Citation
Proceedings of the 12th Regional Wheat Workshop for Eastern, Central, and Southern Africa

Material Information

Title:
Proceedings of the 12th Regional Wheat Workshop for Eastern, Central, and Southern Africa
Series Title:
Proceedings of the ... Regional Wheat Workshop for Eastern, Central, and Southern Africa
Creator:
International Maize and Wheat Improvement Center (CIMMYT)
Place of Publication:
Mexico, D. F.
Publisher:
International Maize and Wheat Improvement Center (CIMMYT) ; Kenya Agricultural Research Institute (KARI)
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Africa ( LCSH )
Farming ( LCSH )
Agriculture ( LCSH )
Farm life ( LCSH )
Genre:
serial ( sobekcm )
Spatial Coverage:
Africa

Notes

Funding:
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services (UFDC@uflib.ufl.edu) with any additional information they can provide.

Downloads

This item has the following downloads:


Full Text
3m. I
* 0 C E E* 0 I0
ofhe I Reioa Wha Worsho
Vm
m
Amm
............ ,i), .... ... .. ....
.. .. ....................................
......................m ::::::::::.........
m~lt
IICIMmmm m mm nmmA
KmAmm




Proceedings of the
12th Regional Wheat Workshop for Eastern, Central and Southern Africa
Nakuru, Kenya, 22-26 November 2004
SBAR Bayer MONSANTO
A im ine,
a, CIMMYI
NOW




CIMMYT (http://www.cimmyt.org/) is an internationally funded, not-for-profit organization that conducts research and training related to maize and wheat throughout the developing world. Drawing on strong science and effective partnerships, CIMMYT works to create, share, and use knowledge and technology to increase food security, improve the productivity and profitability of farming systems, and sustain natural resources. Financial support for CIMMYT's work comes from many sources, including the members of the Consultative Group on International Agricultural Research (CGIAR) (http://www.cgiar.org/), national governments, foundations, development banks, and other public and private agencies.
The Kenya Agricultural Research Institute (KARl) (http://www.kari.org/) was established in 1979 with the express mission of increasing sustainable agricultural production by generating appropriate technologies through research, and disseminating these to the farming community. Inherent to this mission is the protection, conservation, and improvement of the basic resources, both natural and human. Such resources are critical for Kenya's agricultural development and expansion of the nation's scientific and technological capacity. KARI has an extensive history of productive collaborators with national and international institutes and universities, as well as with the private sector.
KARI/CIMMYT 2006. All rights reserved. The designations employed in the presentation of materials in this publication do not imply the expression of any opinion whatsoever on the part of KARI, CIMMYT or their contributory organizations concerning the legal status of any country, territory, city, or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. KARI and CIMMYT encourage fair use of this material. Proper citation is requested.
Correct citation: M.G. Kinyua, J. Kamwaga, J.O. Owuoche, A. C.Ndiema, P.N. Njau, D. K. Friesen, D. Ouya
(Editors) 2006. Proceedings of the 12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakurn, Kenya, 22-26 November 2004. Mexico, D.F.; CIMMYT and KARI
ISBN: 970-648-135-4
AGROVOC Descriptors Triticum durum; Triticum aestivum; Varieties; Plant breeding; Breeding methods;
Plant production; Production increase; Plant diseases; Food production; Food security; Production factors; Economic policies; Pests of plants; Pest resistance; Yield increases; Kenya; Ethiopia; Tanzania; Sudan; Africa
Additional Keywords CIMMYT; KARI
AGRIS Category Codes F30 Plant Genetics and Breeding
E14 Development Economics and Policies
H20 Plant Diseases
Dewey Decimal Classification 633.116




Introduction
Like its predecessors, the 12th Regional Wheat Workshop for Eastern, Central and Southern Africa, held at Merica Hotel in Nakuru, Kenya, 22-26 November 2004 provided a forum where wheat scientists from the region could come together to exchange their results, views, and ideas on the way forward in securing this important food crop.
Ethiopia has an annual production of 1,600,000 metric tons, of both bread and durum wheats, produced mainly on small scale farms. A large gap exists between wheat production and demand in Ethiopia, and the country relies heavily on imports of nearly 800,000 metric tons each year. Kenya's domestic bread wheat requirement, which currently stands at 720,000 tons p.a., is projected to reach 1 million tons by 2010. Like Ethiopia, Kenya produces less wheat than she needs, importing some 400,000 tons each year. Clearly therefore, increased wheat production in the regionthrough approaches such as breeding, agronomy, crop protection and improved economic policies, is not just desirable, but essential for the food security of millions of people.
This proceedings presents valuable new information from research conducted in Kenya (11I articles), Ethiopia (9), The Sudan (2), and Tanzania (1), and is organized into four sections, namely Agronomy, Breeding, Protection, And Economics.
The Agronomy section contains 6 papers, ranging in subj ect mater from water and fertilizer use efficiency in wheat varieties, to a comparison of field pea and barley as break crops in wheat farming. Breeding is the largest section, with 14 articles, including results from a three-year survey of the wheat stem rust situation in Kenya. The Protection section, with 3 articles, reports on the chemical management of weeds, and the Russian wheat aphid Diuraphis noxia-a new and damaging pest in the region. Results from a survey of the natural enemies of D. noxia are also presented, as the starting point for designing a biological control strategy for the aphid. In his economics article, Professor Abbas Elsir M. Elamin deals with the financial and economic profitability of wheat production in The Sudan, and its international competitiveness. The comprehensive analysis discusses government reform policies in the 1990s, and their role in wheat production, consumption and input use from that time to the present.
We believe that this volume represents an important contribution to the regional and global knowledge bases on wheat production and economics, and that a better understanding of the subject matter will contribute to the ultimate goal of improved livelihoods for the people of Africa.
Romano Kiome, PhD
Permanent Secretary, Ministry ofAgriculture




Acknowledgments
The Workshop organizers would like to acknowledge the contributions of several institutions and individuals towards the success of the workshop, and the publication of this proceedings.
We thank Monsanto, Bayer Crop science, and Osho chemicals, for funding the workshop; KARI and CIMMYT for technical and administrative support; and all the workshop participants, for their cooperation during all stages, starting with the preparation and presentation of the papers, to their review/revision for publication.
CIMMYT reviewers Karim Ammar, Hans-Joachim Braun, Jose Crossa, Hugo De Groote, Etienne Duveiller, Augustine Langyintuo, Ivan Monasterio, Wilfred Mwangi, Tom Payne, Javier Pena, Matthew Reynolds, Ken Sayre, Ravi Singh, Richard Trethowan, and Pat Wall, provided thorough and constructive comments on the manuscripts, including suggestions for improvement. This was a valuable contribution towards seeing the proceedings to completion, for which we are very grateful.
Dr. Miriam Kinyua
Chair, Organizing Committee




Contents
Part 1. AGRONOMY
Effect of Nitrogen Fertilizer Levels and Varieties on Gluten Content and Some Rheological Characteristics of Durum Wheat Flour
Bemnet Gashawbeza, Solomon Assefa, Ameha Yaekob, Alemayehu Zemede, Jemanesh
Kifetew and Bekele M ekuria .............................................................................................. 2
Impact of Irrigation Frequency and Farmyard Manure on Wheat Productivity on a SalineSodic Soil in Dongola, Sudan
Elmoiez M. Fadul and Mukhtar A. Mustafa .................................................................... 9
Consumptive Use of Water and Water Use Efficiency by Wheat (Triticum aestivum) in Relation to Irrigation and Nitrogen
Antony M. Kibe, Subedar Singh, and Naveen Kalra ..................................................... 21
Agronomic and Economic Evaluation of Break Crops and Management Practices on Grain Yield of Wheat at Shambo, Western Oromiya, Ethiopia
Tolera Abera and M athewos Belissa................................................................................ 35
Response of Bread Wheat to Nitrogen and Phosphorous Fertilizers at Different Agroecologies of Northwestern Ethiopia
Minale Liben, Alemayehu Assefa, Tilahun Tadesse and Abreham Mariye................... 41
Grain Yield, Water Use and Water Use Efficiency as Affected by Moisture Level Under Rain-Out Shelter
P. A. Ooro, M.G.Kinyua, and J.B.O. Ogola.................................................................. 46
Part 2. BREEDING
Current Status of Stem Rust in Wheat Production in Kenya
R. Wanyera, M.G Kinyua, P. Nau, J.W. Kamundia and S.Kilonzo.............................. 55
Genotype-by-Nutrient Interaction in Wheat Grown in a Marginal Environment in Kenya
J. Kamwaga, H. Okwaro, P. Nau, P. Kimurto, P. Ndungu. and E. Kimani.................. 63
Improving Wheat Productivity for the Drought Prone Areas of Kenya Using the Doubled Haploid Technique
Nau P. N, Kimurto, P.K, Kinyua M. G, Okwaro H. K and Ogolla J.B.0 ................... 67
Grain Yield Stability of Bread Wheat Genotypes in Favorable and Stressed Environments
Desalegn Debelo, Solomon Gelalcha, Balcha Yale, Bedada Girma, Berhanu Mamo, and D ebebe M asresha............................................................................................................... 76
Seedling and Adult Plant Resistance in Ethiopian Wheat Varieties to Local Puccinia graminis Isolates
Emebet Fekadu, Belayneh Admassu and Zerihun Kassaye........................................... 86
Evaluation of Kenyan Wheat (Triticum aestivum L.) Lines for Bread Making Quality
Kimani E.N., J. Ndung'u, M.G. Kinyua and J. Owuoche............................................... 90




