balemi (PhD)
Dagnachew Lule (PhD)

Crop Science Society of Ethiopia (CSSE)

P.O.Box 2003, Addis Ababa, Ethiopia
 Table of Contents

Page

Crop Improvement in the Ever-Advancing Adverse

Agricultural Environment: The Anxiety of the Shock
Asnake Fikre 1

AMMI and G G E Bi-plot Analysis ofSesam e

(SesatnitHt ittdietttn L .) Genotypes in \ortliern Ethiopia
Fiseha Baraki, Yemane Tsehaye and Fe/ien Abay 13

Genetic Diversity and T raits Inheritance in Anehote

[Coccinia abysxinica (L am .) CognJ Accessions of Ethiopia
Desta Fekadu, Derbew Belew, Amsalu Ayana 29

Perform ance of Cassava Clones Under Potential and

liow M oisture Stressed Areas of Ethiopia
Tesfaye Tadesse, Atnafua Bekele, Engida Tsegaye, Geta chew W/Michael,
Tewodros Mulualem, Wubshet Beshir, Mesele G, Shiferaw M 53

Combining Ability Studies for Yield and Yield Components

in Selected Soybean Lines
Tadesse Ghiday, Sentayehu Alamirew Asnake Fikre, Tizazu D., Molla M.,
Asmamaw A., Gezahegn T. 69

Genetic Divergence Analysis on Some Soybean

{G lycin e max L. Merrill) Genotypes Grown in P au e, Ethiopia
Tadesse Ghiday and Sentayehu Alamirew 77

Overlap in the Genetic Basis of Host Basal and Xonhost

Resistances of Barley to L eaf Kusts
AI/o Aman Dido 89

Yield Perform ance and Stability Analysis ofSesam e

(Sesatniti/t inftictnn L .) Genotypes in Northern Ethiopia
Fiseha Baraki, Yemane Tsehay’e and Fetien Abay 99

Vegetable Oil Shortage in Ethiopia:

Its Causes and Solutions
Getinet Alemaw, Mistru Tesfaye, Abush Tesfaye, Wogayehu Worku and Adugna Wakjira 115
 Sebil Vol. 16

Welcome Address
A lem ayehu Assefa
President-CSSE

Distinguished tnembers o f the CSSE,

Ladies and Gentlemen

Established in 1987 and inaugurated in 1991, Crop Science Society of

Ethiopia (CSSE) through its 24 years of journey has continued to bring
professionals in the field o f crop sciences where scientists and
development partners demonstrate their scientific excellence and scientific
thoughts that would eventually contributes towards the development of
agricultural sciences in a way to bring science based solutions to
development challenges. Promoting effective research, disseminating
scientific inform ation, en cou ragin g professional growth, and fostering
interdisciplinary interaction among crop scientists being its main
objectives, I believe, shall also be the headings in its future course but with
interim assessment of how it is contributing to the development of
agricultural sciences and meeting its set objectives. No doubt, however,
that CSSE as a pioneer professional society has also emerged as an
umbrella to all disciplines in the field of agricultural sciences and, through
its biennial conference, served as a forum where interdisciplinary
professional dialogues continued to enrich scientific knowledge.

Big professional societies like Crop Science Society of America (CSSA)

foster plant science for a better world. It is a home of 1000s of professionals
dedicated to advance science and technology for the betterment of human
life.

Let me quote few statements made by one of the 2014 Honorary Fellow
awardees of the Crop Science Society of Philippines (CSSP). He said "Let
us equip ourselves with mindsets and attitudes for facing life. Let us
harness strengths to complain less and act more, to blame other less and
take responsibility more to dwell less on misfortunes and more on
blessings, to learn from failure and learn not to be defeated by it".

In this connection, our engagement as executive committee was not just

walking on a green pasture but we did without complaining. Rather it was
a matter of thinking that it is just committing ourselves and share part of

[i]
 Sebil Vol. 16

our time in a way that we can accommodate committee responsibilities

together with our regular duties. People in various parts of the world are
committing their life be it for good or bad things. Hence, if we believe that
this professional society belongs to us, then there is no reason that we fail
to spare at least our time to effect major committee responsibilities. The
executive committee took over pending issues from the ex-committee and
did its maximum effort with full commitment to finalize. According to the
yearly plan of the society, regular meetings have been conducted; a
number of important issues were raised as agenda and those issues that
require decisions were resolved and those that need improvement were
stated with common understandings among the executive committee.
Some of the major tasks accomplished were:

1) Renewal of the certificate to insure its legality and submitting the

necessary documents to the Charities and Societies Agency was a
bit challenging but we did it and our society will continue to be
recognized as one of the professional societies.
2) Important documents were assembled and took over from the ex­
committee
3) Additional financial supporting documents were
printed/developed based on the feedback we observed in the
previous reports to add value and improve the existing working
documents
4) Annual report was submitted to the agency as one of the
requirements for societies
5) Independent auditing was finalized and submitted to the agency
6) Effort was made to solicit fund and secured some but time was the
major constraint

In addition to the regular meetings there were also urgent meeting held
that enabled to decide on how to handle documents of the society. The
society is in good shape and we need to maintain this.

This 16th biennial conference of the CSSE embraced a number of scientific

papers that will add value to scientific knowledge and agricultural
development at large provided that they are translated in to useable form
and be made integral components of the extension packages. I hope the
conference papers, including the invited papers will be additions to the
existing knowledge. Let all put efforts to make CSSE a forum that will
contribute better than before. It is my wish also that the biennial
[ii]
 Sebil Vol. 16

conferences of the CSSE shall be scientific forums where strategic issues be

identified and dialogue be made and the outcome of which shall be
extended to beneficiaries instead of having always the conventional way of
scientific paper presentations.
In view of this, the executive committee agreed to have four invited papers
1) "V egetable o il S hortage in E th iop ia: cau ses an d solu tion s" that will
be presented by our senior researchers led by Dr Getinet Alemaw
2) M aize L e th a l N crosis D isease (MLND) in E th io p ia : A new m erging
Syndrom T hreatening M aize P rodu ction in E ast A frica by Birhanu
Bekele and his colleagues
3) The research sy stem in C rops w ith U n predictable C hallenges:
E xploring R esearch Issus A h ead (by Dr Asnake -Crops Research
Director)
4) C om m od ity Value C hains to drive in n ov ation p rocesses and
s u sta in a b ility in produ ction , u tiliz ation an d trad e (By Chris from
ICRISAT)
Other scientific papers selected for this conference are also more or less
related to the conference them and sub-themes.

Finally, on behalf of the executive committee, I would like to thank

members of th e c o m m itte e for showing th e c o m m itm e n t to make this
conference happening. Moreover, special thanks go to our major sponsors
(EIAR, FAO, and ICRISAT) for their financial support that helped us a lot
to cover some of the resource demanding tasks (publication of journals
and proceedings, conference packages, and other items). I would like to
thank also EIAR public relations directorate for facilitating the conference
hall and get media coverage.
Please allow me also to introduce you CSSE committee members:
W/o Belaynesh Halilu
Ato Mistru Tesfaye
Ato Wondimu Fekadu
Ato Dagnachew
Dr Dagne Wagari
Dr Solomon Chanyalew
Ato Birhanu Bekele
With this, I would like to say you all welcome to the 16th biennial
conference of CSSE and wish you a pleasant stay. S tay w ith us until the
end o f the conference, since CSSE belon gs to y ou 11
Thank you!!

[iiij
 Sebil Vol. 16

Opening Speech
F a n t a h u n M e ngistu
Ethiopian Institute o f Agricultural Research
Director General

Dear Guests
Conference Participants
Ladies and Gentle Men,

I feel indeed honoured and privileged to have been invited to deliver an

opening speech to members of this august society. As you all know,
Ethiopia after registering a double digit economic growth over the last
couple of years is at the turning point of winning the war against the
ravages of hunger and poverty. Agriculture plays the leading role in this
fight against hunger and poverty. However, the productivity of our
agriculture is low requiring an immediate and dramatic overhaul from its
current level of productivity. Emerging crop production challenges ahead
of us are still challenges that counteract our path to development.

It is saying the obvious that the agricultural sector forms the back bone of
Ethiopian economy with crops contributing to about 65% of the
agricultural GDP and 35% of the country's total GDP. As you are already
aware and are part of the process, the Ethiopian government has set the
Growth and Transformation Plan Two (GTP-II) - our economic
development blue print for the next five years. The core of this economic
development plan is structural shift in the economy from agriculture-
based to industrialization. This structural shift is expected to be
spearheaded by a rapid growth in agricultural production that should be
supported by new technologies in quantity and quality required for
investment in our industries.

The GTP-I has set crop production to increase aggregate production from
about 18 million tons to 39 million tons by the end of the plan period.
Much effort was made to achieve these targets and the agricultural sector
has registered success to meet the set targets but not without any gaps.
Our successes and failures were critically assessed and GTP-II was
planned taking into account lessons from GTP-I but with its own features.

[iv]
 Sebil Vol. 16

Increasing crop productivity requires technologies in the form of

improved crop breeds and crop husbandry, which are developed through
site and system specific experimentations. You as professionals involved
in various trades such as teaching, research, extension, production, etc.
very well know the challenges apparent in technology development and
diffusion all the way from crop production to post harvest processing,
distribution and utilization.

Given the importance of crops in the Ethiopian economy and towards the
achievement of the GTP-II goal, your skills and expertise are vital inputs in
generating and diffusing these technologies and know-how in order to
enhance productivity and value addition along the value chain. I feel that
the theme and sub-themes of the 16th biennial conference are well
articulated and are in line with the focus areas of the government.
Moreover, the invited papers for this conference are also critical areas of
concern and would like to appreciate for being thoughtful in designing the
conference a bit in different way than usual.

I understand that you will, over the next two days, deliberate on important
findings that will be presented by colleagues among you who have made
significant studies on various aspects of crop production, postharvest
processing and utilization. It is my expectation that critical discussions
shall be made on the findings to the effect that useful information shall be
filtered out for demonstration, popularization and scaling up.

On the same note, it is my conviction that this forum will serve as an

important platform for learning and networking among you professionals
to enhance your capabilities in the areas of crops research, education,
production, postharvest processing, marketing, etc.

You being one of premier agricultural professional societies, I urge you to

continue building excellence in all areas of crop sciences and to use this
excellence to serve our agricultural communities, especially smallholder
farmers.

With this, wishing you an enjoyable conference and fruitful deliberations, I

declare the 16th biennial conference of the Crop Science Society of Ethiopia
officially open.

Thank you!!
M
Crop Improvement 111 the Ever-Advancing Adverse
Agricultural Environment: The Anxiety of the Shock
A sn a k e F ikrc
Debere Zeit Research Ceneter. P.O.Box 32,Debre Zeit.
E-mail: fikreasnakeCdvahoo.com o r fa taw 71 (d.gmail. com

A b stract
This view point highlights reflection based on review and observation o f facts focusing on
Ethiopian Agriculture. In what is termed nozv climate change at climax, the politics o f agricultural
environment is spearheading from easily diversity accommodating to diversity erosion where onlif
biological entity with stress smart genes are being advocated. The law o f survival o f the fittest now
appears in a new paradigm shift and getting speedy due partially to the high impact o f human
intervention. Agricultural practices are facing ever unfitness challenges with the varieties or
breeds that have been dei’eloped by level o f scientific. Not more than 10% o f the released 733
varieties are in real impact. Ethiopian crop production process and products have suffered much
from disparate management, enhanced biotic and abiotic shocks, non-reputation, mixed culture
syndrome, environmental illness, inadequacy to administer the sector. Law o f the minimum has
always been operational to suffer the sector as long as piecemeal approach remains in dominance.
Hence, how productivity, production and quality be sustained with a number o f environmental
factors in speedy turn back to unfavorable, and on the other hand the demand side by far is getting
grave? Completely evolving the agrarian society landscape to a different has to be at stake. Still a
resource poor farmer should not be expected to venture in all agricultural produce mix, but to
specialized advantage or not at all. Where is the crop researchers' wisdom, particularly in the third
world to curb the situation right? This review is aimed to instigate a nezo dialogue box necessity
among actors in the sector, spinning on what matters most for the crop development sector in
Ethiopia and assessing the resource for impact path o f the sector.

Key words: Adverse agricultural environment, Crop improvement, Climate change

T h e improved crop version

Landraces were the starting point for the agriculture to start with. A
radical shift in the performance and gains of domesticated species happen
by the time the agriculturists learn variability sources (management,
environment, ecology, gene etc) of the entity; posing heavy combined
selection pressure exercise. Though original to start with, today's scientific
approach, landraces have been almost totally abandoned and put at
distance in favor of overriding derived varieties developed through high
technical knowledge and resource investment. Despite polarization of the
gene shift, the level of stress sink scrambled over modified genes becomes
unable to be tolerated. Any serious threat at this selected varieties or gene
sources is associated with serious socio-economic impact. So mankind has
 Sebil Vol. 16

tried to put all solution options to safeguard his developed technologies,

however, yet agricultural crises derived aid through both emergency and
development-oriented responses in multibillions of dollars (Curtis, 2008) is
alarming case of the globe. Acute interventions (e.g. food shortages),
typically through providing inputs such as seed and fertilizer is the future
of the world from one angle to the other.

Conventional crop breeding has and still is playing prime role in crop
improvement for the mass consumption.

300 | 282

250

.2 150
ra
98
o lO O
o
Z H 68

50
18 15
I i I t

J? d? ^ dT jf jf d? * jf
S"
j 3*
< I <§»
«P'
Figure 1: Varieties developed through conventional crop improvement in Ethiopia (source: Adefris et aI, 2011).

However, critical is to what extent did these 700 plus varieties impacted
the crop culture both in qualitative and quantitative terms? The gene
conversion ratio as compared to existing local cultivars is the highest in
cereals followed by pulses. Yet the impact level at the national level is far
below expectation as can be seen from performances of major crops in
Figure 2.

[2]
 Sebil Vol. 16

Figure 2: Intensification scenario with major field crops over 15 years (CSA, 1996- 2012) against the best yield achieved
under best management combinations
[Note of changes 1996 to 2012: area = 8072k ha to 12086k ha = 50%; production = 9645200MT to 21857080MT = 130%;
productivity (q/ha) =11.95 to 18.08 = 50%; Most crops increment has been less than a double; Most crops remained < 2
t/ha; The rift for the best is about 2/3; Growth and Transformation Plan of Ethiopia has been 18million MT(2011) to
39million MT(2015) ]

Sustainable intensification can be defined as "producing more output from

the same area of land while reducing the negative environmental
impacts and at the same time increasing contributions to natural capital
and the flow of environmental services' (Pretty, 2011). These are
requirements with many implications, but the place to start is yield. Yields
must be raised and yield gaps reduced and closed as part of
intensification. The scope to reduce yield gaps (the difference between
realized productivity and the best that can be achieved with improved
genetics, tech n ology and farm in g practices) is large p recisely becau se yield
gaps are large. In recent studies of yield gaps for major crops in
developing countries, specialists from national and international research
institutions assessed production constraints and their relative importance
(Langeveld, 2009; Waddington et al, 2010). Theoretical yield gaps for
m aize can be as much as 8 t/ha in South Asia due to a range of
constraints including limited water and nutrient availability, inadequate
protection of the crop from pests and diseases, insufficient or inadequate
[3]
 Sebil Vol. 16

use of labor or mechanization and knowledge deficits that result in poor

crop management. The illustration in Fig. 2 of major field crops in
Ethiopia fully supports this assessment. The general gaps between
achieved and best achievement takes more than 2/ 3rd of the area.
Waddington et al (2010) found that in the 6 crops surveyed in Sub-Saharan
Africa and South and East Asia, the yield gaps were greatest for sorghum,
cowpea and chickpea grown by smallholder farmers. Poulton et al. (2006,
p. 244) pointed out " agricultural intensification involves both technical
change and the presence of input, seasonal finance and marketing systems
to increase farm production and deliver it to consumers at a competitive
price". They were particularly pronounced in marginal drier areas of Sub-
Saharan Africa. According to Anthony and Feronni (2012) solutions to
close the yield gap must address the needs of smallholder farmers.
Agricultural biotechnology holds some of the answers, in the illusively
standing perception of the gene based revolution.

Trait switch-off/ switch-on based industrial agriculture have now been in

more of a business and can never be affordable by ordinary global citizen.
In this industrial agriculture phase, International institutions and
governments are exploring funding mechanisms that may help advance
g e n e tic a lly e n g in e e re d (G E ) te c h n o lo g ie s . Barker (2011) claims th e d a n g e r
in view of the biotechnology industry is strongly positioning itself.
Approximately 1,663 patent applications for "clim ate-ready" Crops have
been submitted for approval since June 2008 to June 2010. Three
companies -DuPont, Monsanto and BASF- comprise 66 percentof the
patents. Such proprietary dominance has significant societal and economic
implications and should stimulate robust discussion about the control of
seeds and, ultimately, food supply (Barker, 2011).

A gricultural environm ent turning;: the clim a tic and biosphere

With a probable temperature rise of 1.8 to 4 degrees Celsius (C), and a
possible rise of up to 6.4 degrees C, the impact of global warming on
agriculture will be devastating. According to the Intergovernmental Panel
on Climate Change (IPCC) and other sources, crop productivity will
decline in Central America, South and Southeast Asia and sub-Saharan
Africa. It is particularly troubling that yield declines of 20 to 40 percent
are anticipated for major food crops in Africa well before 2050 (Barker
2011). FAO report states: "Extrem e weather fluctuations present a growing
threat to agriculture. Organic systems appear to be more stable and

[4]
 Sebil Vol. 16

resilient in response to climate disruption based on comparisons with their

conventional counterparts under stress conditions such as severe drought
and flooding."

Barker's review show that the current agricultural practices account for
more than 30 percent of global Green House Gas emissions. When
examining the agriculture sector one finds further cause for alarm. For
example, 60 percent of global nitrous oxide (N 2O) emissions, a greenhouse
gas is, 296 times more potent than carbon dioxide (CO 2 ), is primarily due
to use of synthetic nitrogen fertilizers. Industrial agriculture practices
account for approximately 50 percent of methane emissions, a GHG gas 25
times more potent than CO 2 . In this line Ethiopian agriculture is on a verge
of synthetic fertilization intensification. On the other hand, the biodiversity
embankment against the biotic challenges is appearing on fright do mainly
selection pressure for economic traits.

Crop in ten sification : the will and th e p ractice

Ethiopian economy and future is rested on the performance of the
agricultural sector. A bigger part of the socioeconomic policies and
applications has been strongly and directly linked to agriculture.
Intensification, market orientation and competitiveness are cautiously
embedded within the development roadmap frame works. In Ethiopia,
agriculture, at least in its advanced form, will remain the important sector
even after would be takeover by industrial sector. Nevertheless; the sector
development has remained yet in state of emergency. The practical
application of all big intensification planning/strategies starts by sink to
walk. Over 15 years crop performance evaluation showed nothing but
open gaps and complications that intensification is not easily attainable in
the existing farming setups. And why should be the question to be cleared
ever until success, with its facing Ethiopian agriculture epitome
entertaining.

According to the assessment made bv Betru and Kawashima (2010) on half

a century performances of cereals in Africa; Ethiopia with a population of
about 80 million people in 2005 relied largely on expansion of cereal
harvesting land since 1961. Its cereal yield did not show much
improvement and stays at 1.2 t/ha in 2003. Agricultural productivity in
Sub-Saharan Africa (SSA) is very low and most countries in the region
neither produce enough nor are able to import food for their population
(Diao et alv 2008). This situation has continued to become an important
[5]
 Sebil Vol. 16

challenge for African development program. Most of its easily accessible

and fertile lands have already been over cultivated in the densely
populated highlands of the country. Although, Ethiopian actual arable
land has been cultivated near saturation level, it has enough room for
expansion to hitherto uncultivated lands except for limitation on
availability of renewable water (Table 1). This by itself comes at heavy
environmental price and a great deal of uncertainly. Ethiopia's declining
per capita water has always been under use conflict for thousands of years
among other lower riparian countries like Egypt and Sudan. Although
expansion of agricultural lands seems inevitable in Ethiopia, it could not
sustain continuation of past trends.

Table 1. Data on productivity, econom y and water resource on eleven highly cultivated
countries in Africa.

Cereal yield Per capita Par capita Cereal insecurity

Countries (t/ha)* GDP (U S D )b water (M3)c indicator*
Burundi 1.34 90 460 8.4
Eritrea 0.37 219 1.382 8.4
Rwanda 1.18 208 563 8.0
Tanzania 1.43 288 2,332 7.6
Ethiopia 1.24 106 1.387 7.4
Kenya 1.35 481 860 7.4
Liberia 0.92 152 69,123 7.2
Mauritania 1.45 515 3.610 6.4
Sierra Leone 1.22 202 33,237 6.2
Lesotho 0.94 730 1.687 5.8
Somalia 0.59 100 28,175 5.8
Congo, DR 0.77 119 21.629 5.6
Congo. Rep 0.81 1,118 202,089 5.4
Guinea 1.47 421 23,533 5.2
Egypt 7.52 1,085 773 5.1
Swaziland 1.31 2,317 2785 4.2
Mauritius 3.45 4,893 1,873 2.5
3= f(a, b, c), large avalue meant less food security and small value low food insecurity.
Source: Betru and Kawashima (2010)
The major descriptors in Ethiopian crop culture intensification can be put in dichotomy

M ai p ra ctice s vs best p ra ctice s: Ethiopian agriculture is complex,

unstable and full of mal and best practices. Very smaller proportions (<
l/4th) of the more than 50 million fragmented plots run by 12 million
households undergo 'modern' agricultural practices. And a smaller
proportion of this segment apply full package. And a sub-segment of these
effectively achieve the change. A sub-sub-segment of practitioners

[6]
 Sebil Vol. 16

motivated for the next move. The major division gaps between the mal
and the best practice remained on resource access, knowledge, agro-
ecological suitability, natural and socioeconomic calamities and existence
or absence of stimulus.

Knowledge vs m ission based: the knowledge base on the agricultural

sector is still not practically sufficient to modernize. The arrangement of
the existing knowledge base in the sector even some times is in disarray to
the whole goal. A clear road map has never been followed stably and
steadily to transform from one to the next scope. Over half a century a
number of reforms or approaches have taken place to agriculture that has
a time series link rather than progressive or a buildup linkage. Hence
modernization in agriculture is yet not in state of clearly defined path,
swims in a mixed agricultural syndrome. Development initiatives seem
undermine existing complexity and challenges, and that is why the
knowledge base in the sector is secondary to the mission, and a number of
those initiatives could not shade influence.

Under and over exploitation: Cropping culture in Ethiopia remained

unwisely exploitative of the natural environment. This farming practice
has been extravagant to nutrients and physically damaged the soil of the
ecosystem. This can be evidenced from the fact that today after thousands
of years farming practices; we are left with dead end agricultural lands in
northern part of the country. The concept of sustainability remained
heavily costly in the process of regaining the already lost. On the other
hand specialized agricultural practices have been least exploited, as the
default rule remained a house hold produces all agricultural products for
home use. In this case intensification could be even more complicated
concept and very relative against the will.

Investm ent vs environm ent: The investment component on crop

production has now been boosted to a level high. The nature and practices
of those investment based commercial crop culture will definitely attribute
to productivity and quality of produce. Unlike fragmented land holdings
that have complex nature for development, these farms could be
opportunities of large scale modernization. Yet, since their practice is
based on existing natural potential of those newly explored lands,
regulation has to be integrated to maintain the stats-co on a long term and
multi generation basis.

E7]
 Sebil Vol. 16

Im ported and w orsening th rea ts: the crop culture is by far hardly
skipping unpredictable threats both at wider scale and localized; from
climatic, biotic or abiotic sources. The great yellow rust devastation on
wheat, the heavy mealy bug damage on cotton, pea bruchids, sesame seed
bug, white mango scale, coffee berry diseases, bacterial wilt of potato,
aschochyta blight, bean common mosaic virus, citrus canker, drought
driven crop failure, acidic soil threat on yield, the aggressive weeds of
prosopies, water hycin and parthinium are few of peculiar present
challenges with wider impact on crop yield. Tebikew (2012) has reported
the lack of record for newly introduced bio-threats on crops mainly due to
mal-functionality of quarantine posts established long ago. According to
this author there is almost free movement of bio-agents from one corner to
the other, which put the nation at high cost of threat management.

Along the agricultural investment promotion, technology importation and

other interactions Ethiopian crop sector has been encountering a wildly
threats incomparably hard to locally existing challenges. Smartest
biological entities are continuously introduced and take over existing. The
mitigation underway for challenges in the crop sector usually comes after
damage, which cannot be economically justified.

Technology gaps: The approach of modernization in crop culture at the

fragmented household scale is being threatened by the law of the
minimum. The process from land preparation to table is far from complete
and therefore, seen perplexed at different steps of the chain or the process.
The insurance level to maintain stats co is still at negligence. The stimulus
or subsidy level for that matter is insignificant. The system management
for excess production or deficit; applied technology types; to audit weather
all technologies are interactive, need be a concern. Pushing changes or
improvement in the entire complexity is in synonymous where 'all
agricultural activities towards a single household' that remained un­
monitor-able with poor intersectoral interaction gains. It is also beyond
comprehension that the agro mechanization technology evolvement is
inconsistent and inactivated, like for eg. excercising precision agriculture ,
for instance tef row planting, in the absence of implements to do so.
Obviously production and processing technology gaps is already wide
comparing our crop culture. All this calls for restructuring the entire sector
based on relative advantage and competitiveness and pushing for different
socioeconomic setups that warranty productivity gain and quality.

[8]
 Sebil Vol. 16

In stitu tio n al and sectoral setups and in teraction s for system

productivity: The crop culture of Ethiopia is run under intersectoral
fatigue. The interaction has positive and negative attributes.
Environmental rehabilitation, bumper construction, urbanization,
livestock management, industrialization, mining etc which are undergoing
aggressively in Ethiopia has got huge impact on crop production and
improvement. On the other hand, lead institutional arrangement, capacity
and implementation factors need be critically evaluated on the basis of
synergy, output and reputation. Sectors complimentarily should be the
policy agenda not only for crop but for the entire productivity of the
environment or system productivity. Eg crop land expansion is a threat to
open grazing based sector of livestock; open grazing based livestock sector
is a threat to the environmental productivity by enhancing degradation. So
system need be in place to effectively manage theses economic sectors both
a household and country level.

D ata Imse: agricultural performance particularly yield data base are

generated by Center Statistical Authority (CSA) on zonal or regional or
national basis. However, there is a sense to develop data by environment.
The environment could be a combination of ecological and technological
options. Unlike CSA data basis which explains crop performance based on
both bad and favorable extremes for that particular crop, best technology
adoption combined with ecology brought about 3.5 t/ha for chickpea, 2
t/ha for lowland pulse, 6 t/ha for wheat, 4 t/ha for barley, 2.5 t/ha for tef,
35 t/ha for potato, 3 t/ha for faba bean, 2.4 t/ha for lentils (Kebebew et al,
2011). This simply indicates weather crops intensification should make
advantage of specialization by best adaptation ecologies for competitive
advantage.

T h e challenges and the science based solutions o f the crop secto r in

E th io p ia
Ethiopia now is accommodating different crop improvement initiatives
b o th at r e s e a rc h a n d d e v e lo p m e n t lev els. National a g ric u ltu ra l S y s te m
(NARS) that includes research institutes of both federal and regional,
universities, and private sectors that have all enhancing research based
technology use (technology development and adaptation). International
institutions (CGIARS) and some NGOs are also undertaking key crop
improvement based research integrating with national research systems.
In particular, Ethiopian Agricultural Research Institute has remained lead
in influencing the agricultural system through its role of multi-crop
[9]
 Sebil Vol. 16

technology development as explained in earlier section. Agricultural

extension system has also adaptive trail based technology promotion that
verifies the technology. There are different crop production improvement
initiatives viz., The Comprehensive Africa Agriculture Development
Programme (CAADP) to accelerate agricultural growth, improve food
security and strengthen environment resilience across African agrarian
economies, FAO, other bilateral project based initiatives, which can all be
added to the pool of science based crop improvement and productivity
warranting efforts. The point should be how far all those efforts gone to
their objectives by the end of the day?

T h e seed system and development

In the process of technology pack of crops seed is central. There are about
50 crop spp in the Ethiopian crop production culture. Only few crops are
blessed to have improved seed system, while the mass is constrained by
lack of formally qualified seed supplies. Except maize and wheat other
crop spp have either inadequate or totally absent seed system. Both formal
and informal seed system operates in Ethiopia. According to Getenet et al
(2001) the national demand for seed is estimated 480/000tons/annum, of
which only 4% is served by formal seed system. Asnake at al., (2012) after
a decade has confirmed that even 3% contribution. Both groups agreed
that the informal seed sector, which is derived by the farmers in different
arrangements (modifications) is and the would be reliable sources of seed
in the sector for long time to come. In the presence of supporting policy
environment of both sectors, the informal seed system is yet poorly
organized and least backup, but most contributors. Asnake et al.(2012)
argue the fact that the informal seed system should not be expected to pass
stringent qualification in the absence of getting basic transformation
inputs. Being a source of diverse crop seeds with huge total amount and
act in decentralized manner, the informal seed system need be shared
basic services like extension, early generation seed and marketing scheme.

Sum m ary and Conclusion

In the ever advancing challenge of crop production environment, there has

to be a shift of approaches in crop technology design, development, and
promotion. There is a need to combine contemporary attributes to the
outputs of crop technologies that warranty resilience. Crop productivity in
[10]
 Sebil Vol. 16

particular and agricultural system productivity in general consistently

achieved only if inter-sectoral complementarity and synergy system is well
developed. Institutional capacity in agriculture is a matter of great concern
to improve the sector contribution further.

R e feren ces

Adefris Teklewold and Daniel Mekonnen. 2012. Varietal development and

release for enhancing the seed system in Ethiopia. In Adefris T/Wold,
Asnake Fikre, Dawit Alemu, Lemma Desalegn and Abebe Kirub
(Eds).The defining moments in Ethiopian Seed system. 147-167.
Anthony, V.M. and Ferroni, M. 2012. Agricultural biotechnology and
smallholder farmers in developing countries. ELSEVIER Syngenta
Foundation for Sustainable Agriculture, Schwai'zwaldallee 215, 4002
Basel, Switzerland.
Asnake Fikre. Adugna Wajira, Frew Mekibib, Setegn Gebeyehu. 2012.
Practices and developments in the informal seed system of Ethiopia.
In Adefris T/Wold, Asnake Fikre, Dawit Alemu, Lemma Desalegn
and Abebe Kirub (Eds).The defining moments in Ethiopian Seed
svstem.
J
237-252.
Curtis, M., 2008. The Crisis in Agricultural Aid: how aid has contributed to
hunger.Background Paper for ActionAid p. 44.,
http://curtisresearch.om/ Agricultural.Aid.pdf.
Diao, X., Headey, D., Johnson, M., 2008. Toward a green revolution in
Africa: what would it achieve, what would it require? Agricultural
Economics 39, 539-550.
Getinet Gebeyehu, Gurmu Dabi and Gudissa Shaka, and Zewdie Bishaw.
2001. Focus on Seed Programs The Ethiopian Seed. Pp 22. Seed Unit,
1CARDA, P.O. Box 5466, Aleppo, Syria.
Hengsdijk H, Langeveld JWA. 2009. Yield Trends and Yield Gap Analysis
of Major Crops in the World. Werkdocument: Wageningen
University.
Kebebew Assefa, Asnake Fikre, Dawit Alemu and Adefris Teklewold
(eds). 2011. Mitigating crop technologies and seed gaps: addressing
the unaddressed. Proceedings of technology prescalingup mission;
E1AR, Addis Ababa, Ethiopia.
Pretty J: Sustainable intensification in Africa. In Sustainable Intensification:
Increasing Productivity in African Food and Agricultural Systems,
vol. 9. Edited by Pretty J, Toulmin C, Williams S. Earthscan; 2011:1.

[11]
 Sebil Vol. 16

Poulton, C., Kydd, J., Dorward, A., 2006. Overcoming market constraints
on pro-Poor agricultural growth in Sub-Saharan Africa. Development
Policy Review 24, 243-277.
Shawel Betru and H. Kawashima 2010.African cereal demand and supply
analysis: Past trends and future prospect . African Journal of
Agricultural Research Vol. 5(20), pp. 2757-2769.
Tebikew Damite. 2012. Free movement of seeds and propagative materials
and the spread of crop pests in Ethiopia. . In Adefris T/Wold, Asnake
Fikre, Dawit Alemu, Lemma Desalegn and Abebe Kirub (Eds).The
defining moments in Ethiopian Seed system. 22-234.
Waddington, S.R, Li, X., Dixon J, Hyman G, de Vicente MC. 2010.Getting
the right focus: production constraints for six major food crops in
Asian and African farming systems. Food Sec. 2:27-48.
 Sebil Vol. 16

AM M I and G G E Bi-plot Analysis of Sesame

(Sesam um in flirt tat L.) Genotypes
in Northern Ethiopia
F ischa B a r a k i 1 , Y c m a n e T s e h a y e 2a n d Fetien A b a y 2
' Crop research case team. Humera Agricultural Research Center, Tigray. Ethiopia.
E-mail:fish051har@gm ail.com :~Department o f Crop anti Horticultural science. Mekelle University

A b stract
The experiment zoas conducted in Northern Ethiopia (Humera, Dansha and Sheraro) under
rainfed condition from 2011-2013. Thirteen sesame genotypes were evaluated to identify stable and
adaptability sesame genotypes and the design znas Randomized Complete Block Design with three
replications. The combined ANOVA for grain yield showed significant effects o f the genotypes,
environments and their interaction. The average grain yield o f the genotypes across t)ie seven
environments was 743 Kg/ha. Genotypes G4, G1 and G12 outperformed the rest; 927kg/ha,
895kg/ha and 833kg/ha yield, respectively. Whereas, genotype G9 was the lozvest yielding genotype
(614.3 kg/ha). Additive main effect and multiplicative interaction bi-plot and Genotype x
Environment interaction bi-plot rez’ealed that G12 was the most stable, but G7, GS and G9 ivere
unstable genotypes. Furthermore, the genotype main effects and GGE bi-plot shozoed E5 as the
most discriminating and representative environment. The GGE bTplot also identified two different
grozving environments. The first environment containing E4 and E6 zvith the zoining genotype
G l; and the second environment encompassing El, E2, E3, E5 and E7 zvith zoinning genotype o f
G4.