Physiological Traits Associated with Drought Tolerance in Bread Wheat (Triticum Aestivum L.) under Tropical Conditions
Kimurto, P.K., M.G. Kinyua, J.B.O. Ogola, J.M. Macharia, andPN. Njau .............. 96
Allelism of Resistance Genes to Phaeosphaeria nodorum in Wheat
C. A. Kuwite and G. R. H ughes..................................................................................... 109
Evaluation of Kenyan Breadwheat (Triticum aetivum L.) Varieties for Resistance to Russian Wheat Aphid in Multi-location Trials
J. Maling'a, M. G. Kinyua, A. Kamau, J. K. Wanjama ,P. Njau, and
J. K am undia .................................................................................................................... 117
On-Farm Evaluation and Comparison of New and Old Wheat Varieties
R. V Ndondi, C.A. Kuwite andR.Shekibula ............................................................... 125
Evaluation of Improved Wheat Varieties Under Different Management Practices in Eastern Wallagga Highlands
ToleraAbera, Daba Feyisa, Girma W. Tsadik, Hasan Yusuf and Gemechu Keneni 134
Cell Membrane Stability (CMS) as Screening Technique for Drought Tolerance in Bread and Durum Wheat Genotypes
Alemayehu Zemede, H. Martens, and M.T. Labuschagne .......................................... 140
Physiological Races and Virulence Diversity ofPuccinia graminis f sp. tritici on Wheat in Ethiopia
Belayneh Admassu, Emebet Fekadu and Zerihun Kassaye......................................... 145
Participatory Evaluation of Bread Wheat Varieties in the Central Highlands of Ethiopia
Kassa Getu, Kassahun Zewdie, Yeshimebet Gebrehiwot andAddisu Alemayehu......... 151
Part 3. PROTECTION
Survey of Natural Enemies of the Russian Wheat Aphid, Diuraphis Noxia (Kurdijimov) in Kenya
M .M acharia, M Njuguna, and L Koros........................................................................ 161
Evaluation of the Herbicide Monitor, Alone and in Tank-Mixes for Weed Control in Wheat (Triticum aestivum L.)
D.O.K. Amadi, J.N. Kamwaga and G.O. Mbanda......................................................... 166
Control of the Russian Wheat Aphid, Diuraphis Noxia (Kurdijumov) in Wheat Using Systemic Insecticides in Kenya
M M acharia, M Njuguna andL Koros........................................................................ 175
Part 4. ECONOMICS
Analysis of Marketing and Pricing Policies on Technology, Input Use and Production of Wheat in the Sudan
A bbas Elsir M Elam in.................................................................................................... 181
Appendixes
Opening Speech ............................................ ................................... 195
Closing Speech ................................................ ...... ...................... 197
Question and Answer Sessions ........................................................................................199
Participants E-mail Addresses ........................................................................................231




Part 1. AGRONOMY
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




2
Effect of Nitrogen Fertilizer Levels and Varieties on Gluten Content and Some
Rheological Characteristics of Durum Wheat Flour
Bemnet Gashawbeza', Solomon Assefa', Ameha Yaekob', Alemayehu Zemede', Jemanesh Kifetew'and Bekele Mekuria2
'Ethiopian Agricultural Research Organization, Debre Zeit Agricultural Research Center, P.O. Box 32, Debre Zeit, Ethiopia;
2 Kality Food Share Company, P.O.Box 1819, AddisAbaba, Ethiopia
Abstract-Improved gluten and rheological characteristics of durum wheat
(Triticum durum Desf) have recently received increased attention due to their positive
effect on the quality of pasta products. An experiment was conducted at Debre Zeit and Akaki in the Central Highlands of Ethiopia, during the 2000/01 cropping season, to study the influence of nitrogen fertilization levels (0, 30, 60, 90 and 120 KgNha-1) on gluten content and some rheological characteristics (dough resistance, expansion and extensibility) of five dumm wheat varieties at Akaki and Debre Zeit. At Debre Zeit, gluten content significantly varied (P<0.05) depending on wheat genotype (variety) and nitrogen fertilizer levels, but at Akaki only varietal differences were significant. Nitrogen level increment consistently increased wet gluten content of varieties at Debre Ziet and; the highest nitrogen level, 120 Kgha-', gave significantly higher wet gluten percent 29.5 % than all other treatment levels, except 90 Kgha-' (27.3%). Tob 66 (30.8%) had the highest wet gluten content followed by Foka (27.5%). At Akaki, Tob 66 was the highest (24.9%), followed by Boohai (22.4%).
The studied rheological characteristics at Debre Zeit showed that nitrogen fertilizer application had no effect on either of the traits (dough resistance and extensibility), but varietal differences were observed in dough resistance. Therefore, durum wheat quality improvement should concentrate both on selection of appropriate varieties and management practices such as nitrogen fertilizer application. Further study on
input-responsive semi dwarf varieties is recommended.
Introduction
Durum wheat is used for pasta production because of its hard grain texture, amber color, and other grain quality traits related to endosperm protein. Increasing grain protein concentration
and improving other grain quality traits have been the major objectives in most varietal
development processes (Metho, Hammes and Beer, 1997). This is because the processing
quality of wheat is largely dependent on the amount and quality of endosperm protein. Grain protein content affects milling and other industrial qualities of wheat. As a result, premiums
are commonly paid for protein levels above base line (Woolfolk et al., 2002).
Gluten is an important endosperm protein that affects pasta quality. It is a visco-elastic
component of wheat dough responsible for physical dough properties (C'uric et al., 2001).
Due to its strong relation with greater cooked firmness and increased tolerance to
overcooking, strong gluten varieties are preferred (Josephides et al., 1987). Besides gluten
content, theological characteristics that measure gluten quality are also important criteria of
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




3
selection for quality durum wheat. D'Egidio et al. (1990) have shown that Alveogrphic
parameters are excellent indicator of durum wheat gluten properties.
Quality durum wheat grain used for production of pasta is achieved by selection of
appropriate variety, environment and management practices (Rharrabti et al., 2001). Under high rainfall and low soil fertility conditions grain protein production is limited (Simmonds,
1989). The effect of environment and management methods, especially availability of nitrogen in the soil during ripening, on vitreousness of durum wheat has been reported
(Mosconi and Bozzini, 1973; Ottman et al., 2000). Numerous authors have obtained
improved grain protein upon intensive nitrogen fertilization (Geleto et al., 1996;
Gashawbebeza et al., 2002; Virga et al., 2003). The work of Hadjichristodoulou (1979)
showed significant effects of nitrogen fertilization, genotype and location on vitreousness, a physical grain quality correlated with protein content, of durum wheat. Dexter et al. (1982)
also found significant difference between varieties for several quality traits at different
nitrogen fertilizer levels.
Durum wheat is a traditional crop for the Ethiopian farmers. Its production however has
been limited to locally made foods. The current increasing of food processing industries and
importance of pasta products by the urban people of the country has increased the importance
of quality durum wheat production.
The objective of this experiment was, therefore, to evaluate the effect of nitrogen
fertilizer and varieties on gluten content and some rheological properties of durum wheat
flour.
Materials and Methods
A field experiment was conducted during the 2000/01 cropping season under rain-fed
conditions at two locations in the Central Highlands of Ethiopia (Debre Zeit and Akaki).
Debre Zeit (80 44 N, 39 0 02'E) is mid-highland (1900 m.a.s.l.) characterized by moderate
rainfall (85 1mm average annual rainfall); 17.90C average mean temperature and Pellic
Vertisol soil; and Akaki (8 052'N, 38 0 47'E) mid altitude area (2100masl) characterized by
average annual rainfall of 1086mm and 15.60C average mean temperature. Five medium tall
to tall durum wheat varieties viz (Kilinto, Tob66, Foka, Assasa and Boohai) that were
selected for their industrial quality by the local pasta industries were planted at five different
nitrogen levels (0, 30, 60, 90 and 120 kgha' N) with uniform basal application of l0kgha'
phosphorus in the form of Triple Super Phosphate (TSP). The experiment was laid out in
Randomized Complete Block (RCB) in factorial arrangement with three replications. The plot
size was 3m x 4m (12m2) and data was recorded from 10.4m2. Seeding rate was 150 kgha'.
Nitrogen was split applied half at planting and the remaining half at full tillering. Wet gluten
content was determined from flour and was determined by gluten wash method (ICC standard No. 106/2) while rheological characteristics viz. extensibility (L in mm) and dough resistance (P=height X 1.1 in mm) were measured using Chopin Alveograph (ICC No. 121) (ICC, 2000).
Rheological data was taken for Debre Zeit only. Data was analyzed by ANOVA using
MSTATC.
Results
Wet gluten. ANOVA showed that wet gluen content significantly varied depending on
varieties, locations and; location X nitrogen level interaction was significant (P<0.05) but no
marked difference between the nitrogen levels in the combined analysis. However, the
difference between varieties and fertilizer levels were significant (P<0.05) at Debre Zeit. At
Debre Zeit, the highest nitrogen level (120 Kgha') had significantly higher wet gluten percent
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