Key words: AMMI bi-plot, Environment, GEI, GGE bi-plot

In trod u ction

Sesame (Sescimum indicum L.) is an ancient oil seed crop cultivated in

tropical regions around the world. It is not clearly known where sesame is
originated, however many scholars declare different regions as the centers
of origin of this crop. Ethiopia is recognized as a center of diversity for
sesame due to the presence of highly diverse types in the country.
According to the FAOSTAT (2012), Ethiopia is the sixth largest sesame
producer in the world following Myanmar, India, China, Tanzania and
Uganda and third in Africa. The productivity of sesame in Northern
Ethiopia (in the study area) (525 kg/ha) is very low compared to the
national average yield of about 757 kg/ha (CSA, 2013); as well as with the
world average yield. Countries like Mozambique can produce up to 1500
kg/ha indicating that there is a potential to increase the yield of this crop
[13]
 Sebil Vol. 16

(Buss, 2007). Apart from the lower yielding problem the quality of sesame
seeds is deteriorating from time to time which may negatively affect the
important traits like seed color, aroma, size and uniformity. Sesame
genotypes grown in Ethiopia, including the released varieties, are highly
variable in their performance when grown across locations. Hence, it is
important to test newly introduced genotypes or released varieties across
locations respectively before releasing a variety and/or recommending a
variety to a certain location.

Consequently, testing genotypes or varieties across different environments

often face a challenge of genotype by environment interaction. GEI
(Genotype x en viron m en t interaction) occurs when different genotypes
respond differently to different environments (Allard and Bradshaw,
1964). In the presence of significant GEI, there are a number of univariate
and multivariate stability measures used to identify stable and high
yielding genotypes. Additive main effect and multiplicative interaction
(AMMI) is one of the important methods to an a ly ze multi-environment
trials (METs) data and it interprets the effect of the genotype (G) and
environments (E) as additive effects and the GEI as a multiplicative
component (wnich are sources of variation) and submits it to principal
component analysis (Zobel et al., 1988). Another multivariate stability
measure called Genotype main effects and Genotype x Environment
interaction (GGE) biplot is also important to identify mega-environments,
the "which-won-where" pattern, and to evaluate genotypes and test
environments (Yan et al., 2007).

Like other crops, the yield and yield attributes of sesame is affected by a
number of biotic and abiotic factors leading to unstable and variable
performances when grown over a wide range of environments. Hence, this
experiment was undertaken to identify stable and high yielding sesame
genotype(s) and recommend the best genotype(s) for the different sesame
growing areas.

M aterials and M ethods

L o catio n s and experim ental M aterials

The experiment was conducted in Northern Ethiopia (Humera and
Dansha areas) under rain fed condition from 2011-2013, and at Sheraro in
2013 cropping season. The two locations (Humera and Dansha) with the

[14]
 Seb il Vol. 16

three years (2011, 2012 & 2013) combined to form six environments (El -
E6), respectively and Sheraro in 2013 was the seventh environment (E7).
Thirteen sesame genotypes (G1-G13 (brought from Werer Agricultural
Research Center (Table 1) and sown in RCBD with three replications. Each
genotype was randomly assigned and sown in a plot area of 2.8 m by 5m
with lm space between plots and 1.5 m between blocks. Inter and intra
row spacing used was 40 cm and 10 cm, respectively. Each experimental
plot received all management practices equally and properly as per the
recommendations for the crop.

Table 1: Agro-climatic and soil characteristics of the experimental sites

Location Latitude Longitude Altitude Annual Min - Max Soil texture

(°N) (°E) (m) Rainfall (mm) Temp (°c) Clay (%) Silt (%) Sand (%)
Humera 14°15' 36°37' 609 563.2 18.8-37.6 35.66 25.66 38.66
Sheraro 14°24' 37°45' 1028 676.7 18.8-34.9 21 27.28 51.71
Dansha 13°36' 36°41' 696 888.4 28.7(mean)

S ta tistica l Analysis:
Homogeneity of residual variances was tested prior to combined analysis
over locations in each year as well as over locations and years (for the
combined data) using Bartlet's test (Steel and Torrie, 1998). Accordingly,
the data collected indicated homogenous variance. Normality test was also
conducted and all data showed normal distribution. A combined analysis
of variance was performed using GenStat 16th edition (GenStat, 2009)
statistical software. The model employed in the analysis was;
Yijk = p + Gi + Ej + Bk + GEij + £ijk
where: Yijk is the observed mean of the ith genotype (Gi) in the jth
environment (Ej), in the kth block (Bk); p is the overall mean; Gi is effect of
the ith genotype; Ej is effect of the jth environment; Bk is block effect of the ith
genotype in the j th environment; GEij is the interaction effects of the ith
genotype and the jth environment; and £ijk is the error term.

From the combined ANOVA result the presence of GEI and the variation
due to genotype, environment and genotype x environment interaction
were partitioned. Moreover, mean comparison was performed using
Duncan's Multiple Range Test (DMRT) to identify genotypes significantly
differ in their mean.

AMMI and AMMI bi-plot analysis, showing the genotype and

environmental means against Interaction Principal Component analysis
 Sebil Vol. 16

one (IPCA1), and interaction Principal Component analysis one (IPCA1)

against Interaction Principal Component analysis two (IPCA2) were also
performed using Meta-analysis procedure I using the same statistical
software. GGE bi-plot was also executed using the Meta-analysis of
GenStat 16th edition.).

R esu lts and D iscussion

Combined ANOVA and estim ation of*variance com ponents

The results obtained from the combined analysis of variance of all the
evaluated traits and genotypes is illustrated in table 2. The genotype,
environment and genotype x environment interaction (GEI) variance were
decomposed to provide a general overview in relation to the evaluated
traits and overall performance of the genotypes (tables 2 and 3).
Accordingly, the genotypes, the environments and the genotype x
environment interaction components showed highly significant variation
(p<0.001) for all agronomic traits. On top of the genetic variability, the
ANOVA (table 2) also revealed that the environments (both locations and
growing seasons) on which the experiments were conducted were
different from one another in treating the genotypes. Moreover, it also
indicates that the response of the genotypes were unstable and fluctuated
in their trait expression with change in the environments. These all
phenomenon clearly confirms the existence of GEI in this study.

Table 2: Mean Squares for different agronomical traits recorded on sesame genotypes across locations

Source of YLD
Variation d.f (kg/ha) DF DM LCBZ (cm) NB NC PH (cm)

Rep 2 732 7.1 4.9 82.2 0.1 31.3 74.8

Genotype 12 208413** 16.5** 27.9** 120.1** 1.4** 81.1** 429.7**
Env 6 329874** 60** 221.8** 279 2 .2 " 9.4** 1268.2** 16284.7**
t

Gen*Env 72 24149** 4.1** 8.3** 64.0** 60.9** 189.1**

o

Residual 180 2707 1.3 2.4 25 0.1 16.1 50.7

NB: *,** statistically significant at (p<0.05 and p<0.01) respectively ns= non-significant, d.f= degree of freedom,
YLD=Grain Yield, DF=Days to 75% flowering, DM= Days to 75% maturity. LCBZ= Length of capsule bearing zone,
N6 =Number of branches, NC=Number o f capsules and PH=Plant height.

For most of the traits the contribution of environment for the overall
variance was higher (ranging from % for grain yield to for plant height)
followed by genotype x environment interaction and genotype
respectively (table 3). Similar results were reported by (Hagos, 2009;
[16]
Ahmed and Ahmed, 2012). With respect to grain yield, the greatest source
of variation was mainly the inherent genetic component meaning
genotypic effect (37.3 %) (table 3) which is similar to the results reported
by Zenebe and H ussien (2009) and John et a l (2001).
 Sebil Vol. 16

Table 3: Combined Sum of Squares for agronomic traits of Sesame genotypes evaluated during 2011 -2013

Source of
variation d.f YLD DM DF LCBZ NB NC PH
1464(0.0) 62.6(0.4) 149.7(0.1)
Replication 2 14.2(1.3) 10.0(0.4) 164.4(0.6) 0.2(0.2)
Genotype 12 2500959(37.3) 197.6(17.9) 334.7(12.4) 1441.0(5.2) 16.8(14.4) 973.2(6.1) 5157.0(4.1)

Env 6 1979243(29.5) 359.8(32.5) 1331.0(49.1) 16753.4(61.0) 56.3(48.1) 7609.0(47.7) 97708.7(77.7)

G en'Env 72 1738701(25.9) 295.6(26.7) 596.1(22.0) 4609.3(16.8) 32.0(27.4) 4383.6(27.5) 13617.1(10.8)

Residual 180 487308(7.3) 238.5(21.6) 438.0(16.2) 4486.6(16.3) 11.6(9.9) 2907.5(18.2) 9125.1(7.3)

Total 272 6707676 1105.7 2709.9 27454.7 116.9 15935.9 125757.5

¥ No out of parenthesis and inside parenthesis are SS and % SS of traits respectively; Gen=genotype; Env=environment

[181
 Sebil Vol. 16

A gronom ic p erform an ce o f Sesam e Genotypes

The average grain yield of the tested sesame genotypes over the seven
environments was 742.9 kg/ha. G4 had the highest average grain yield
(926.8 kg/ha) followed by G I (895.1 kg/ha) while G9 was the lowest
yielding genotype (614.3 kg/ha) (table 4). G4 was also early flowering

Table 4: Combined mean yield and related traits of sesame genotypes over all environments

Genotype YLD DF DM LCBZ NB NC PH

Acc#031 895.1b 4 2 .2 a 86.7bcd 51.3a 2.3a 2 8.88a 119.5a
Oro(9-1) 638.1w 42.1a 89.6a 44.1“ * 1.78f 21.42e 106d
NN-0079-1 740.4e 41 0 * 86.5* 49 J at. 1.8* e 2 4 .5 *d 118.4a
Acc-034 926.8a 39.7f 84.7e 48.8ab 2.2a 27.2ab 108.3*
Abi-Doctor 662.69* 40ef 85.7d 51.1a 1.6*9 23.3cde 112.4*
Serkamo 711,5ef 41.4b 87.0150 47.7ab 1 .7 " 23.5cde 119.9a
Acc-051-020sel-14 687.5* 4 1 .2 * 87.4bc 47.1 1 * 1.8* * 25.11)0 108.8*
Tate 655.29* 42.8a 87.8b 43.01d 2.1a 24.6bcd H 0 .4 *
Acc-051-02sel-13 614.3' 40.2def 8 7 .2 * 4 9 .0 4 * 1.6*9 24.2cd 1 1 2 .5 *
Adi 697.6f 40.5°-f 8 7 .2 * 48.5ab 1.59 22.1de 108.8*
Hirhir 791.5d 41 .0 ** 87.8b 47.9ab 2.0bc 2 4 .6 *d 111.8*
Setit-1 832.7C 40.7b_e 8 6 .4 * 49.5ab 1 9bc 2 4 .4 * 112.1*
Humera-1 805.1a1 40.9bcd 8 6 .9 * 47.3b 2.0b 25.8bc 116.4ab
Mean 742.9 41.07 86.98 48.1 1.9 24.6 112.7
LSD 83.83 1.8 2.5 8.04 0.409 6.47 11.5
CV(%) 7 2.8 1.8 10.4 13.7 16.4 6.3
NB: Means followed by the same letter are not statistically different from each other (DMRT, at 5%)

V arian ce estim ate for grain yield o f th e genotypes

The combined ANOVA for grain yield revealed that there were highly
significant variation (p<0.01) among the genotypes, environments (year,
ocation, year x location) an genotype by environment interaction
(Genotype x Year, Genotype x Location and Genotype x Year x Location)
(table 5). These significant variations of the genotypes, environments and
the GEI indicated that the response of the genotypes were unstable and
fluctuated in their grain yield with change in environment and these
phenomenon clearly declared the presence of GEI in this study.

[193
 Sebil Vol. 16

Table 5: Combined ANOVA for grain yield (kgI ha) of sesame genotypes

Source of Variation d.f SS MS

Rep 2 1464 732
Genotype 12 2500959 208413**
Year 2 1180209 590104**
Location 2 722088 361044**
Genotype x Year 24 388067 16169**
Genotype x Location 24 702189 29258**
Year x Location 2 76946 38473**

Genotype x Year x Location 24 648446 27019**

Residual 180 487308 2707
Total 272 6707676 24661
Key: df= Degree of freedom, SS=Sum of squares, MS= Mean squares

Figure 1 clearly visualizes the inconsistent performances of the genotypes

across the environments. The grain yield of the thirteen genotypes were
highly variable over the seven environments showing highest grain yield
cross-over interaction from environment to environment. Among the
environments the highest seed yield (1131 kg/ha) was observed from
genotype G1 in environment seven (E7) and the lowest seed yield (395.9
kg/ha) was recorded from genotype G9 in environment two (E2) (Table 6,
figure 1).

~ 1200
I 1000 ■ El
■ E2
x 800 ■ E3
>• ■ E4
T3 600
■ E5
1 400 ■ E6
a ■ E7
2 200
£
0
G1 G2 G3 G4 G5 G6 G7 G8 G9 G10 G il G12 G13
Evaluated Sesame Genotypes

Key: E1, E2, E3...refers to the Environments and G1, G2, G3... refers to the Genotypes
Figure 1: Mean seed yield of 13 sesame genotypes across seven environments

[20]
 Sehil Vol. 16

Table 6: Grain yield recorded from 13 sesame genotypes in each of seven environment and overall genotypic mean

Genotype Gen Test Environments

Name Code E1 E2 E3 E4 E5 E6 E7 Mean
Acc#031 G1 736.8c 840 ab 8 0 1 .8 ^ 852.2a 911.5 a 992.5a 1131a 895.1b
Oro(9-1) G2 612.6d 470.9'9 711.8e 582.6e( 701.2e 725.79 661.7' 638.1hi
NN-0079-1 G3 606.1d 610.6de 721.7 e 787.9abc 832.3bc 750'9 873.9c 740.4e
Acc-034 G4 964a 930.6a 846.4a 868.4a 890.3ab 98 7 ab 1001b 926.8a
Abi-Doctor G5 562.21 698cd 752.7d 453.19 684.5e 805.2ef 682.4ef 662.69h
Serkamo G6 470.4e 577.9 e 793.3c 709.7cd 717.5 e 887.1cd 8 2 5 cd 711.5ef
Acc-051-020sel-14 G7 815.3 b 465.3'9 738.7de 501.5 f9 712.3 e 850.3de 728.8e' 687.5f9
Tate G8 454.7e 510ef 552.69 775.9 .abc 699.2e 923.3abc 670.4®' 655.29h
Acc-051-02sel-13 G9 438.5e 395.99 618.9f 4759 751.8de 924.5abc 695.4 e' 614.3'
Adi G10 558.5d 686.8 cd 596.2' 646.6de 604.6' 946.7abc 843.6c 697.6'
Hirhir G11 587.2 d 794.6bc 827.1 ab 742.6 «>c 828.1 bc 913.2ted 848.1c 791.5d
Setit-1 G12 746.5bc 811.2 b 820.5abc 806.9 ab 812.3 cd 932.3abc 898.8< 832,7C
Humera-1 G13 630.1d 770.4 be 803.3 ^ 839.9a 876.3abc 960.9abc 755de 805.1cd
Mean 629.5 658.6 737.3 695.5 770.9 892.2 816.6 742.9
LSD ±) 69.7 100.5 26.4 86.7 69.5 65.2 78.1 83.83
CV(%) 6.6 9.1 2.1 7.4 5.3 4.3 5.7 7
NB: Means followed by the same letter are not statistically different from each other (DMRT, at 5%), bolded yield is highest seed yield of genotypes in their respective environments, and
Underlined yield is lowest seed yield of genotypes in their respective environments

[21]
 Sebil Vol 16

A M M I analysis
In addition to the usual ANOVA the ANOVA from the AMMI model for
grain yield also detected significant variation (p<0.001) for both the main
and interaction effects indicating the existence of a wide range of variation
between the genotypes, years (seasons), locations and their interactions.

AMMI 1 bi-plot analysis: The AMMI bi-plot analysis provides a graphical

representation to summarize information on main effect and interaction
effects of both genotypes and environments at the same time. The AMMI1
bi-plot containing the genotype and environment means against
interaction principal component analysis one (EPCA1) scores is illustrated
in Fig.2. As indicated in the figure the displacement along the abscissa
reflected differences in main effects, whereas displacement along the
ordinate exhibited differences in interaction effects. Genotypes and
environments with EPCA1 greater than zero classified as high yielding
genotypes and favorable environments whereas those with IPCA1 lower
than zero classified as low yielding genotypes and unfavorable
environments (Yan and Thinker, 2006).

Accordingly, genotypes G l, G4, G il, G12 and G13 were the genotypes
with above average mean grain yield as they laid-down on the right side
of the vertical line (grand mean of the genotypes and environments).
Conversely, genotypes yG2, G5, G6, G7, G8, G9 and G10 had yield below
the grand mean because of they laid down to the left side of the vertical
line. Exceptionally, G3 laid very close to the vertical line, indicating the
mean yield of this genotype was similar to the overall environment mean.
G4 followed by G l had higher mean yield in the favorable environments,
whereas G9 and G2 had lower mean yield in the unfavorable
environments. Regardless of their contribution for the interaction, G8 and
G5 fall on the same vertical line (ideal) showing their similarity in their
mean yield. G l and G10 which laid on the same horizontal line had similar
contribution in the interaction component despite of their yield
performance. With regards to the environments, E5, E6 and E7 had grain
yield above the grand mean and were considered as favorable
environments. In the other hand, E l, E2 and E4 had below average grain
yield and were considered as unfavorable environments. E3 laid very
close to the grand mean line indicating that genotypic yield in E3
represents the overall genotypic mean across all environments.

[22]
 Sebil Vol. 16

*E 4 |
10 -
oG 8

5 - ° C j6 :
uT oG K ) 1 '3 c?j6l
O *E 2 403 oG I*fe7
© 0 - oG 9 i <*E5
ON ................j...........................o G I 2 ..............................
rn
m
-5 -
<; cG2>G5 oG 4
O
c_
-10 -

oG 7
-1 5 -
*E1
----1------------1------------1------------1------------1------------1------------1------------
600 650 700 750 800 K50 900 950

M ain effect
Figure 2: AMMI1 bi-plot showing Genotype and Environmental means against IPCA1. Where the environments are
represented by (E) and the genotypes by (G) with in detail description of the environments and the genotypes in the
material and method part.

AMMI 2 bi-plot: The AMMI 2 bi-plot with IPCA1 in the X-axis and IPCA2
in the Y-axis, is plotted in figure 3. The first interaction principal
component (IPC1 or PCI) explained 33.59% and the second interaction
principal component (IPC2 or PC2) about 23.96% of the sum of squares of
the genotype by environment interaction . The two interaction principal
components cumulatively explained about 57.55% of the sum of squares of
the genotype by environment interaction (Fig. 3). Purchase (1997) stated
that the closer the genotypes to the origin are the more stable they are and
the furthest the genotypes from the origin are the more unstable they are.
In addition the closer the genotypes to the given vector of any
environment is the more adaptive to that specific environment and the
farthest the genotypes to the given vector of any environment is the less
adaptive to that specific environment. Accordingly, genotypes G7, G8, G9
and G4 are far apart from the bi-plot origin indicating these genotypes as
the more responsive and contributed largely to the interaction component
and considered as specifically adapted genotypes. On the other hand, G il,
G12, G13 and G3 were the genotypes with least contribution to the
interaction component as they located near to the bi-plot origin indicating
[23]
 I

Sebil Vol. 16

their wider adaptability (fig. 3). Regarding to the adaptability of the

genotypes in the environments; genotypes G I, G3, G10 and G i l were
adaptive to E7; and genotypes G2, G5, G12 and G13 were adaptive to
environments E3, E l, E2 and E4 respectively.

On
rn

P C I - 33.59%
Figure 3: AMMI2 bi-plot showing PC1 versus PC2 indicating the stability of the Genotypes. Where the environments are
represented by (E) and the genotypes by (G) with in detail description in the material and method part.

GGE Bi-plot
The first two principal components in the GGE bi-plot of this study
constituted 74.44% of total variance of. As indicated by Yan and Thinker
(2006), the similarity between two environments as well as genotypes is
determined by both the length of their vectors and the cosine of the angle
between them and the relations is illustrated in figure 4. The angle
between E l and E4, E6 is about 90° indicating there was no correlation
between these environments and produce different information about the
tested genotypes (figure 4). The rest of the environments had vectors with
less than 90° indicating that, these environments were positively correlated
to each other. E2 had longest vector and small IPCA2 and that was
relatively the most representative and discriminating environment and
considered as the ideal environment fni widely adapted genotypes. Hence,

[24]
 Sebil Vol. 16

Genotypes with above average yield in this environment had above

average yield in all environments. El and E4 were the most discriminating
but least representative environments which were with little information
of the genotypes and favorable for specifically adapted genotypes.
Exclusively, E6 was neither discriminating nor representative
environment. To clearly display graphically, the 'which-won-where'
pattern of a polygon view of GGE bi-plot is exhibited in figure 5. The
polygon was formed by connecting the vertex genotypes that were
furthest away from the bi-plot origin such that all other genotypes were
included in the polygon. The polygon view of bi-plot analysis (.Fig.5),
showed there were two different sesame growing environments. The one
environment includes the high yielding environments (E4 and E6), which
were in the Dansha area with the winning genotype G l; the second
environment included the low to medium yielding environments (El, E2,
E3, E5 and E7), which were under Humera and Sheraro areas with a vertex
genotype G4. The other vertex genotypes (G7, G8 and G9) without any
environment in their sectors were not the highest yielding genotypes at
any environment rather they were the poorest genotypes at all or some
environments.
S c a t t e r plot (T o ta l - 7 9 . 4 4 % )

PCI - 6 5 .6 9 %
Figure. 4: The environment-vector view of the GGE bi-plot to show similarities among test environments

[25]
 Sebil Vol. 16

Scattcr plot (Total - 7 9 .4 4 % )

P C I - 6 5 .6 9 %

Figure 5: The which-won-where view of the GGE bi-plot

Conclusion and Recom m endation

The combined ANOVA showed significant differences among the sesame

genotypes in this study for mean grain yield across environments. The
results also showed that the environments were highly variable with
respect to climatic and/or edaphic factors. This GEI in turn indicated that,
the performance or ranking of the genotypes was variable across
environments and it was difficult to identify superior genotype for all
environments or locations. The GGE bi-plot identified two sesame
growing environments; the area of Dansha (E4 and E6) with G I as a
winning genotype, and the other environment encompassing Sheraro (E7)
and Humera and Dansha (E1,E2, E3 and E5) with G4 as a Wining
genotype. The AMMI bi-plot and GGE bi-plot of grain yield data
identified G 12 as the most stable and widely adapted genotype for grain
yield while, G4 and G I were specifically adapted in the environments.

[26]
 Sebil Vol. 16

A cknow ledgem ents

The authors cordially acknowledge the research members of crop

department in Humera Agricultural Research Center and the Public and
Private Partnership Organization project (PPPO) for financial support.

R e feren ces

Zobel, R.W. Wright, M.J. Gauch, H.G. (1988). Statistical analysis of a yield
trial. Agron. J., 80:388-39
Ahmed M.B.S and Ahmed F.A. (2012). Genotype X season interaction and
characters association of some Sesame (Sesamum indicum L.) genotypes
under rain-fed conditions of Sudan. African Journal of Plant Science,
6(l):39-42.
Allard, R.W, Bradshaw A.D. (1964). Implication of genotype by
environmental interaction in applied plant breeding, Crop Science
60:503-506.
Buss, J. (2007). Sesame production in Nampula: Baseline survey report,
Mozambique, pp.2-20.
Ceccarelli, S. (2012). Plant breeding with farmers - a technical manual.
ICARDA, Syria.
CSA (2013). Agricultural Sample Survey. Report on Area and Production,
Volume III, Addis Ababa, Ethiopia.
FAOSTAT. (2012). (Food and Agriculture Organization of the United
Nations), http://faostat.fao.org/
GenStat. (2009). GenStat for Windows (14th Edition) Introduction. VSN
International, Hemel Hempstead.
Hagos Tadesse. (2009). Genotype by Environment Interaction and yield
stability of Sesame (Sesamum indicum L.) genotypes under North
Western and Western lowland Tigray, M.Sc. thesis, Mekelle
University, Ethiopia.
John, A. Subbaraman, N. Jebbaraj, S. (2001). Genotype by Environment
Interaction in Sesame (.Sesamum indicum L.): Sesame and safflower
newsletter no. 16, Institute of Sustainable Agriculture, FAO, Rome.
Steel, R. and Torrie, J. (1980). Principles and Procedures of Statistics a
Biometrical Approach. 2nd ed. Me Graw-Hill, Inc. pp 471-472.
Van, W. Thinker, N.A. (2006). Bi-plot analysis of multi-environment trial
data: Principles and applications, Can. J. Plant Sci., 86:23-645.

[27]
 Sebil Vol. 16

Yan, W. Kang, M.S. Ma, B. Woods, S. and Cornelius, P.L. (2007). GGE bi­
plot vs. AMMI analysis of genotype-by-environment data, Crop Sci.,
47:643-653.
Yan, W. (2001). GGE bi-plot windows application for graphical analysis of
multi-environment trial data and other types of two-way data, Agron.
J. 93:1111-1118.
Zenebe, Mohammed, Hussien Mohammed (2009). Study on Genotype X
Environment Interaction of Oil Content in Sesame (Sesamum indiciim
L.), Middle-East Journal of Scientific Research, and 4:100-104.
Zobel, R.W. Wright, M.J. Gauch, H.G. (1988). Statistical analysis of a yield
trial. Agron. J., 80:388-393.

[28]
 Sebil Vol. 16

Genetic Diversity and Traits Inheritance in

Anchote [C occiniu a b yssinica (Lain.) Cogii]
Accessions of Ethiopia
Desta F e k a d u 1, D erb ew Belew 2, A m salu Ay a n a 3
'EJAR, D ebreZ eit Research Center,desdar200H@ginail.com. P.O.Box 32, Bishoftu, Ethiopia
'Jim nia University, College o f Agriculture and Veterinary Medicine, dbelew2002(a vahoo.com, Jimma.
’ Oromia Institute o f Agricultural Research. avana6a(ci vahoo. com. Addis Ahaba. Ethiopia

A b stract
Anchote [(Cocciniu abysinica (him.) Cogn.] is a tuberous root crop grown mainly in xvestern and south
western parts o f Ethiopia. The need to promote this under-utilized crop and safeguard its diversity is
paramount because o f its nutritional, agronomic, socio-cutturat and socioeconomic importance for the
growers. There was no research attempt made so fa r to characterize anchote accessions o f Ethiopia. For
morphological characterization o f anchote a total of thirty six accessions of anchote have been used in the
study. The accessions were collected from four regions o f Ethiopia, namely: Oromia, Benishangul Guniuz,
SNNP, and Amhara and were conserved ex situ at Institute o f Biodiversity Conservation. Descriptors
prepared for cucurbits and sweet potatoes were used as there unis no descriptor developed fo r anchote due to
the scanh/ o f information and research on the crop. The study was conducted at Debre Zeit Research Center
(DZARC) Research Field on vertisols in 2010 and 2011 cropping seasons. The objectives o f the study were to
characterize anchote accessions and to estimate the extent o f variability and evaluating the diversity that
exists in anchote accessions, to cluster the test genotypes in to different homogenous groups, and to build
suitable selection parameters fo r anchote using the most important traits. Significant differences (p<0.05 and
p<0.01) were observed among the accessions for most quantitative traits. The principal component (PC)
analysis fo r quantitative characters plotted on hoo dimensions using the first two principal components (PC j
(53 %) and PCz (25 %)) with a total o f 78 % variation explained. The cluster analysis fo r 79 quantitative
characters demonstrated high morphological diversity (45-75 %) and they were grouped into four major
cluster classes. The genetic advance and herilability o f the quantitative traits shoxved high values for root
yield, number o f seeds per fruit, fruit weight, and leaf area. Leaf area was highly and positively correlated with
root yield and fruit weight but root number per plant was negatively correlated with the root yield. The most
important traits that could be used in anchote selection would be those traits that recorded higher values o f
heritability and genetic advance except fo r number o f h a d e s per fruit.

Key words: Anchote, Accessions, Locules

In tro d u ctio n

Anchote, Cocciniu abysinica (Lam.) Cogn., is an annual trailing vine

belonging to the Cucurbitaceae family best known and grown principally
for its tuberous root even though its tender leaves are also widely used as
food (Abera, 1995). The name 'anchote' refers to the edible tuber of the
cultivated races of Cocciniu abyssinica. Anchote is one of the most important
indigenous root and tuber crops to Ethiopia, it is a good source of
vitamins, minerals protein and calcium compared to other root crops.

[29]
 Sebil VoL 16

Therefore, it can be used as a strategic crop to alleviate protein deficiency

in areas of nutrition with low protein source (Amsalu et al., 2008).

The genus Coccinia is made up of 30 species of which eight are reported to

occur in Ethiopia. The species recorded in Flora of Ethiopia since 1995
include: Coccinia abysinica (Lam.) Cogn., C. adoensis (Hochst. Ex. A. Rich.)
Cogn.)/ C. grandis{L.) Voigh(Syn. C. indica Wight and Arn.), C. megarrhiza,
C. Jeffrey and C. schliebenni Harms. The remaining three species have not
so far been described and named. According to Amare (1973), anchote is
cultivated in areas between 1300-2800 m above sea level where the annual
rainfall is 762-1016 mm.

In addition, constraints like absence of well characterized cultivars

morphologically and nutritionally, absence of information on suitable
planting time for each agro ecologies, lack of improved varieties having
wider adaptation and with high yield and other merits, lack of awareness
of the crop on production and utilization in non growing areas of the
country, etc. could be mentioned as a bottle necks for the under­
development of anchote in Ethiopia. Accordingly, 36 anchote accessions
collected from the producing areas of the country were evaluated for their
variability in quantitative traits to identify the desirable traits for selection.
The present investigation is an attempt to estimate the extent of variability
among the accessions and evaluating the diversity that exists in anchote
accessions and to construct suitable selection parameters for anchote using
the most important traits.

M a terials and M ethods

Thirty six anchote accessions (29 obtained from the Ethiopian Institute of
Biodiversity Conservation (IBC) and seven from own collections) (Table 1)
were used in this study. The field trial was conducted at Debre Zeit
Agricultural Research Center (DZARC) Research Site, Bishoftu( 1860
m.a.s.l). It receives an annual average rainfall of 851 mm and the mean
minimum and maximum temperatures are 8.9°C and 24.3°C, respectively
on vertisols in 2009/2010 to 2010/2011cropping seasons.

[30]
 Seb il Vol. 16

Table 1. Description of Anchote accessions used for the experiment

Accession Genus Species Region Zone Woreda 1 district Altitude

no. name name /province (ma.sl)
90801 Coccinia abyssinica Oromia Horro Guduru Wollega Abbay Chomen
90802 Coccinia abyssinica Oromia Horro Guduru Wollega Abbay Chomen
207984 Coccinia abyssinica Benishangul Asosa Asosa 1400
Gumuz
223085 Coccinia abyssinica Oromia East Wollega Digga Leka
223086 Coccinia abyssinica Oromia East Wollega Digga Leka
223087 Coccinia abyssinica Oromia W est Wollega Gimbi
223088 Coccinia abyssinica Oromia West Wollega Gimbi
223090 Coccinia abyssinica Oromia West Wollega Gimbi
GM* Coccinia abyssinica Oromia W est Wollega Gimbi/Abba Sena 2400
223092 Coccinia abyssinica Oromia East Wollega Sibu Sire
223093 Coccinia abyssinica Oromia East Wollega Sibu Sire
223094 Coccinia abyssinica Oromia East Wollega Sibu Sire
NJ* Coccinia abyssinica Oromia West Wollega Nejo 1909
223096 Coccinia abyssinica Oromia East Wollega Guto Wayu
223097 Coccinia abyssinica Oromia East Wollega Guto Wayu
223098 Coccinia abyssinica Oromia East Wollega Guto Wayu
223099 Coccinia abyssinica Oromia East Wollega Jimma Arjo
223100 Coccinia abyssinica Oromia East Wollega Jimma Arjo
223101 Coccinia abyssinica Oromia East Wollega Jimma Arjo
223104 Coccinia abyssinica Oromia Jimma Dedo
223105 Coccinia abyssinica Oromia Jimma Dedo
DD* Coccinia abyssinica Oromia Dembi Dollo Gidda Gebo 2359
223108 Coccinia abyssinica Oromia llu Ababor Ale
223109 Coccinia abyssinica Oromia llu Ababor Ale
223110 Coccinia abyssinica Oromia llu Ababor Ale
223112 Coccinia abyssinica Oromia llu Ababor Bedelle
223113 Coccinia abyssinica Oromia Jimma Manna
229702 Coccinia abyssinica Amhara Misirak Gojam Hulet lju Enese
220563 Coccinia collyssini Oromia East Shoa Bako Tibe
DIGGA* Coccinia abyssinica Oromia East Wollega Digga 2123
230565 Coccinia collyssini Oromia East Wollega Guto Wayu
230566 Coccinia collyssini Oromia West Wollega Gimbi 1820
240407 Coccinia abyssinica SNNP Keficho Shekicho Decha
KICHI* Coccinia abyssinica Oromia East Wollega Gute 1821
KUWE* Coccinia abyssinica Oromia East Wollega Sibu Sire 1987
SODDU* Coccinia abyssinica Oromia East Wollega Sibu Sire 1823
* own collections
The experiment was laid out in randomized complete design replicated
three times with spacing of 40 cm x 20 cm between rows and plants
respectively. Seeds were used as direct planting material at 5 cm depth in
the soil for all accessions. International Plant Genetic Resources Institute
(IPGR1 (2002)) descriptor list for sweet potato and European Cooperative
Program for Plant Genetic Resources (ECPGR (2008)) for Cucurbita species
were adopted with some modifications to suit with the crop characteristics
to record and describe quantitative traits(Table 2). The data were recorded
on randomly selected plants from each plot for all quantitative characters.
The variability parameters; genotypic variance (o2g), phenotypic variance
[31]
 Sebil VoL 16

(c^p), genotypic coefficient of variation (GCV), phenotypic coefficient of

variation (PCV), broad sense heritability (H2), and genetic advance (GA)
(Johnson et al., 1955) and genetic advance as percent of mean (GAM) were
employed for each traits as suggested by Jim et al. (2003). The pooled data
of two years were subjected to statistical analysis using SAS statistical
package (SAS, 2002) SAS 9.2. Mean separation was done using Least
Significant Difference (LSD) at 5 % significance level.