4
than all other treatment levels except 90 Kgha'. At Akaki, on the other hand, only varietal
differences were significant (Table 1).
Varietal differences were evident at both locations. In the combined analysis Tob 66
gave the highest wet gluten content (27.9%) while the lowest was Kilinto (21.18%). The
response to applied nitrogen was higher at Debre Zeit and Tob 66 also gave the highest wet
gluten percent (30.8) while the least was Kilinto (22.55%). But no significant difference was
observed between Boohai, Foka and Assassa.
Dough resistance. Varietal difference was significant for Dough resistance. But no
marked difference between nitrogen levels was observed. Boohai (113.9) gave the highest
dough resistance followed by Foka (101.5) while the least was Assassa.
Dough extensibility. This trait was not affected by the applied nitrogen fertilizer levels
(Table 2). Besides, there was no statistically significant difference between varieties.
Table 1. Nitrogen fertilizer level effects on wet gluten (%) of durum wheat varieties at two
locations (2000/01)
Wet gluten
Treatment Debre Akaki Mean
Zeit
Nitrogen (Kg ha-') 0 24.4c 22.2 23.3
30 26.5bc 21.8 24.2
60 26.5bc 22.8 24.6
90 27.3ab 21.29 24.3
120 29.5a 20.95 25.2
LSD (0.05) 2.543 NS NS
CV (%) 12.91 15.57 14.1
Variety Boohai 27.02b 22.4ab 24.73b
Foka 27.5b 20.65b 24.1 lb
Kilinto 22.5c 19.8b 21.18c
Tob 66 30.8a 24.93a 27.9a
Assassa 26.2b 21.29b 21.29b
Means within a column with similar letters are not significantly different at P<0.05 by DMRT. NS =
no statistically significant difference
Table 2. Effect of nitrogen levels on theological characteristics of durum wheat flour at Debre
Zeit (2000/01)
Treatment Dough resistance Expansion Extensibility P/L
(P) (G) (L)
Nitrogen 0 95.6 19.35 78.8 1.1
30 97.8 18.09 68.6 1.4
60 93.7 19.1 76.9 1.2
90 88.8 18.5 73.4 1.2
120 101.7 19.27 70.1 1.4
Variety Boohai 113.9a 20.3 78.8 1.4
Foka 101.5ab 18.7 73.3 1.4
Kilinto 92.5bc 18.8 74.6 1.2
Tob66 91.8bc 18.0 68.7 1.3
Assassa 78.lc 18.6 72.1 1.1
LSD(0.05) 15.8 NS NS
CV(%) 22.59 11.6 18.64
Means within a column with similar letters are not significantly different at P<0.05
by DMRT. NS = no statistically significant difference
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




5
Generally, nitrogen level increment increased wet gluten contents of durum wheat varieties at
Debre Zeit but not on rheological characteristics (dough resistance and extensibility). The
response of gluten contents was linear with the respective applications of 0, 30, 60, 90, and
120 kgha-'. In similar studies, the positive effect of nitrogen increment on grain protein
(Prima et al., 1982 and Geleto et al., 1996) and also on gluten content of wheat (Prima et al.,
1982 and Varga et al., 2003) was reported.
At Akaki, on the other hand, the response to the applied nitrogen was not consistent. The
work of Campbell et al. (1993) also showed significant contribution of environmental factors
on grain nitrogen content of wheat in addition to nitrogen application. As in most grain
quality traits in wheat, protein content is known to be affected by genetic and environment
mainly location, fertilizer application and other management practices. The effect of nitrogen
fertilizer on whole meal protein content of durum wheat was reported (Gashawbeza et al.,
2002). Thus this difference could be ascribed to the difference between the two locations in
air temperature and degree of water-logging. Akaki is more waterlogged than Debre Zeit and
could have inhibited efficient utilization of the applied nitrogen (Gashawbeza et al., 2002).
According to Spiertz and Devos (1983), under high temperature especially during grain filling
period protein synthesis is favored. Therefore, the environment at Debre Zeit could favor
protein synthesis due to its relatively higher mean temperature and less water-logging stress.
Besides, no correlation between gluten content and grain yield at Debre Zeit (r0. 15)
suggesting improving gluten content couldn't risk grain yield. From the result of the study, for high rainfall and waterlogged environments like Akaki, it appears that other crop management practices such as drainage of excess water and adjustment of planting dates should be studied,
in addition to nitrogen application, to increase gluten content of durum wheat.
Gluten content of varieties was significant at both locations and Tob 66 had the highest
while kilinto had the lowest (Table 1). As shown in Fig. 1, at Debre Zeit, there is a general
linear increasing trend of gluten content of varieties with increasing nitrogen level. However, the increase was sharp beyond 90 N kg/ha for Tob 66, which gave 37 % gluten content at the
highest nitrogen fertilizer level. Besides, Tob 66 had consistently high wet gluten starting
from the nil nitrogen treatment to the highest level and, this could suggest the efficiency of
the variety in converting assimilates to protein. At Akaki on the other hand, there is no
definite trend for the applied nitrogen levels and gluten content (Fig.2).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




6
40
35
30
25
. 20
Assassa
-- Boohai 15 -9-Foka
*-Kilinto ,,6Tob 66 10
5
0
0 30 60 90 120
Nitrogen Level(kglha)
Fig. 1. Gluten Content of durum wheat varieties at different nitrogen levels at Debre Zeit
30
25
20
15
---- Kilinto
-2 -4 Tob 66
Foka 10
Boohai
-0Assassa
5
0
0 30 60 90 120
Nitrogen level (kglha)
Fig. 2. Gluten Content of durum wheat varieties at different nitrogen levels at Akaki
According to ICC (2001), the rheological characteristics of a dough are expressed as the
resistance of the dough to stretching and its extensibility until it begins to rupture. (Curic et
al., 2001) measured the physical properties of the dough of wheat flour, which primarily
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




7
depend on the gluten quality. In this study, the observed non-significant difference in
rheological properties due to increasing application of nitrogen fertilizer could suggest that
the traits are affected more by genotype rather than by environment in which varieties are
grown. Unlike this, however, Virga et al. (2003) reported the use of the intensive production system in bread wheat resulted in a significant improvement in dough resistance. Curic et al.
(2001) reported gluten quality of wheat to be dependent on location and region of cultivation.
In this study, varietal differences were observed in dough resistance. Assassa showed the least performance in dough resistance but it was within the acceptable range for pasta industry use.
From the alveograph curve, the resistance to stretching is taken from the maximum height
reached by the curve (P) while extensibility is taken from the total length (L) in the horizontal direction of the curve. Alveograph strength is taken from the area under the alveogram (W) X
I0-J ergs and the greater the reference area shows the higher the value of W. Alveograph
strength (W X 10-' ergs) and the ratio of dough resistance to extensibility (P/L) are the two
commonly used indicators of quality of durum wheat rheological characteristics. When W is greater or equal to 250 and P/L ratio is greater than 0.8, the flour is considered to be obtained from strong grains (Professional pasta website, 1999). Considering P/L ratio, all varieties fall
under strong gluten category.
Discussion
Increasing grain protein, especially that of gluten content of durum wheat and its strength has recently received higher attention due to the premium product produced out of it. In Ethiopia,
growing selected varieties like Boohai, Foka and Tob 66 with high gluten content and
acceptable dough characteristics under appropriate growing environment and; could be a good option for durum wheat growers to produce high quality durum wheat. Considering the varied agro-climatic conditions of the country, which varies within a short distance, the durum wheat
growing environments should be characterized for suitability to produce industrial quality
durum wheat. Regarding the studied rheological properties, the observed non-significant difference between nitrogen fertilizer levels could show selection of appropriate varieties
could be more important factor than the growing environment. Although early to conclude, the observed significant differences in gluten content between 60kgha', recommended for
grain yield at Debre Zeit, and 120Kg ha' N levels in this study could suggest the need to
consider economically important quality traits besides grain yield in fertilizer rate recommendations. The durum wheat breeding program should also concentrate on
development of varieties with desirable chemical grain quality as well as physical dough
properties required by Pasta industries. Such type of study on input responsive semi dwarf
durum wheat varieties is recommended.
Acknowledgments-The authors are thankful to the technical staff of Durum Wheat Research
Project of the Debre Zeit Agricultural Research Center. The authors would like to
acknowledge Kality Food Share Company for quality analysis.
References
Campbell, C.A., Selles. F., Zentner, R.P. and MacConkey B.G. 1993. Nitrogen management for zero
till spring wheat: disposition in plant and utilization efficiency. Commun. Soil Sci. Plant Anal
24:2223-2239
Curic, D., Karlovic, D., Tusak, D., Petrovic,B.and Dugum, J.2001. Gluten as a standard of wheat flour
quality. Biotechnol.39: 353-361
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