Table 2. List o f quantitative traits and methodology o f data recording

T rait Code Methodology

Days to 50% emergence DE Date of 50% emergence from planting
Germination type Epigeal or hypogeal
Days to maturity DM Date of leaf senesce start to occur
Leaf area LA Leaf area measurement was done using a scientific leaf area meter
(Area Meter AM 200-002 ADC Bioscientific Ltd) at full maturity stage
from 20 sampled plants
Powdery mildew severity PM Scored in percentage basis with two weeks interval for two months
after four months of planting when started to occur
Number of root per plant NRP The clustered number of roots per plant were counted
Root weight per plant (kg) RWP 20 sample roots was measured using a sensitive balance (BP
16000-S) and the average was recorded
Marketable root yield(t/ha) MRY Total weight from 20 plants from three middle rows
Total root yield (t/ha) RY Total weight from 20 plants from three middle rows
Root dry matter (%) RDM The dry weight of root after oven drying at 105 °C for 24 hours
Fruit length(cm) FL Length of the fruit measured for twenty samples from top to bottom
using ruler
Fruit diameter (cm) FD Width of the 20 sampled fruit was measured using a caliper
Number of strip lines on fruit surface NST The strip lines on the surface were counted
Number of locules per fruit NL The fruit was cut longitudinally and the locules extracted and
counted
Number of seeds per fruit NSF After cutting the fruit longitudinally the seeds were extracted,
washed and counted
Fruit weight (g) FW 20 sampled fruits were weighed using a sensitive balance
Thousand seed weight (g) TSW Seeds collected from the twenty sampled plants were counted to
1000 and weighed using a sensitive balance
Number of sepal and petal NS and Counted from the flowers
NP

R e su lts and D iscussion

The analysis revealed consistent and highly significant (p< 0.01) variation
among the tested anchote accessions for the quantitative traits evaluated
except for the number of sepals and petals, and number of locules per
fruit(Table 3). Days to 50 % emergence ranges from nine to twelve days
after planting and with the average of nine days.
 Sebil Vol. 16

Table 3. Mean and R-square values of 20 quantitative traits of 36 anchote accessions

Quantitative traits Mean R-square

Days to emergence (no.) 9 0.99
Days to maturity (no.) 177 0.88
Leaf area (cm2) 122.09 0.92
Powdery mildew (%) 41.42 0.92
Root number per plant 3.60 0.71
Root weight per plant (kg) 1.18 0.96
Total root yield (t/ha) 73.51 0.96
Root dry matter (%) 20.03 0.96
Fruit length (cm) 6.86 0.87
Fruit diameter (mm) 42.61 0.97
Fruit weight (gm) 74.37 0.99
Number of strips on fruit 9.89 1.00
Number of locules per fruit 6 0
Number of seeds per locule 19.72 0.90
Total seed number per fruit 118.20 0.99
Thousand seed weight(gm) 34.26 0.97
Number of sepals 4.92 1.00
Number of petals 4.92 1.00

L e a f M orphological T r a its

There is highly significant (p<0.001) difference among the accessions for

leaf area (Table 4, 12). According to the result, from high yielder
accessions, larger leaf area value was recorded; accession number 223096
(156.7 cm2 area and 130.8 t/ha root yield), KUWE (147 cm2 area and 104
t/ha root yield), and 223097 (174.0 cm2 area and 79.8 t/ha). Lower leaf area
was recorded from accession number 223092 (72.47 cm2 and 70.42 t/ha).
The range of the leaf area was wider (72.47 cm2 to 174.07 cm2) among the
accessions. According to Lebot (2009), an early development of the root
system in root and tuber crops, occurs during the period of vine growth
and leaf area acquisition and finally tuber development.

In addition, leaf area was highly and positively significantly correlated

with root weight per plant (r=0.29), marketable root yield (r=0.29) and fruit
weight (r=0.32) (Table 9). This result is in conformity with the works of
Lebot (2009) that there is a high correlation between leaf area and tuber
yield in vine tuber crops such as yam and sweet potato. There is no
correlation between the leaf area and root number per plant even though
Lebot (2009) reported a strong correlation between leaf area and the
number of roots per plant in yam and sweet potato. Even though leaf area
index (LAI) is an important component of the ideotype and the yield
potential is closely related to its value, measuring it in vine crops is
particularly difficult due to the complexity of their canopy (Chipungahelo
[33]
 Sebil VoL 16

et al., 2007), thus was not considered in this study due to the runner
growth nature of the crop and the matured lower leaves get senesce while
the new leaves flush with increased vine length and growth period. The
broad sense heritability for the leaf area was 81.34% with 32.25 genetic
advance. The phenotypic and genotypic coefficients of variation were
15.76 and 14.25, respectively with genetic advance as percent of mean
26.30 (Table 7) showing wider variability among the tested accessions
which exhibits the crop's potential to exploit through further research and
improvement.

F ru it T r a its

There was highly significant (p<0.05) difference among the tested

accessions for most of the traits except number of locules per fruit (Table
5). The length of the fruits ranged from 5 cm to 9 cm. According to the
results of this study, leaf area and size of the fruits is directly proportional
since some accessions with small leaf area were found with shorter fruit
length. This result is in agreement with the findings of Koller (2009), that
the direct proportionality of the leaf and fruit sizes in Coccinia indica and
Coccinia grandis. Fruit diameter is larger for those with longer fruits. The
highest root y ie ld e r a c c e s s io n w a s w ith s m a lle r fruit d ia m e te r c o m p a re d
with the lower root yielders associated with bigger fruit diameter. The
fruit shape index showed pronounced variability among accessions (1 .2 to
2.9). According to Alice and Peter (1995), fruits with high shape index have
fewer locules in tomato. In contrary there was no difference in number of
locules in the tested anchote accessions even though there was a wide
difference in fruit shape index among the accessions.There was a narrow
variability among accessions for number of strip lines on the fruit surface
ranging from 8 to 11 . All the tested accessions were with equal number of
locules in their fruits but the number of seed per locule(ll-66) was highly
variable among accessions. According to the findings of Solmaz and Sari
(2008) the large number of locules in anchote is in contrary with the
number of locules of most cucurbits which is one for inferior ovary and
three for superior ovaries. Fruit weight ranged from 42.68-107.46 grams
and thousand seed weight ranged from 22.87 to 47.75grams.

[34]
 Sebil Vol. 16

Table 4. Leaf area (cm2) values of 36 anchote accessions

Accession number Leaf area (cm2)

90801 97.434*
90802 97.24»r
207984 89.887r
223085 102.45W
223086 142.54cd,e
223087 114.35 mn°P
223088 123.969hiiklm
223090 111,64n°P
GM 109.06°w
223092 116.95klmn0
223093 72.47s
223094 97.27V
NJ 91.85r
223096 156.69b
223097 174.07^
223098 131.17<waw
223099 119.52iklmn0
223100 133.49def9'
223101 143.52bcde
223104 133.35def9hl
223105 102.54W
DD 1 2 6 .2 9 9 ^
223108 130.10e^ hi
223109 144.09bcd
223110 116.50,mno
223112 135.01cdef9f1
223113 108.89°P9
229702 129.749hiikl
220563 110.33°^
DIGGA 115 13mn°p
230565 140.00cdef
230566 135.61cdef
240407 139.32cdef
KICHI 120.53®*"
KUWE 147.13bc
SODDU 135.40cdef9h
LSD (0.05) 13.55
CV (%) 6.81

[35]
 Sebil Vol. 16

Table 5. Mean values of fruit morphological traits and shapes

ACC No. FL (cm) FD(mm) FSI NST NLF NSL NSF FW (g) TSW (g) FS
90801 8.6ab 60.8a 1.4 10b 16* 96k 99.82b 26.25°p Round
90802 6.8ef9 57.50 1.2 10 b 23* 132* 73.34kl 26.25°p Round
1.7 Plum
207984 7ef 41.5ldm 10b 18*9* 108 67.67" 42.70bcd shaped
223085 69W 35<»r 1.7 10 b 14* 84m 69.24™ 42.12cd Spherical
223086 9a 50.4*° 1.8 10b 22«*> 132* 99.1b 23.119 Round oval
223087 69*' 42.5)klm 1.4 10» 32* 1926 57.120 30.87™ Round
223088 7a 32.5’ 2.2 11a 18*9* 1081 75.2>k 32.44klm Spherical
223090 6.6*9 46.7** 1.4 10b 18*9* 108J 78.29*' 34.93* Spherical
1.8 Round
GM 7.2def 40.2 m"° 10b 2 0 *' 121h 62.96° 38.18*9* oval
1.6 Round
223092 69*' 37.9°p 10b 169*' 96k 48.14' 24.18P9 oval
1.3 Round
oval

O
42.68u 22.929

r-
13*

cO
223093 5i 38.8"°p 9°
223094 7 ef 41.5klm 1.7 10 b 14* 84m 53.92r 39.14ef9 Spherical
NJ 6 .4 ^ 38.1°p 1.7 9s 12* 72° 49.17s1 35.75*5 Spherical
223096 7ef 26.9s 2.6 10b 15*1 90' 95.1d 28.57"° Round oval
1.7 Round
223097 6.5*9* 37.9op 10b 18*9* 1081 77.329*'! 45.19b oval
223098 7ef 43.5>kl 1.6 10b 19®*9 114' 66.82" 34.92* Round
223099 5.6** 42.5iklm 1.3 9° 22°* 132* 53.08' 27.07° Spherical
223100 69hl 4 2 .2 ^ 1.4 10b 22 132* 59.66” 41.17“ *® Spherical
1.5 Plum
223101 6 ghi 41.2lmn 10b 21cdef 1269 72.12H" 31.49"" shaped
223104 7.6cde 50.9<* 1.5 10* 20^ 120" 107.46s 37.019W Spherical
223105 8.4abc 29.3s 2.9 10b 28b 168c 105.83 43.05bc Spherical
1.8 Round
8bcd 44.2i*k 10b 24c 144<j 98.251* 28.76"° elongated
DD
223108 9* 36.2w 2.5 8d 24 e 144d 79.76*9 30.33"m Round oval
223109 9a 38°p 2.4 10b 169*' 96k 59.82” 22.879 Round oval
223110 6.8 <*9 36.9W 1.8 10b 33a 198a 79.12*9 41.06“ * Oval
223112 6.5*9* 43iM 1.5 10b 22«te 132* 63.09° 32.32"" Round oval
1.4 Plum
223113 7ef 49.60de' 10b 19*8 m 78.399* 26.6 8°p shaped
7ef 47.3*9* 1.5 10b 23cd 138® g&ee* 41 14cde Round oval
229702
220563 5i 32.6r 1.5 10b 13* 78* 51.83re 33.66>kl Round
DIGGA 6.8 ef9 48.1ef9 1.4 10b 22c*> 132* 60.6op 47.75a Round
1.3 Plum
230565 69hi 46.89*' 10 b 114 75.91** 31.75"" shaped
1.3 Round
230566 5.4*i 42.8iklm 10b 21cdef 1269 71.92"" 36.879*' round
240407 6.8 e{9 51c 1.3 10b 11k 66p 92.1® 40.16** Round oval
KICHI 8.4abc 48.2^9 1.7 10b 20* / 120" 99.82b 38.08^ Round oval
KUWE 6.6*9 44.6 1.5 10b 23ed 138® 81.78* 37.389*' Oval
1.3 6 Plum
SODDU 69*' 47.1 10b 19®f9 114' 75.29* 37.249*' shaped
SD 1.14 7.25 0.46 0 5.24 29.55 17.56 6.83
SEM 0.11 0.69 0.04 0 0.50 2.84 1.68 0.65
LSD(0.05) 0.90 2.70 ns ns 3.60 4.72 2.96 2.55
CV(% ) 8.04 3.84 0 0 11.19 2.45 2.44 4.55
Key: FS-Fruit Shape, FL-Fruit Length, FD-Fruit Diameter, NST- Number of Strips on a fruit, NLF-Number of Locules per
Fruit,NSL-Number of Seed per Locule, NSF-Number of Seed per Fruit, FW-Fruit Weight, TSW-Thousand Seed Weight

[36]
 Sebil Vol. 16

Table 1. Mean values of storage root quantitative traits

Accession Root dry matter Mean Root Mean Root weight Root Yield
number (%) number per plant per plant (kg) (t/ha)
90801 16.28mno 4.7a 0.639h)' 78.139hii
90802 19.64'ik 4.33ab 0.399 47.929
207984 20.31h'i 3.67abcd 0.60h'ik 75.08h'ik
223085 22.62f9h S.OO^ 0.55kln 68.04kln
223086 25.90° 3.33cbde 0.77cd 95.94°d
223087 22.96ef9h 4.33ab 0.67f9h 83.79f9h
223088 10.97P 0.591“ 73.130”
223090 33.33a 3.33tK<ie 0.7 5 ^ 93.96de
GM 10.71p 3 3 bcde 0.18s 22.71s
223092 26.82c 3.33bcde 0.60hiik 74.58h'ik
223093 18.21iklm 3 6 7 ab°d 0.57'ikl 70.42i“
223094 22.40f9h S.OO"1® 0.389 46.889
NJ 22.60f9h 3.33baie 0.389 47.299
223096 25.81cd 4.00abc 1.05a 130.833
223097 18.59klml 3 0 7 abcde 0.649hi 79.809hi
223098 25.40cde S.OO^ 0.40P9 49.38P9
223099 21.809hi 4.00abc 0.76cde 94.79cde
223100 24.85cdfe 4.00abc 0.52n 64.38ln
223101 13.74° 3.67abcd 0.60hiik 74.38h'ik
223104 22.73efsh 3.67abcd 0.77°d 96.46cd
223105 18.41 iklm 0.74def 92.50def
DD 23.19def9 2.33e 0.26r 32.71f
223108 19.39'iw 4 00ab° 0.56ikl 69.79^
223109 23.02ef9 3.67abcd 0.69ef9 86.04ef9
223110 17.05k™ 3.67abcd 0.56ikl 69.79*'
223112 5.709 3.33 **** 0.69ef9 85.42ef9
223113 6.229 3.33bC(ie 0.88b 109.17b

Table 6. Mean values of storage root quantitative traits (continued)

Accession Root dry Root number Mean Root weight Root Yield
number matter (%) per plant per plant (kg) (t/ha)
229702 20.629N 4.33ab 0.80bcd 99.60bcd
220563 25.34cde 2.67de 0.51°° 63.19no
DIGGA 14.97n0 3.33M * 0.47n°p 58.75"°?
230565 29.89b 4.67a 0.44P9 54.17°P9
230566 19.441“ 3.00cde 0.51'n° 63.75lno
240407 16.85'™ 4.33ab 0.50"° 61.67no
KICHI 16.87'™ 4 00abc 0.53kln 65.83kln
KUWE 17.29ki™ 3.00°de 0.83b° 103.96*
SODDU 2 1.27*' 4.00abc 0.50°° 62.29no
L S D (0 .0 5 ) 2 .6 7 1 .2 6 0 .1 6 9 .7 5
C V (% ) 8 .2 2 1 .6 1 8 .1 3 8 .1 4
E R C B D *(% ) 92 65 95 95
Efficiency relative to RCB Design beyond 100%

[37J
 Sebil Vol. 16

Storage Root T raits

There is a wider diversity among the accessions for both quantitative traits
of anchote root (Table 6 and 7). There was highly significant (p<0.01)
difference among accessions for all quantitative traits of the storage root.
Growth, dry matter production and partitioning pattern in anchote
accessions is not known so far. According to the result, the root yield and
root dry matter (%) were directly proportional to each other in that higher
root yielding accessions also had higher percentage of root dry matter and
the low root yielding ones had lowest values of dry matter (Table 6). The
storage root yield of 36 accessions ranged from 22.71 to 130.83 t/ha with
phenotypic and genotypic coefficient of variance, genetic advance, genetic
advance as percent of mean, and heritability (broad sense) of 32, 30.88,
45.13, 61.30, and 93 respectively (Table 7). The wider variability in storage
root traits revealed in anchote accessions is in agreement with the root
traits diversity exhibited in dry matter and root yield variability of sweet
potato germplasm collection from Tanzania (Elameen et al., 2010). Root
yield (ton/ha) and mean root weight per plant (kg) were highly and
positively significantly correlated; with r=0.29 for both (Table 9). The
significant correlation of root yield and leaf area could be due to the
indeterminate growth nature of anchote plant which in turn increases the
assimilate accumulation to the root and increases the final root yield.
Storage root number per plant ranged from 2 to 5 and it was not
significantly correlated with total root yield and other quantitative traits
could be due to the difference in storage root size and then their storage
root weight per plant among accessions.

The variability estimates for storage root quantitative traits showed that
most of the traits resulted in narrow difference of PCV and GCV except
number of roots per plant (Table 7). The broad sense heritability (H2)
values also higher and evidence the reliability of using these traits for
selection in anchote improvement.

Table 7. Phenotypic and genotypic coefficients of variation, genetic advance, genetic advance
as percent of mean and heritability of four quantitative storage root traits
Root traits PCV GCV GA GAM H2
Root no./plant 32.03 21.83 1.10 30.40 46.00
Root weight/plant (kg) 32.01 30.89 0.72 61.30 93.13
Root yield(t/ha) 32 30.88 45.13 61.30 93.00
Root dry matter (%) 29.52 28.34 11.22 56 92.14
PCV-phenotypic coefficient o f variation, GCV-genotypic coefficient of variation, GA-genetic advance,
GAM-genetic advance as percent of mean, H2- broad sense heritability

[38]
 Sebil Vol. 16

Cluster and Correlation Analysis

Cluster analysis classified accessions into four major groups (Figure 1 and
Table 8). Group I with two sub groups: la and lb ; group II two subgroups:
2a and 2b; group III two sub groups: 3a and 3b; group IV two sub groups:
4a, 4b and two sub subgroups, 4ai, and 4 a i Generally, the quantitative
traits are clustered in to four major groups and eight sub groups. Accession
number 223096 (14) was found outlier from the others and was not
included in the cluster. Within these four major groups, all accessions
clustered randomly with no specific clustering linked to agro-ecological
zone but with their quantitative traits and majority of the individuals
clustered in a major group four with genetic dissimilarity range of 0.45-
0.75 indicating high diversity within members. Among the two sub groups
of group four, sub groups 4a and 4b had many accessions clustered
together in the same branches though recorded in different names from
different agro-ecological zones. Among the sub groups of g rou p fou r, in
sub group 4b many accessions clustered together in the same branch of 4bi
than 4a2 and lower number of accession clustered together in 4ai and 4b2.
(n contrast to group four, individuals in major group one and three were
few in number and in group two more number of accessions clustered
together than group one and three. The higher root yielding accessions
were clustered in to cluster I and II regardless of the places from where the
accessions were collected. Collections from Jimma Zone (Dedo and Mana
woredas) were clustered into cluster I and II. This may be because of the
seeds share among farmers through the markets.

[39]
 Sebil Vol. 16

A
v
e
r
a
g
e

i
s
t
a
n

2 5 21 14

Accession numbers Clustered

Table 8. Cluster groups for quantitative trails
Cluster I Cluster II Cluster III Cluster IV
223087 223086 90802 90801
223110 223104 GM 240407
223105 229702 DD 207984
KUWE 23085
223090 230563
223099 223092
223113 223093
223094
NJ
223088
KICHI
223098
230565
223100
230566
SODDU
DIGGA
223101
223112
223108
223097
223109
Total = 3 Total= 7 Total= 3 Total= 22

[40]
 Sebil Vol. 16

Table 9. The Pearson Product Moment Correlation Coefficient(r) for quantitative traits

LA PM RNP RWP MRY TRY RDM FL FD NST NL NSL NSF FW TSW NS NP

LA 0.999” 0.032 0.286" 0.286" 0.286" 0.056 0.128 -0.061 0.191 (a) 0.108 0.122 0.324" 0.06 0.136 0.136
PM 0.03 0.286” 0.286" 0.286" 0.058 0.128 -0.06 0.191* (a) 0.108 0.123 0.324 0.061 0.136 0.136
RNP 0.095 0.095 0.095 0.001 0.099 0.248 -0.092 (a) 0.088 0.055 0.121 -0.097 -0.144 -0.144
RWP 1.00" 1.00" 0.013 0.099 -0.115 0.041 (a) 0.096 0.111 0.323" -0.189' 0.08 0.08
MRY 1.00" 0.013 0.099 -0.115 0.041 (a) 0.096 0.111 0.323" -0.189- 0.08 0.08
TRY 0.013 0.099 -0.115 0.041 (a) 0.096 0.111 0.323" -0.189" 0.08 0.08
RDM -0.08 -0.071 -0.104 (a) -0.122 -0.132 -0.032 -0.103 0.231* 0.231*
FL 0.168 -0.029 (a) 0.192’ 0.173 0.575" -0.118 -0.102 -0.102
FD 0.095 (a) 0.026 0.022 0.276" -0.12 -0.287" -0.287"
NST (a) 0.022 0.023 0.237* 0.206’ -0.047 -0.047
NL (a) (a) (a) (a) (a) (a)
NSL 0.934" 0.234' 0.117 0.051 0.051
NSF 0.247" 0.126 0.054 0.054
FW 0.117 -0.007 -0.007
TSW 0.076 0.076
NS 1.00"
NP
LA= leaf area, PM=powdery mildew, RNP=root number per plant, RWP=root weight per plant, MRY, TRY= marketable and total and root yield respectively, RDM= root dry matter,
FL=fruit length, FD=fruit diameter, NST=number of strips on a fruit, NL= number of locules per fruit, NSL= number of seed per locule, NSF= number of seed per fruit, FW= fruit weight,
TSW= thousand seed weight, NS=number of sepals, NP=number of petals
“ Significant at p <_0.01 and * at p £0.05

[41]
 Sebil Vol. 16

Correlation analysis was used to look into the patterns of variations of the
traits associations and their relation patterns. Leaf area showed positive
significant correlation with root weight per plant, root yield and fruit
weight. Root number per plant was not significantly correlated with the
root yield even though there was positive correlation; rather the root
weight per plant showed perfect positive significant correlation. Fruit
length was correlated significantly with number of seeds per locule and
the fruit weight but negatively correlated with the number of sepals and
petals. With increasing fruit length, the number of seeds and the weight of
the fruit increases. The numbers of strips on the fruit were strongly
correlated with the leaf area, the weight of the fruit, and the thousand seed
weight. For all of 36 accessions, the number of locules per fruit was equal;
six locules. The number of seed per locule in the fruit was positively and
significantly correlated with the fruit length, number of seed per fruit, and
the weight of the fruit. With increasing size of fruit, the number of seeds
within the fruit increases. The fruit weight showed perfect significant
positive correlation with the leaf area, root weight per plant, total root
yield, fruit length, number of strip line of the fruit surface, number of
seeds per locule and number of seeds per fruit.

V ariab ility estim ates

The quantitative traits considered in the study showed the presence
and/or absence of wider variability that exists in the tested accessions
(Table 10 and Table 11). The heritability (broad sense) value for root
number per plant was lower as compared to other traits but most of the
traits recoded higher value. This shows that anchote accessions tested in
this study had wider variability and they would be potential materials for
future study on the crop. Relatively, a very wide difference between PCV
and GCV was observed in root number per plant which indicated more
influence of the environmental factors in determining such trait.

On the other hand, slight differences between phenotypic and genotypic

variations were observed in the rest of the traits, which indicated genetic
variation existed in the expression of these characters and can be exploited
by selection (Karuri et al., 2010). Heritability (in broad sense) (Haynes et al.,
1995); all quantitative traits exhibited more than 81 % except the number of
roots per plant (Table 10).

[42]
 Sebil Vol. 16

Genetic advance (as percent of mean) was computed to compare the extent
of predicted genetic advance of different traits under selection since the
estimates of heritability and genetic advance should always be considered
simultaneously as high heritability is not always associated with high
genetic gain (Johnson et al., 1955). The higher estimated heritability values
or all of the quantitative traits except for number of roots per plant index
indicating that phenotypic selection for these traits could be highly
efficient. These results were in harmony with those obtained by Parmar
and Lai (2005) and Singh and Lai (2005) who reported the same higher
values of heritability in snake gourd genotypes of Egypt. This indicated
that selecting the top 5 percent of the base population could result in an
advance of 0 to 61.30 percent over the population mean (Table 11).
Comparatively high expected genetic advances were observed for root
weight per plant (kg), root yield (ton/ha), root dry matter (g), number of
seed per fruit, number of seed per locules, fruit weight (g), thousand seed
weight (g), fruit diameter (mm), root number per plant, leaf area (cm2),
fruit length (cm), number of sepals, and number of petals (Table 11).
Hence, selection for these characters is likely to be TYiore effective.

Table 10. Ranges, components of variance, mean and heritability (in broad sense) for quantitative
traits of anchote accessions

T raits Range C om ponents o f variance Mean H2

Phenotypic Genotypic
LA (cm2) 72.47-174.07 370.49 301.36 122.09 81.34
PM (%) 24.86-58.76 41.27 33.57 41.42 81.30
RNP 1.67-6.30 1.33 0.62 3.60 46.00
RWP(kg) 0.36-2.09 0.142 0.13 1.18 93.13
MRY(ton/ha) 22.71-130.83 56.67 52.77 73.51 93.00
RY(ton/ha) 22.71-130.83 56.67 52.77 73.51 93.00
RDM (%) 5.70-33.33 34.96 32.21 20.03 92.14
FL(cm) 5.00-9.03 1.14 0.84 6.86 73.00
FD(cm) 26.94-60.83 46.60 43.90 42.61 94.00
NST 8-11 0.22 0.22 9.88 100.00
NL 6 0.00 0.00 6.00 100.00
NSL 11-33 26.70 21.80 19.72 82.00
NSF 66-198 848.00 839.00 118.20 99.00
FW(g) 42.68-107.46 264.00 261.00 74.37 99.00
TSW(g) 22.87-47.75 49.00 46.60 34.26 95.00
NS 3.0-6.0 0.21 0.21 4.92 100.00
NP 3.0-6.0 0.21 0.21 4.92 100.00

[43]
 Sebil Vol. 16

Table 11. Phenotypic and genotypic coefficients of variation, genetic advance, and
genetic advance as percent of mean for 18 quantitative characters in 36 anchote accessions

Traits PCV GCV GA GAM

Leaf area 15.76 14.22 32.25 26.30
Powdery mildew 15.51 13.99 10.76 25.90
Root number per plant 32.03 21.83 1.11 30.40
Root weight per plant 32.01 30.89 0.72 61.30
M jrketable root yield 32.00 30.88 45.13 61.30
T )tal root yield 32.00 30.88 45.13 61.30
Foot dry matter 29.52 28.34 11.22 56.00
Fruit length 15.56 13.32 1.61 23.50
Fruit diameter 16.02 15.54 13.23 31.10
Number of strip lines on fruit 4.77 4.77 0.97 9.83
Number of locules 0.00 0.00 0.00 0.00
Number of seeds per locule 26.21 23.70 8.71 44.10
Number of seed per fruit 24.63 24.51 59.39 50.20
Fruit weight 21.84 21.70 33.04 44.40
Thousand seed weight 20.43 19.92 13.71 40.00
Number of sepals 9.22 9.22 0.93 19.00
Number of petals 9.22 9.22 0.93 19.00
Shannon Weaver Diversity Index (H’) and frequency classes

The Shannon Weaver Diversity Index (SWDI), was determined as a

measure of diversity in each accession (Table 13). Phenotypic classes and
percentage frequency for five quantitative traits ( number of locules,
number of seed per locule, number of seed per fruit, number of sepals and
petals, and number of strips on the surface of the fruit) were calculated (
Table 12) and there was a wide distribution in quantitative characters
among the accessions. The value of SWDI (H') ranged from null to
0.98(vine tip pubescence) and more than 65.95 % of the traits exhibited H'
values greater than 0.50. Generally, the high level of diversity observed
among anchote accessions in most of the traits evaluated, in this study
(Table 13) is in agreement with Gichimu et al. (2009) on morphological
diversity of watermelons of Kenya.
 Sebil Vol. 16

Table 12. Phenotypic classes and frequency (%) of five quantitative traits

C haracters Code/class N um ber of classes Frequency o f classes (% )

Number of locules per fruit 6 1 100.00
Number of seeds per locule 11 1 2.78
12 1 2.78
13 2 5.56
14 2 5.56
15 1 2.78
16 3 8.33
18 4 11.11
19 4 11.11
20 3 8.33
21 2 5.56
22 5 13.89
23 3 8.33
24 2 5.56
28 1 2.78
32 1 2.78
33 1 2.78
Number of seeds per fruit 66 1 2.78
72 1 2.78
78 2 5.56
84 2 5.56
90 1 2.78
96 3 8.33
108 4 11.11
114 4 11.11
120 2 5.56
121 1 2.78
126 2 5.56
132 6 16.67
138 2 5.56
144 2 5.56
168 1 2.78
192 1 2.78
198 1 2.78
Number of sepals and petals 3 1 2.78
4 2 5.56
5 32 88.89
6 1 2.78
Number of strips on the surface of 8 1 2.78
fruit 9 3 8.33
10 31 86.11
11 1 2.78

Table 13. Estimates of Shannon Weaver Diversity (H') Index based on six quantitative traits of 36 anchote accessions

Traits H' Values

Number of locules per fruit 0.00
Number of seeds per locule 0.93
Total number of seeds per fruit 0.94
Number of sepals and petals 0.33
Number of strip lines on the fruit surface 0.39

[45]
 Sebil Vol. 16

P rin eiiial (oni|M>iient Analysis

Principal component analysis is a multivariate technique for examining
relationships among several quantitative variables that are correlated
among each other by converting into uncorrelated characters (Crossa et al.,
1995). In Figure 2, the 36 anchote accessions grouped into two main
components, PC I and PC II and 16 quantitative traits evaluated (Table 14).
Accession number 223096 was the highest storage root yielding and was
outlier from other accessions.

The first principal component (PC I) explained about 53 % of the gross

accessions variance. About 31.58 % of the total variation accounted for by
PC I alone was due mainly to variations in leaf area, fruit weight, number
of seeds per fruit, and total root yield. Variations in total root yield
constituted a large part of the total variance explained by PC I. Likewise,
the second PC (PC II) accounting for about 24.99 % of the total variance of
the accessions originated mainly from fruit weight, number of seeds per
locule and number of seed per fruit. The two components: PCi and PC 2
explained a total of 78% variation. Some quantitative traits contributed
higher contribution in that total root yield, leaf area, number of seeds per
fruit, and fruit weight were larger contributors for PC]. Fruit diameter and
thousand seed weight contributed negatively to PCi and number of seeds
per locule and fruit, and fruit weight had the biggest contribution to PC 2.

The highest root yielding accession 223096, was located at the right end of
the graph and the lowest root yielder accession, GM, also located at the left
end even though not outlier (Figure 2 and Table 14). On the other hand, all
of the higher roots yielding accessions were separated from the rest on the
right upper side of the graph.

[46]
 Sehil Vol. J 6

100

223110
75

223087

50
DD
223105

223108
25
GIVI 90802
D IG G A 2 2 3 1 0 0
230566
KICHI 223101 223112 ;> 2 2 9 7 D 2 k u w e
223098 230565
223099,231 0 4
PC2 223088
223097
207984
223090 223113
90801
-25
. 223094 223092 223109
223085
NJ 2205£$)4o7
-50 223093 22309)

-75

I r

-75 -50 -25 25 50 75

-100 100

PC1
Figure 2. Principal Components for variation among 36 anchote accessions

[47]
 Sebil Vol. 16

Table 14. Eigenvectors and eigenvalues of the first two principal components of quantitative
traits of 36 anchote accessions

Trait Eigenvectors
PCi PC2
Leaf area (cm2) 0 .2 0 5 0 .0 8 8
Fruit length (cm) 0 .0 0 5 0 .0 0 7 8
Fruit diameter (mm) -0 .0 1 5 0 .0 1 9 8
Fruit weight (g) 0 .1 7 6 0 .1 4 6
Number of locules (no.) 0.000 0.000
Number of seeds per locule(no.) 0 .0 2 8 0 .1 5 8
Number of seeds per fruit (no.) 0 .1 6 8 0 .9 4 5
Thousand seed weight (g) -0 .0 2 2 0 .0 4 7
Number of strips line on the fruit surface (no.) 0.001 0.001
Root number per plant (no.) 0 .0 0 3 0.001
Root weight per plant(kg) 0 .0 0 8 -0 .0 0 2
Total root yield (tones/ha) 0 .4 9 4 -0 .1 1 5
Root dry matter (%) 0.0001 -0 .0 2 8
Downy mildew infestation (%) 0 .0 6 8 0 .0 2 9
Number of petals(no.) 0.001 0.001
Number of sepals (no.) 0.001 0.001
Eigen value 4 .1 2 0 2 .1 4 1
Percent of total variance expected 5 3 .0 5 2 4 .9 9
Cumulative percent of total variance explained 5 3 .0 5 7 8 .0 4

Sum m ary and Conclusion

This study is the first in characterizing the Ethiopian anchote accessions

for their diversity and trait inheritance. There is no adequate information
on the clonal variations in Coccinia abyssinia and there was no variety of
anchote developed so far. The attempts made so far to collect, conserve
and characterize C. abyssinia growing in different parts of the country, and
document the indigenous knowledge related to the use and management
of the crop is scant. Ethiopian Institute of Biodiversity Conservation has
made collection expeditions from different parts of the country mainly
Oromia, SNNP, Benishangul Gumuz and Amhara regional states.