D'Egidio,M.G., Mariam, B.M., Nardi,S., Novaro,P. and Cubadda,R.(1990). Chemical and technological
variables and their relationships: A predictive equation for pasta cooking quality. Cereal
Chem., 67,275-281
Dexter, J.E., Crowle, W.C. Matsuo, R.R. and Kosmulax, F.G. 1982. Effect of N fertilization on quality
characteristics of five North American amber durum wheat varietys. 1. Can.J. Plant Sci.
(Canada), 62: 90-902
Gashawbeza, B., Mekuria, B., Yaekob, A., Mitiku, D., Tadesse, T., and Kifetew, J. 2002. Effect of
nitrogen fertilization on grain protein content of durum wheat varietys. Proceedings of 6f
Annual Conference of Ethiopian Soil Science Society. (in press)
Geleto, T.,Tanner, D.G., Mamo, T. and Gebeyehu, G.. 1996. Response of rainfed bread and durum
wheat to source, level and timing of nitrogen fertilizer on two Ethiopian Vertisols: II. N
uptake, recovery and efficiency.
Hadjichristodoulou, A. 1979. Genetic and environmental effects on vitreousness of durum wheat.
Euphytica 28:711-716
International Association for Cereal Science and Technology. 2000. ICC Standards 106/2,Viena,
Austria
Josephides, C.M.; Joppa, L.R. and Youngs, V.L. 1987. Effect of chromosome 1 B on gluten strength
and other characteristics of durum wheat. Crop Sci. 27:212-216
Metho, A.L, Hammes, P.S. and De Beer,J.M. 1997. Effect of cultivar and soil fertility on grain yield,
yield components and grain nitrogen content of wheat. African Crop Science Proceedings,
Vol. 3: 695-709
Mosconi, G and Bozzini, A. 1973. Effects of application of late nitrogen fertilizer to durum wheat.
Revista di Agronomia:75 (Cited in Field Crop Abstr. (1975) 4226)
Ottman, M.J., Doerge, T.A. and Edward, E.M. 2000. Durum grain quality as affected by nitrogen
fertilization near anthesis and irrigation during grain fill. Agron. J. 92:1035-1041.
Prima, G.Di., Sarrino, R. and Stringi, L. 1982. Nitrogen, its role in controlling yield and quality of
durum wheat in the warm arid zone of Sicily. Inst. Agron. Gen. Erbace, Italy (Cited in An
Annoted Bibliography in Durum Wheat 1972-1984, ICARDA)
Rharrabti,Y., Villegas,D., Moral,L.F.G., Aparicio, N., Elhani, S. and Royo, C. 2001. Environmental
and genetic determination of protein content and grain yield in durum wheat under
Mediterranean conditions. Plant Breeding 120,381-388
Simmonds, D.H. 1989. Inherent quality factors in wheat. Wheat and wheat quality in Australia.
CSIRO, Australia, P31-61
Spiertz, J.H.J. and Devos, N.M. 1983. Agronomical and physiological aspects of the role of nitrogen
in yield formation of cereals. Plant and soil 75:379-391
Woolfolk, C.W., Raun, W.R. Johnson. G.V. Thomason, W.E., Mullen, R. W., Wynn, K. J. and
K.W.freeman.2002. Influence of late-season foliar nitrogen application on yield and grain
nitrogen in winter wheat. Agron.J. Agron.J. 94:429-434
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




9
Impact of Irrigation Frequency and Farmyard
Manure on Wheat Productivity on a SalineSodic Soil in Dongola, Sudan
Elmoiez M. Fadull and Mukhtar A. Mustafa2
'Dongola Agric. Research Station (ARC), Sudan. 2Faculty ofAgriculture, University of Khartoum, Sudan
Abstract-A field experiment was conducted in January 2001 and December 2002, at Dongola University Farm to investigate the effects of irrigation frequency and farmyard manure application on salt leaching and on wheat (Triticum aestivum L.) growth on a saline-sodic sandy loam soil classified as fine loam, mixed, hyperthermic, sodic haplocalcids. Each experiment consisted of three irrigation frequencies: 7, 14, and 21 days, and three levels of farm yard manure (F.Y.M): 0, 4.8 and 9.7 ton/fed. The quantity of water applied was proportional to the irrigation frequency and estimated from knowledge of reference evapotranspiration as predicted by Jensen and Haise equation, a crop factor and an irrigation efficiency value. Each treatment was replicated thrice in a split-plot design. Irrigation treatments accommodated main plots (7 x 18m) and manure application sub-plots (7 x 6m). Data, collected at harvest, showed that all irrigation treatments caused salt leaching, which decreased with increase of soil depth. The data did not reflect the salt distribution between irrigation intervals. The 'irrigated weekly' (F 1) was the superior treatment; all forage and grain yields and their components increased with irrigation
frequency.
INTRODUCTION
Northern Sudan is dominated by hyper-arid, arid and semiarid ecological zones that favor the
formation of salt-affected soils (Nachtergaele, 1976; Mustafa, 1986). Dongola, in the
Northern State, has two main soil orders: Entisols in the first terrace and Aridisols in the
upper second and third terraces. Entisols, at the close proximity of the Nile, are fertile, nonsaline, non-sodic and highly productive soils. However, they are endangered by gully erosion
at the riverside and sand encroachment from the adjacent desert. Furthermore, the land is intensively cultivated and fractionated due to land tenure laws. The accessibility of good
quality Nile water (Mustafa, 1973) and Nubian aquifer water prompted horizontal expansion of agriculture in salt-affected soils of the upper terraces (Aridisols). The productivity of these
soils is constrained by osmotic and specific ion effects and nutritional imbalance.
Wheat (Triticum aestivum) is a strategic field crop in Sudan, since it constitutes the main
staple food for most of the urban population. It is grown in Northern and Central Sudan.
However, the new trend is to restrict its production to the northern State because of its
favorable climatic environment and consequent higher yield as compared to Central Sudan.
The national need for more wheat production necessitated the expansion of wheat production
to salt-affected areas. The proper use of these soils requires appropriate soil and water
management.
The share of the Nile water of Sudan is limited because of the large expansion of
irrigated agriculture in the arid and semi-arid regions. Thus, there is need to economize on
water use by increasing its efficiency.
The present research was undertaken to investigate the effect of irrigation frequency and
farm yard manure on salt leaching and wheat growth on a saline-sodic Aridisol in Dongola.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




10
Materials and Methods
A field experiment was conducted in two successive seasons (Jan. 2001 April 2001 and Dec.
2001- April 2002) on an old alluvium saline-sodic sandy loam soil classified as fine loam,
mixed, hyperthermic, sodic Haplocalcids (Soil Survey Staff 1996) at Dongola (19 N' 290 30'
E), 228 above sea level. The characteristics of this soil are presented in Table 1.
The treatments consisted of three irrigation frequencies:7 (F1), 14 (F2) and 21 (F3) days,
and three levels of farm yard manure: 0 (MO), 4.8 (M1) and 9.7 (M3) ton/feddan (1 feddan =
0.42 ha).Each treatment was replicated three times in a split-plot design. Irrigation
frequencies were designated to main plots and farmyard manure levels were accommodated in subplots. The main plots were arranged in completely randomized block design. The land was disc ploughed to 20-cm depth, and leveled using a long-span blade leveler. Nine main plots (7
x 18 in), each subdivided into three sub-plots (7 x 6 m) were constructed using earth
embankments. The main plots were 1-meter apart to check lateral water movement. In the
second season, the experiment was repeated in the same plots, which were harrowed before
seeding.
Wheat (Wadi El Niel variety ) was sown on 1 January 2001 in the first season and on
27th December 2001 in the second season. Seeds were hand dibbled in rows 20 cm apart in each subplot with a seed rate of 60-kg/ feddan. The quantity of water applied per irrigation
(Q, mm) was estimated by the following relationship:
Q = kc x ETp x F x Ei
Where: kc = a crop coefficient, ETp = potential evapotranspiration (mm/day), F = irrigation frequency (days) and Ei = efficiency of irrigation water application (0.7). kc was estimated using the FAO (1984) procedure, ETp was estimated using the following Jensen and Haise
(1963) equations:
ETp = CT (T Tx) Rs
CT = 1/ (C1+7.6 CH)
CI = 38-2E/ 305
CH= 50 mb/ (e2-el)
T, = 2.5 0.14(e2-e1) E/550
Where: T= mean air temperature (C), E = the site elevation (in), el and e2 = the saturation
vapor pressure of water (mb) at the mean monthly minimum and maximum air temperature of
the warmest month in the year, respectively, Rs = short wave incoming solar radiation. Thus,
the water applied at each irrigation event was proportional to the irrigation frequency and
varied with the month. However all treatments received the same total amount (seasonal) of water by the time of harvest. These amounts were 856.6 and 774.7 mm in the two successive
seasons. The estimated irrigation water requirement (IWR) in each month expressed in
mm/day is presented in Table 2. The quantity of water applied per irrigation, Q is the product
of IWR and F. Thus using Table 2, a predetermined quantity of water was applied to each
subplot using Parshall flume. Three irrigations at 7-day frequency were applied to all subplots
for establishing the crop before subjecting it to different irrigation regimes. The FYM was
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




heaped, mixed with water, and composted for one month before application. The average EC
of the composted FYM was 20.7 dS/m. It was applied at a predetermined rate at the third
irrigation. In all treatments, P fertilizer (Triple super phosphate) was applied before sowing in
the order of 80 kg/feddan. Twenty-one and 63 days after sowing, urea at the rate of 20 kg N/
feddan was applied to all treatments. Fertilizers were applied at the same rate and stage in
both seasons. Manual weeding was done whenever needed.
Soil samples were collected, before sowing and at harvest, by an auger from each subplot
at an increment of 0.2 m from the surface to a depth of one meter. Gravimetric moisture
content, electrical conductivity (ECe) and sodium adsorption ratio (SAR) of the saturation
extract were determined using standard procedures (Richards, 1954, Black, 1962).
At harvest, ten plants were selected at random from each subplot and their height, head
length and leaf area index were determined. One-meter square was taken at random and the number of heads was counted. Ten heads were selected at random and threshed collectively and the number of seeds per head was counted. Thousand grains were taken randomly from
each treatment and weighed. The total biomass enclosed in an area of 5 x 6 m2 in each subplot
was cut at ground level and the weight was measured and expressed in ton/feddan. The
harvested crop was left to dry for a week threshed and the total gain yield was determined in
ton/feddan. The crop water use (CWU) was determined by the following water balance
equation:
CWU = I + P + AM
where: I = Amount of irrigation water (mm), P = Amount of rainfall (mm) and AM = The
difference between cumulative water content before sowing minus that at harvest in one meter depth (mm). The crop water use efficiency (CWUE) was calculated by dividing grain yield by
crop water use (kg/m3)
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




12
Table 1. Seedling and adult plant reactions (0-4 scale) of wheat varieties when tested with to P.
graminis isolates from Ambo
Isolate 1 Isolate 2
No. Variety Seedling reaction Adult plant reaction Seedling Adult plant
reaction reaction
Durum wheat
1 Assassa 2 3 2 2
2 Tob-66 1 1 1 1
3 Kilinto 1 3 2 2
4 Foka 3 2 1 1
5 Gerardo 1 1 2 2
6 DZ1640 2 1 2 2
7 DY1050 2 2 2 2
8 Cocorit-71 1 3 1 3
9 Yielma 2 1 1 1
10 DZ1928-2 1 1 1 2
11 DZ 1691 2 2 2 2
12 DZ 1695-5 2 2 3 2
13 DZ1543 1 2 1 3
14 Boohai 2 2 2 2
15 Bichena 3 2 1 3
Susceptible 4 4 4 4
check
Bread wheat
1 Kubsa 2 4 2 3
2 Wabe 3 3 0 1
3 Galama 2 2 2 2
4 Tuse 3 3 0 1
5 Katar 3 2 2
6 Shina 3 3 0 2
7 Hawi 0 0 0
8 Simba 0 0 0
9 Wetera 0 0 0
10 HAR 2192 1 2 0 2
11 HAR 2508 3 1 1
12 F-H-6-1-7 1 3 1
13 Abola 2 0 0
14 Tura 2 1 0
Susceptible 4 4 4 4
check
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




13
Table 2. Estimated irrigation water requirements (IWR) of wheat for the two seasons in
Dongola.*
Month ETp kc ETcrop IWR
(mm/day) (mm/day) (mm/day)
First season
(2001)
January 3.50 0.70 2.45 3.50
February 4.48 1.2 5.38 7.70
March 6.70 1.2 8.04 11.4
April 8.69 0.70 6.08 8.70
Second season
(2001-2002)
December 3.64 0.60 2.18 3.10
January 3.50 0.70 2.45 3.50
February 4.48 1.10 4.93 7.00
March 6.70 1.10 7.37 10.5
* ETp = Potential evapotranspiration estimated by Jensen and Haise equation, kc = crop factor,
ETcrop = Crop (actual) evapotranspiration
Results and Discussion
The results of the two seasons showed that irrigation per se caused significant salt leaching.
Furthermore, salt leaching was not significantly affected by irrigation frequency (IF),
farmyard manure (FYM) or by their interaction. Thus, the ECe data of each season were averaged over the three levels of FYM and plotted to reflect the main effects of irrigation
frequency (Fig. la and Fig. lc), and averaged over the three levels of irrigation frequency and
plotted to reflect the main effects of FYM (Fig. lb and Fig. Id). In general, the effectiveness of
salt leaching decreased with increase in the depth of the soil layer. It may be cautioned that
this data was collected at harvest. It indicates the overall effect of treatments at harvest, but it
does not reflect the salt distribution between irrigation intervals.
For the first season, the ECe redistribution profiles depicted a top desalinized zone (0
60 cm) and a salt accumulation zone (60-100 cm). The main effect data of IF showed that F1, F2 and F3 reduced the initial ECe of the 0 20 cm by 89, 90 and 79 %, respectively (Fig. la).
The reductions of the initial ECe (20-40 cm) by the same treatments in sequence were 56, 50 and 59 %, respectively. The reductions of the initial ECe (40-60 cm) by the same treatments
in sequence were 22, 36 and 19 %, respectively. Irrigation with good quality Nile water
(Mustafa, 1973) caused desalinization of these layers. It is evident that the efficiency of salt
leaching decreased with increasing soil depth. This may be attributed partly to the gradual decrease of water available for salt leaching and consequent decrease in downward water
movement with increase in soil depth, and partly due to the rise of salt by capillarity during
the drying cycle.
The initial ECe (0-20 cm) was reduced by 85, 84 and 88% due to the application of M0,
M1 and M3, respectively (Fig. 1 b). The same treatments in sequence reduced the initial ECe
(20-40 cm) by 53, 58 and 54 %, respectively. At the 40 -60 cm depth, the initial ECe was reduced by 28, 26 and 23, respectively. The reductions of the initial ECe in the successive
depths due to FYM were of nearly similar order of magnitude as that due to IF. The effect is
conceivably an indirect effect of irrigation. FYM is expected to promote water movement,
enhance root development and ramification, and thereby enhance salt leaching. However, its effect was not significant. Although, the FYM was washed before its application, its residual
salinity limited its effectiveness in salt leaching. Furthermore, vaporization during
composting reduced its potential as a nitrogen source.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