The present study looked at the genetic variability and trait inheritance of
36 accessions o f anchote currently conserved ex situ in the Ethiopian gene
bank and at Debre Zeit Agricultural Research Center. The cluster analysis
grouped the accessions into four clusters. The analysis also showed that
only those accessions that had root yield of greater than 95 tons per hectare
were grouped in the same cluster, cluster II. Most of the accessions were
grouped in to different clusters regardless of the collection region. This
indicates that there is no significant relationship between phenotypic
diversity and geographical origins. One explanation could be that anchote

[48]
 Sebil Vol. 16

traditionally has been subjected to exchange over long distances. The

study also revealed that some of the accessions currently maintained in the
field gene bank might be duplicate accessions originating from the same
region. Thus, further research including molecular characterization need
to be carried out so as to verify the findings. The information generated in
this study, however, can be used to facilitate the conservation and
utilization o f the ex situ conserved accessions of anchote and further
studies on the crop. The study result also increases the awareness of
agronomists, breeders, nutritionists and producers about the diversified
uses of anchote as it is a high yielding crop with less management and
possibilities of developing better varieties with regard to yield, nutritional
quality, medicinal values and industrial uses.

Eighteen quantitative traits were considered and there was high variability
among accessions.
Further characterization should be done at molecular and biochemical
levels for the same set of accessions to ascertain the findings of this study.
In addition, this work demonstrated the importance of employing other
reliable methods such as DNA based markers to confirm the identified
traits and groups and to answer whether the groups are stable or have
■inks to other attributes.

R eferen ces

Abera Hora, 1995. ANCHOTE: An Endemic Tuber Crop. Jimma College of

Agriculture. Jimma, Oromia, Ethiopia.
Alice, K. and K. Peter, 1995. Association of fruit shape index and quality
characters in tomato. Vellanikkara, Kerala.
Amare Getahun, 1973. Developmental Anatomy of Tuber of Anchote; A
Potential Dry Land Crop. In: Godfrey-Sam-Aggrey,W. and Bereke
Tsehai Tuku (Eds.). Proceedings: First Ethiopian Horticultural
Workshop, Feb.20-23,1985, Vol.II. Pp.313-323. Addis Ababa, Ethiopia.
Amsalu Nebiyu, Weyessa Garedew, Assefa Tofu, Wubishet Abebe, Asfaw
Kifle and Edosssa Etisa, 2008. Variety development of taro, cassava,
yam, and indigenous root and tuber crops of Ethiopia. Pp.303-315. In:
Gebremedhin Woldegiorgis, Endale Gebre and Berga Lemaga (Eds).
Root and Tuber Crops: The Untapped Resources. EIAR, Addis Ababa,
Ethiopia.

[49]
 Sebil Vol. 16

Chipungahelo, S., A. Nigereza, P. Kawamala and T. Kwileka, 2007. Effects

of light regimes on different crops, sweet potato (Ipomoea batatas L.
Lam), cowpea ( Vigna unguiculata 1 walp) and pineapple (Ananas
comosus L. Merr.). African Crop Science Conference Proceedings, Vol. 8:
467-471.
Crossa, J., H. Deiacy and S. Taba, 1995. The use of multivariate methods in
developing a core collection. In: Hodgkm, J., Brown, A .H.D., Van
Hintum, Th.j.L. and Morals, E.A.V (eds.) Core Collections of plant
Genetic Resources, Pp.77-92.
ECPGR, 2008. Minimum descriptors for Cucurbita spp., cucumber, melon
and watermelon. Pp.1-15.
Elameen, A., A. Larsen, S. Klemasdal, S. Fjellheim, L. Sundheim, S. Msolla,
E. Asumba and O. Rognli, 2010. Phenotypic diversity of plant
morphological and root descriptor traits within a sweet potato,
Ipomoea batatas (L.) Lam., germplasm collection from Tanzania.
H ogskoleveien, Norway.
FAO, 2010. Food Comes First: FAO and the eight Millennium
Development Goals. Rome, Italy. Pp.1-21
FAO, 1990. Roots, tubers, plantains and bananas in human nutrition.
Rome, Italy.
Gichimu, B.M., B.O. Owuor, G.N. Mwai and M.M. Dida, 2009.
Morphological Characterization of Some Wild and Cultivated
Watermelon Accessions in Kenya. ARPN Journal o f Agricultural and
Biological Science, 4 (2). Nirobi, Kenya.
Gomez, K.A. and A.A. Gomez, 1984. Statistical Procedure for Agricultural
Research. 2nd ed. John Wiley and Sons.inc., New York.
Haynes, K.G., D.R. Wilson and M.S. Kang, 1995. Genotype x Environment
Interactions for Specific Gravity in Diploid Potatoes. Crop Science,
35(4):9 7 7 -981.
IPGRI, 2002. Descriptors for sweet potato. Rome, Italy. 52p.
IPGRI, 1993. Deversity and Development. The strategy of the international
plant genetic resources institute, IPGRI, Rome, Italy.
Jim, H., C. Martinez and N. Wyman, 2003. Estimating and Interpreting
Heritability for Plant Breeding: An Update. Estimating and
interpreting heritability for plant breeding Reviews. Vol.22. Internet
Document, http'.//Genotypic, phenotypic Heritability \ARS Publication
request Estimating and Interpreting Heritability fo r Plant Breeding An
Update.htm.

[50]
 Sebil Vol. 16

Johnson, H.W., H.F. Robinson and R.E. Comstock, 1955. Genotypic and
phenotypic Correlations in soybeans and their implications in
selection. Agronomy Journal, Vol.47:477-483.
Karuri, H.W., E.M. Ateka, R. Amata, A.B. Nyende, A.W.T. Muigai, E.
Mwasame and S.T. Gichuki, 2010. Evaluating diversity among Kenyan
sweet potato genotypes using morphological and SSR markers. Int. }.
Agric. B iol, 12: 33-38.
Koller, E., 2009. Javanese medicinal plants used in rural communities.
Wien, Java. 216p.
Lebot, V., 2009. Tropical root and tuber crops: Cassava, sweet potato, yam
and aroids. Cambridge, UK.
Parmar, A.M. and T.Lal, 2005. Variability studies in melon. Research on
crops 6 (2): 314-317, Hisar, India.
Singh, G. and T. Lai, 2005. Genetic variability, heritability and genetic
advance for yield and its contributing traits in muskmelon. Journal o f
research. Punjab agricultural Univ, 42(2): 168-174.
Solmaz, I. and N. Sari, 2008. Characterization of watermelon (Citrullus
lanatus) accessions collected from Turkey for morphological traits.
Genet Resour Crop Evol, Vol.56:173-188.
 Sebil Vol. 16

Performance of Cassava Clones Under Potential

and Low Moisture Stressed Areas of Ethiopia
T esfaye T ad e sse1, A tn afu a B ekele1, E ngida T seg ay e1, G etachew W /M ich ael2, T ew odros
M u lualem 2, W u b sh et B e sh ir3, M esele G 1, S hiferaw M 1
1Hawassa Agricultural Research Center, 2Jimmu Agricultural Research Center.
3= Sekota Research center. E-mail: tesfayet2@yahoo.com

A bstract
Cassava is one o f the most important food crops that constitute a considerable portion o f the daily
diet of the people and also serves as one of the major source o f carbohydrate. Despite its importance,
production of cassava in Ethiopia has different constraints as well as opportunities. Among which
shortage o f improved varieties is the first and the most important one. It is mainly cultivated by
small holder resource poor farmers on small plots o f land. Average storage root yield obtained is as
low as 100 quintals per hectare despite the potential yield o f 600 quintals per hectare per year. This
low yield might be due to the cultivation o f local, low yielding, and late maturing cultivars. To
contribute to the alleviation o f the problem and provide farmers other alternative varieties,
Hawassa agricultural research center, in collaboration with Jima and Sekota agricultural research
centers, conducted evaluation o f cassava clones for potential and moisture stressed agroclimatic
conditions o f the country. Jima and Hawassa sites represent potential agroclimatic condition while
Amaro and Sekota represent moisture stressed areas. A total o f seven cassava clones namely AWC-
1 (MM 96/5280), AWC-2(MM 90/5280), AWC-3 (MM 96/7151), AWC-4 (MM96/1871), AWC-
5(MM96/3868) and Kello (standard check) were evaluated by using randomized complete block
design in three replications. The evaluation was carried out for two consecutive years from 2012 to
2014. The combined analysis result indicated that there ivas statistically significant difference
among the clones tested and the experimental locations. The clone AWC-1 (368.3 q/ha) gave the
highest storage root yield followed by AWC-2(351.1 q/ha) and AWC-5 (339.5 q/ha) but there was
no statistically significant difference on the total storage yield among the clones AWC-2, AWC-3
and AWC-5. In the same way, the highest dry matter content was recorded from the clones AWC-
2 (51.8%), AWC-3 (48.5%) and Kello (49.1%). Among the locations tested, the best result ivas
obtained from Amaro (351.4 q/ha), which is characterized by its low moisture stress indicating that
cassava can resist/tolerate low moisture stress and gave comparative yield provided that other
factors are not limiting . Thus, those clones with the highest storage root yield and dry matter
content were promoted to variety verification and will be released for wider production.

Key w ords: Cassava, Moisture Stress, Potential, Storage Root, Dry Matter

Introduction

Cassava is a monoecious perennial shrub having variable height ranging

between 1 and 5m, although maximum height usually does not exceed 3m
(Bernardo and Heman, 2012). But it is extensively cultivated as an annual

[53]
 Sebil Vol. 16

crop in tropical and subtropical regions for its edible starchy tuberous
roots (MoC, 2014). Cassava is a very important food crop in the tropics,
that is, at latitudes of 30 degrees and from sea level to 1800 meter above
sea level. Although the principal economic products are its roots, cassava
leaves are also extensively being used in Africa and Asia, as human food
or animal feed. Cassava is the fourth most important commodity after rice,
wheat and maize, and is a basic diet of many millions of people (FAO and
IFAD, 2000). In addition to the economic value of the products and
byproducts obtained from cassava, it offers other recognized advantages:
tolerance of drought, capacity to produce considerable yield in degraded
soil, resistance to insect pests and diseases, tolerance of acid soils (which
are predominant in most of the world's tropical plains), and flexibility in
planting and harvesting time (Bernardo and Hernan, 2012).

Despite its enormous production potential such as, adaptation to diverse

environments, recognized tolerance to biotic and abiotic constraints, and
diversity of uses, cassava has not yet managed to fully develop its
potential in tropical agriculture due to numerous factors. Among the
factors that constrained the production of cassava is lack of early
maturing, high yielding, low hydrogen cyanide containing varieties. In
Ethiopia, it is mainly cultivated by small holder resource poor farmers on
small plots of land. It is both a food security crop and a source of
household income. It is increasingly becoming a source of industrial raw
material for production of starch, ethanol, waxy starch, bio-plastics,
glucose, bakery and confectionery products, glue among others (Tesfaye et
al., 2013). Cassava produces bulky storage roots with a heavy
concentration of carbohydrates, about 80 percent. The shoots grow into
leaves that constitute a good vegetable rich in proteins, vitamins and
minerals. New knowledge of the biochemistry of the crop has proved that
the proteins embedded in the leaves are equal in quality to the protein in
egg. Cassava leaves and roots, if properly processed, can therefore provide
a balanced diet protecting millions of African children against
malnutrition.

According to FAO estimates, 276,721,584 tons of cassava were produced

worldwide in 2013. Africa accounted for 57%, Asia for 32%, and others for
11% of the total world production. In 2013, Nigeria produced 54 million
tones making it the world's largest producer followed by Thailand,
Indonesia and Brazil with 30.2, 23. 94 and 21.23 million tons respectively.

[54]
 Sebil Vol. 16

In terms of area harvested, a total of 20,732,192 hectares was planted with

cassava throughout the world in 2013; about 64% of this was in sub-
Saharan Africa. The average yield in this year was 11.3 tons per hectare,
but this varied from 1.3 tons per hectare in Burkina Faso to 35 tons per
hectare in India. In the largest producer, Nigeria, the average yield was 14
tons per hectare (FAOStat, 2013).

The average total c o v e ra g e and production of cassava per annum in

Southern region of Ethiopia is 4,942 hectares with the yield of 53,036.2
tones indicating the average productivity of cassava in the country is not
more than 10 ton per hectare (SNNPR, Bo A, 2000), which is by far lower
than the world average (11.3) and other tropical countries such as India,
China, Brazil and Nigeria which recorded a yield of 35.00, 24.6,14.00 and
14.03 tons per hectare per year, respectively. The yield is still lower than
that of the east African countries like Kenya (15.8 t/ha) and Uganda
(12.2t/ha) (FAO State, 2013)

In Ethiopia, cassava generally can be grown/is being grown in almost all

parts of the country. But bulk of its production is situated in south, south
western and western parts of the country. Most of the varieties produced
were local farmers' varieties which are low yielding, late maturing, bitter
type and containing high hydrogen cyanide (Anshebo, et al., 2004). To
alleviate these problems a number of research activities focusing on crop
variety improvement were conducted in different federal and regional
research centers that led to the release of two out performing varieties in
2005 (MoA, 2005). But the varieties were late maturing and the numbers
were low to provide additional alternative to the farmers and increase
genetic diversity. Hence evaluation of seven cassava clones including
standard and a local check were conducted at different agroclimatic
condition of the country. As a result, promising varieties with regard to
storage root yield per a given period from a given area of land were
obtained. Therefore, this paper is aimed to show the performance of
cassava clones under different agro ecological conditions of the country.

M aterials and M ethods

Evaluation of cassava clones for their storage root yield and other
agronomic traits was conducted in four location of the country namely
Hawassa, Amaro, Jima and Sekota. Two of the locations (Hawassa and

[55]
 Sebil Vol. 16

Jima) are classified as potential areas for the production of cassava

whereas the other two arbitrary classified as low moisture stressed dry
land areas. The overall description of the locations is given in the table
below.
Table 1: Mean annual temperature, rainfall, altitude, longitude and latitude of Hawassa, Amaro, Sekota and Jima

Locations Mean annual Temptarure °C Mean annual Altitude Longitude Latitude

Minimum Maximum rainfall (mm) (masl)
Hawassa 7.2 33.5 1024.2 1708 38° 28’ 34"E 7° 3’ 43" N
Amaro 12 25 800 1477 37°32"1 O': 38°E 5° 3" 55’: 60N
Jima* 11.3 26 1597 1753 36° E 7046'
Sekota 28 40 474.5 1300 38° 58' 50" E 13° 14' 06' N
* Amsalu Nebyu, 2006.

A total of seven cassava clones (five introduced, one standard check, and
one local farmer's variety) were tested in the experiment. The treatments
were arranged in randomized complete block design with three
replications and conducted for two consecutive years, 2012-2014 except
Jima where only one season data was availed. Gross and net plot size
where the experimental units were planted were 4 m x 6 m and 2 m x 4 m,
respectively. Storage root yield and other yield related data such as root
length, root girth, number of roots per plants and growth rate were taken
from the net plot at harvesting except the growth rate which was taken in
three months interval since planting. Data on the root length, root girth,
number of roots per plants and growth rate were taken from randomly
selected five plants. Whereas the storage root yield data was taken from
each plot in kilogram and converted in to yield per hectare in quintal by
using the following formula:
Yield per hectare = yield per plot (kg) X 10000 m2
8 m2 X 100 (kg/q)

Note that 1 quintal is equals to 100 kg. The dry matter content of the clones
were taken after oven dry for 24 hours at 110°C for consecutive dates until
the weight gets constant. The clones were planted by using lm x 1m plant
and row spacing. The spacing between plots were 2 m whereas the space
between reps were 3 m. Before carrying out the combined analysis,
homogeneity of variances test for total storage root yield across locations
and years was conducted by using Levene's, Welch's and Bartlet's tests.
The collected data were analyzed by using SAS statistical software, 2002
version 9.0 and IRRISTAT statistical soft ware.

[56]
 Sebil Vol. 16

R esu lts

Cassava clones performed differently at different environmental

conditions. But the combined analysis of variance table indicated that
there was highly statistically significant difference among clones (P<0.01)
for their root girth (RD), root length (RL), marketable and total storage root
yield. There was also statistically significant difference among clones for
their number of storage roots per plant (RNPP). But there was no
statistically significant difference among the clones for their unmarketable
yield (table 2). The test for equality of variances showed no significant
difference for all Levene's, Weltch's and bartlett's test at p value <0.05
(table 3)

Table 2: Analysis of variance

Variable df SS MS MSe F value P r> F

RD 6 26.3908779 4.3984796 0.7283698 6.04** <.0001
RL 6 982.246564 163.707761 52.22382 3.13** 0.0079
MRYQ 6 172106.9671 28684.4945 6295.768 4.56** 0.0005
MRYT 6 1721.069671 286.844945 62.95768 4.56** 0.0005
UNRYQ 6 9826.8046 1637.8008 987.5142 1.66NS 0.1411
UNRYT 6 98.268046 16.378008 9.875142 1.66 NS 0.1411
TRYQ 6 199418.9061 33236.4844 9276.680 3.58** 0.0032
TRYT 6 1994.189061 3 32.364844 92.76680 3.58** 0.0032
RNPP 6 73.1054296 12.1842383 4.048105 3.01* 0.0102
*= significant at a value less than 0.05, **= highly significant at a value less than 0.01
RD= average root girth (cm), RL=root length (cm), MRYQ= Marketable root yield per hectare (quintals), MRYT=
Marketable root yield per hectare (tones), UNRYQ= Unmarketable root yield per hectare (quintals), UNRYT =
Unmarketable root yield per hectare (tone), TRYQ= Total root yield per hectare (quintals), TRYT= Marketable root yield
per hectare (tone), RNPP=Average number of roots per plant.

Table 3: Test for homogeneity o f variances

Levene's Test for TRYQ

Mean
Source DF Sum of Squares Square F Value P r> F
trt 6 3.82E+09 6.37E+08 1.48 0.189
Error 134 5.77E+10 4.30E+08
Welch's ANOVA for TRYQ
Source DF F Value P r> F
trt 6 1.91 0.0941
Error 57.64
BARTLETT TEST FOR EQUALITY OF VARIANCES TRYQ
CHISQ PVALUE
8.6778491 0.1925226

[57]
 Sebil Vol. 16

Storage root yield and yield related components

As far as storage root performance is concerned, the highest marketable
and total storage yield was obtained from the clone AWC-1 followed by
AWC-2 and AWC-5. The least score was recorded from the local farmers'
cassava variety followed by kello (the standard check) and AWC-4 in the
increasing order. The local varieties shown the largest storage root
diameter but there was no statistically significant difference among the
clones for this particular trait. The cassava variety Kello followed by the
clone AWC-5 and AWC-2 gave the highest storage root length, 40.67, 38.25
and 37.77 cm, respectively. The highest number of roots per plant was
recorded from the clone AWC-2 next to AWC-5 (table 4).

Table 4: Storage root yield of cassava clones combined over location

Cassava Marketable Unmarketable storage Total root RD (cm) RL (cm) RNPP

clones storage root yield root yield (q/ha) storage yield
(q/ha) (q/ha)
AWC-1 307.5 60.82 368.3 4.831 32.69 6.879
AWC-2 299.8 56.06 351.1 4.524 37.77 8.115
AWC-3 275.4 43.95 319.3 4.826 33.81 7.873
AWC-4 254.9 33.95 288.9 4.751 35.55 6.517
AWC-5 287.2 52.42 339.5 4.557 38.25 8.515
Kello 227.5 53.41 281 4.756 40.67 7.45
Local 199.1 52.06 251.1 6.059 38.26 6.59
CV(% ) 29.69 62.45 30.40 17.59 19.71 26.99
LSD 50.05 NS 60.75 0.5383 4.55 1.27
Note: Marketable storage roots =those roots weighting 100-500gms. Unmarketable storage root yield (q/ha)=those roots
weighting more than 500gm and less than 100gms, miss shaped, infected by disease and infested by insect pests, rotten,
Total storage root yield=the sum of marketable and unmarketable root yields.

All cassava clones recorded higher yield advantage over both the local and
standard checks. The clone AWC-1 showed 47 and 31% yield advantage
over the local and standard checks followed by the clone AWC-2, which
recorded a yield advantage of 40 and 25% over the local and standard
checks, respectively. However, the yield advantage of the clone AWC-4
over the local and the standard checks was very minimal. Similarly, the
yield o f the standard check was better than that of the local farmer's
varieties (Local Check). The yield advantage of the standard check over
the local check was 12% (Fig. 1).

[58]
 Sebil Vol. 16

casava clones

Fig.1: Yield advantage of cassava clones over the standard and local checks

The combined analysis result of the dry matter content of storage root of
cassava clones under investigation showed the presence of statistically
significant differences. The clone AWC-2 gave the highest dry matter
content followed by the clones AWC-3 and the standard check (Kello)
(Table 3). The dry matter content also varied with the locations. At
Hawassa there was no statistically significant difference among cassava
clones tested as opposed to what was observed at Amaro for the same
clones (Fig. 2).
Table 5: Percent dry matter content of cassava clones tested across locations

Cassava clones Hawassa Amaro Mean

AWC-2 50.2 53.4 51.8
AWC-3 47.0 50.0 48.5
Kello 46.5 51.6 49.1
Lcheck 46.2 50.4 48.3
AWC-5 45.1 29.9 37.5

AWC-4 45.0 43.5 44.3

AWC-1 44.2 48.4 46.3
LSD NS* 20.3 10.5

CV 10.58 24.45 18.9

*NS= Non significant

[59]
 Sebil Vol. 16

C h aracters’ association
The correlation coefficient of most of the traits indicates positive and
significant association among each other with some exceptions. Cassava
storage root diameter significantly correlated with root length (r=0.24),
marketable root yield (r=0.50) and total storage root yield (r=53). But not
significantly correlated with unmarketable storage root yield and number
of roots per plant. Root length also showed significant positive correlation
with marketable and total storage root yields with r value of 0.40 and 0.25,
respectively. To the contrary, root length showed significant negative
correlation with unmarketable storage root yield (r=-0.34) and no
significant correlation with number of roots per plant (r= -0.1). Even
though marketable and unmarketable storage root yields were not
statistically correlated with each other, both of them were significantly and
positively correlated with total storage root yield with correlation
coefficient of 0.92 and 0.41, respectively. The total storage root yield also
showed significant positive correlation with root numbers per plant with
r=0.416. It was also positively and significantly correlated with leaf yield
per plant. Root weight per plant also showed significant positive
correlation with dry matter content and leaf yield per plant (Table 6).

Table 6: Pearson Correlation Coefficients of storage root yield and yield component of cassava clones

Variables RD RL MRYQ UNRYQ TRYQ RNPP RW PP DM LYPP

1 0.34337 0.5053 0.08437 0.492 0.089 0.557 0.08 0.377
RD 0.0036 <.0001 0.4874 <.0001 0.465 <.0001 0.508 0.001
1 0.4251 -0.40108 0.235 -0.113 0.072 -0.481 0.547
RL 0.0002 0.0006 0.05 0.35 0.551 <0001 <.0001
MRYQ 1 0.03987 0.926 0.267 0.486 -0.288 0.503

[60]
 Sebil Vol. 16

0.7431 <.0001 0.025 <.0001 0.016 <.0001

1 0.415 0.457 0.337 0.652 -0.41
UNRYQ 4E-04 <.0001 0.004 <.0001 4E-04
1 0.416 0.57 -0.016 0.303
TRYQ 3E-04 <.0001 0.899 0.011
1 0.508 0.046 -0.145
RNPP <.0001 0.704 0.231
1 0.312 0.037
RWPP 0.009 0.76
1 -0.514
DM <.0001
1
LYPP

Yield stability o f cassava clones across locations

As indicated in table 7, the performance of cassava clones tested across
locations varied with agroclimatic conditions. The clone AWC-1 gave the
highest total storage yield (quintals)/hectare at Hawassa and Amaro
locations. At Sekota, the other clone AWC-2 gave the highest storage root
yield while at Jima the clone AWC-5 followed by the clone AWC-2 gave
the highest yield (Table 7). With regard to the performance across location,
the best average mean root yield of the two seasons indicated that the
location Amaro is the best area for cassava production followed by
Hawassa. The root yield at Jima is higher but since the experiment was
only for one season it was not included in the comparison. The Additive
Main effects and Multiplicative Interaction (AMMI) stability analysis of
seven genotypes on seven environment also indicated the existence of the
variability of cassava clones performance under different environmental
conditions. The clones AWC-1 and AWC-5 seems to have possessed wider
adaptability as they are found closer to the origin of the plot. To the
contrary, farmer varieties (local check) were able to adapt only to specific
environmental conditions as it was found far away from the origin of the
plot.

[61]
 Sebil Vol. 16

The interaction principal component axis (IPCA) score also indicated the
stability of a clone across environments. The more the IPCA approximate
zero, the more stable the clone is over all the environments tested.
According to IPCA1 of table 7, clone AWC-1 and AWC-5 had
approximately zero score (0.1 and 0.27, respectively) and hence could be
considered as most stable clones.

[62]
 Sebil Vol. 16

Table 7: Mean total storage yield (quintal/ha) clone ranks in seven environments
IPCA1 IPCA2
Environments SCORE SCORE
Hawassa Hawassa Ra TRT
Clone name 1 Rank 2 Rank Amarol nk Amaro2 Rank Jimal Rank Sekota 1 Rank Sekota2 Rank MEANS Rank
AWC-1 460.4 1 456.2 1 416.7 1 444.6 3 351.2 5 238 3 210.8 3 368.3 1 -0.10 0.10

AWC-2 299.2 5 367.5 3 2892 3 507.9 1 478.9 2 241 2 274.3 1 351.1 2 0.49 -0.75

AWC-3 320.8 3 327.9 5 257 9 5 474.6 2 453.1 3 190.4 5 210.7 4 319.3 4 0.37 0.81

AWC-4 281.2 6 250.4 6 269.2 4 410.8 4 352.1 4 232.3 4 2259 2 288 9 5 0.92 -0.55

AWC-5 352.1 2 416.7 2 236.7 6 396.2 6 531.3 1 247 1 196.2 6 339.4 3 0.27 0.11

Kello 318.8 4 347.9 4 326.2 2 302.9 7 285.9 7 182.6 6 202.3 6 281 6 0.69 -0.46

Lcheck 156 2 7 161.7 7 179.2 7 407.5 5 351 6 251.1 7 -0.88 -0.98

Environment mean 312.67 332.61 282.16 42064 400.50 22188 220 03 31416
IPCA1 -0.67 -0.55 -0.70 0.71 0.10 0.95 0.12
IPCA2 025 07 0 -053 -0.56 0.64 -0.83 -0.41

[63]
 Sebil Vol. 16

Table: analysis of variance for the AMMI model

SOURCE • D.F. S.S. M.S. F FPROB

TREATMENTS 6 98987.1 16497.8
LOCATIONS 6 277624. 46270.6
TREATMENT X SITES 34 123804. 3641.29
AMMI COMPONENT 1 11 78078.0 7098.00 3.570 0.005
AMMI COMPONENT 2 9 30019.5 3335.50 2.973 0.033
AMMI COMPONENT 3 7 10469.3 1495.61 1.999 0.190
AMMI COMPONENT 4 5 3374.80 674.960 0.725 0.666
GXE RESIDUAL 2 1862.41
TOTAL 46 467901.

Discussion

Storage nx>t yield and yield related components

The total storage yield obtained from the clone AWC-1 (371.1 q/ha), AWC-2 (329.8
q/ha) and AWC-5 (307.5 q/ha) observed in the current study was by far higher than
the yield obtained from most of cassava growing countries in the world in general and
East Africa in particular. FAOstat (2013) indicated that the average yield obtained from
India, China, Brazil and Nigeria was 349.59, 245.45, 139.15 and 140.26 quintals per
hectare per year. Average storage root yield obtained from East African countries such
as Kenya, Uganda and Tanzania in 2013 was 158.92, 120.18, 75.00 quintals per hectare,
respectively. Differences in cassava tuber yield are determined by several factors, such
as number of tubers, tuber length and tuber weight per plant. Ntawuruhunga and
Dixon, (2010) concluded that storage root number, storage root size and storage root
diameter were the main yield components contributing to yield enhancement in
cassava. As far as yield related traits were concerned, the value obtained in the current
study was in agreement with the report o f Kenneth, (2011). In his study, the highest
storage root length obtained from the variety Cuban White Stick was 40.46 cm. In the
same way, the highest storage root number per plant was obtained from the variety
John LaMotte (7.78), which is similar to the value recorded from the current study
(8.515).

Characters' association
The total storage root yield showed significant positive correlation with root numbers
per plant, storage root length, storage root weight/plant, leaf yield per plant but
negatively correlated with dry matter content This shows that those traits which are
positively and significantly correlated with storage root yield were important
components of yield across locations. The findings of the current study is also in line
with the report of Ntawuruhunga et al., (2001). They indicated that storage root weight
(r=0.53) and storage root number (r=0.45) as the main component of total yield per a
given area per a given time. Dry matter content was negatively correlated with storage
root weight, suggesting that when the storage root weight is high, the dry matter
content tends to be low, which is in line with the study conducted by Kenneth, (2011).

[64]
 Sebil Vol. 16

Yiclfl stability of cassava clones across locations

The Additive Main effects and Multiplicative Interaction (AMMI) and the IPCA scores
indicated the clones AWC-1 and AWC-5 seems to have possessed wider adaptability as
they are found closer to the origin of the plot. To the contrary, farmer varieties (local
checks) were able to adapt to specific environmental conditions,which is far away from
the origin of the plot and larger absolute value scores of IPCA. This indicates that the
clones, AWC-1 and AWC-5 are not affected by environmental conditions. In the same
wry, a variety performance trial conducted in Indonesia by using 15 genotypes at
di'ferent range of altitudes, some of the clones' storage root yield was stable across
locations. The clone Malang 4 (G3) and CMM 03038-7 (G8) are adaptive clones to
environment at medium altitude up to 800 masl (Kartika Noerwijatia and, Rohmad
Budionob, 2015). Kartica Noerwijati, (2014) indicated that environment gives the most
effect (64.69%), followed by genotype-by- environment interaction effect (6.53%), and
genotype effect (4.94%) on performance of a given genotype of cassava. He also
indicated the most stable cassava genotype by using that GGE biplot with high yield
which is in line with this study showing the possibility of obtaining the most stable
varieties across locations.

Conclusion and Recom m endation

The clone A w e-1 gave the highest yield, followed by AWC-2 compared with standard
and local checks. The clone AWC-1 gave a yield advantage of 47 and 31% over the local
and the standard check followed by the clone AWC-2. Three clones (AWC-1, AWC-2
and AWC-3) showed no significant difference in terms of dry matter content (more than
50%). Those clones having stable and higher root yield combined with higher dry
matter content were proposed for variety verification trial for a wider dissemination as
wrell as production. One of the most important problem of cassava production is the
lack of early maturing varieties. In Ethiopia, cassava generally grows in almost all parts
of the country. But bulk of its production situated in south, south western and western
parts of the country. Most of the varieties produced were local farmers' varieties, which
are low yielding, late maturing, bitter type and containing high hydrogen cyanide. The
existing improved and farmer's varieties take more than 18 months for full maturity.
Therefore, continues breeding and selection program is required so as to fill the gap due
to the production of late maturing varieties.

Acknowledgement

The authors would like to thank South Agricultural Research Institute, Hawassa
Agricultural Research Center for facilitating cassava research works. Our
acknowledgment extends to East African, Agricultural Productivity Program for
funding cassava research activities in general and cassava variety development in

165]
 Sebil Vol. 16

particular. The authors also kindly acknowledge Mr. Ayalneh Tilahun and Dr. Agdew
Bekeie for supporting in stability analysis by using IRR1STAT software.

R eferen ces

Amsalu N (2006). Phenotypic Diversity of Cassava (Manihot esculenta Cranz.) in

Ethiopia. In: Kebebew Asefa and Lijalem Korbu eds. Proceedings of the 12th
Annual Conference of the Crop Science Society of Ethiopia, 22-24 May 2006, Addis
Ababa, Vol. 12, Pp 23-29.
Anshebo T., A. Tofu, E. Tsegaye, A. Kifle and Y. Dagne, 2004. New cassava varieties for
tropical semi arid climate of Ethiopia. In: proceedings of the 9th ISTRC-AB
symposium 2004, pp. 526-530 N airobi, Kenya.
Bernardo Ospina and Hernan Ceballos, 2012. Cassava in the third millennium: Modern
production, processing, use and Marketing Systems. CIAT publication No. 377,
574pp, CIAT, Colombia.
Food and Agricultural Organization (FAO) STAT, 2013. Statistical data base of the food
and agricultural organization of the united nations accessed from
http://faostat.org on August 10, 2015.
FAO and IFAD (Food and Agriculture organization of the United nations and
International Fund for agricultural development), 2000. The world Economy of
cassava: Facts, Trends and outlook. Rome, Italy, 59pp.
1RRISTAT (international rice research institute statistical soft ware), 2007. For windows
version 7, Metro Manila, Philippines.
Kartika Noerwijatia and , Rohmad Budionob, 2015. Yield and Yield Components
Evaluation of Cassava (Manihot esculenta Crantz) Clones in Different Altitudes.
Energy Procedia, Conference and Exhibition Indonesia - New and Renewable
Energy and Energy Conservation (The 3rd Indo EBTKE-ConEx 2014), Vol. 65, Pp.
155-161
Kartica Noerwijati, 2014. Fresh Tuber Yield Stability Analysis of Fifteen Cassava
Genotypes Across Five Environments in East Java (Indonesia) Using GGE Biplot -
Research Gate. Available from:
http://www.researchgate.net/publication/260030897_Fresh_Tuber_Yield_Stabilit
y_Analysis_of_Fifteen_Cassava_Genotypes_Across_Five_Environments_in_EastJ[
ava_(lndonesia)_Using_GGEJBiplot [accessed Aug 10, 2015].
Kenneth VA Richardson, 2011. Evaluation of three cassava varieties for tuber quality
and yield. Gladstone road agricultural centre crop research report no. 4, 12pp.
MOA (Ministry of Agriculture), 2013. Plant variety release, protection and seed quality
control directorate, Crop variety register issue No. 16, Addis Ababa, Ethiopia
MoC (Ministry of commerce of the people's republic of China), 2014. Reference material
for china aid training program. The 2014 seminar on cassava industry
development for English speaking countries in Africa, Beijing, China, 351pp.

[66 ]
 Sebil Vol. 16

Ntawuruhunga, P. and Dixon, A. (2010). Quantitative variation and interrelationship

between factors influencing cassava yield. Journal of Applied Biosciences Vol., 26,
pp.594-1602.
Ntawuruhunga P., P.R. Rubaihayo, J.B.A., Whyte, A.G.O. Dixon and D.S.O. , 2001.
Inter-relationships among traits and path analysis for yield components of cassava:
a search for storage root yield indicators. African crop science journal vol. 9, No. 4,
pp. 599-606.
SAS Institute Inc., 2002. SAS Software Version 9.00, NC, USA.
SNNPR BoA (Southern Nationalities and Peoples Region, Bureau of Agriculture), 2014.
Basic Agricultural Information Planning and programming Service, Hawassa,
Ethiopia. Unpublished row data.
Southern Nationalities and Peoples Region, Bureau of Agriculture (SNNPR, BoA), 2000.
Basic Agricultural Information Planning and programming Service, Hawassa,
Ethiopia.
Tesfaye Tadesse, Getahun Degu, Ermias Shonga, Shiferaw Mekonen, Temesgene Addis
and Birhanu Yakob, 2013. Current status, Potentials and challenges of Cassava
production, processing, marketing and utilization: Evidence from Southern
Ethiopia. Greener Journal of Agricultural Sciences, Vol. 3 (4), pp. 262-270.

[67]
 Sebil Vol. 16

Com bining Ability Studies for Yield and Yield _

Com ponents in Selected Soybean I^ines

T adesse G hiday, "Sentayehu Alam irew. 'A snake Fikre, 'T izazu D., 'M olla M.,
A sm am aw A., 'G ezahegn T.
Pawe Agricultural Research Center. Ethiopia. P.O.Box 25. Pa we. Ethiopia.
E-mail: txhidav20 12(a:gmuil. com
Jimma University College o f Agriculture and Veterinary Medicine
' Ethiopian Institute o f Agricultural Research

A bstract

C om bining ability is m ostly used by breeders to select appropriate parental cultivars to produce the
larger progeny o f new com binations through their hybridization. The objectives o f this research
w ere to estim ate general com bining abilitx/ o f parents an d specific com bining abilities o f FI
hybrids, to identify suitable parents and hybrids f o r yield a n d its contributing traits. In this study,
18 FI hybrids obtain ed by crossing six lines (Ethiopia) ivith three testers (Brazil) in line * tester
m ating system during 2013/14 and were planted in random ized com plete block design with two
replications during 2014/15. A m ong the lines, BELESA 95 proved to be a good general com biner
fo r num ber o f pods per plant, grain yield and pod weight. M ajority o f the best specific com bination
f o r differen t characters resulted fro m the crosses am on g the parents w ith high x low and low x lozv
GCA effects. N on-additive effects w ere predom inant for num ber o f pods p er plant an d grain yield,
but appreciable additive effects w ere noted fo r hundred seed weight, pod w eight and prim ary
branches. The breeding m ethod w hich can exploit n on-additive as well as additive types o f gene
action is su ggested for soybean im provem ent.

K ey w ords: Soybean, GCA, SCA

1ntroduction

in self-pollinated crops like soybean (Glycine max (L.) Merrill),

recombination breeding has been extensively used to develop the
variability reservoir for exploitation in breeding program (Yang et al.,
2009). In a systematic breeding program, it is essential to identify the elite
parents for hybridization, and superior crosses to expand the variability
reservoirs for selection of superior genotypes (Sharma et al., 2007).
Combining ability studies help in such endeavor (Darwish et al., 2007). In
the present investigation, line x tester design with well adapted and
widely grown varieties of soybean (tester) was used to obtain information
on combining ability of elite lines for five characters of economic
importance in soybean.

[6 9 ]
 Sebil Vol. 16

M aterial and M ethods

o

Nine cultivar lines were selected on the basis of their geographical

adaptation and morphological diversity. Out of them, three were testers
(males), each crossed with six genotypes used as lines (females). All the
testers (PSB2005-03, PSB2005-04 and PSB2005-06) were obtained from
Brazil as introduced varieties (which are currently under production in
Brazil). All the six lines (GIZO, AFGAT, GISHAMA, BELESA-95,
ETHIOUGOSLVIA, and WOGAYEN) are high yielding Ethiopian varieties
in the Southern West and Northern West parts of the country released for
commercial production.

Eighteen FIs along with nine parents were planted in randomized

complete block design with two replications during 2014/2015 at
experimental station of Pawe Agricultural Research Center. The parents
were randomized among themselves. Each replication have single row of 5
meters length at 60cm x 5cm spacing. The mean data was recorded on ten
random plants for five quantitative characters number of pods per plant,
grain yield, hundred seed weight, pod weight and number of primary
branches that was used for statistical analysis. The combining ability was
done adapting Kempthorne (1957) and Kapila et al., (1994) procedure. The
total variance among FI hybrids was further partitioned into variance due
to lines, testers and interaction component, which was used to estimate the
additive and non-additive components of variance. Also, the contribution
of lines, testers and their interaction towards total variability for each
character was computed for assessing their relative importance (Singh and
Chaunhary, 1977).

R esu lt and D iscussion

Analysis of variance indicated the presence of significant difference among
the treatments for all the five characters studied (Table 1). The parents
differed significantly for all the characters except grain yield; however,
mean squares due to lines were significant only for hundred seed weight
and pod weight. The testers differed significantly for number of pods per
plant, grain yield and hundred seed weight. The hybrids showed
significant differences only for pod weight. Further partitioning of
variance among the hybrids showed that the mean square due to lines

[70]
 Sebil Vol. 17

were significant for hundred seed weight and pod weight, due to testers
for number of pods per plant, grain yield and hundred seed weight, and
number of pods per plant in case of line x tester interactions (Table 1).

Table 1: Analysis of variance for five characters in Soybean at Pawe on station (2014/2015).

Source of variation Number of Grain Hundred seed Pod No. of primary

d.f pods per plant yield weight weight branches
Replication 1 239** 63.3* 693.9** 0.05 41.1**
Treatments 26 71.7** 26.2* 127** 0.04** 3.3**
Parents 8 110.1** 24.4 307.1** 0.05** 8.0**
Parents vs crosses 1 190.4** 157.3** 0.3 0.1 0.1
Hybrids 17 50.8 21.9 55.6 0.03* 1.4
Lines 5 59.1 23.1 83.7* 0.07** 1.9
Testers 2 140.4* 63.6* 208.1** 0.02 1.1
Lines x testers 10 35.5** 16.1 22.5 0.02 1.1
error 26 29.3 15.1 43.6 0.02 1.7
*, "significant at 5% and 1% levels, respectively

Table 2: Proportional contribution (%) to total variance for five characters in Soybean at Pawe on station (2014/2015).

Number of Grain yield Hundred seed Pod weight No. of primary

Source of variation pods per plant weight branches
Lines 35.8 32.5 46.3 63.7 43.6
Testers 21.3 22.4 28.8 5.6 6.1
Lines x testers 42.9 45.1 24.9 30.7 50.3

This was also clearly illustrated when the proportional contribution of

each character studied. Lines and their interaction with testers contributed
more than 70% of the total variance for all the characters. Except for
number of pods per plant (21.3%), grain yield (22.4%) and hundred seed
weight (28.8%), the contribution of testers was very little. The contribution
of lines varied from 32.5% for grain vield to 63.7% for pod weight (Table
2).

V arian ce com ponents

The estimate of variance components (GCA and SC A) indicated that non­
additive components were dominant for number o f p o d s per plant and
grain yield characters, though appreciable additive effects were noted for
huncired seed weight, pod weight and primary branches (Table 3), as
reported earlier (Mamta at el., 2010 and shanty et al.,2008) .These
observations suggest that in soybean breeding, the methodology that can
exploit both the additive as well as non-additive effects would be of
immense value. Diallel selective mating (Shanti et al., 2003 and Sharma et

[71]
 Sebil Vol. 16

a l, 2007), which provides better opportunity for recombination,

accumulation of desirable genes and selection would help in concentrating
most of such genes in a pure line (Bastawisy et al.,1977). A judicious
integration of the classical approach (Pedigree and Bulk) with diallel
selective mating may be of great help in achieving the quantum jump in
soybean improvement (El Sayad et al., 2005, Sharma et al., 2007 and Sajata
et al., 2011)

Table 3: Estimates of variance components for five characters in soybean at Pawe on station (2014/2015).

Source of Number of Grain Hundred Pod weight No. of primary

variation pods per plant yield seed weight branches
GCA 0.48 0.18 1.02 0.0001 0.08
SCA 3.07 0.47 -0.69 0.0001 -0.05
GCA/SCA 0.15 0.38 -1.47 1.000 -1.60

The estimates of GCA effects (Table 4) showed that among lines and
testers, BELESA 95 was superior, as it showed positive and significant
GCA effects for number of pods per plant and grain yield, beside pod
weight. The lines ETHIOUGOSLAVIA and GIZO were good combiners for
hundred seed weight (Table 4). None of the testers was found to be a good
general combiner. Association between per se performance and GCA
effects was not evident in the present study. In fact, in many cases, the
lines or testers with high mean had low GCA effects, indicating the
ineffectiveness of choice of parents based on per se performance for
hybridization (Table 5).

Table 4: General combining ability effect of the lines and testers for five characters in soybean at Pawe on station
(2014/2015).
Source of variation Number of pods Grain Hundred seed Pod weight No. of primary
per plant yield weiqht branches
Lines
AFGAT 2.47 0.89 0.57 0.08 0.41
WOGAYEN 1.11 0.48 0.55 -0.01 0.33
ETHIOUGOSLAVIA 4 .4 3 -1.67 4.27* -0.03 0.93
GIZO -0.49 -0.64 6.48* -0.16* -0.03
BELESA 95 5.89* 4.24* 2.00 0.11* -0.37
GISHAMA 2.47 0.89 0.57 0.08 0.41
SE± 2.21 1.59 2.70 0.05 0.53
SE(gi-gj) 3.13 2.25 3.81 0.07 0.75
Testers
PSB2005-03 1.48 1.30 3.07 -0.04 -0.24
PSB2005-06 1.74 0.86 0.59 0.03 -0.02
PSB2005-04 -3.22* -2.16* -3.66* 0.01 0.26
SE± 1.28 0.92 1.56 0.03 0.30
SE(gi-gj) 1.8 1.30 2.20 0.04 0.43
‘ Significant at 5% level

[72]
 Sebil Vol. 17

Fourteen out of eighteen crosses occupied the first five ranks for five
characters (Table 5). The four top ranking for grain yield proportionate to
the order of number of pods per plants indicates the close association
between number of pods per plant and grain yield. Of these fourteen
crosses, eleven crosses were between low x low, two crosses between high
x low and only one cross involved high x high gca parents. The largest
number of low x low and high x low gca crosses in top ranks for different
characters is of great interest, as such combinations could result in
desirable transgressive segregants if the additive effects of one parent and
the complementary epistatsic effects (present in the cross) act in the same
direction and m ax im ize the expression of plant attributes under selection.
Table 5: Specific combining ability of the best five crosses based on per se performances at Pawe on station (2014/2015).

Gca STATUS OF
Sea
CHARACTER CROSS MEAN PARENT
EFFECT
P1 P2
NUMBER OF BELESA 95 X PSB2005-03 44.2 6.57 High Low
PODS PER PLANT ETHIUGOSLAVIA X PSB2005-06 39.5 5.92 Low Low
BELESA 95 X PSB2005-06 37.2 -0.69 High Low
AFGAT X PSB2005-03 37.2 2.99 Low Low
G IZ O X PSB2005-06 34.1 2.59 Low Low
GRAIN YIELD (g) BELESA 95 X PSB2005-03 30.0 4.82 High Low
ETHIUGOSLAVIA X PSB2005-06 24.5 3.18 Low Low
BELESA 95 X PSB2005-06 24.3 2.42 High Low
AFGAT X PSB2005-03 22.4 -2.35 low Low
WOGAYEN X PSB2005-06 22.3 1.28 Low Low
HUNDRED SEEDS WOGAYEN X PSB2005-03 44.5 4.48 Low Low
WEIGHT BELESA 95 X PSB2005-03 42.0 2.80 Low High
(g) G IZ O X PSB2005-03 41.1 -2.80 High High
G IZ O X PSB2005-06 40.5 3.59 Low Low
G IZ O X PSB2005-04 40.2 -1.12 Low Low
POD WEIGHT (g) ETHIOUGOSLAVIA X PSB2005-06 1.1 0.07 Low Low
BELESA 95 X PSB2005-03 1.1 0.09 High Low
WOGAYEN X PSB2005-03 1.0 0.04 Low Low
AFGAT X PSB2005-06 1.0 0.03 Low Low
W OGAYEN X PSB2005-03 1.0 0.07 Low Low
PRIMARY ETHIOUGOSLAVIA X PSB2005-03 7.3 1.32 Low Low
BRANCHES GISH AM AX PSB2005-03 7.3 0.75 Low Low
G IZO X PSB2005-04 7.2 1.18 Low Low
AFGATX PSB2005-04 7.0 0.61 Low Low
WOGAYEN X PSB2005-04 6.7 -0.34 Low Low

[73]
 Sebil VoL 16

Conclusion
Two crosses Blessa 95 x PSB2005-03 (high x low) and Ethio-yugoslavia x
PSB2005-06 (low x low), exhibited high mean values for number of pods
per plant and grain yield. These crosses may be further exploited for
isolating the desirable segregates for economic traits of number of pods
per plant and grain yield.
i
Acknowledgement
The authors greatly thank the Ethiopian Institute of Agricultural Research
for financial support. Pawe Agricultural Research Center, staff members of
the Soybean Research Team is greatly acknowledged for their active and
unlimited cooperation in the execution of the experiment.

References
Bastawisy, M. H.; Ibrahim, A. M. and Mansours, S. H. 1997. Combining
ability and heterosis studies for yield and its components in some top
crosses of soybean [Glycine Max (L.)] Merril. Annals of Agricultural
Sciences, Moshtohor. 35(1): 93-106.
Darwish, I. H. I. 2007. Heterosis and enheritance of some quantitative
character in soybean (Glycine Max L.). Egyptean Journal of Plant
Breeding. 11(1): 131-142.
El Sayad Z. S.; Seliman, M. M.; Mokhtar, S. A.; El Shaboury, H. M. G. and
ElHafez, G. A. A. 2005. Heterosis, combining ability and gene action in
FI and F2 diallel crosses among six soybean. Genotypes. 43(2): 545-
559.
Kampthrone, 0 . 1957. An introduction to genetic statistics. Johnwiley and
Sons, Inc., New York. Kapila, R. K.; Gupta, V. P. and Rathore, P. K.
1994.
Mamta - Arya; Kamendra Singh; Push Pendra and gupta, M. K. 2010.
Heterosis study for yield, its components and quality characters in
soybean [Glycine Max (L.) Merril]. Soybean Research. 8: 75-79.
Shanti Patil, Vandana Khambalkar; Khedikar, Y. P. and Menna Wankheda.
2003. Study of FI crosses of soybean. Journal of Soils and Crops. 13(1):
112-115.

[74]
 Sebil Vol. 17

Shanti Patil; Vandana Kalamkar; Sable, N. H. and Maheshwari, J. J. 2008.

Study of heterosis in soybean. Journal of Soil and Crops. 18(1): 208-
211 .
Sharma, B.; Singh, B. V.; Kamendra Singh; Pushpendra; Gupta, M. K. 2007.
Selection criteria for improvement of grain yield in soybean [Glycine
Max (L.) Merrill]. 5(1): 43-44.
Singh,R.K. and Choudhary, B.D.1977.Line x Tester Analysis In :
Biometrical methods in quantitative genetic analysis Kalyani
Publishers, New Delhi. ppl78-85.
Sujata Bhat; Basavaraja, G. T and Salimath, P. M. 2011. Studies on genetic
variability in segregating generation of soybean [Glycine Max (L.)
MerrilJ. Crop Research, Hisar. 42(1/2/3): 251-254.
Yang - Jia Yin and Gai Jun Yi. 2009a. Heterosis, combing ability and thin
genetic basis of yield among key parental materials of soybean in
Huang - Huai valleys. Acta. Agronomica Sinica. 35(4): 620-630.

[75]
 Sebil Vol. 16

Genetic Divergence Analysis on Some Soybean

(G ly c in e m a x L. Merrill) Genotypes Grown in
Pawe, Ethiopia
T adesse G h id a y 1a n d S en ta y eh u A lam irew 2
'Pawe Research Center. P.O.Box 25. Pawe. Ethiopia; E-mail: Ivludav20/2 ale/nail.com
2 College o f Agriculture and Veterinary Medicine. Jimma University>

A bstract
A total o f 4 9 soy bean genotypes xvere evaluated to assess genetic diversity fo r grain yield and
yield related traits. D -square statistics (D 2) has been used to classify the divergent genotypes
into different groups. The gen otypes w ere evaluated fo r 13 characters and show ed m oderate
zwriability fo r the com ponents studied. The cluster an alysis grou ped the 49 soybean genotypes
into fiv e different clusters. This indicates the presence o f m oderate diversity am ong the tested
genotypes. From clu ster mean values, genotypes in clu ster III and V deseri’e consideration fo r
their direct use as parents in hybridization program s to develop high yielding soybean varieties.
The results o f the principal com ponent analysis revealed that fiv e principal com ponents (PC I to
PC5) accou nted nearly fo r 79.06% o f the total variation. The differentiation o f the gen otypes into
different clusters was because o f relatively high contribution w ithin the first principal
com ponents such as num ber o f pods p er plant, biological yield, grain yield p er p lot and grain
yield per plant. Therefore, the above m entioned characters w hich load high positive contribution
more to the diversity and they w ere the ones tluit m ost differentiated the clusters. It w as also
noted that differentiation o f getioh/pes into differen t clusters urns because o f the small
contribution o f fexv characters rather than the cum ulative effect o f a num ber o f characters. The
inform ation obtained from this study can be used to plan crosses an d m axim ize the use o f gen etic
diversity an d expression ofheterosis.

K ey w ord s: Soybean, hybridization, genotypes, genetic divergence,

Ethiopia

In tro d u ctio n

Soybean [Glycine max (L.) Merrill] is the most important vegetable food
sources in the world. In Ethiopia, soybean is an introduced crop and had
a higher expansion of cultivated area in recent years, with a crop
production of 636531.01quintal of harvest with an average of
productivity 19.98 quintal per hectare in 2012/2013 cropping season
(FAO, 2012). National average yield is very low compared with its
potential, and yields obtained in other soybean producing countries. It is
largely grown in the lowlands of the country and constitutes roughly 2-
3% of the annual pulse production and plays an appreciable role in

[77 ]
 Sebil Vol. 16

human nutrition and health, edible oil, livestock feed and many other
industrial and pharmaceutical applications (CSA, 2012).

For a successful breeding program, the presence of genetic diversity and

variability play a vital role. Genetic diversity is essential to meet the
diversified goals of plant breeding such as breeding for increasing yield,
wider adaptation, desirable quality and pest disease resistance (Gemech
et al., 2012). Genetic divergence analysis estimates the extent of diversity
existed among selected genotypes (Mahalanobis, 1936). Precise
information on the nature and degree of genetic diversity helps the plant
breeder in choosing the diverse parents for purposeful hybridization.

Genetic diversity plays an important role in plant breeding either to

exploit heterosis or generate productive recombinants. The choice of
parents is of paramount importance in breeding programs. Thus, the
knowledge of genetic diversity and relatedness in the germplasm is a
pre-requisite for crop improvement programs. Reduction in the genetic
variability makes the crops increasingly vulnerable to diseases and
adverse climatic changes. So, precise information on the nature and
degree of genetic diversity present in soybean introductions from
principal areas of cultivation would help to select parents for evolving
superior varieties. The aim of this study was to identify genetically
divergent soybean parents with desirable traits for hybridization
particularly for yield.

V latcrials and M ethods

Forty-nine soybeans nationally released and introduced varieties were

used in the experiment (Table 1). The experiment was conducted at Pawe
Agricultural Research Center (PARC), North West Ethiopia from 2013 to
2014 summer seasons in a simple lattice design with two replications,
each plot with four rows of 0.40m width and 5 m row length. Sow ing was
done by hand drilling at a seed rate of 70 kg/ha. The spacing between
plots and replication were 0.40 m and 1 m, respectively. Di ammonium
phosphate (DAP) fertilizer was applied at the rate of 100 kg/ha. All the
cultural practices were performed as recommended. The plant data
during the cropping season and after harvesting were noted.
Observations recorded on a plot basis included days to flowering, days to
m aturity, grain filling period, 100 seed weight, biological yield per plot,

[78 ]
 Sebil Vol. 16

seed yield per plot and harvest index per plot. The seed yield per plot
and biological yield per plot were measured by taking a net plot of 0.8 m
x 5 m or 4 m2 and this was used to determine harvest index. All other
characters were recorded on a single plant basis by randomly taking five
plants from each experimental plot. Climatic conditions for the cropping
seasons have been indicated in Figures 1 and 2.

Table 1: Forty-nine soybean genotypes of different crosses (hybrids), nationally released and introduced varieties used
in this study (2013/2014).

Entry Entry
Genotype Source Year Genotype Source Year
No No
1 AFGAT (TGX-1892-1 OF) IITA/Nigeria 2007 26 TGX-1987-34F IITA/Nigeria 2007
2 AWASSA 95 (G 2261) USA 2005 27 TGX-1987-11F IITA/Nigeria 2007
3 BLACK-HAWK USA 2005 28 TGX-1740-2F IITA/Nigeria 2007
4 CLARK 63K USA 2005 29 TGX-1987-9F IITA/Nigeria 2007
5 PROTONA-2 USA 2005 30 TGX-1987-23F IITA/Nigeria 2007
6 NYALA Turkey 2003 31 TGX-1987-64F IITA/Nigeria 2011
7 ETHIO-YOGOSLAVIA USA 2005 32 TGX-1987-62F IITA/Nigeria 2011
8 WOGAYEN USA 2005 33 TGX-1987-15F IITA/Nigeria 2011
9 GIZO USA 2005 34 TGX-1986-3F IITA/Nigeria 2011
10 GISHAMA USA 2005 35 TGX-1987-35F USA 2011
I 11 AGS 7-1 USA 2005 36 TGX-1987-19F IITA/Nigeria 2011
12 NOVA USA 2005 37 TGX-1935-10E IITA/Nigeria 2011
13 WELLO USA 2005 38 TGX-1987-40F IITA/Nigeria 2011
14 GOZILLA USA 2005 39 TGX-1987-38F IITA /N igeria 2011
15 EAZ-3600 USA 2005 40 TGX-1987-37F IITA/Nigeria 2011
16 BELESSA 95 (PR-149) USA 2005 41 TGX-1987-14F IITA/Nigeria 2011
17 LOTTUS USA 2005 42 TGX-1987-1 OF IITA/Nigeria 2011
18 PARC-1 USA 2005 43 TGX-1987-65F IITA/Nigeria 2011
19 PARC-2 USA 2005 44 CROWFORD IITA/Nigeria 2007
20 PARC-3 USA 2005 45 WILLIAMS IITA/Nigeria 2007
21 PARC-4 USA 2005 46 COCKER-240 IITA/Nigeria 2007
22 PARC-5 USA 2005 47 BOSHE IITA/Nigeria 2007
23 PARC-6 USA 2005 48 JALELE IITA/Nigeria 2007
24 TGX-1987-18F IITA/Nigeria 2007 49 TGX-1989-59F IITA/Nigeria 2007
25 TGX-1987-20F IITA/Nigeria 2007
Source: Pawe Research Center 2011

[79 ]
 Sebif Vol. 16

Figure 1, Monthly total rain fall (mm) of Pawe Research Center, 2013 and 2014

Statistical analysis
The statistical package SAS version 9.2 was used for genetic divergence
calculation, Cluster mean analysis and principal component analysis
'SAS Institute, 2008).

[80 ]
 Sebil Vol. 16

R esu lts and D iseussion

Mean squares o f the 13 characters from analysis of variance (ANOVA)

are presented in Table 2. Significant differences among genotypes
(P<0.001) were observed among the 49 genotypes for 13 of the traits
studied, indicating the presence of inherent variation among the
materials. Desirable genes from this germplasm can effectively be
utilized to develop high performing pure line varieties after crossing.
These results are agreement with earlier findings of Banger et al. (2000)
on soybean.

G en etic divergence
Differences in morphological and quantitative traits have been
considered as simple indicator of genetic variability in crop species and
varieties. Divergence analysis is a technique used to categorize genotypes
that are similar as possible into one group and the other into different. D-
square statistics (D2) developed by Mahalanobis (1936), has been used to
classify the divergent genotypes into different groups. The extent of
diversity present between genotypes determines the extent of
improvement gained through selection and hybridization. The more
divergent the two genotypes are the more will be the probability of
improving through selection and hybridization.

The D-square statistics (D2) resulted in classifying the 49 soybean

genotypes into five distinct clusters (Table 3 and Fig. 3). This indicates
the presence of moderate diversity among the tested genotypes. The
cluster analysis based on the pooled mean of genotypes resulted in
classifying the 49 genotypes into four groups and one solitary (Table 3).
This indicates that tested soybean genotypes were moderately divergent.
The chi-test for the five clusters indicated that there was statistically
accepted difference between clusters (Table 4). The genotypes were
distributed (Table 3) in such a way that 24 genotypes were grouped into
Cluster-I (49.98%), 14 genotypes in to Cluster-II (28.57%), 3 genotypes
into Cluster-111 (6.12%), 7 genotypes into Cluster-IV (14.29%) and one
genotype into Cluster-V (2.01%). Cluster-I contains moderate value of
characters. Cluster-II contains high m ean number of d ay s to m aturity
(116.90days), stand count at harvest (97) and plant height (71.76 cm).
Cluster-111 contains high days to flowering (64.17days), number of seeds
per pod (2.62) and biological yield 1972.87). In Cluster-Ill there were

[81 ]
 Sebil Vol. 16

characters such as grain filling period (44.5 days), hundred seed weight
(11.43g) and harvest index (19.44 %) of lower value. Cluster-IV contains
high grain filling period (61.14days) and hundred seeds weight (12.97g).
Cluster IV contains low number of branches per plant (2.23), number of
seeds per pod (2.34), biological yield (527.99g), seed yield per plot
(166.6g), number of pods per plant (12.35) and seed yield per plant
(2.15g). Cluster V contains high number of branches per plant (3.45),
number of pods per plant (33.00), harvest index (39.51) and grain yield
per plant (7.51). In Cluster V there were characters such as days to flower
(49.5days), days to maturity (96days), plant height (44.6cm) and stand
count at harvest (75.5) of lower value. The maximum inter cluster was
between Cluster-Ill and V (252.6) followed by Cluster III and IV (250.71)
and Cluster II and V (220.89) (Table 2).The minimum being Cluster I and
IV (16.72) followed by Cluster I and II (23.67). Generally, this study
showed that the genotypes included in this study are moderately
divergent. Therefore, the results of the distance of genotypes between
clusters has shown that there is a room for the genetic improvement of
soybean varieties and the information generated can be used to plan wide
crosses, to exploit genetic diversity and maximize the expression of hetrosis
[61.

Cluster III and V, Cluster III and IV and Cluster II and V exhibited the
greatest inter cluster divergence from all other cluster in this study.
According to Gemechu and Ghaderi [5,7], increasing parental distance
implies a great number of contrasting alleles at the desired loci, and to
the extent that these loci recombine in the F 2 and F 3 generation following
a cross of distantly related parents, the greater will be the opportunities
for the effective selection for yield factors. Thus, crossing of genotypes
from these clusters with other clusters may produce higher amount of
hetrotic expression in the first filial generations (Fi's) and wide range of
variability in subsequent segregating (F2) populations. Thus, crosses
involving cluster III and V with any other cluster is suggested to exhibit
high heterosis and could result in segregates with higher seed yield, i.e.
transgressive segregation.

[8 2 ]
 Sebil Vol. 16

Table 3: The distribution of genotypes into 5 clusters based on D ' analysis for 49 soybean genotypes tested at Pawe
(2013 and 2014).

No of Percentage
Cluster Genotypes (%) Genotypes
GIZO, GISHAMA, TGX1987-11F, TGX1987-15F, TGX1987-64F,
TGX1987-14F, TGX1987-62F, NYALA, TGX1740-2F, LOTUS, TGX1935-
24 48.98 10E, AWASSA-95, WEGAYEN, GOZELLA, CROWFORD, TGX1989-59F,
I
BLACK HAWK PARC-3, TGX1986-3F, PARC-5, PROTONA-2, TGX1987-
35F, TGX1987-18F and PARC-2
WELLO, PARC-4, TGX1987-20F, TGX1987-9F, PARC-1, JAKELE,
14 28.57 TGX1987-34F, TGX1987-38F, GISHAMA, NOVA, TGX 13-3-2644, AGS 7-
II
1, PARC-6 and TGX1987-1 OF
III 3 6.12 TGX1987-37F. TGX1987-10F and EAZ-3600
IV TGX1987-23F, WILLIAMS, TGX1987-19F, COKER-240, CLARK-63K
7 14.29
,BOSHE and AFGAT
V BELESA-95
1 2.01

J.*
! *1*
r
9
J
P l. i
|%1
I
I
I %9

tP

f*

f l. i
*

Hint «f lkl(ir||l|i »> ( I m I w

Figure 3. Figure showing the clusters to which the genotypes belong and average distance between clusters (2013
and 2014).

Principal component analysis (PCA) reflects the importance of the largest

contributor to the total variation at each axis of differentiation (Sharma,
2005). The Eigen values are often used to determine how many factors to
retain. The sum of the Eigen values is usually equal to the number of
variables. Therefore, in this analysis the first factor retains the information

[83 ]
 Sebil Vol. 16

contained in 3.341 of the original variables. The principal components of

these data are given in Table 5. Five principal components PCI to PC5
which are extracted from the original data and having latent roots greater
than one accounted nearly 79.06% of the total variation (Table 5),
suggesting that these principal component scores might be used to
summarize the original 13 variables in any further analysis of the data. Out
of the total principal components retained, PCI, PC2, PC3, PC4 and PC5
with values of 23.86, 20.74, 14.23%, 11.78, and 8.45%, respectively
contributed more to the total variation. According to Chahal and Gosal
(2002) characters with largest absolute value closer to unity within the first
principal component influence the clustering more than those with lower
absolute value closer to zero. Therefore, in the present study,
differentiation of the genotypes into different clusters was because of
relatively high contribution of few characters rather than small contribution
from each character.
Table 4: Mean value of 13 characters for the 5 clusters of 49 soybean genotypes tested at Pawe (2013 and 2014)

Character Cluster I Cluster II Cluster III Cluster IV Cluster V

DF(days) 61.08 61.04 64.17** 54.79 49.5*
DM(days) 113.77 116.90** 108.67 115.93 96*
PH(cm) 58.60 71.76** 66.43 55.11 44.6*
GFP(days) 52.69 55.86 44.5* 61.14** 46.5
NBP(N°) 2.35 2.64 2.78 2.23* 3.45**
NPP(N°) 15.49 20.50 22.18 12.35* 33**
NSP(N°) 2.41 2.52 2.67** 2.34* 2.55
BY(g) 859.94 1356.4 1972.87** 527.99* 1434.57
GYP(g) 264.99 367.77 386.98 166.6* 566.8**
SCAH(N°) 91.06 97** 83.5 81.57 75.5*
HSW(g) 12.62 11.79 11.43* 12.97** 12.6
Hl(%) 30.52 27.33 19.44* 31.07 39.51**
GY(g) 2.89 3.80 4.77 2.15* 7.51**
* and **low and high value of a trait respectively
DF= days to flowering, DM~days to maturity, GFP= grain filling period, PH=plant height. NBP=number o f branches per
plant. NPP= num ber o f pods per plant, NSF*= number o f seeds per plant, BY= biological yield. GYP= grain yield plot.
SCAH= stand count at harvest, HSW= hundred seed weight. Hl= harvest index and G Y -grain yield per plant

[84 ]
 Sebil Vol. 16

Table 4: Pair wise generalized squared distance (D2) among 49 soybean genotypes in five clusters at Pawe (2013 and
2014)

Cluster 1 II III IV V
1 23.67169* 163.7231" 16.72047ns 181.2203**
II 82.2458** 75.68003** 220.8855**
III 250.7112** 252.604**
IV 172.9535**
V

NB: x2= 21.03 and 26.22 at 5%, 1% probability level, respectively, “ significant and highly significant at5% and1%
probability level respectively

Principal com ponent analysis

In Table 5, the first principal component had high positive component
loading from grain yield per plot (0.501), number of pods per plant (0.473),
grain yield (0.471) and biological yield (0.350). The positive loading shows
the presence of positive correlation trends between the components and
the variables. Therefore, the above mentioned characters which load high
positive contributed more to the diversity and they were the ones that
most differentiated the clusters. The major contributing characters for the
diversity in the second principal component (PC2) have high positive
component loading from plant height (0.439), stand count at harvest
(0.403) and days to flowering (0.341). The major contributing characters for
the diversity in the third principal component (PC3) had high positive
component loading from grain filling period (0.504), harvest index (0.453),
hundred seed weight (0.351) and days to maturity (0.320); and negative
loading from biological yield (-0.356) and number of seeds per pod (-
0.326). In principal component four (PC4) high positive component
loading from days to maturity (0.349), number of branches per plant
(0.339) and grain filling period (0.308) and high negative loading from
stand count at harvest (-0.474). In principal component five (PC5) high
positive component loading from hundred seed weight (0.543) and
number of seeds per pod (0.514) and high negative loading from number
of branches per plant (-0.352) and harvest index (-0.341). The positive and
negative loading shows the presence of positive and negative correlation
trends between the components and the variables. Therefore, the above
mentioned characters which load high positively or negatively contributed
more to the diversity and they were the ones that most differentiated the
clusters.
Usually it is customary to choose one variable from these identified
groups. Hence, for the first group grain yield per plot (0.501 ) is best
choice, which had the largest loading from component ones, plant height
[85]
 Sebil Vol. 16

(0.439) for the second, grain filling period (0.504) for the third group, stand
count at harvest (-0.474) for the fourth group and hundred seed weight
(0.543) for the fifth group.
Table 5: Eigen vectors and Eigen values of the first five principal components of 49 soybeans genotypes evaluated at
Pawe (2013 and 2014)
Principal component analysis
Trait PC 1 PC 2 PC 3 PC 4 PC 5
DF -0.042 0.341 -0.223 0.139 -0.239
DM -0.011 0.423 0.320 0.349 0.080
GFP 0.014 0.264 0.504 0.308 0.242
PH 0.178 0.439 0.010 0.153 0.014
NBP 0.226 -0.008 -0.098 0.339 -0.352
NPP 0.473 -0.135 0.066 0.073 0.024
NSP 0.076 -0.139 -0.326 0.156 0.514
BY 0.350 0.101 -0.356 0.149 0.272
GYP 0.501 -0.020 0.0335 -0.092 0.007
SCAH 0.145 0.403 -0.069 -0.474 0.032
HSW 0.002 -0.130 0.351 -0.217 0.543
HI 0.227 -0.166 0.453 -0.241 -0.341
GY 0.471 -0.182 0.060 0.078 -0.039
Eigen Value 3.340632 2.903523 1.992596 1.648593 1.T83332
Difference 0.437109 0.910926 0.344003 0.465261 0.216324
Proportion 0.2386 0.2074 0.1423 0.1178 0.0845
Cumulative 0.2386 0.446 0.5883 0.7061 0.7906

D F - days to flowering, O F-days to maturity, G FP- grain filling period, PH=plant height, NBP=number o f branches p e r
plant, NPP= number o f pods per plant, NSP= number o f seeds per plant, BY= biological yield, GYP= grain yield plot,
SCAH= stand count at harvest, HSW= hundred seed weight, H I- harvest index and GY=grain yield per plant

vSiiininai^ «iii(l Conclusion

There were differences in the performance of the genotypes as there were

statistically supported significant differences among genotypes for most of
the 13 characters and relatively wide range of the mean values for most of
the characters. Nevertheless, the level of the genetic differences for many
traits, including grain yield, may not be sufficient to expect progress in
selection. Therefore, in order to improve the diversity of soybean in
Ethiopia, subsequent crossing program aimed at developing soyb ean
varieties of better diversity by crossing between highly divergent varieties
needs to be carried out. The cluster analysis based on D2 analysis on
pooled mean of genotypes classified the 49 genotypes into five clusters,
which makes them to be moderately divergent. There were statistically
significant differences between all of the clusters, except between cluster I
and IV.