14
For the second season, irrigation reduced the salinity throughout the profile (0 100 cm)
(Fig. Ic and Fig. Id). Treatments F1, F2 and F3 reduced the initial ECe (0-20 cm) by 59, 80 or
84 %, respectively. M0, M1 and M3 reduced the initial ECe (0-20 cm) by59, 81 and 84 0,
respectively. It seems that the reduction of salinity was due mainly to irrigation and not to the
application of farmyard manure. This is because the top layer is permeable and did not need
FYM to promote water movement. It is evident that salt leaching was greater in the first
season because of the initial higher salt level in the virgin uncultivated soil. However, trends
of salt leaching were similar.
Fig. la indicates that F2 was relatively more efficient in salt leaching than F, or F3, since
it desalinized the top 70 cm whereas F, and F3 desalinized the top 60 cm. However, the effect
was not significant.
The second season data showed that irrigation reduced the initial ECe throughout the soil
profile (0-100 cm). This was a cumulative effect, since the experimental plots of the first
season were irrigated in the second season. Treatments F1, F2 and F3 reduced the initial ECe
(0-20 cm) by 59, 80 and 84 %, respectively. M0, M, and M3 reduced the initial ECe at the
same depth by 59, 81 and 84%, respectively. The results indicate that the effect of FYM was an indirect effect of IF. FYM did not enhance salt leaching by water because the inherent salt
content of the manure was not reduced greatly and it was not incorporated in the soil.
Furthermore, salt leaching indicated, as percentage decrease of the initial ECe, was greater in
the first season. This is because the initial ECe in the first season was greater than in the
second season.
SAR redistribution
The main effects of treatments on SAR are presented on Figs. 2a, b, c, d. Irrigation per se
significantly reduced the initial SAR in the top 50 cm in the first season and through out the
whole profile in the second season. In the first season, F1 and F2 reduced the initial SAR
throughout the profile; however, the percentage decrease in SAR was minimal below 50 cm
depth. F3 increased the initial SAR in the third layer. In general, the effectiveness of
dealkalization decreased with increase in soil depth. This trend was similar to desalinization
and could be explained on the same manner. Dealkalization was due to dissolution of cabearing compounds, replacement of Na+ by Ca++ ions on the exchange complex and leaching
of the more mobile Na ions (Mustafa and Abdelmagid, 1981). It is evident that the initial
SAR of the top layer was rapidly and markedly reduced merely by leaching with water, this
may attributed to the low SAR of the irrigation water and decrease in soil solution SAR values during leaching because of dilution and faster leaching of sodium compared with
calcium and magnesium. Similar results were obtained in India (Leffellar and Sharma, 1977;
Dahiya etal., 1981,1982).
For the first season, the initial SAR of the 0-20 cm was reduced by 72, 80 and 64 % by
F1, F2 and F3, respectively. The reductions in the second layer caused by the same frequencies
in sequence were 35, 27 and 39 %, respectively. The initial SAR values of the third layer were
reduced by 6 and 7 and increased by 12 % by the same frequencies in sequence. The
differences between treatments were not significant. The effect of FYM on SAR
redistribution profile was qualitatively similar to that of irrigation frequency. The trend of
FYM effect on SAR profile was similar to that of salt leaching and it may be explained in the
same manner.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




15
ECe, dS/m
0 5 10 15 20 25 30 35 40 45 50
Initial Fig. (la) Mean electrical conductivity
(dS/m) profile as affected by irrigation frequency (days) at the end of the first season (Dec. 2000- April 2001)
20
40 14
Soil
depth, cm 21
60
7
80
M2I
20
100 ECe, dS/m
0 5 10 15 20 25 30 35 40
. ._ Fig. (lb) Mean electrical conductivity
(dS/m) profile as affected by farm yard 40 manure (ton) at the end of the first season
(Dec. 2000- April 2001)
60
soil80
depth,
cm
0
100
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




ECe, dS/m 16
0 5 10 15 20 25 30
Fig. (1c) Mean electrical conductivity (dS/m)
1 trial profile as affected by irrigation frequency (days) at
the end of the second season (Dec. 2001- April 20 2002)
40
60
Soil
depth, F2
cm
80
3
100
Mo Initial
20
40 ECe, dS/m
0 5 10 15 20 25 30 35 40
Fig. (1d) Mean electrical conductivity (dS/m) 60 profile as affected by farm yard manure (ton)
at the end of the second season (Dec. 2001M1 April 2002)
80
M2
100
Soil
depth,
cm
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




17
SAR 10 20 30 40 50 60 70
20 7 Initial Fig. (2a) Mean sodium adsorption ratio (SAR)
profile as affected by irrigation frequency (days) 40 at the end of the first season (Dec. 2000- April
2001) 60
80
Soil 14 21
depth, cqpD0
Initial 20
40 SAR
0 10 20 30 40 50 60 70
Fig. (2b) Mean sodium adsorption ratio (SAR) 60 profile as affected by farm yard manure (ton) at
the end of the first season (Dec. 2000- April 2001) 80
2 0
100
Soil
depth,
cm
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




18
SAR
0 10 20 30 40 50 60
20
F1 Initial Fig. (2c) Mean sodium adsorption ratio
(SAR) profile as affected by irrigation 40 frequency (days) at the end of the second
season (Dec. 2001- April 2002)
2
60
80
Soil
depth,
cm
100
3
Mo Initial
20
M, SAR
40 0 10 20 30 40 50 60
Fig. (2d) Mean sodium adsorption ratio (SAR) 60 profile as affected by farm yard manure (ton) at
the end of the second season (Dec. 2001- April 2002)
80
Soil
deptO
cm
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




19
Yield components and total grain yield
Since irrigation had the predominant impact on growth and yield of wheat, only the main
effects of irrigation frequency are reported in Table 3. In general, all variables increased with
increase in FYM, but the effect in many cases was not significant. In the first seasons, LAI,
head length, number of grains per head, and total biomass increased with increase of irrigation
frequency; but the difference was only significant between F1 and F3;. Plant height, 1000
grains weight total grain yield and water use efficiency increased significantly with increase
of irrigation frequency. The number of heads/m2 increased with increase of irrigation
frequency but the effect was not significant. In all cases, F1 was the superior treatment. In the
second season, all forage and grain yields and their components increased with increase in irrigation frequency but the effects were only significant in the case of LAI and number of
grains per head.
For the first season, the impact of irrigation frequency and FYM on grain yield, and
consumptive water use efficiency, which were significant, is presented in Table 4. It is
evident that treatment F1M2 was the superior treatment and treatment F3M0 was the inferior
treatment. In comparison to the inferior treatment, the total grain yield was significantly
increased by 170 % and the CWUE was increased by 100 %. Increasing the irrigation interval
would prolong both osmotic and water stress, reduce cell elongation and decrease plant
growth and grain yield (Heyn, 1940; Mustafa and Abdelmagid, 1981). Irrigating weekly will
increase the water potential between irrigation and alleviate both stresses. Treatment F1M2
was also superior in the second season, but the effect was not significant. In comparison to the
inferior treatment, the total grain yield was significantly increased by 46 % and the CWUE was increased by 31 %. The beneficial impact of the superior treatment was reduced in the
second season because was relatively ameliorated in the first season.
It is recommended that in these higher terrace saline-sodic sandy loam Aridisols, wheat
should irrigated weekly at a rate equivalent to the potential evapotranspiration. Farm Yard
manure may be washed first properly composted and incorporated into the plough layer after
the first month.
Table 3. Main irrigation effect (means at harvest for the three levels of F.Y.M) on the growth and
yield of wheat in two successive season (2000/01- 2001/02)
Type of data Irrigation frequency /day LSD
F1 F2 F3
First season (April 2001)
LAI 1.6 1.2 1.0 LSD 0010 = 0.30
Plant height (cm) 65.7 56.1 49.5 LSD 0.0022= 5.06
Head length (cm) 6.8 6.2 6.0 LSD 0.0042 =0.41
No. of heads//m2 378.6 339.7 307.1 NS
No. of grains / head 30.6 25.6 20.3 LSD OO115= 5.63
1000 grains weight (gm) 33.3 30.7 29.7 LSD 0.0002= 0.64
Total biomass ton/fed 2.9392 2.2050 1.5563 LSD 0.0225 = 0.8073
Total grain yield ton/fed 1.0834 0.8220 0.5220 LSD 0.0045 = 0.20963
Water use efficiency kg/r3 0.301 0.228 0.145 LSD 0.0046 = 0.06
Second season (April 2002)
LAI 1.5 1.4 1.1 LSD 0.041 = 0.29
Plant height (cm) 61.7 58.8 56.4 NS
Head length (cm) 6.30 6.00 5.70 NS
No. of heads//m2 422.9 398.1 391.6 NS
No. of grains / head 31.3 26.0 24.1 LSD 0.00,6 = 3.2
1000 grains weight (gm) 34.0 33.1 31.9 NS
Total biomass ton/fed 2.3092 2.0549 1.9795 NS
Total grain yield ton/fed 0.9303 0.7966 0.7060 NS
Water use efficiency Kg/m3 0.285 0.244 0.217 NS
F1, F2 and F3 mean 7, 14 and 21 days respectively.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