[8 6 ]
 Sebil Vol. 16

The principal component analysis extracted five principal components

PCI to PC5 from the original data and having Eigen value greater than one
accounting nearly 79.06% of the total variation. Characters with largest
absolute value closer to unity within the first principal component such as
number of pods per plant, biological yield, grain yield per plot and grain
yield per plant influence the clustering. The differentiation of the
genotypes into different clusters was because of relatively high
contribution of these characters. Therefore, the above mentioned
characters which load high positive contributed more to the diversity and
they were the ones that most differentiated the clusters. The present
investigation provided considerable information useful in genetic
improvement of soybean. Genotype grouped into cluster V showed
maximum inter cluster diversity. From cluster mean values, genotypes in
cluster III and V deserve consideration for their direct use as parents in
hybridization programs to develop high yielding soybean varieties. There
is significant genetic variability among tested genotypes that indicates the
presence of better opportunity to bring about improvement through wide
hybridization by crossing genotypes in different clusters. Further studies
on the soybean genotypes have to be tested for more locations and seasons
to recommend highly performed ones. More genotypes and more number
of characters have to be included for sound recommendation. Crossing of
genotypes from distant clusters enables to have more variability for
desirable traits for improvement. Influence of environment and an
agronomic practice on the genetic potential of the varieties in different
soybean environments is necessary. This is helpful to stratify the
environments based on quality and yield suitability. Generally, the
development of soybean varieties possessing higher grain yield, higher
protein and oil content is important.

Recom m endation

Further studies on the soybean genotypes have to be tested for more

locations and seasons to recommend highly performed ones. More
genotypes and more number of characters have to be included for sound
recommendation. Crossing of genotypes from distant clusters enables to
have more variability for desirable traits for improvement. Influence of
environment and an agronomic practice on the genetic potential of the
varieties in different soybean environment is necessary. This is helpful to
stratify the environments based on quality and yield suitability. Generally,

[87 ]
 Sebil Vol. 16

the development of soybean varieties possessing higher grain yield, higher

protein and oil content is important.

R eferen ces

[1] Bangar, N.D., Mukhekar, G.R., Lad, D.B & Mukhekar,D.G, 2000.
Genetic variability, correlation and regression studies in soybean, /.
Mah. Agric. Univ., 28(3): 320-321.
[2] Chahal, G.S.,& Gosa), S.S.2002. Principles and Procedures o f Plant Breeding:
Biotechnology and conventional approaches. Narosa Publishing House,
New Delhi.
[3] CSA. 2012. Agricultural Sample Survey Statistical Bulletins. Central
Statistical Authority, Addis Ababa, Ethiopia.
[4] FAO. 2012. FAOSTAT, FAO statistical data bases-agriculture
(available at http:// apps. fao. org.). http://faostat.fao.org/
FAOSTAT
[5] Gemechu, K., Musa, J., Tezera, W. & G etnet, D. 2005. Extent and pattern
of genetic diversity for morph-agronomic traits in Ethiopian highland
pulse landraces: I. Field Pea (Pisum sativum L.). Genetic Resources and
Crop Evolution (2005)52:539-549
[6] Gemechu, K., Endashaw, B., Mohammed, A.„ Kifle, D., Emana, G. &
Fassil A. 2012. Genetic diversity and population structure of Ethiopian
Chick Pea (Cicer arietinum L.) Germplasm accessions from different
geographic origins as revealed by microsatellite markers. Plant Mol.
Boil. Rep., 30, 654-665.
[7] Ghaderi, A., Adams, M.W., Nassib, A.M. 1984. Relationship between
genetic distance and heterosis for yield and morphological traits in
dry edible bean and faba bean, Crop Sci., 24, 37-42.
[8] Mahalanobis, P.CV 1936, On the generalized distance in statistics. In:
Proc. of the Nation. Acad. Sci., (India), 2:49-5[9] Ramigiry, S.R., 1999,
Genetic divergence in soybean. Madras Agricultural Journal. 8, 3,
5167-770
[10] Sharma, S.S. 2005, Genetic divergence in Indian varieties of soybean.
Soybean Research, 3, 9-16
[11] Sharma, J.R. 1998. Statistical and Biometrical Techniques in Plant
Breeding. New Age International (P) Limited Publishers, New Delhi.

[8 8 ]
 Sebil Vol. 16

Overlap in tlie Genetic Basis of* Host Basal and

Nonhost Resistances of Barley to Leaf Rusts
Alio Anian Dido
Oromia Agricultural Research Institute, Adami Tullu Agricultural Research Center. Plant Biotechnology
Research Team. P. O. Box 35, Batu (Ziway), Oromia, Ethiopia. E-mail: ulloam an20l tXagmail. com

Abstract
Barley host basal resistance (partial resistance) and nonhost basal resistance (nonhost resistance)
to leaf rusts are based on prehaustorial mechanism associated with papillae formation. Both are
mainly governed by genes with relatively small, quantitative effects, located on quantitative trait
loci (QTL). The genes for host basal resistance seem to play similar roles in basal resistance as
those governing non host basal resistances. We presume that these two resistance types are based
on shared principles. Quantitative trait loci-near isogenic lines (QTL-NILs) were developed for
basal resistance QTLs o f our interest using SusPtrit as recurrent parent. SusPtrit is an
experimental line which is exceptionally susceptible to leaf rusts for which normally barley is a
nonhost. Three host basal resistance QTL-NILs and a nonhost basal resistance QTL-N1L were used
in this study. They were challenged with one homologous and three heterologous leaf rusts. The
result showed that the 3 QTLs fo r host basal resistance and the QTL for nonhost basal resistance
have a significant effect on both homologous and heterologous rusts. This gives an indication that
indeed, host and nonhost basal resistance are associated.

K ey w ords: Partial resistance, non-host resistance, near isogenic lines

t | -‘

In trod u ction

Barley host basal resistance (alias partial resistance) and nonhost resistance
to leaf rusts are based on a prehaustorial mechanism associated with
papilla formation at sites of cell wall penetration attempt. Both types of
resistance are typically governed by genes with relatively small,
quantitative effects, located on quantitative trait loci (QTL). Jafary et al.
(2006 and 2008) reported that the QTLs for host basal resistance and no
host resistance of barley tended to co-locate on a consensus map of barley.
Based on the similarity in resistance mechanism and the co-localisation of
QTLs for host basal resistance and non-host resistance, we presume that
these two resistance types are based on shared principles. Five QTL-near
isogenic lines (QTL-NILs) were developed by introgression four host basal
resistance QTLs and a nonhost resistance QTL into SusPtrit genetic
background (need reference). SusPtrit is an experimental line which not
only susceptible to the homologous ru st P. hordei, but also exceptionally
[89]
 Sebil Vol. 16

susceptible to leaf rusts for which normally barley is a nonhost (Figure 1).
The QTL-NILs developed allow us to study the possible association
between the QTLs for host basal resistance and nonhost resistance.

Crop/Variety Puccinia h ordei Puccinia triticina

Barley-Vada

Barley-SusPtrit

Wheat-8860

Figure 1. Susceptibility of SusPtrit to homologous rust P. hordei (barley is a host) and heterologous rust P. triticina (barley
is a nonhost) at seedling stage (Atienza et.al, 2004).

Therefore, this study was conducted to evaluate whether relatively Iarge-

effect QTLs show association in their reaction to homologous and
heterologous rust isolates. Also, to evaluate their mechanism of resistance
at tissue and cell level, especially whether the resistance is based on
hypersensitivity or non-hypersensitivity.

M a terials and M ethods

This experiments was conducted in Plant Research International (PRI)

laboratory and greenhouse at Wagenin^en University, The Netherlands.
 Sebil Vol. 16

P lan t m aterials
Near isogenic lines (NILs) with SusPtrit genetic background (Table 1) and
having resistan ce QTLs, Rphq2, Rphq3, R phqll, R phql6 and Rnhq (Rnhq-V
and Rnhq-L), were used for this study.

Table 1. QTL-NlLs used in this study

QTL-NILs Donor line
Vada-Rphq2 Vada
Vada-Rphq3 Vada
L-94-Rphq2 L-94*Vada
L-94-Rphq3 L-94*Vada
Su-Rphq2 SusPtrit*Vada
Su-Rphq3 SusPtriPVada
Su-Rphq11 SusPtrit'Steptoe
Su-Rphq16 SusPtrit*Dom
Su-Qnh.L SusPtrit*L94
Su-Qnh.V SusPtrit*Vada

The parental lines for each respective NILs were used as a reference. In
addition, L94-NILs (L94-Rphq2 and -Rphq3) and Vada-NILs (Vada-rphq2,
and -rphq3) were included. For the histology assays, host plants
corresponding to the rust species under observation were added as a
reference.

Inoculum
Four isolates of rust fungi were used in infection studies (Table 2). They
were multiplied on their respective host species. Ured ini os pores were
collected and dried in desiccators for 5-7 days before used for inoculation.

Table 2. Rust isolates used in this study

Pathogens Host plant Common name

P. hordei isolate 1.2.1 Hordeum vulgare Barley leaf rust
P. hordei-murini H. murinum Wall barley leaf rust
P. hordei-secalini H.secalinum Meadow barley leaf rust
P. triticina isolate “Flamingo'’ T.aestivum W heat leaf rust

Phenotyping Q T L r X I L s with homologous rust isolate

Seeds of QTLs-NILs were sown in 37 x 39cm boxes in two rows along with
reference lines. Depending on the availability of seeds, 1-2 seeds were
sown for each NIL. The secondary leaves were clipped out and the fully
grown primary leaves were fixed horizontally with adaxial side up in an
inoculation tower and inoculated with 3.5 gram of freshly collected spores
of P. hordei isolate 1.2.1 diluted 10 times with lycopodium spores to obtain
uniform spore distribution. The inoculated boxes were placed in a
[91]
 Sebil Vol. 16

humidity chamber to incubate the spores overnight for eight hours at

100% relative humidity in the dark at 18°C. After incubation, the
inoculated boxes were transferred to a greenhouse compartment where the
temperature is set at 14 ± 3°C with 30-70% relative humidity. The
experiment was carried out in two replications.
The latent period (LP) was measured one week after inoculation. The
LP50S was then calculated with the following formula:
LP50 = T l + (T2- T l) x (N100-2-N1)
(N2-N1)
T l = the time just before 50% of the pustules are mature
T2 = the time just after 50% of the pustules are mature
N1 = number of mature pustules at Ti
N2 = number of mature pustules at T 2
N100/2 = half of the total mature pustules number
To normalize the results, the relative latency period (RLP50S) was
calculated relative to the LP50S of SusPtrit which set at 100.

Phenotypiiig Q T I r M L s with heterologous ru st isolates

The QTL-NlLs were grown as described above except that in this case
susceptible host plants were included. Ten to tw'elve days after sowing,
completely unfolded primary leaves were fixed horizontally with the
adaxial side up, and inoculated with lOmg of spores per box using a
settling tower (Atienza et al.2004). For each QTL-NIL and reference lines,
two seedlings were inoculated of which one seedling was sampled for
histological studies. The second seedling was used for macroscopic
phenotyping. To avoid cross contamination of rusts, the settling tower and
other tools were cleaned with 70% ethanol before and after use. Latency
period (LP) was measured six days (Phs) and eight days (PJnn and P.
triticina) after inoculation. Additionally, the level of infection was
quantified by estimating the following traits: infection frequency (IF,
pustules/cm2), flecks (F, non- sporulating infection sites/cm2), frequency
of visible infection sites (VIF, IF/total amount of visible infection
sites/cm2), and TotF (total amount of visible infection sites/cm2) by using
a metal frame with 1 cm2 window. RLP50S and Relative Infection
Frequency (RIF) were calculated by setting the RLP50S and RIF of SusPtrit
to 100 as stated above. Analysis of variance for both RLP50S and RIF was
carried out using GenStat statistical software (11.1th edition). All
genotypes were grown in one box and the experiment was carried out in
two replications.

[92]
 Sebil Vol. 16

H istological evaluation o f pathogenesis

Histology is the study of the microanatomy of cells and tissues of plants. It
is commonly performed by examining cells and tissues under microscope.
The specimen having been sectioned (cut into a thin cross section), stained,
and mounted on a microscope slide.

P rep aratio n o f le a f sam ples for flu orescence m icroscopy

Seven days after inoculation, leaf segments of approximately 2-3cm long
cut out from the middle of primary leaf. These leaf segments were
prepared as whole mount for fluorescence microscopy, except that Uvitex
2B (Ciba-Geigy) was used instead of Calcofluor. The leaf segments were
immediately fixed and bleached by boiling for 1.5 minutes in a water bath
in lactophenol-ethanol (1:2 v/v). After the leaves were bleached, the
lactophenol-ethanol was poured off and they were washed lx 30 minutes
in ethanol (50%) and in 0.05N NaOH (2g/l), respectively one after the
other. The washed leaf segments were rinsed 3x in water and soaked for 30
minutes in 0.1 M Tris/HCl buffer (pH 8.5). After 5 minutes of staining in a
solution of 0.1% Uvitex in the same buffer, they were rinsed thoroughly 4x
in water and then washed for 30 minutes in a solution of 25% glycerol.
Finally, to prepare the slides, small drops of glycerol were added on the
slide b efore the leaf sam p les w ere put along the lon gitu d in al axis of the
slide and the leaves samples were embedded on slide with the adaxial side
facing up. Then the slide cover was carefully placed on the samples.

O bservation o f infection units under I Y-m ieroscope

For ease of inspection, different classes of infection units were set based on
status of infection unit, where an infection unit is described as non-
penetrating (NP), early aborted (EA) and established (Jafary et al.,2006,
Niks, et al., 2000, Niks and Marcel, 2009 ). The detailed observation and
scoring were done with a 10x10 and 40xl0magnification. The preparations
were screened starting from one of the corners and moving horizontally
along longitudinal axis of the leaves. The outmost stomatal rows were
excluded from observation to avoid possible border effects. Also
overlapping and infection points close to air bubble were ignored. The
infection hyphae were scored as "established" type (more than six
haustorial mother cells) or "early aborted" type (having six or less
haustorial mother cells). The data collected were analyzed using GenStat
statistical software.

[93]
 Sebil Vol. 16

R esu lts and D iscussion

Host basal resistan ce Q T L s and a nonhost resistan ce Q T L confer

longer latency period for 1*. hordei, a homologous rust
P. hordei had a longer latency period on QTL-NILs introgressed
respectively with, host basal resistance QTLs Rphq2, Rphq3, R p h q ll,
Rphql6, and nonhost resistance QTL, Rnhq.v compared to the control
SusPtrit. The host basal resistance QTLs prolonged the latency period
more than the nonhost resistance QTL (Figure 2).

p h r-
H r± ~
r h
1

i
Sus-Rphq2 Sus-Rphq11 Sus-Rphq16 Sus-Rphq3 Sus-Rnhq.v SusPtrit
QTL-NILs

Figure 2. The relative latency period of P. hordei (RLP50S) on the QTL-NILs introgressed with host basal resistance QTLs
Rphq2, Rphq3, R p h q ll. R phq16and nonhost resistance QTL, Rnhq.v, respectively.

Host Imsal resistan ce Q T L s and a nonhost resistan ce Q T L confer

higher levels o f p re haustorial resistan ce to heterologous ru sts
Both host basal resistance (Rphq2, Rphq3, R phqll, Rphql6) and nonhost
resistance (Rnhq.v) QTLs significantly reduced the infection frequency of
heterologous rusts P. hordei-murini (Plirn), P. hordei-secalini (Phs) and P.
triticina (Pf) (Figure 3a). Histology demonstrated a significantly higher
percentage of early aborted colonies and smaller size of established
colonies on all QTL-NILs compared to SusPtrit (Figure 3b and 3c). These
results suggest that less colonies are established on QTL-NILs and that the
growth of the established colonies is restricted on the QTL-NILs. All the
QTLs of both host basal and nonhost resistances had higher effect against
Phs and Phm than against Pt.

[94]
 Sebil Vol. 16

c g r ~
s. ®° ■
3 ------ -----------------------"■------ FH 1

' "20--“S™--
M l L i i i i i 1
S im -A p h q d S u t* R p h q 3 8 u «-R p h q l1 S u «-R | > h q ie S u »* R n h q V S u sP trit
cm- ft ■

Phs Phm

Figure 3. The effect of host basal resistance QTLs Rphq2, Rphq3, Rphq11, Rphq16 and nonhost resistance QTL,
Rnhq.v, respectively on heterologous rusts, P. hordei murini, P. hordei secalini and P. triticina. (a)

The infection frequency of the heterologous rusts on the QTL-NILs in

relative to SusPtrit. (b) The percentage of early aborted colonies on QTL-
NILs and SusPtrit. (c) The colony size of established colonies with
saprogenic tissues on QTL-NILs and SusPtrit.

[95]
 Sebil Vol. 16

Conclusion

The present results suggests that most barley genes conferring a basal
resistance level to the homologous rust P. hordei may also confer resistance
to heterologous rusts that normally do not infect barley. To our
knowledge, it is the first time that such a clear overlap in the genes
involved in the natural variation of host basal and nonhost resistances is
obtained. Those results will have important implications in the
development of strategies to achieve durable resistance against such
fungal pathogens.

Acknowledgments
This work is part of M.Sc. thesis of an author and the study was financially
supported by Netherlands Fellowship Program (NFP-AP).

References

Atienza, S.G., Jafary, H. and R.E. Niks. 2004. Accumulation of genes for
susceptibility to rust fungi for which Barley is nearly a non-host
results in two barley lines with extreme multiple susceptibility. Planta.
220: 71 -79
Jafary, H, Szabo, LJ, Niks, RE (2006). Innate non host immunity in barley to
different heterologous rust fungi is controlled by sets of resistance
genes with different and overlapping specificities. Molecular Plant-
Microbe Interactions 29 (11): 1270 - 1279.
Marcel TC, Varshney, RK, Barbieri, M, Jafary, H, de Kock, MJD, Graner, A,
and Niks, RE (2007). A high-density consensus map of barley to
compare the distribution of QTLs for partial resistance to Puccinia
hordei and of defence gene homologues. Theor Appl. Genet 114:487-
500.
Marcel, TC, Gorguet, B, Ta, MT, Vels, A, Niks, RE (2008). The verification
of QTLs for partial resistance to Puccinia hordei in NILs of barley
confirms an isolate-specific effect. New Phytol.177: 743-755.
Niks, RE, Fernandez, E, Van Haperen, B, Bekele, AB, Martinez, F (2000).
Specificity of QTLs for partial and non-host resistance of barley to leaf
rust fungi. Acta Phytopathol.Hun.35:13-21.
[96]
 Sebil Vol. 16

Niks, RE, and Marcel, TC (2009). Nonhost and basal resistance: how to
explain specificity? New Phytologist 182: 817-828.
Qi, X, Niks, RE, Stain, P, and Lindhout P (1998). Identification of QTLs for
partial resistance to leaf rust (Puccinia hordei) in barley. TheorAppl Genet
96:1205-1215.
Qi, X, Jiang, G, Chen, W, Niks, RE, Stam, P, and Lindhout, P (1999).
Isolate-specific QTLs for partial resistance to Puccinia hordei in barley.
Theor. Appl. Genet. 99: 877-884

[97]
 Sebil Vol. 16

Yield Performance and Stability Analysis of w m7

Sesame (Sesamum indictmi Genotypes in X j . )

Northern Ethiopia
F iseh a B a r a k i '\ Y em a n e T se h a y e 2 an d F eticn A b a y 3
vCro\7 Research Core Process, Hinnera Agricultural Research Center, Ethiopia
E-mail:fish051bar@gnuiil.com
Department o f Crop and Horticultural Science, Mekelle University, Ethiopia
’ Departm ent o f Crop and Horticultural Science, Mekelle U niversity, Ethiopia

Abstract
The study ivas carried out in three locations (a total o f 7 environm ents) o f N orthern E thiopia from
2011-2013 cropping seasons and thirteen sesam e gen otypes w ere eiuiluated. T he objective o f this
study ivas to estim ate the stability o f sesam e gen otypes an d to determ ine the association o f the
stability param eters and sesam e seed yield. The experim ent was laid out in random ized com plete
block design w ith three replications and w ith a total plot size o f 14m 2. The A dditive M ain effects
an d M ultiplicative Interaction (AM M I) m odel fo r grain yield detected significant effects o f the
gen otypes (37.3 % sum o f squares (SS)), environm ents (29.5 % sum o f squares) and G enotype x
Environm ent interaction (25.9 % SS). T he model also extracted fiv e sign ifican t interaction
principal com ponent analysis (IPCA) with a total o f 96.9 % SS and 90.3% corresponding degrees
o f freed om . A ccording to the total rank value (TR), G12, G i l and G4 w ere the most stable and
w idely adapted gen oty pes respectively, w hereas G8 and G 9 were the most unstable genotypes.
Grain yield was positively associated with all o f the ranks o f stability p aram eters at different
sign ifican ce levels. Grain yield (GY) had significant and positive correlations w ith YSI, Pi, S I and
CV. Regarding to the inter-param eter correlation coefficients, neither o f the stability param eters
zvere negatively correlated am ong each other. Environm ents E l, E2, and E4 w ere unfavorable
en vironm ents ; w hile E5, E6 and E7 w ere fav ora ble environm ents and E3 was m oderately fav orable
en vironm en t fo r m ost o f the sesam e genotypes.

Key words: AMMI, Environment, GxE interaction, IPCA,

Introduction
Sesame {Sesamum indicum L.) is an annual plant that belongs to the
Pedaliaceae family. It is an erect herbaceous annual plant with either
single stemmed or branched growth habits and two growth characteristics
of indeterminate and determinate type, reaching up to 2m height and
with a large tap root of reaching 90 cm (Pham, 2010). Most of the sesame
seeds which are rich in fat, protein, carbohydrates, fibre and some
minerals are used for oil extraction and the rest are used for edible
purposes (El Khier et ill., 2008). Among the different varieties of sesame
Sesamum indicum is the most usually cultivated variety all over the world.

[99]
 Sebil Vol. 16

Sesame, mainly grown for its seeds, contains about 50-60% oil content is
also rich in fat, protein, carbohydrates, fibre and some minerals (Caliskan
et a l, 2004).

About 7.8 million hectares of the total world crop area is under sesame
cultivation with about 3.83 million metric tons'of total production. Of the
world production of sesame, Asia and Africa account for 2.29 and 1.38
riillion tons, respectively. In Northern Ethiopia Sesame is the most
important cash crop and it also uses for local oil extraction. So to improve
the production and productivity of sesame in Ethiopia evaluating different
genotypes across different environments or the study of genotype x
environment interaction (GxE interaction) might be important to estimate
the stability of sesame genotypes so as to identify area specific or widely
adapted improved sesame seeds.

GxE interaction (genotype by environment interaction) refers to the

deviation in performance of any attributes of genotypes within the various
growing environments (across locations and years). The presence of GxE
interaction complicates the varietal selection process as it reduces the
usefulness of genotypes by confounding their yield performance through
minimizing the association between genotypic and phenotypic values
(Farshadfar et al., 2012). However, it is possible to develop genotypes with
low GxE interaction via sub-division of heterogeneous area into smaller,
more homogeneous sub-regions; and by selecting genotypes with a better
stability across a wide range of environments (Farshadfar, et al., 2011b). So
GxE interaction may be considered both as an opportunity and a challenge
for breeders.

In multi-environment trials different researchers are using different

stability parameters to select the stable genotype for its area of
adaptability. Among which, AMMI Stability Value (ASV), Yield Stability
Index (YSI), Sum of Interaction Principal Component (SIPC), Cultivar
performance Measure (Pi), Wricke's Ecovalence (Wi ), Francis and
Kannenberg's Coefficient of Variability (CV) and Nassar and Htihn's Mean
Absolute Rank Difference (SI) are the most common parameters to
identify stable genotype.

[100]
 Sebil Vol. 16

Materials and Methods

Experim ental M aterial and Method

Plant M aterial and Methods

The experiment was conducted in seven environments that comprise
location (three testing locations), variety (13 varieties) and year. The soil
and climatic description of the locations is given in table 1. The 13
genotypes (see table 2) were evaluated for three years in both Humera and
Dansha (2011-2013) and for one more year in Sheraro (2013), All the field
evaluation were undertaken under a rainfed condition and the
environments were designated as El (Humera 2011), E2 (Humera 2012), E3
(Humera 2013), E4 (Dansha 2011), E5 (Dansha 2012), E6 (Dansha 2013) and
E7 (Sheraro 2013). The thirteen sesame genotypes were sown in a
randomized complete block design (RCBD) with three replications in a
plot area of 14 m2, Each experimental plot received the same rate of DAP
(100kg/ha) and urea (50kg/ha).

S tatistical Analysis:
About six stability statistics currently in use are examined in this study.
Homogeneity of residual variances was tested prior to a combined
analysis over locations in each year as well as over locations and years (for
the combined data) using Bartlet's test (Steel and Torrie, 1980).
Accordingly, the data collected were homogenous and all data showed
normal distribution.

Table 1: Agro-climatic and soil characteristics of the experimental sites

Location Latitude Longitude Altitude Anual RF Min - Max Soil texture

(°N) (°E) (m) (mm) Temp(°c) Clay (%) Silt (%) Sand
(%)
Humera 14° 15' 36°37' 609 576.4 18.8-37.6 35.6 25.6 38.6
Sheraro 14°24' 37°45’ 1028 676.7 18.8-34.9 21 27.3 51.7
Dansha 13°36' 36°41' 696 888.4 28.7(mean)

[1 0 1 ]
 Sehil Vol. 16

Table 2: Description of the sesame genotypes

Genotype name Gen code Status Seed color Source
A cc#031 G1 Advanced line White WARC
Oro (9-1) G2 Advanced line White WARC
NN-0079-1 G3 Advanced line White WARC
Acc-034 G4 Advanced line White WARC
Abi-Doctor G5 Advanced line White WARC
Serkamo G6 Released Brown WARC

Acc-051 -020sel-14 G7 Advanced line Brown WARC

Tate G8 Released Brown WARC
Acc-051-02sel-13 G9 Advanced line White WARC
Adi G10 Released White WARC
Hirhir G11 Farmers seed (local check) White HuARC
Sett-1 G12 Released (standard check) White HuARC
Humera-1 G13 Released (standard check) White HuARC
WARC-Werer Agricultural Research Center, HuARC-Humera Agricultural Research Center

S ta tistica l Analysis
Homogeneity of residual variances was tested prior to a combined
analysis over locations in each year as well as over locations and years
using Bartlet's test (Steel and Torrie, 1980). Accordingly, the data collected
were homogenous and all data showed normal distribution.

A M M I Model Analysis
The grain yield data were subjected to AMMI analysis, which combines
analysis of variance (ANOVA) with additive and multiplicative
parameters in to a single model (Gauch, 1988). After removing the
replicate effect when combining the data, the genotypes and environments
observations are portioned in to two sources: Additive main effects for
genotypes and environments; and non-additive effects due to genotype by
environment interaction.

Stability Analysis using th e VMM I Model

o Wricke's (1962) ecovalence and Francis & Kannenberg's (1978)
Coefficient of variability were performed using the SAS program
developed by Hussein et al. (2000).
o Lin & Binns's (1988) cultivar superiority performance and Nassar &
Huhn's (1987) absolute rank difference were also carried out using
the GenStat 16th edition (Gen stat, 2009).
o AMMI stability value (ASV) was calculated in the excel spread
sheet using the formula developed by Purchase et al. (1997):

o ASV = J [ | | ^ ( I P C A l score)]2 + (IPCA2score) 2

[102]
 Sebil Vol. 16

Where: SS is sum of squares; IPCA1 is interaction of principal

component axis one; and IPCA2 is interaction of principal
component axis two.
o SIPC (Sum of interaction principal component) was also calculated
in the excel spread sheet using the formula developed by Snelier, et
al. (1997):
N

Where: A05nYin is the interaction principal component (1PC) scores

for the ith genotype; n is number of IPC; and N is number of
significant IPC retained in the model via F-test.
o Similarly yield stability index (YSI) was also computed by summing
up the ranks from ASV and mean grain yield :
YSI= RASV+RGY
Where: RASV is rank of AMMI stability value and RGY is rank of
mean grain yield
To statistically compare the seven stability analysis procedures used in the
study, the Spearman's rank correlation coefficient (rs) Steel and Torrie
(1980) was estimated using SPSS version 16 statistical software.

R esu lt and Discussion

1. Stability M easures from A M M I (Additive M ain effects and

’M ultiplicative In tera ctio n ) Model
The AMMI model for grain yield detected significant variation (p<0.001)
for both the main and interaction effects indicating the existence of a wide
range of variation between the genotypes, years (seasons), locations and
their interactions (table 3). Genotypes had a lion's share in grain yield
variation and accounted about 37.3% of the total sum of squares indicating
that the greatest source of variation for grain yield among the genotypes
were mainly the inherent genetic component Similar result were reported
in sesame (Zenebe and Hussien, 2009). Environments and interaction
effects had 29.5% and 25.9 % contribution for the total sum of squares
respectively. The AMMI model extracted five significant (p<0.001) IPCAs
from the interaction component (table 3). These five IPCAs accounted a
total 96.9% of the interaction sum of squares with 90.3% corresponding
degrees of freedom with a remaining 3.1% considered as noise (Table 3).

[1 0 3 ]
 Sebil Vol. 16

The extracted IPCAs are capable of providing an information on the

interaction effect although their degree decreases from the first to the last
IPCAs. However, the first two IPCAs could best explain the interaction
sum of squares (Zobel et al., 1988). Accordingly, the first two IPCA's with a
total of 57.6% sum of squares and 44.4% of corresponding degrees of
freedom used to explain the interaction effect.

Table 3: Combined AMMI analysis of variance for grain yield of Sesame genotypes

GXE
explained Cumulative
Source of variation DF TSS TSS (%) <%) (%) MS
Genotypes 12 2500959 37.3 208413***
Environments 6 1979243 29.5 329874***
Block (within Env) 14 90120 1.3 6437ns
Interactions 72 1738701 25.9 24149***
IPCA1 17 583954 33.6 33.6 34350***
IPCA2 15 416566 24.0 57.6 27771***
IPCA3 13 260364 15.0 72.6 20028***
IPCA4 11 240543 13.8 86.4 21868**’
1PCA5 9 181706 10.5 96.9 20190***
Residuals 7 55568 7938
Error 168 398652 2373
Total 272 6707676 24661
***highly significant at (P<0.001), ns= non-significant

The IPCA scores of the genotypes is presented in table 4. Genotypes with a

greater IPCA score are the more responsive ones for the interaction effect
and the more specifically adapted genotypes to a certain environment or
location. In contrast to this, the genotypes with smaller IPCA scores are
with lower interaction and are considered as widely adapted genotypes.
Accordingly genotypes with greater magnitude of IPCA1 such as G7
(13.7), G8 (9.2), and G5 (5.7), were the more responsive and contributed
largely to the interaction component and may be considered as a
specifically adapted genotypes. On the other hand, G12 (0.57) followed by
G9 (0.68) and G3 (2.3) were the genotype with least contribution to the
interaction component as they are with lower IPCA indicating their wider
adaptability or stability (see table 4), which was also similar to YSI and
SIPC stability ranks.

A M M I Stability Value (A SY ) Analysis

The ASV is the distance from the coordinate point to the origin in a two-
dimensional scatter gram of IPCA1 scores against IPCA2 scores in the
AMMI model (Purchase et al., 1997). The genotypes with larger IPCA
[104]
 Sebil Vol. 16

score, either negative or positive, are the more specifically adapted to

certain environments and those with smaller IPCA scores indicate a more
stable genotype across environments. Accordingly, G3 with lowest ASV
(3.9) followed by G il (4.5) and G12 (4.8) were the most stable genotypes;
whereas, G7 (19.9) followed by G8 (13.8) (table 4) were ranked as less
stable and more sensitive genotypes to environmental change.

Yield S ta b ility Index ( Y S I ) Analysis

Yield stability index (YSI) Farshadfar et al. (2011a) is recommended as a
measure of stability, which is calculated by summing the rank of mean
grain yield across environments and rank of AMMI stability value of
genotypes. The genotypes with lowest value of this parameter are
desirable genotypes with high mean yield and stability. Hence, YSI
identified G12 and G i l as the most stable genotypes respectively whereas
G9 was identified as the least stable genotype. Both ASV and YSI were also
used by Hagos and Fetien (2011) to describe stability of sesame genotypes
in Northern Ethiopia.

Sinn o f in tera ctio n P rin cip al Com ponent (S II* C )

Sum of interaction principal component (SIPC) is another stability
statistics from AMMI model developed by (Snelier et al., 1997). It is sums
of the absolute value of IPC scores (SIPC) of the genotypes that were
retained in the AMMI model via F-tests. The genotypes with smaller SIPC
are considered as the most stable and widely adapted otherwise
specifically adapted. With respect to SIPC G12 (8.3) was the most stable
genotype and considered as a widely adapted; and G8 (27.9) and G7 (27.2)
as unstable genotypes with a highly variable performance across
environments. Similar report has been made by Zali et al. (2012) in chick
pea using SI PC.

2 . I n ivariate S tab ility Analysis M ethods

Irrespective of how a stability parameter is measured, one of the
most critical question is whether it is genetic or not. If the characteristic
measured by the parameter is non- genetic, it is not heritable and thus
selection for such a parameter is fruitless. However, Farshadfar, et ai.
(1999), have proved that stability indices are genetic and hence heritable.
The above discussions confirmed the ranking order of grain yield and
other traits of the genotypes was varying from environment to
environment. Hence, for better selection and further adaptation of high
yielding and stable genotypes for the locations distinguishing their
[1 0 5 ]
 Sehil Vol. 16

stability is critically important and the stability of the genotypes based on

different measures is depicted in table 4.

Cultivar perform ance M easure (I*i)

Lin and Binns (1988) defined the performance measure (Pi) of the Ith test
genotype as the distance mean square between the cultivar's response and
the maximum response over locations. In this stability measure pair-wise
genotype by environment interaction computation is performed between
each cultivar and the maximum yield at each location. Hence, the less is
the distance of the genotype with maximum yield the smaller the value of
Pi and the more stable is the genotype. The genotype coded as G4 was the
most stable as to the Pi stability parameter. The Pi based ranking of the
genotypes was very similar with that of the mean yield ranking (with few
exceptions) and often considered as a measure of performance than a
measure of stability (Alberts, 2004).

W rick e ’s Eoovalence (W i )
According to Wrick (1962) genotypes with a low Wi value have smaller
deviations from the overall mean across environments and are thus more
stable. According to the meaning of ecovalence, the stable genotype
possesses a low ecovalence. Hence, G7 followed by G9, which possessed
high ecovalence, was marked as the least stable genotypes.

F ra n eis and K aiu ien berg ’s Coefficient o f Variability (C V )

This stability parameter developed by Francis and Kannenberg (1978)
measures the performance and coefficient of variation (CV) for each
genotype over all environments and the genotype that provides a high
yield performance and consistent low coefficient of variation is considered
to be stable genotype. Therefore, G9 was identified as the least stable
genotype.

N assa r and Ilid in ’s M ean Absolute R a n k D ifferen ce ( S I )

Nassar and Hiihn (1987) described non-parametric measures of stability
based on ranks of the genotypes across locations and provide a viable
alternative to existing parametric analyses. The mean absolute rank
difference (SI) estimates are all possible pair wise rank differences across
locations for each genotype. This gives equal weight to each location or
environment and genotypes with less change in rank are expected to
be more stable. According to this SI stability parameter G4 was identified

[1061
 Sebil Vol. 16

as the most stable genotype and G10 and G3 identified as unstable

genotypes.

Despite the lack of consistency among the stability measures, G12, G il

and G4 were the most stable and widely adapted genotypes respectively,
whereas G8 and G9 were the most unstable genotypes equally according
to total rank value (TR).

[107]
 Sebil Vol. 16

Table 4: Mean grain yield (GY), various stability measures and their ranking order

Gen GY R IPCA1 IPCA2 ASV R YSI R Pi R S1 R SIPC R CV R Wi R TR

G1 895.1 2 -5.71 1.81 7.0 6 8 3 4433 2 1.8 3 23.2 9 14.7 7 39728.3 7 37

G2 638.1 12 3.31 5.29 8.3 9 21 8 55022 11 2.7 7 17.4 4 14.2 6 28506.8 3 48

G3 740.4 6 -5.22 8.69 3.9 1 7 2 27403 6 3.5 12 22.3 6 13.9 5 38856.6 6 38

G4 926.8 1 -13.70 -5.22 11.4 10 11 4 1239 1 0.6 1 22.9 7 6.5 1 48249.6 9 33

G5 662.6 10 0.68 -12.13 8.2 8 18 7 48904 10 3.1 8 23.2 10 17.9 8 52922.5 10 61
G6 711.5 7 3.53 3.94 7.6 7 14 6 38459 8 1.9 4 16.6 3 20.4 9 31195.4 5 42

G7 687.5 9 2.78 2.24 19.9 13 22 9 43319 9 3.5 11 27.2 12 21.6 11 100929 4 13 78

G8 655.2 11 4.00 -0.69 13.8 12 23 10 56649 12 3.1 9 27.9 13 25.0 12 66710.4 11 79
G9 614.3 13 2.28 2.24 12.2 11 24 11 70599 13 3.3 10 23.1 9 31.2 13 68370.6 12 79
G10 697.6 8 -5.53 -2.84 6.3 5 13 5 36750 7 4.0 13 24.0 11 20.6 10 45063.7 8 59
G11 791.5 5 4.88 -3.34 4.5 2 7 2 19282 4 2.7 6 16.0 2 13.1 4 18713.3 2 22
G12 832.7 3 -0.57 4.77 4.8 3 6 1 9525 3 1.3 2 8.3 1 7.5 2 10063.2 1 13
G13 8051 4 9.25 -4 77 5.7 4 8 3 20244 5 2.0 5 17.9 5 12.9 3 30245 5 4 29
Where: GY= grain yield; ASV = AMMI stability value: YSI= yield stability index; Pi = Lin & Binns's cultivar superiority performance; S1= Nassar & Hiihn’s absolute rank difference; SIPCA=
Sum of interaction principal component; CV = Francis & Kannenberg's Coefficient of vanability; Wi = Wncke's ecovalence, TR= total rank; Rnk= Rank

[108]
 Sebil Vol. 16

3 . A ssociations between (ira in Yield and the Stab ility P a ram eters
Spearman's coefficient of rank correlation Steel and Torie (1980) was
executed for possible pair wise comparisons of the ranks of different
stability parameters and yield ranks (table 9). The correlation between
mean grain yield (GY) and the seven stability parameters varied
considerably. Grain yield was positively associated with all of the ranks of
stability parameters at different significance leveis. Grain yield (GY) had
significant and positive correlations with YSI, Pi, SI and CV, but its
association with ASV, SIPC and Wi was non-significant although it was
positively associated. Grain yield was highly and positively correlated (at
p<0.01) with Pi (r = 0.98) as the rank-correlation coefficient was near to
unity (table 9). This indicates that choosing a genotype based on Pi
stability parameter could lead to selecting a genotype with highest grain
yield. As reported by Issa (2009), the association between GY and Wi was
positive but weak.

I able 5: Spearmen rank correlation among ranks of grain yield and other stability parameters

GY AS V YSI Pi S1 SIPC CV Wi

GY
ASV 0.522ns
YSI 0.818" 0.890"
Pi 0.984” 0.560' 0.846"
S1 0.670’ 0.121ns 0.419ns 0.626’
SIPC 0.41ns 0.630* 0.64V 0.432ns 0.569’ -
CV 0.775" 0.588' 0.807" 0.797" 0.654’ 0.682’
Wi 0.44ns 0.764” 0.741" 0.478ns 0.473ns 0.894" 0.698"
ns= non significant, "significant (P<0.05). ** highly significant ( P<0.01)

The inter-parameter correlation coefficients indicated that neither of the

stability parameters were negatively correlated among each other (table 9).
The highest positive correlation among the parameters (at p<0.01) were
recorded between Wi and SIPC (r = 0.894), ASV and YSI (r = 0.890), and Pi
and YSI (r=0.85). This also partially concurred with the findings of
Farshadfar (2011b). On the other hand, the weakest correlation was
observed between YSI and SI (r=0.12) indicating that the genotypes
selected according to the ranking order of these parameters may be quite
different.
[109]
 Sebil Vol. 16

4 . Environm ental P erfo rm an ce ami Stability

The environments had different mean grain yields (table 5) and this
indicates that the different environments were not equally favorable or
unfavorable for the genotypes grown under them. Environments often
classified as favorable and unfavorable ones based on the environmental
index (El) where environments with a negative index considered as
unfavorable and those with positive regarded as favorable (Farshadfar,
2008). Accordingly, E l had a negative environmental index (-113.4), and
was classified as the least favorable environment while E6 had the highest
positive environmental index (149.2) and considered as the most favorable
environment (table 5). In general E l, E2, and E4 both with negative
environmental index had below average mean yield and considered as
unfavorable environments. Whereas, E5, E6 and E7 with positive and
significant environmental index had above average mean yield
performance and classified as favorable environments. Exceptionally, E3
which had negative but non-significant El was considered as moderately
favorable environment for most of the genotypes.
The environments were also described for their stability based on their
ASV. Hence, E l (22.2) and E4 (16.6) with highest ASV were the least stable
environments; whereas E7 (5.4) followed by E5 (7.1) were stable
environments and which may be better for further breeding program.

Table 6: IPCA scores, Environmental Index (El), and AMMI stability value of seven environments
Env.
Env. code Env. Mean IPCA1 IPCA2 El AS V
Humera-2011 E1 629.5 -15 705 2.69668 -113.44** 22.2
Humera-2012 E2 658.6 2.1842 12.9542 -84.343** 13.3
Humera-2013 E3 737.3 -5.3562 -4.7389 -5.6429ns 8.9
Dansha-2011 E4 695.5 11.7269 2.10177 47.443** 16.6
Dansha-2012 E5 770.9 0.54909 -7.0447 27.9571** 7.1
Dansha-2013 E6 8922 416373 -10.169 149.257*’ 11.7
Sheraro-2013 E7 816.6 2 43698 4.19989 73.6571** 5.4
"sign ifica nt at (P 0 .0 1 ). ns= non-significant

Conclusion and Recom m endation

The AMMI model, that detected significant variation (p<0.001) for both the
main and interaction effects, extracted five significant (p<0.001) IPCAs
from the interaction component. The existence of such significant GxE
[110]
 ♦

Sebif Vol. 16

interaction in varietal selection may be both a challenge and an

opportunity for plant breeders and breeding program. According to total
rank value (TR) genotypes G12, G il and G4 were the most stable and
widely adapted genotypes respectively, whereas G8 and G9 were the most
onstable genotypes.
Grain yield was positively associated with all of the ranks of stability
parameters at different significance levels: Grain yield (GY) had significant
and positive correlations with YSI, Pi, S I and CV, although it was non-
significantly and positively associated with ASV, SIPC and Wi. Regarding
the inter-parameter correlation coefficients, neither of the stability
parameters were negatively correlated among each other. The highest
positive correlation among the parameters (at p<0.01) were recorded
between Wi and SIPC (r = 0.894), ASV and YSI (r = 0.890), Pi and YSI
(r=0.85).

W ith respect to the environments E l, E2, and E4 were considered as

unfavorable environments; while E5, E6 and E7 were classified as
favorable environments. Exceptionally, E3 which had negative but non­
significant El was considered as moderately favorable environment for
most of the genotypes.

Acknowledgements

The first author would like to sincerely acknowledge for research members
of crop department in Humera Agricultural Research Center and Public
and Private Partnership Organization project (PPPO) for financial support.

R eferen ce

Aremu, C.O., Adewale, D.B. and Adetunji, I. A. (2011). Cause and effect
variations and trait selection ndex for indigenous Sesame (Sesamum
Indicum L.) genotypes, International Journal of Applied Agricultural
and Apicultural Research, 7 (2), pp. 64-71.
Alberts, M.J. (2004). A comparison o f statistical methods to describe genoh/pe x
environment interaction and yield stability in multi-location maize trials,

[111]
 Sebil Vol. 16

M.sc. thesis, Department of plant science, The University of Free State,

Bloemfontein, RSA.
Caliskan, S. M. Arslan, H. Arioglu, and Isler N. (2004). Effect of planting
method and plant population on growth and yield of Sesame
(.Sesamum indicum L.) in Mediterranean types of environment, Asian J.
Plant Sci., 3: 610-613.
El Khier, M. K. S., Ishag K.E.A. and Yagoub A. E.A. (2008). Chemical
composition and oil characteristics of Sesame seed cultivars grown in
sudan, Research Journal o f Agriculture and Biological Sciences, 4(6),
pp.761-766.
Farshadfar E. (1999). Path analysis of genotype * environment interaction
in wheat substitution lines, Iranian J. Agric Sci., 30 (4), pp.665-671.
Farshadfar, E. (2008). Incorporation of AMMI stability value and grain
yield in a single non-parametric index (GSI) in bread wheat, Pak. J.
Biol, sci., 11:1791-1796.
Farshadfar, E., M. Geravandi, and Vaisi Z. (2012). Chromosomal
localization of QTLs controlling genotype x environment Interactions
in barley, Inti J Agri Crop Sci, 4: 317-324.
Farshadfar, E., N. Mahmodi, and Yaghotipoor A. (2011b). AMMI stability
value and simultaneous estimation of yield and yield stability in bread
wheat (Triticum aestivum L.), Aust. J. Crop Sci., 5:1837-1844.
Farshadfar, E., Z. Vaisi, and Yaghotipoor A. (2011a). Non parametric
estimation of phenotypic stability in Wheat-barley disomic addition
lines, Annals of Biological Research, 2:586-598.
Francis, T.R. and Kannenberg, L.W. (1978). Yield stability studies in short-
season maize. A descriptive method for grouping genotypes, Canadian
Journal o f Plant Science, 58, pp.1029-1034.
Gauch, H.G. (1988). Model selection and validation for yield trials with
interaction, Biometrics, 44:705-715.
GenStat, (2009). Gen Stat for Windows (14th Edition) Introduction. VSK
International, Hemel Hempstead.
Hagos, T. and Fetien A. (2011). Additive Main Effects and Multiplicative
Interactions analysis of yield performance of Sesame genotypes
across environments in Northern Ethiopia, Journal of the Drylands,
4:259-266.

[1121
 Sebil Vol. 16

Hussein, M., Bjornstad, A. and Aastveit, A. (2000). SASG x ESTAB: A SAS

Program for Computing GEI Stability Statistics, Agronomy Journal,
92,pp.454-459.
Issa, B. A. (2009). Genotype by Environment Interaction and yield stability o f
Maize hybrids evaluated in Ethiopia, M.Sc. thesis, Department of Plant
Sciences/Plant Breeding, University of the Free State, Bloemfontein,
South Africa.
Lin, C.S. and Binns, M.S. (1988). A superiority measure of cultivar
performance for cultivar x location data, Canadian Journal o f Plant
Science, 68, pp.193-198.
Nassar, R. and Huhn, M. (1987). Studies in estimation of phenotypic
stability: Test of significance for non-parametric measures of
phenotypic stability, Biometrics, 43, pp.45-53.
Pham, D.T., T T .D Nguyen, A.S. Carlsson, and M.T. Bui (2010).
Morphological evaluation of Sesame (Sesamum indicum L.) varieties
from different origins, Aust. J. Crop Sci., 4:498-504.
Purchase, J.L. (1997). Parametric analysis to describe Genotype x
environment interaction and yield stability in Winter Wheat, Ph.D.
Thesis, Department of Agronomy, Faculty of Agriculture of the
University of the Free State, Bloemfontein, South Africa.
Sneller, C.H., L. Kilgore-Norquest, and Dombek D. (1997). repeatability of
yield stability in soybean, Crop Sci., 37: 383-390.
Steel, R. and Torrie J. (1980). Principles and Procedures of Statistics a
Biometrical Approach. 2nd ed. Me Graw-Hill, Inc,:471-472.
Wricke, G. (1962). Ubereine Method zur Erfassung der Okologischen
Streuberite in Feldversuchen. Z.Pflanzenzuchtg, 47: 92 - 96.
Zali, H., Farshadfar, E., H.S. Sabaghpour, and Karimizadeh R. (2012).
Evaluation of Genotype x Environment interaction in chickpea using
measures of stability from AMMI model, Annals of Biological
Research, 3: 3126-3136.
Zenebe, M. and Hussien M. (2009). Study on Genotype X Environment
Interaction of yield in Sesame (Sesamum indicum L.), J. Phytol., 2009, 1:
199-205.
Zobel, R.W., M.J. Wright, and Gauch H.G (1988). Statistical analysis of a
yield trial. Agron. J., 80: 388-393.

[1 1 3 ]
Vegetable Oil Shortage in Ethiopia:
Its Causes and Solutions
G etin et A le m a n 1, M istru T esfaye3, A bush T esfaye, W ogayehu W o r k u ' a n d A d u g n a W a k jira 2
Oilseed Breeders. M elkassa. EIAR Head Office3,HolettCT .Jimma4 and Kulumsa' Agricultural Research
Centers. Ethiopian Institute o f Agricultural Research. P. O. Box 2003 Addis Abeba.

A bstract
This paper examines the current oilseeds production globally and nationally, the causes, effects and solutions
o f vegetable oil shortage in the countn/ and outlines the solutions in a long and short term basis. The
nutritional and industrial value o f vegetable oils is dependent on its fatty acid composition. Oils (noug,
sesame, sunflower peanut, soybean) containing long chain fatty acids (oleic and linoleic) are nutritionally
preferred while those containing short chain fa tty acids (palmitic and stearic) are resistant to oxidation and
hence are suitable fo r frying. The global production o f oilseeds reaches 500 million metric tons in 2013. Most
o f the produce was oil palm cultivated in Indonesia and Malaysia and soybean from USA, Brazil and
Argentina. During 20T.4, the total production o f oilseeds in Ethiopia, was 7.5 million quintals with half o f it
being tioug and sesame. N mg is mostly consumed locally while sesame is exported as foreign exchange
earnings. Although Ethiopia is a major exporter o f sesame, the country is also a net importer o f vegetable oils
for food, detergent and other industries. The shortage o f wgetable oil could be due to many confounded
reasons but the most important once are stagnant production and productivity, lack o f the research to come up
with breakthrough technology and under developed and underfinanced processing sector. YJe believe that the
country has the right ecology and sufficient production package to increase production and productiinty o f oil
seeds and output o f vegetable oils. Self-sufficiency in vegetable oil can be achieved through scaling up o f
soybean, groutidnut and sunflmver best practices in the short term. Recently, locally manufactured and
refined soybean and groundnut oils is available in the market. In addition, research on noug and high yielding
perennial crop oil palm can alleiuate the crisis in the long term. Noug seed bears oil containing high linoleic
acid with nutty taste and pleasant odor. Its loiv yielding and poor partitioning o f dry matter capacity can be
increased through genetic improvement o f economically important traits, selection for self-compatible lines,
photosynthesis efficiency and utilization o f appropriate breeding methods. In addition, Ethiopia has suitable
ecology to grow oil palm in Keffa-Sheka Zone and Gambella region. Hence, adoption o f hybrid cold tolerant oil
palm varieties along with their cultivation and processing technology can contribute significantly towards
vegetable oil self-sufficiency for food, detergent, cosmetic etc. industries.

Introd u ction

Oilseeds are important raw materials for food, feed, detergent,

pharmaceutical, cosmetic, ink, polymeretc industries. The Food and
Agriculture Organization of the UN defines oil crops as those from which
oils and fats are extracted for use for human food products and industrial
purposes. FAO classifies oil crops in two temporary and permanent crops.
Temporary are annual crops whose seeds are used mainly in cooking
(culinary) and industry such as soybeans and sesame. The permanent oil
crops are perennial plants whose seeds, fruits or mesocarp, and nuts are
used for the extraction of cooking oil or industrial oils and fats. Examples
are the coconuts, olives and oil palm. Vegetable oils have vast ranges of
[115]
 Sebil Vol. 16

uses. In cooking they are used to provide texture and flavor and when
heated they are used to cook other foods. They are also inputs in variety of
industries. Cosmetic, paints, insulators etc industries.

In Ethiopia, oilseeds can be classified into highland (linseed and

gomenzer), midlands (noug, sunflower and safflower) and lowland
(sesame, groundnut and castor) based on ecological adaptation. In
addition, oilseeds can be also divided into those that can be used as food
and other purposes and those that are industrial only. The industrial as
well as nutritional value of vegetable oils is dependent on their fatty acid
composition. Most vegetable oils contain 18 carbon atoms with the
exception of gomenzer that had erucic acid with 22 carbon atoms (Table 1).
Among oilseeds noug, safflower and sunflower contain high level of
iinoleic acid while castor contains the highest oil in its seed with 90% of
ricinoleic acid which makes it the most versatile industrial oil.

G lobal Production o f O ilseeds

The global area coverage of ten major oilseeds reached about 260 million
hectares with total production of over 500 million metric tons in 2013
(Figure 1 and 2). Globally oil palm, soybean, cotton seed, sunflower,
rapeseed, olive, castor, coconut and sesame are the major oil seeds. Among
these, castor is non-edible and used of industrial applications and cotton
seed is a byproduct of cotton production. Among these major oilseeds, oil
palm and olive are permanent crops. The global production and trade of
oil seeds is dominated by oil palm and soybean production. Palm oil
production and supply exceeded all other oil seeds since 2005. Most of
the palm oil supply is coming from Indonesia and Malaysia with other
African and Latin American countries contributing little (Figure 3). In
addition South East Asia produces almost all of the palm oil supplied to
the world market.

[1 1 6 ]
 Sebil Vol. 16

Table 1. Range of fatty acid composition of food and industrial oils grown in Ethiopia

Fatty add Oilseed Palm Oil

Noug2 Linseed' Gomenzer3 Sesam e’ Castor1 Vernon ia4 Cotton Kernel Mesocarp
i
Oil Content 27-47 30.0-45.8 39.8-46.4 49.5-51.3 36-58a 15-36 17.5-27.0
Leuric (C 1 2 0 ) 45.0
Meristic (C14:) 18.0 9.0

Palmitic 7.6-8.7 4.3-12.3 3.0-3.7 8.4-10.3 2 .5 4 .0 2.0-2.5 9.0 44.0

Stearic 5.6-7.5 1.9-6.3 4.5-5.8 3.3-5.1 2 .5 -3 3 3.0 24.0
Oleic 4.8-8.3 11.3-29.4 10.6-12.1 39.5-43.0 2.6-7.6 4.7-13.5 20.0-25.0 15.0 40..0

Linoleic 74.8- 10.0-17.4 17.3-19.9 4 1 .0 4 5 .0 12.7-21.6 50.5-55.0 10.0

79.1
Linolenic 0.0-0.9 40.5-68.3 11.7-16.1 3.4-8.0
Arachidic 0.4-0.8
Ecosenoic 7.5-8.3
Erucic 39.9-43.3
Ricinoleic 81.2-91.8

Vemoleic 56.1-78.2
No of 241 33 11 21 120 122 5
Accessions

Source 'Seegler 1980, 2Getinet and Adefris 1995. 3 Getinet et al 1996 and ‘Tesfaye 2003, oil content of castor is based on 120
accessions grown at Melkassa Agricultural Research Center in 2009.

African continent or the origin of the oil palm contributes insignificant

amount of palm oil rather is a market destination for South east Asia palm
Oil. The productivity of oil palm is by far higher than all other major
oilseeds (Table 2).

The global production of soybean is dominated by USA, Brazil and

Argentina (Figure 4). Most of the soybean varieties cultivated in these
countries is herbicide resistant and the production is fully mechanized.
Soybean contains 40% protein and 20% oil in its seed and the protein
content increases to 60% after the oil extraction. The meal remaining after
the oil extraction is an excellent raw material for food fortification as well
as poultry feed. In Brazil, it is used as poultry feed and the country is the
largest exporter of poultry globally.

[117]
 Sebil Vol. 16

y 400 1
§ 300
H*
■Cotton S eed
•J 200
5 100 - >Palm k e rn e l a n d
F ru it
R apeseed

Year

Figure 1. Global Production of major oilseeds in million metric tons during 2004-2013 (FAO Stat 2014).

“C a s to r

■ Cocunut

S esam e

Year
G lobal p ro du ctio n of castor, co cu nu t and sesame d u rin g 2004-13
Figure 2. Global Production of Castor, Coconut and Sesame production in thousand metric tons during 2004-2013 (FAO
Stat 2014).

C o lo m b ia Re st 3 9 Largest Producers of palm Oil 2012

2% C o u n trie s
N ig e ria
a th a ila n d
3%

Figure 3. Major Producers of oil palm globally during 2013 (2015: International oil palm Conference. Bogota, Colombia).

[1181
 Sebil Vol. 16

Million Metric Tons

Figure 4. Major soy bean producers around the world (FAO Stat 2014).

Table 2. Productivity of major oilseed crops.

Oil Crop Oil production % of total Average oil Planted area

million tones production yield million hectare
Soybean 33,58 31.69 0.36 92.10 42.24
Sunflower 9.66 9.12 0.42 22.90 10.50
Rapeseed 16.21 15.30 0.59 27.30 12.52
Oil palm 33.73 31.64 3.68 9.17 4.2
Total 105.94 218.02

N ational Production Oilseecls

The major oilseeds in Ethiopia are linseed, noug and gomenzer in the
highlands and sesame groundnut and castor in the low lands. Cotton seed
which is the major byproduct of ginneries contributes significant amount
of raw material for the oilseed crushing industries. During 1997 the
amount of oilseed produced in the country reached a little over 500 million
quintals and increased only to 700 million quintals in ten years (Figure 5).
The increased in production comes from the contribution of sesame than
other oilseeds. Although the total production has shown a little increase
the production of noug, linseed, groundnut and gomenzer did not show
significant increase.

[119]
 Sebil Vol. 16

Im p ort o f Vegetable O il and E x p o rt o f O il Seeds

During the last ten years almost all brands of vegetable oil were imported
to Ethiopia. These include soybean, sunflower, rapeseed, palm oil, Olive,
corn etc oils from abroad. However the bulk or most of the oil imported to
Ethiopia is palm oil from south East Asia countries particularly from
Malaysia and Indonesia. The amount of oil palm and cost in USD has
been increasing for the last five years (Figure 6 and 8). Similarly, the
amount and cost of virgin olive oil during 2010-2014 was increasing
(Figure 7). Olive tree oil is grown in the Mediterranean region including
Syria, Lebanon, Morocco, Spain and Italy.

During 2010-2014 Ethiopia has been also exporting sesame seed

particularly to China, japan and Israel (Figure 9). Export of sesame has
been the second agricultural export commodity next to coffee. The value
of total oilseed export and palm oil import in Million USD is shown on
Figure 8. The total oil exported was mostly sesame seed of grade 2 and
three. Grade one is described as 99% pure, snow white in color and having
at least oil content of 50%. As shown in Figure 8, the revenue from
oilseeds export was higher than the expense to import vegetable oils. In
2014-15, the total production of oilseeds was 7.6 million quintals of which
2.9 million quintals was sesame. The balance 4.7 million quintals was
consumed locally, hence one can assume that 4.7 million quintals of
[1 2 0 ]
 Sebil Vol. 16

oilseed or about 1.41 million quintals of vegetable oil was consumed in the
country from local sources. On the other hand 3.9 million quintals of
vegetable oil was imported in the same period. Therefore, the share of
local production was 26.4% of the demand and the per capita consumption
of vegetable oil is estimated at 5.63kg/person. However the value of
vegetable oil import is only for legal import and does not include the
amount imported as food aid and contraband.

• W t in T h o u s e n d T ones

•V a lu e in M illio n USD

Year

Figure 6. The amount of palm oil imported and in thousand tones and its value in Million USD during 2010-2014 in
Ethiopia (ERCA 2014).

[121]
 Sebil Vol. 16

•W t in Q

V a lu e in th o u s e n d USD

Y ear

Figure 7. Volume and value of virgin Olive oil imported in Ethiopia during 2010-2014 (ERCA 2014).

-P a lm O il In p o r t

■Total V e g e ta b le O il
Im p o rt

V a lu e o f O il S eed
E x p o rt

Year

Figure 8. Value of total vegetable oil and palm oil imported and oilseed export in million USD during 2010-2014 (ERCA
2014).

[122]
 Sebil Vol 16

Japan Yemen -O th e r*
Saudi Arablt
2% ,
Jordan
3%
Turkey
6%
United States
6%

Vietnam
7%

Figure 9. Destinations of Ethiopian Sesame seed export (ERCA 2014)

Causes of Vegetable Oil Shortage

Although there may be many reasons for the shortage of vegetable oils, the
major factors are stagnant production and productivity, the inability of the
oilseeds research to show a breakthrough and under financed and under
developed oilseed processing industry.

Stagnant Production and Productivity of Oilseeds

The area coverage and productivity as well as total production of oilseeds
namely noug, sesame and linseed was compared with maize and wheat.
During the 2005-2014 the area under maize and bread wheat has increased
substantially (Figure 10) while the area under sesame, noug and linseed
has remained stagnant. During the same period, the productivity of the
three oilseeds remained almost constant. The productivity in quintals per
hectare of maize increased from 17 to 33 during the last ten years while
that of bread wheat increased by 9 quintals (Figure 11). It should be
emphasized that the population growth and the requirement of industry
for raw materials is increasing faster while production and productivity of
oilseeds was not increasing as at a similar rate. Therefore the stagnant
production as well as productivity of oilseeds and the ever growing
appetite of the industry for raw material is increasing. Similarly the
growth of industries the economic growth of the country exerts its own
demand.
[123]
 Sebif Vol. 16

Figure 10. Area in Million hectares covered by maize, wheat, and three oilseeds in Ethiopia during 2005-2014 (CSA
20105-2014).

Y ea r

Figure 11. Productivity of maize, bread wheat, noug, linseed and sesame in quintals/ha nationwide in Ethiopia during
2005-2014(CSA 20105-2014).

[124]
 Sebil Vol. 16

T h e R e se a rch on O ilseeds lias not achieved a B reakth rou g h

The research on noug, linseed, castor, sunflower, safflower, soybean and
gomenzer started at Debre Zeit Research Station during the early 1960;s.
In 1966 the Institute of Agricultural Research was established and over
time recommendations on seed rates, weeding frequencies, fertilizer rates,
sowing dates harvesting stages etc for noug, linseed, gomenzer, sunflower,
safflower, soybeans, sesame, groundnut and castor were generated. In
addition, a number of improved varieties were released during the last 50
years (Figure 12). However improved seed of these varieties is not readily
available as it is in cereals and pulses. In addition a breakthrough in terms
of genetic improvement such as hybrid varieties in sunflower and noug
was not realized.

Figure 12. Number of released varieties of oilseeds in Ethiopia (MoA 2014).

Currently the oilseeds research in EIAR is organized under two case teams
namely Pulses, Oilseeds and Fiber crops for noug, linseed, gomenzer,
sesame and ground nut and Aromatic Medicinal and Bio-energy for castor,
oil palm and physic nut. This shows that the research lacks focus and
definite direction. In addition, oilseeds research is not supported by
CGIAR Centers or simply donors are not interested to support oilseeds
research. The technology generated from research was not scaled up as
much as in cereals and pulses. There is also very weak seed multiplication
and supply scheme of oilseeds.

Underdeveloped and U nderfinanced O il Seed P rocessin g S e c to r

Ethiopia has a total of 7 public and 19 private large to medium and, 834
small scale oil processors (Wijnands 2009). Most of these mills are very old
and out dated and only one has a solvent extraction and refining facility.
[125]
 Sebil Vol. 16

It was also indicated that two third of the vegetable oil locally is produced
by small scale processors. Although the government regulation states that
vegetable oil for cooking must be refined, it has never been reinforced. The
amount of palm oil imported is about 10-15 times greater than the amount
of vegetable oil processed locally. This was also the result of a study
indicated by Wijnands et. al. (2009). The ever increasing of import of
vegetable oil and its prices has lowered the per capita income of vegetable
oil in the country. The ever increasing of vegetable oil indicates that there
is large vegetable oil market in Ethiopia. The vegetable oil supply in
Ethiopia is faced with multi factual problems. The market is flooded with
imported vegetable oil mainly palm oil. On the other hand the local
industries are faced with high cost and low quality of raw material on one
hand and stiff competition from imported cheap pahn oil. However the
ocal vegetable oils have higher nutritional value and an added advantage
of the meal that can be used as feed or fortification. In addition some
oilseeds are versatile and can be used for other purposes such as peanut
for peanut butter.

The Ethiopian vegetable oil industry is confronted with shortage and very
high price of low quality raw materials and old technology on the one
hand and stiff competition from imported refined palm oil on the other.
Hence the local factories find it very difficult to compete with imported
vegetable mostly palm oil. Whereas the local oil processors are obliged to
pay VAT palm oil is imported tax free, hence the local oil processors were
driven out of business and became palm oil importers themselves.
Nevertheless, recently there is refined and semi refined packed and well
contained noug and soybean oil in the market. The meal remaining after
the extraction of soybean oil is used as food fortification and formulation.

Solu tion s for V egetable Oil C risis

Ethiopia has suitable climate and soils for the cultivation of several
oilseeds with ample amount of technology for scaling up. These include
scaling up of soybean, groundnut and sunflower technologies in the short
term. In addition to scaling up of available technologies, basic and
adaptive research on noug and oil palm would assist the supply of
vegetable oil. Noug bears very high quality oil that can be branded as
prime product. On the other hand oil palm is very productive, has low
production cost and can be a very good solution to the local vegetable oil
supply. However the adaptive research should include processing and

[126]
 Sebil Vol. 16

packaging. The solutions for the vegetable oil crisis in Ethiopia in a short
and long term basis are discussed below.