20
Table 4.Mean grain yield (ton/feddan) and consumptive water use efficiency of wheat (kg/m3) as
affected by irrigation frequency and farm yard manure.*
Frequency (day) MO M1 M2 Mean
Grain yield
7 0.828 1.121 1.302 1.083 a
14 0.622 0.929 0.915 0.822 b
21 0.482 0.511 0.573 0.522 c
Mean 0.644 a 0.854 b 0.930 b
Water use efficiency
7 0.230 0.311 0.361 0.301 a
14 0.172 0.258 0.254 0.228 b
21 0.134 0.141 0.159 0.145 c
Mean 0.179 a 0.237 b 0.258 b
*MO = Zero farm yard manure, M1= 4.8 ton / feddan farmyard manure, M2 = 9.7 ton/feddan.
Main frequency effect (F) LSD 0.0045 = 0.20963
Main farmyard manure effect (M) LSD 0.0003 = 0.11100
Main frequency effect (F) LSD 0.0046 = 0.06
Main farmyard manure (M) LSD 0.0003 = 0.03
References
Black, C. A. (1962). Methods of Soil Analysis. Part I. Agronomy Monograph No. 9. American Society
of Agronomy, Inc., Publisher, Madison, Wisconsin, USA.
Dahiya, I.S., Malik, R.S. and Singh, M. (1981). Field studies on leaching behaviour of a highly saline
sodic soil under two methods of water application in the presence of crops. J. of Agric. Sci.
Cambridge. 97: 383-389.
Dahiya, I.S., Malik, R.S. and Singh, M. (1982). Reclaiming a saline -sodic, sandy loam soil under rice
production. Agricultural water management. 5: 61-72.
FAO (1984). Irrigation and drainage in crop water requirements. Paper No. 24. Rome, P. 35-44.
Heyn, A.N.F. (1940). The physiology of cell elongation. Bot. Rev. 6: 515.
Jensen, M. E. and Haise, H. R. (1963). Estimating Evapotranspiration from Solar Radiation. Proc. Am.
Soc. CN. Engr., J. 89:5-41.
Leffellaar, P.A. and Sharma, R.P. (1977). Leaching of a highly saline sodic soil J. of hydrology. 32:
203-219.
Mustafa, M. A. (1973). Appraisal of the water quality of the Blue and the White Niles for irrigation
use. African Soils. 18: 113-124.
Mustafa, M.A. (1986). Salt affected soils in the Sudan their distribution, properties and management.
Reclamation and revegetation research. 5 : 115 124.
Mustafa, M.A. and Abdelmagid, E.A. (1981). The effect of irrigation interval, urea -N, and gypsum on
salt redistribution on a highly saline-sodic montmorillonitic clay soil under forage sorghum.
Soil Sci. 132 4t: 308-315.
Nachtergaele, F.O.F. (1976). Studies on saline and sodic soils in Sudan. FAO/ UNDP project for
strengthening the soil survey administration: SUD/71/553. Technical Bulletin N. 24. May
1976. Wad Medani, Sudan.
Richards, L. A. (1954). Diagnosis and improvement of saline and alkali soils. Hand book 60., USDA.
Soil Survey Staff (1996). Soil Taxonomy (key). Seventh Edition.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




21
Consumptive Use of Water and Water Use Efficiency by Wheat (Triticum aestivum) in
Relation to Irrigation and Nitrogen
Antony M. Kibel, Subedar Singh2, and Naveen Kalra3
'Egerton University, Agronomy Department, P.O. Box 536 Njoro, Kenya (akmwangi(5,vahoo.com); 2fndian Agricultural Research Institute, Water Technology Center, 110012 N. Delhi, India; 3lndian Agricultural Research Institute, Center for Agriculture Systems Simulation, 110012 N. Delhi, India
Abstract-A field experiment was carried out on a sandy loam soil to study the use relationships to yields of a late sown wheat cultivar, HD-2285 under limited and adequate water and nitrogen regimes. The experiment was laid out in a split-plot design with four irrigation levels (lo, no post-sowing irrigation; I1, one irrigation at CRI; 12, two irrigations, each at CRI and flowering; 13, four irrigations each given at CRI, jointing, flowering and dough stages) in main plots and a combination of three N levels, viz. No, N50 and N100 and two zinc levels, Zo and Z5 in sub-plots, (subscript numbers signifying N and Zn quantities in kg/ha), in three replicates. The consumptive use of water (CU) by wheat increased with every additional irrigation level to a maximum of 328.4 mm and 301.7mm. Water use efficiency (WUE) was maximum (1.38 kg grain/m3 water use) with two irrigation treatments given at the crown root initiation (CRI) and flowering stages. This did not correspond with the above ground biomass and grain yield production, which were highest under four irrigation treatments (13). The moisture use rate increased with increase in irrigation water to a maximum of 2.74 and 2.51 mm/day in the first and second seasons, respectively. Moisture extraction was maximum (59.4% 65.8%) from the 0-30 cm soil layer. Water use efficiency increased markedly with increase in nitrogen application attaining a maximum (1.42 and 1.52 kg grain/3 water use) under 100 kg N/ha application. Maximum WUE did not correspond to the highest grain yield. The rate of increase in both grain yield and WUE started to decline as ET further
increased beyond 270 mm.
Introduction
The importance of irrigation water to wheat is to provide the water for enhanced transpiration.
This permits higher leaf area indices and stomatal apertures and therefore, a greater rate of
biomass production which often lead to higher yields (Davis, 1994). Improving water use efficiency (WUE) in crop production and sustainable use of water resources are the need of the day. Fully satisfying crop water requirements may be prohibitive in terms of sustainable
utilization of limited water. The solution therefore, is to limit water application to specific
stages thereby minimizing loss of yield from water stress. Thus, it is imperative to study and
know which critical growth stages of wheat; a small amount of water application would
results in optimal WUE and minimum loss of yield.
Apart from water, other factors such as essential nutrients play important roles in
optimizing growth and yield of wheat. Major wheat growing regions of the world are poor in
available nitrogen. Nitrogen fertilization is essential for obtaining reasonable yields.
Improvement of plant nutrition often leads to increased leaf area index (LAI). As the LAI increases, transpiration increases and evaporation from the soil surface declines because of shading and canopy closure. Nitrogen fertilization can enhance new leaf growth (increased
LAI and CGR) and delay plant senescence (increased leaf area duration), resulting in
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




22
increased transpiration demand. Nutrient deficiencies however often lead to more rapid
senescence (Davis, 1994). Thus, a better understanding of the response and relationship of
wheat growth (i.e., dry mater accumulation and LAI) to nutrient and water inputs is necessary
in our assessment of a cultivar. This will also enable us make appropriate decisions for
purposes of optimizing the productivity of wheat under various environments.
Seasonal consumptive use of water by wheat is universally acknowledged to increase
with increasing available soil moisture. This is commensurate with the increase in growth
and yield of wheat. Water use efficiency is however, known to decrease with increase in
irrigation water applied beyond an optimum (Zhang, et al. 1999). The quantities of moisture used and the efficiency of it's utilization differs however, as influenced by factors such as the
environment, cultivars, duration of the cultivars on the land and the rooting pattern of the
crop.
The effects of irrigation on crop production are usually quantified using crop production
functions that relate crop yield to the amount of water applied (English and Raja, 1996).
These functions are used to optimize on-farm irrigation and economic evaluation of irrigation
water application. Water production functions are mathematical equations relating crop
response in terms of biomass or grain yield with water availability or its uptake by the crop.
These functions can be used in managing water resource for achieving maximum returns with minimum amount of water application as irrigation. The use of water-yield relationships has been employed in the development of models that are used for prediction purposes. This type
of work has mostly been done successfully in the developed countries. There is a need for more work in the developing countries, where water is limited, for purposes of determining
the optimum amounts of water required in producing appreciable and sustainable crop yields.
In light of the considerations mentioned above, the present investigation was undertaken
at the Water Technology Center, Indian Agricultural Research Institute, New Delhi during the
winter seasons of 1999-2000 and 2000-2001 with the following objectives:
1. To determine the influence of irrigation, nitrogen and zinc on the growth of wheat
2. To determine the seasonal consumptive water use, water use efficiency and moisture
use rate of wheat under varying irrigation and nutrient levels
3. To compute production functions for wheat under varying water supply conditions.
Materials and Methods
Four irrigation levels (Io, no post-sowing irrigation; 1, one irrigation at CRI; 12, two
irrigations, each at CRI and flowering; 13, four irrigations each given at CRI, jointing,
flowering and dough stages) were allotted to main plots and six fertilizer levels (NoZ0, no
nitrogen and zinc; NoZ5, no nitrogen and 5 kg Zn/ha; N50Z0, 50 kg N/ha and no Zinc; N50Z5, 50 kg N/ha and 5 kg Zn/ha; N100Z0, 100 kg N/ha and no zinc; N100Z5, 100 kg N/ha and 5 kg
Zn/ha) were allotted to the sub-plots of a split plot design replicated thrice.
Dry matter production (DMP), Leaf Area Index (LAI), grain and straw yield data was
recorded. Periodic soil samples from 0-120 cm soil depth were collected layer-wise to
compute water use by crop as described by Dastane (1972). Crop evapo-transpiration to yield
relationships was determined through regression analysis.
Soil samples were taken from plots at depth intervals of 0-15, 15-30, 30-60, 60-90,and
90-120 cm soil profiles and dried to, respectively, to determine the soil moisture. Samples
were taken at sowing time; 48 hours before; and after irrigating; and at harvest, with the help
of a tube auger. Samples were dried in the oven at 1050 C for over 24hr till constant dry
weight. The moisture content was measured thermo-gravimetrically. Seasonal consumptive
water use (CU) and the water use efficiency (WUE) of the crop, daily moisture use rate
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