Sealing Up o f B e st P ra c tic e s o f Soybeans

Soybean is the second most important oilseed crop globally. The seed
grown in Ethiopia contains 20% oil and 40% protein. The meal remaining
after the oil extraction is raw material for fortification of children and
mothers food. In USA and Brazil the meal remaining after the oil
extraction is feed to poultry. Soybean is grown from lowlands to mid
altitudes 1300-1700 meters of elevation with annual rainfall of 500-1700
mm. Fekadu et al (2011) tested 20 genotypes of soybean at Awassa, Bonga,
Goffa, Inseno and Areka and yields were 31 q/ha at Gofa and lowest at
Areka with llq / h a. Oil content was about 20% with protein content of
40%. Among genotypes, AGS-115-1 yielded an average of 25
quintals/hectare with 32 quintals at Gofa. Similar results were also
reported by Agdew in 2010. Soybean variety improvement programs has
so far registered a number of varieties as shown on Table 3. The proximate
analysis of some of the soybean varieties is shown on Table 4. The oil
content of these varieties is about 20% with protein content of 40%. The
total carbohydrate is about 30% with crude fiber content of 5%. Scaling up
of best practices of soybean can increase the availability of cooking
vegetable oil for home purposes with additional advantage of food
fortification. The nationwide seed yield of soybean during 2014 was 20
q/ha. There are ten soybean processing companies in Ethiopia that
produce soy oil, soy fortified food and soy based feed for poultry. The
total annual raw material requirement of these factories is about 25, 700
tones. Recently one can see refined and semi refined soybean oil in super
markets, groceries and kiosks. The oil is labeled and neatly contained and
distributed in three brands. Therefore, the oil extraction and refining
technology is already adopted and availability of the vegetable oil shall
increase with increase of soybean production.

Scalin g Up o f lle s t P ra ctice s o f G roundnut

Groundnut is cultivated in lower altitudes; below 1600 m such as in
Eastern Hararghe, Lower WabeShebele particularly Gode area, Dedessa
valley, Gambella and Benishangul Gumz region.Variety development in
groundnut is manly evaluation of introduced materials from International
Crop Research Institute for the Semi-Arid Tropes and USA either for oil or
confectionery. Confectionery types are large seeded with attractive color

[127 ]
 Sebil Vol. 16

and contains low oil in their seed. In addition to high yield and high/low
oil content, selection for high yield and resistance to rust and leaf spot are
the major selection factors. Research has developed and released 17
varieties, out of which 4 are confectionery and 13 oil types (Table 5).
Productivity of released varieties varied based on production system.
Improved varieties yield 40-80 q/ha of unshelled groundnut under
irrigation, 30-50 q/ha under sufficient rainfall and less than 20 q/ha under
marginal or stress (Adugna 1992). Seed yields of groundnut under
irrigation and high rainfall are high with satisfactory oil contents.
Agronomic and crop protection package practices have been developed
largely at Werer Agricultural Research Center. Scaling up of these
technologies in Eastern Hararghe and along Wabe Shebelle , Dawa Genale
and Awash rivers. Under irrigation and Western lowlands particularly in
Gambella and Benishangul Gumz regions and lower and upper Beles can
increase the supply of oilseeds domestically. In addition to cooking oil,
groundnut can also be used as peanut butter.

Table 3. Soybean Varieties Released in Ethiopia.

Variety Yearof Yield in q/ha Reference

Release Research Farmers
Field
Clark 1998 30 20 N ISA 1998
Coker 240 1998 30 20 NISA 1998
Davis 1998 30 20 NISA 1998
Williams 1998 30 20 NISA 1998
Crawford 1998 30 20 NISA 1998
Kaland 1998 30 20 NISA 1998
Hell 1998 30 20 N IS A 1998
Belesa-95 2003 29 20 MoRAD 2003
Jalele 2003 22 15 MoRAD 2003
Cheli 2003 22 15 MoRAD 2003
Awassa-95 2005 18 16 MoRAD 2005
AFGACT 2007 15 13 MoA 2007
Boshe 2008 16-30 14-28 MoA 2007
Wegagen 2010 20 18 MoA 2010
KOME 2011 12-37 12-32 MoA 2010
KATTA 2011 14-32 13-28 MoA 2010
N ova 2012 22.5 12-20 MoA 2010
Wello 2012 19-32 17-22 MoA 2010
Nyla 2014 18-24 10-16 MoA 2014

[128]
 Sebil V o l.16

Table 4. The nutritional quality of some of the released soybean varieties

Unit (%/
Rcal/kg) Williams Clark-63k Crowford Davis G - 2261 Cocker-240
Moisture % 9.31 9.07 9.44 7.97 8.09 8.26
Crude Protein % 30.60 27.86 33.51 28.25 32.95 29.07
Oil content % 20.70 20.67 19.48 21.55 19.60 21.06
Crude Fiber % 5.25 5.43 5.11 5.44 5.26 5.58
Ash % 5.50 5.50 5.55 5.44 5.26 5.58
Salt as Nacl % 0.38 0.38 0.36 0.36 0.33 0.33
Me for Poultry Rcal/kg 3504 3499 3436 3583 3508 3552
T.carbohydrate % 28.64 30.47 26.91 31.11 29.17 30.47
Aspartic % 3.63 3.40 4.18 3.44 4.04 3.42
Threonine % 1.26 1.16 1.37 1.21 1.36 1.17
Serine % 1.64 1.54 1.83 1.60 1.80 1.56
Glutamic % 6.77 6.15 786 6.34 7.44 6.25
Glycine % 1.28 1.30 1.38 0.91 1.13 1.18
Alanine % 1.38 1.27 1.62 1.57 1.71 1.36
Valine % 1.38 1.24 1.54 1.33 1.51 1.29
Methionine % 0.43 0.45 0.51 0.45 0.46 0.43
Isoleucine % 1.27 1.14 1.41 1.22 1.37 1.18
Leucine % 2.31 2.12 2.60 2.22 2.50 2.16
Tyrocine % 1.12 1.05 1.27 1.09 1.22 1.04
Phenylalanine % 1.43 1.33 1.61 1.40 1.59 1.32
Ustidine % 0.83 0.76 0.92 0.77 0.92 0.74
Lysine % 1.72 1.83 2.18 1.93 2.14 1.78
Arginine % 2.07 1.94 2.59 2.03 2.58 1.99
Cysteine % 0.26 0.26 0.24 0.21 0.26 0.27
Tryptophan % 0.49 0.42 0.39 0.41 0.39 0.40

Table 5. Agronomic parameters of released groundnut varieties in Ethiopia

Seed yield Days to Year of Source
Variety Oil content (%)
(kg ha-1) maturity Release
Manipinter NA 48-50 150-155 NA N IS A 1998
Shulamith 50-65 44-49 150-155 1976 N IS A 1998
NC-343 40-60 45-50 160-165 1986 NISA 1998
NC-4X 50-70 42-49 150-155 1986 N IS A 1998
Roba 20-22 43-45 160-165 1989 N IS A 1998
BatiSedi 15-20 44-46 155-160 1993 NISA 1998
Bulqi-01 20-23 44-45 140-145 2002 MoRAD 2004
Lote-01 20-22 45-46 130-135 2002 MoRAD 2004
Werer-961 26-28 44-45 127-130 2004 MoRAD 2004
Werer-962 29-30 4648 130-135 2004 MoRAD 2004
Werer-963 20-22 4546 130-135 2004 MoRAD 2004
Werer-964 20-21 4547 160-165 2004 MoRAD 2004
ICGV-94205 26-70 Confectionery 145-160 2008 MoRAD 2008
ICGV-94222 20-80 Confectionery 150-160 2008 MoRAD 2008
ICGV-93164 26-80 Confectionery 130-155 2008 MoRAD 2008

ICGV-93370 60 Confectionery 115 2009 MoRAD 2009

Fetene 60-62 50-52 115-120 2009 MoRAD 2009

Eta 20-22 Confectionery 125-135 2010 MoRAD 2011
Fenta 18-21 4243 130-140 2010 MoRAD 2011
Source: Adugna Wakjira 1991, Getinet et al 1997, NA = not available

[129]
 Sebil Vol. 16

S ealin g up o f Sunflow er
Sunflower is an introduced oilseed crop in Ethiopia. It has been cultivated
in Northern Ethiopia since the 19th century. In Ethiopia, sunflower has
similar ecology as maize and soybean and is adapted to most state farms.
During the 1980; s sunflower variety trials were tested in Arsi, Bale, and
North West and Western State farms. Sunflower was found to be adapted
from lowlands such as Werer (750 masl) to highlands such as Herero (2400
masl). Seed yields and oil contents were satisfactory as compared to most
oilseeds. The problems associated with sunflower were diseases
particularly downy mildew, head and stem rot, leaf spot and rust.
Although sunflower has been a m ajor cro p for state farms, it was pulled
out of production due to diseases mostly Scloretina wilt, head and stem rot
as a result of poor crop management and crop rotation. The crop has not
entered in to peasant holdings probably due to bird problem.

Sunflower variety trials started at Alemaya with testing of 18 varieties

during 1960 through 1967. Although, disease incidence and oil content
data were not reported seed yields were 11 to 31 quintal/ha with an
average of 19.8 q/ha (Bantayehu 1968). Sunflower research in crop
management, disease and pest as well as multi-location variety testing
continued during 1968 to 1976 with Awassa as coordinating center.
Variety testing during 1968-76 resulted in the recommendation of an
introduced variety Russian Black for production in Awassa area. During
1982-85, ten sunflower genotypes were tested at seven locations including
state farms (Table 6). Seed yields were 16-19 q/ha with an average of 18
quintals/ha. Among locations, seed yields were 29 quintal/ha at Birr
Valley and 10 quintal/ha at Sheneka. Oil contents were > 40% at most of
the locations. Furthermore adaptation tests indicated that sunflower can be
cultivated from lowlands such as Werer (750 masl) and to high lands as
high as Herero (2400 masl). Since then coordinated and concerted effort
was interrupted and recently testing of introduced hybrid varieties by
foreign companies is being conducted in the rift valley and
BeneshangulGumz. Currently there are six open pollinated and three
hybrid varieties registered in Ethiopia (Table 7). The hybrid varieties were
registered by International Companies such as Pioneer Hybrid and Ashiraf
, Vibha Seeds, and Minerva PLC.

[130]
 Sebil Vol. 16

Table 6. Seed yield in q/ha of ten sunflower varieties grown in the national variety trial at seven locations during 1982-85
cropping season with fertilizer rate of 41/46 kg/ha of N/P2O5(Solomon 1988).
Treatments Locations
Debre
Herero Sheneka Kulumsa Zeit Hawassa Upper Birr Bako Mean
Hemus 10.9 10.3 23.4 19.3 20.0 33.8 15.9 19.2
Argentario 14.4 10.0 24.0 18.3 18.1 30.1 14.7 19.6
Sunhi 301-A 14.2 11.0 23.3 20.6 17.9 30.0 15.1 18.9
Sungaro 380-A 15.0 13.0 23.0 18.0 23.3 28.1 16.6 19.4
Amiata 14.4 9.0 22.3 19.9 23.4 29.1 18.4 18.8
Vinimik 14.2 9.9 22.0 19.6 18.7 30.2 16.8 18.7
L.C.I 15.9 10.7 20.9 17.2 20.1 29.1 15.9 18.9
Novisad 20 16.9 9.3 23.8 16.1 19.1 29.1 13.2 18.4
Russian Black 15.5 9.6 22.7 14.7 19.8 29.6 20.3 18.3
Eliodoro 12.1 8.7 21.2 15.6 17.1 23.2 12.8 16.6
Mean 14.3 10.0 22.7 18.0 19.7 29.2 15.9 18.6
Altitude in Meters 1432 2330 2200 1900 1700 1690 1650

Table 7. Agronomic parameters of released sunflower open pollinated and hybrid varieties in Ethiopia (MoA 2015).

Seed Yield q/ha

Farmers Year of
Variety Days to Mature Plant height Research plot Oil Content Release
Ayehu 125-160 138-272 25-29 23-24 32.7 2014

X6859 150 190 21-25 38.0 2014

Camara II 151 204 2-25 42.5 2014
MLN 11037 151 209 17-20 40.4 2014
Vincenzo 155 207 18-22 38.7 2014
KAZANOVA (hybrid) 110 160-170 45 20-30 43.4 2013
NS-H-45 (hybrid 125-135 180-195 55 25-35 43.1 2013
NS-H-111 (hybrid 105-115 165-185 50 25-35 46.9 2013
Oissa (NSH-25) 110-150 183 18 - 2005

Sunflower production in Ethiopia is limited by disease infestation and

breeding for disease resistance is a priority. Among insect pests African
ball warm, cabbage saw fly, linseed flea beetle and leaf hopper are
important. In addition 22 minor insects were identified.

D iseases: Survey of diseases showed four major diseases and eleven

minor diseases (Geremew et al. 2012). Among these disease downy
mildew (Plasmopharahalstedii), rust (Pucciniahelianthi), Sclerotinia wilt,
stem and head rot (Sclerotiniasclero riorum) and leaf spot (Septoriahelkianthi)
were economically important. These diseases particularly wilt, stem and
head rot were responsible for the reduction of sunflower cultivation in the
state farms.

[131]
 Sebil Vol. 16

L o n " T erm

In ten sificatio n o f G en etic and Physiological R e se a rch to Develop

Su p er Noug
Noug is indigenous and major oil seed to Ethiopia. Noug is inherently low
yielder with low harvest index. When fertilizer is applied or under high
fertility condition the crop grows luxuriously having massive vegetative
growth rather than economic response. The breeding methods of noug are
not well developed and we have very little knowledge of the genetics and
physiology of noug. It appears that research on basic genetics and
photosynthesis efficiency is required to select an efficient plant type. In
the short term availability of genetically, pathologically and physically
pure seed of improved noug varieties would help to increase the
production and productivity of noug. Scaling up of noug production
technologies in major growing areas improve its production. The object of
seed production is to produce genetically, pathologically and physically
pure seed of a certain variety of a crop. In noug isolation distances, the role
of insects in noug pollination, seed renewal rates etc remains to be studied.
In this strategy we believe that the seed production system employed on
sunflower by Russian plant breeders during the 1920 and 1930 can be
adopted. The method of T. D. Lysenko continuously improves and
genetically maintains a registered variety (Figure 14). Honey bees can be
utilized to improve the seed production and productivity in noug
(Mehmet et al 2009). Initially, an experiment is required to assess the
impact of honey pollination on the seed set of noug. Four treatments
namely open pollination, pollination with honey bees in cages, hand
pollination in cages and pollination in cages without hand and bee
pollination should be planted using two genotypes in three replications.
Bee keeping is a major activity of Ethiopian farmers and noug production
can be incorporated with bee keeping.

|1323
 Sebil Vol. 16

select 100010 2000 heads on and Hul1

Analysis 1000-1200 heads selected for
first year Nursery

First Year Nursery

Select 400 to 600 heads and

Orobanche resistance evaluation for every plant

plants with the

seed the seed nursery. In this

nursery
3-4 intensive rouging
is made with
mutual inter pollination,

d is grown

from the super elite

Figure 14. T. D. Lysenko method of sunflower seed production (Postovoit 1964 )

(■crmpla.Mii C h aracterization
The object of the germplasm characterization must be to isolate variants
for various agronomic traits such as earliness, kernel size, disease
resistance, seed size and oil content. Those variants can be utilized to
develop inbred lines that can be utilized in various breeding procedures.

The available noug germplasm has been intensively characterized 241

before, however, valuable variants were not carefully advanced due to in
availability of techniques such as double haploids etc. It appears that the
available noug germplasm should be intensively characterized for days to
flowering and maturity, plant height, number of heads, branches, and
seeds per head and per plant; size of leaf, head and seed; disease score

[133]
 Sebil Vol. 16

(leaf spot, blight, powdery mildew etc);insect incidence (noug fly, black
pollen beetle); self-compatibility score, 1000 seed weight, hull percentage,
kernel percentage, oil content and fatty acid composition. In some cases
development of techniques to measure a trait such as kernel and hull
percentage may be required. The methods of disease and pest score
should be carefully designed. During the course of characterization,
important variants for various agronomic traits can be isolated or tagged.

In bred L in e Development
Inbreeds can be used in various genetic studies, recurrent selections,
synthetic variety development and as well as hybrid variety using hand
pollination. IBL are homozygous for the trait of interest. IBL can be
developed using conventional techniques through selfing and advancing
generations in plant to row method. Selection can be made within and
among rows up to the third generation and then selection within rows can
be terminated. Although, the level of inbreeding in noug has not been
established it is likely that selfing can be continued up to the fifth
generation.

The second method and shortest and reliable method would be using
double haploids (Mistiru et al. 2010). The technique of double haploids
can be utilized to shorten the selfing generation. Double haploids are
homozygous lines and can be utilized for genetic studies as well as
recurrent selections. IBL can be evaluated in the field through
coziventional and Marker Assisted Selection (MAS) techniques. Molecular
techniques such as AFLP, SSR and TRAP can be used to assess the genetic
variability of noug IBL to a maximum precision. These techniques were
effectively utilized to assess the genetic variability of sunflower (Bert etal.
2004, You etal. 2009) Maize IBL. Once IBL are developed and evaluated,
they should be evaluated for their general or specific combining ability
using testers for hybrid variety development. It is mandatory that IBL
should be disease resistant containing high oil in their seed.

G en etic Stu d ies o f Econom ically Im portant T r a its

The inheritance of agronomic traits such as oil content is important to
design breeding programs towards their improvement. Therefore, genetic
studies for days to flower, plant height, oil content, seed size, hull
percentage, resistance for diseases should be carried out using IBL

[1341|
 Sebil Vol. 16

preferably developed using microspore culture. The prerequisite for

genetic study is that contrasting homozygous genotypes should be
available so that their Fi, F 2 and BCi can be generated to estimate their
mode of inheritance and number of genes controlling the trait. Once Fi, F 2
and BCi are produced they should be grown together along with their
parents to arrive at conclusions. If parents are not homozygous for the
trait of interest, the result would be doubtful or inconclusive. Some traits
such as oil content that are controlled by several loci with low heritability
may be difficult to study their inheritance. In such cases biometrical
techniques can be utilized to estimate their heritability from multi location
tests data.

Method of Reserve
The method of reserves developed by Posovoit in the USSR for improving
cultivars was widely used and highly successful in improving oil content,
Orobanche resistance, and disease resistance of sunflower. The method is a
form of recurrent selection that includes progeny evaluation and
subsequent cross pollination among superior progenies (Fig 15).
According to Postovoit (1967) from 10,000 to 15 000 plants are selected
from a heterogeneous population and their seed analyzed for hull and oil
percentage. Based on this analysis the progeny from 1000 to 1200 heads
are evaluated for agronomic, disease and seed quality traits in single row
plots in two replications. A check, consisting of the best cultivar most
similar to the lines being evaluated, is included in every third plot as a
control. On the basis of the first year observation, 15 to 20% of the best
plants are evaluated for second year the original seed of the best
remaining 20 to 50 plants is planted in a replicated isolation nursery for
cross pollination. Undesirable plants are removed during the season seeds
from individual heads are again in the next cycle of selection, for cultivar
testing and analyzed for hull and oil percentage, and the seed from the
best plants is mixed for use in the next cycle of selection, for cultivar
testing and for seed production.

Recurrent Selection
Recurrent selection methods have been used to improve the performance
of populations for quantitatively inherited traits mainly in maize. It is a
cyclical process which except for mass selection includes three phases
namely development of progenies, progeny evaluation and recombination
of selected families of progenies. Although most recurrent selection

[135]
 Sebil Vol. 16

methods include these three phases they vary in types of progenies

(inbreds, full-sib, half-sib etc), number of progenies evaluated, number of
selected families , parental control and the type of progeny intermated. In
all recurrent selection methods, two goals remain constant namely
maintaining the genetic variability of the populations and increasing the
mean of the population to facilitate long term selection. Generally seven
recurrent selection methods namely
l.M ass, 2. Modified ear to row 3. Half-sib with inbred tester 4. Full-
sib^. Si progeny 6. S 2 progeny and Reciprocal full sib selection
methods are used

[136]
 Sebil Vol. 16

Studies on the efficiency of these methods indicated that, all selection

methods were successful in significantly improving the population per se
performance for grain yield. S 2 progeny selection had the greatest and
mass the least impact in improving grain yield.
10,000 to 15,000 single plant selections
Oil and Hull Analysis

Nursery
First Year Nursery

Orobanche resistance evaluation 150-200 heads

selected, for second year nurserv
I Second Year Nursery*
Second ?rs 20-50
Heads selected for
cross TTtnrrtrrt 1i u i i c r

I
Seeds from cross pollination nursery are used
for preliminary and competitive strain
testing

Preliminary evaluation of best

Cultivars conducted on the basis of competitive
testing

STATE WISE STRAIN TESTING

Figure 15. Postovoit (1967) method of sunflower breeding.

[137]
 Sebil Vol. 16

Hybrid V arieties
Probably no other biological technology contributed to the world's food
supply than hybrid varieties such as in corn. Hybrids are Fi generation
between two inbred lines that have good combining ability or hybrid
vigor. Hybrid varieties have shown impressive yield advantage over their
parents in corn, sunflower, rice and sorghum. Much of the hybrid vigor
exhibited by Fi hybrid varieties is lost in the next generation.

It has been suggested that the low seed yield in noug is due to its self­
incompatibility (Riley and Hiruy 1989). However, some populations do set
seed upon selfing indicating that there exists some degree of self­
compatibility. Recently, Mulatu and Bryngelsson (2010) suggested that
there are a minimum of ten self-incompatible alleles and oneself
compatible allele at the S-locus in noug. Furthermore, cross and self-
pollination in several populations of noug helped in the identification of
self-compatible lines. The Fi hybrid between self-compatible lines
exhibited hybrid vigor with synchronized flowering and outperformed
their parents. This has opened the possibility of hybrid variety
development once sufficient number of self-compatible lines with
desirable agronomic traits is available. In this regard, the available
germplasm may be a good source of self-compatible lines that can be used
for cross fertilization and consequently to hybrids.

Adaptation O il Palm Production and P rocessin g Technology

Oil palm is very efficient plant that produces 50% oil in its mesocarp or
fleshy part of its fruit. Oil palms are found in the genus Elaeis.
Elaeisguineensis originates in West Africa and Elaeisoleifera originated in
Central and South America. Both Species are high yielders and efficient
utilizes of resources. Mature palm tree reaches to a height of 20 meters
with leaf length of five meters. It has no branches and can live up to 50
years. However trees are replaced 20-25 years due to declaiming yield.
Plants start fruiting in three years. At harvesting each Fruit Bunch can
weigh up to 50 kg and their processing/crushing has to be done soon after
harvesting within 48 hours.

Ecology o f Oil palm

Oil palm is the most important vegetab'e oil source globally due to its low
cost of production and high yield. ’ Ience, the price of palm oil is the

[1T3J
 Sebil Vol. 16

lowest even it is imported. The natural habitat of oil palm is 10° N and 10 °
S of the equator. Corley and Tinker (2008) and Harley (1988) summarized
the most suitable ecology of oil palm. These include: Annual rain fall of
2000 mm per year evenly distributed throughout the year without any
marked dry period and preferably 100 mm in each month, mean®
maximum temperatures of about 29-33° C and or mean minimum
temperatures of about 22-24° C, sunshine of 5-7 hours/day in all months
and solar radiation of 15MJ/m2 per day, relative humidity above 85 %, low
vapor pressure deficit, no extreme temperature and wind speed, soil
neither excessively nor poorly drained and slope not more than 20%.

S u itab le A reas o f O il P ahn in Ethiopia

The areas identified as the most suitable for oil palm cultivation are shown
on (Figure 16 Most identified potential areas failed in the SNNP regional
state). There are some pocket areas with more than 1800 mm of total
annual rainfall in the Benshangul, Oromia , Amhara and eastern parts of
SNNP regions but the distribution of the rainfall is concentrated to five to
six months.

Figure 16. Suitable areas of oil palm in Ethiopia.

O il P ahn V arieties
The most important traits of oil palm are slow growth, fresh fruit bunch,
mesocarp to fruit, shell to fruit and oil to bunch ratio. The most important

[139]
 Sebil Vol. 16

factor in the success of oil palm industry is the discovery of a single gene
responsible for thin shell thickness (Stephen 2009). Genotypes that have
thin shell are Tenera, and contain up to 30% higher oil than thick shell
Duras. The hybrid between them is Pisiferas and are female sterile.
Utilization of hetrosis of Fi hybrids in oil palm has been utilized
extensively. Fi hybrids are first generations offspring of two distinct
different but homozygous parents each with two sets of chromosomes.

Oil palm is adapted in lower altitudes and warmer areas. However, high
altitude adapted or cold tolerant oil palm hybrid teneras bred in Costa Rica
have been successfully adapted in higher altitude areas up to 1000 masl in
east Africa including Cameroon, Kenya, Ethiopia, Uganda and Tanzania,
Malawi beginning in Ethiopia (Chapman et al 2003). The plants can
tolerate temperature lower than those suitable for classic oil palm hybrids
and still produce more oil in cooler condition than other cultivars. FAO
has been pioneering this development of oil palm in poor rural
communities in Africa with ASD de Costa Rica as a source of vitamin E
and A, energy and protein as well as providing small scale portable
presses easily adapted to village activities. Additional benefit of cold
tolerant hybrids oil palm is complete cover of the ground and stabilize the
environment. Cold tolerant hybrids or altitude adapted oil palm hybrid
teneras seed is available from ASD Costa Rica where the palm was bred.
Twenty years of breeding using DAM1 deli crosses with Cameroon and
Tanzania selections have led to the development of precocious bearing
cold tolerant oil palms.

Cold tolerant hybrids were planted on 100 ha of land at Gelesha in

Gambella Region in 1987 at 1000 masl. In addition a hybrid variety test
that included 13 hybrids in Randomized Complete Block Design was
planted. However data was not collected. The preliminary data have
benefited other neighboring countries that are expanding and growing
very fast towards self-sufficiency. Around Gelesha farmers are growing
oil palm F2 and F3 generations. Pilot Production of oil palm plantation on
100 ha is available at Gelehsa in Gambella. At Gelesha a hybrid oil palm
trial was planted by FAO experts during 1987. Unfortunately, data was
not collected. Currently both the plantation and the trial have passed their
economic life. Observations at Gelesha indicated that fresh fruit bunch
yield of 8-12/tree with up to 40 kg was obtained. As the demand for
vegetable oil increases for food and raw material for various industries

[140]
 Sebil Vol. 16

such as detergent, cosmetics and pharmaceuticals, the production of palm

oil locally has become vital Oil palm is being planted by Indian Companies
along Baro river. The production system of oil palm in Asia is by private
companies while in Africa small scale farmers are the major producers.

N utritional con ten t o f palm oil

The nutritional content of palm oil is explained by its fatty acid
composition and carotene content. The red pigment of the oil is linked
with the carotene content. The high carotene and vitamin E content offered
antioxidant property that is supposed to resist cancer. As source of energy,
it also supplies calorie requirement of our body. The typical blend in palm
oil is 45% palmitic, 40% oleic, 10% linoleic and 5% stearic (Sumathi et al.,
2008). The presence of higher percentage of palmitic acid contributes to
increased degree of saturation (high melting point). This allowed the palm
oil (and its products) good resistance to oxidation and heat so that it is an
ideal ingredient in frying oil blends (Sumathi et al., 2008). However, some
literatures reported health complications with respect to higher
consumption of saturated fatty acids. On the other hand, inclusion of
unsaturated fatty acids in our diets has health benefits for their ease in
metabolism and assimilation. In this regard, those oleaginous crops with
higher percentage in oleic and linoleic fatty acids (Table 1) do have
nutritional advantage in our daily diets.

Improving the biochemical, chemical, physical activity of the fruit and

physico-chemical property of the oil determines the nutritional content, oil
quality, storage stability and processing. Therefore, involvement of
multidisciplinary teams and stakeholders has paramount importance in
the productive chain of oil palm.

R eferen ces

Adugna Wakjira 1992. Groundnut Breeding in Ethiopia. P 51-56.

Proceedings of The First National Oilseeds Workshop. Dec. 3-5, 1991.
Addis Ababa, Ethiopia.
Agdew Bekele, Getinet Alemaw and HabtamuZeleke 2012. soybean
Genetic variability under southern Ethiopia condition,, agronomic
condition and their association with yield Lambert Academic
Publishing, Germany.

[141]
 Sebil Vol. 16

Central Statistical Agency 1988-2010. Report on Area and Production of

Major Crops, Private Peasant Holdings Meher season. Statistical
Bulletins, Addis Ababa, Ethiopia.
Bantayehu Gelaw 1968. Progress Report on Cereals, Oilseeds and Pulses,
Haile Selassie I University, Collage of Agriculture, Experiment Bulletin
64
Ethiopian Statistical Authority 2006-2014. Statistical abstracts Addis Abeba
Ethiopian Customs Authority 2006-2014, Addis Abeba.
FAO 2010. Statistical Abstract 2005-2007.
Fekadu Gurmu, Hussen Mohammed and Getinet Alemaw 2011. Genotype
x environment interaction and stability of soybean for grain yield and
nutritional quality. African Crop Science journal 17:87-99.
Getinet Alemaw and Adefris TekleWold 1995. An agronomic and seed
quality evaluation of niger (Guizotinabyssinica CASS) germplasm grown
in Ethiopia. Plant Breeding 114:375-376.
Kushairi, A., Rajanaidu, N., Mohd Din A. Z. A., Noh and junaidah J. 2003.
PS5 breeding populations selected for thin shelled teneras, MPOB
information series 183.
Champman, K. R. Escobar Recardo and Griffee Peter 2003. Cold Tolerant
or Altitude Adapted Oil Palm Hybrid Development Initiatives in Asia
and African Region, AU. J. T 6(3)134-138.
Corley, R.H.V. and Tinker P. B. 2003. The Oil Palm, Fourth Edition,
Blackwell Publishing O xford England.
International oil palm Conference, 2015: Bogota, Colombia
LeClercq 1969. Un sterilite male cytoplasmique chez le tournesol Ann.
Amelior. Plantes 19(2)99-106.
Ministry of Water and Energy 2007. The Biofuel Development and
Utilization Strategy of Ethiopia September 2007 Addis
Abeba.
Ministry of Agriculture and Rural Development 2004. Crop Variety
Register, Issue no.7. Addis Abeba Ethiopia.
Ministry of Agriculture and Rural Development 2005. Crop Variety
Register, Issue no.8. Addis Abeba Ethiopia.
Ministry of Agriculture and Rural Development 2007. Crop Variety
Register, Issue no.10. Addis Abeba Ethiopia.
Ministry of Agriculture and Rural Development 2008. Crop Variety
Register, Issue no.11. Addis Abeba Ethiopia.
Ministry of Agriculture and Rural Development 2009. Crop Variety
Register, Issue no.12. Addis Abeba Ethiopia.

[142]
 Sebil VoL 16

Ministry of Agriculture and Rural Development 2010. Crop Variety

Register, Issue no.13. Addis Abeba Ethiopia.
Ministry of Agriculture and Rural Development 2011. Animal and Plant
health regulatory Directorate, Crop Variety Register No 14, Addis
Abeba, Ethiopia
Ministry of Agriculture and Rural Development 2014. Animal and Plant
health regulatory Directorate, Crop Variety Register No 15, Addis
Abeba, Ethiopia
Mistiru Tesfaye, Telaye Feyissa and LikeleshGugsa 2010. Embrayogenic
callus induction and regeneration in anther culture of noug
(Guizotiaabyssinica (LF).CASS) Ethiopian J. of Sci. 33(1)49-58.
Morcillo, F.; Cros, D.; Billotte,N.; Ngando-Ebongue, G.-F.; Domonhedo,H.;
Pizot, M.; Cuellar ,T.; Espeout, S.; Dhouib, R.; Bourgis, F.; Claverol,S.;
Tranbarger, T.J.; Nouy, B.; Arondel, V. 2012. Improving palm oil
quality through identification and mapping of the lipase gene causing
oil deterioration. DOI: 10.1038/ncomms 3160 / www.nature.com
/nature communications.
Mulatu Geleta and Tomas Bryngelsson 2010. Population genetics of self­
incompatibility and developing self-compatible genotypes in niger
(Guizotiaabyssinica) Euphytica
National Seed Industry Agency 1998. Crop Variety Register Issue no. 1
Addis Abeba Ethiopia.
National Seed Industry Agency 2001. Crop Variety Register Issue no. 4
Addis Abeba Ethiopia.
Pustovoit V.S. (ed) 1967. Hand book of selection and seed growing of oil
plants, Izdatelstvo "Kolos" Moscow, (Translated by Israel Program for
Scientific Translation, Jerusalem 1973).
Poetsch, J., Haupenthal, D.;Lewandowski, I.; Oberlander,D.; Hilger,T.
2012. Acrocomiaaculeata -a sustainable oil crop. Rural 21 - 03/2012.
Science World 41-44.
Riley,K.W. and Hiruy Belayneh 1989. Niger. In: Oil Crops of the World,
pp.394-403, (Robbelen, G., Doweny, R.K. and Ashri, A., eds). Me Graw
Hill Publishing Company, New York.
Seegeler, C. J. P. 1983. Oil Crops in Ethiopia. Their Taxonomy and
Agricultural Significance. Centre for Agricultural Publication and
Documentation, Wageningen, The Netherlands. 368 pp.
Scrimgeour, C. 2005. Chemistry of fatty acids: composition and structure.
Shahidi, F. (Ed.). Sixth edition. Bailey's Industrial Oil and Fat Products,
John Wiley & Sons, Inc.

[143]
 Sebil Vol. 16

Sumathi, S., Chai, S.P. and Mohamed, A.R.. 2008. Utilization of oil palm as
a source of renewable energy in Malaysia. Renewable and Sustainable
Energy Reviews (12): 2404-2421.
TenawWorkayehu, ZenebeMekonen, TarekegneLirie, and BerhanuEriso
2011. Oil crop Research in the Southern Region of Ethiopia. In:
Geremew Terefe, Adugna Wakjira and Dereje Gorfu (eds). Oilseeds:
Engine for Economic Development. Ethiopian Institute of Agricultural
Research (EIAR), P.O. Box: 2003, Addis Ababa, Ethiopia.
YusofBasiron 2007. Palm Oil Production Through Sustainable Plantations
Eur. J. Lipid Sci. Technol. 109: 289-295
You B., Cai., B. A. Vick and J. Hu 2009. Genetic diversity and relationships
among 177 public sunflower inbred lines assessed by TRAP Markers,
Crop Science Vol.48:1242-1249.

[144]