23
(MUR) and soil moisture extraction pattern of the crop were computed from the periodic soil
moisture content data.
The seasonal consumptive water use by the crop under different treatments was worked
out using the formula described by Dastane (1972):
N n
CU Z (Epx0.6)+Z (MIi-M21xAsixDi+ER+GWC
1 i=1 100
where
CU = Consumptive use of water (mm)
Ep= Pan evaporation value from the USWB class A open pan evaponimeter for the period
from the date of irrigation to the date of soil sampling after each irrigation (mm)
A = constant used for obtaining ET value from Ep value for the given period of time
M = Per cent soil moisture (w/w) of the ith layer of the soil at the time of sampling after
each irrigation
M2i= Per cent soil moisture (w/w) of the ith layer of the soil at the time of sampling before
each irrigation
Asi = Apparent specific gravity of the ith layer of the soil
Di = Depth of ith layer of soil (mm)
ER = Effective rainfall, if any during the period under consideration, mm
GWC =Ground water contribution to the root zone moisture during the given period of time
(mm)
n = Number of soil layers
N = Number of days between pre-irrigation and post-irrigation soil moisture samplings
The ground water contribution (GWC) was considered negligible as the ground water table
during crop growth periods remained below 4.0 from soil surface during in both the seasons.
Crop water use efficiency was computed using the following formula and was expressed as kg
grain /m3 water used.
Grain yield (Economic yield)
WUE =
Consumptive Water Use (Seasonal)
Leaf area index was measured using the leaf area to dry leaf weight relationship. At each dry mater sampling date, a sub sample of fresh leaves from the main sample (Im row length) was
taken from one replication and leaf area was recorded with the help of LI-3 100 Area Meter,
LI-COR, Inc. The leaves were then dried at 60 'C for about 24 hr until a constant weight was obtained. Specific leaf area (SLA) was calculated as the ratio of leaf area (cm2) to leaf weight
(g). This SLA value for each treatment at each sampling time was multiplied by the
corresponding dry leaf weight values, to get leaf area index. Leaf area index is expressed as
total leaf area of the crop (one side only) per unit ground area occupied by the crop. The area
occupied by the plants under consideration in the field was taken as 0.25 m x 1.00 m.
The experimental data pertaining to each character was analyzed statistically by using
Analyses of Variance technique (ANOVA) for split plot. Standard error of mean difference
and significant difference (LSD) at 5% level of significance were worked out for each
character. Pooled analysis of the two years data was done only for grain and straw yield. The
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




24
computed parameters such as data on soil moisture studies were explained on the basis of
comparative performance.
Results and Discussion
Growth of Wheat (Dry matter and Leaf area index)
The four irrigation treatment applied at CRI, jointing, flowering and dough stages along with
one pre-sowing irrigation increased the dry matter production of wheat at harvest significantly
over other irrigation treatments in both the 1999-00 and 2000-01 seasons.
Total above ground dry matter (yields) production (TDMP) at 30, 60 and 90 days after
sowing was significantly more with added irrigation, compared to the no post-sowing
irrigated TDMP production in second season (Table 1). In the first season however, only the
90 days after sowing (DAS) above ground DM showed significant differences due to
irrigation treatments. This was attributed to the 68.2mm rainfall (Fig. 1) during the initial part
of the first season that mitigated irrigation effects on DMP up to the 60 DAS. Similarly,
application of irrigation increased the leaf area index (LAI) of wheat significantly over the
non-irrigated control treatment at 60 and 90 DAS, in the second year (Table 2).
Table 1: Influence of I, N and Z on total dry matter production (g/m2), at different growth stages
Treatment 30 DAS 60 DAS 90 DAS Harvest
1999-00 2000-01 1999-00 2000-01 1999-00 2000-01 1999-00 2000-01
10 304.9 222.5 1299.2 999.4 5711.8 5192.3 7061.0 6852.5
I1 316.5 218.9 1306.5 1082.9 6966.1 6181.1 8739.2 8354.4
12 316.6 224.4 1309.6 1139.9 7518.3 6695.2 9876.2 9553.1
13 317.8 236.0 1308.0 1238.7 7915.8 7300.5 10454.2 10524.1
SEm+ 4.34 1.77 10.37 7.57 54.56 42.77 89.60 75.26
CD(p=0.05) NS 5.3 NS 22.8 164.3 128.9 269.9 226.8
NO 292.2 202.9 1086.3 929.5 5423.3 5174.9 7998.4 7744.8
N50 306.6 224.2 1295.2 1085.9 7098.1 6228.7 9029.8 8889.2
N100 343.1 246.4 1536.0 1330.3 8562.6 7623.2 10069.7 9829.6
SEm+ 2.11 2.2 12.34 8.64 65.24 52.89 97.7 87.2
CD(p=0.05) 15.4 6.6 37.2 26.1 196.6 159.4 300.5 262.2
ZO 313.5 225.2 1303.9 1114.6 7018.6 6331.7 8982.2 8800.3
Z5 314.5 225.7 1307.9 1115.9 7037.4 6352.8 9083.1 8842.0
SEm+ 4.17 1.79 10.05 53.26 43.18 85.4 81.4 71.2
CD(p=0.05) NS NS NS NS NS NS NS NS
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




25
60 90
-80
50 0
70-.
cc)
,. E 40 60 c
50 0 ~ 2
C0. 30 0.CL-.
=c 40 E -- 2 0 30 ~E
20 e
20 "0c
10 00
0- 0
Standard meteorological weeks
Rainfall Evaporation -a-Average temperature -u-Average humidity.
60 90
80
50 -a
o 70 U
o 40 60 .2CU E
= 50 3
E 30
-aE
-o 40 E"
S 20 30
CC
-T n20 >
10
0
S Rainfall Evaporation
-- Average temperature --j-Average humidity.
Fig 1:Pattern of rainfall, evaporation, average temperature and average relative humidity during
Late sown wheat growth periods of 1999-00 and 2000-01
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




26
Table 2 :Influence of irrigation, Nitrogen and Zinc on the LAI of late sown wheat at 30, 60 and 90
DAS
Treatment 30 DAS 60 DAS 90 DAS
1999-00 2000-01 1999-00 2000-01 1999-00 2000-01
1o 0.164 0.291 0.595 0.561 2.349 2.007
11 0.163 0.289 0.614 0.564 3.178 2.697
12 0.163 0.293 0.607 0.566 3.467 2.922
13 0.162 0.292 0.613 0.576 3.603 3.11
SEm 0.01 0.05 0.007 0.008 0.036 0.028
CD(p-0.05) NS NS NS NS 0.01 0.09
No 0.14 0.226 0.519 0.493 2.696 2.314
N50 0.159 0.278 0.616 0.525 3.272 2.77
Nioo 0.19 0.368 0.687 0.682 3.48 2.963
SEm+ 0.001 0.005 0.008 0.008 0.047 0.034
CD(p-0.05) 0.003 0.014 0.024 0.025 0.141 0.103
Zo 0.163 0.291 0.606 0.566 3.142 2.676
Z5 0.163 0.291 0.608 0.568 3.157 2.693
SEm+ 0.001 0.004 0.006 0.007 0.038 0.028
CD(p-0.05) NS NS NS NS NS NS
Grain and Straw yield of Wheat
The maximum average grain yield of 4.0 tons in first season and 3.9 tons per hectare (t/ha) in
the second season was obtained with the four irrigation treatment (13) given at CRI, jointing,
flowering and dough stages of the crop (Fig. 2). Minimum yields of 2.3 t/ha in both the
seasons were obtained with no post sowing irrigation (lo). The per cent increase in grain yield
due to one (11), two (12) and four irrigation levels over the no irrigation treatment was 36.2,
58.0 and 71.5 in 1999-2000, and the corresponding per cent increase in 2000-2001 was 32.0, 57.8 and 73.9, respectively. Each and every addition of either irrigation or nitrogen resulted
in a progressive and significant increase in grain yield up to 13N100 level in first and second year of experimentation, respectively (Table 3). The minimum grain yield of 2.00 t/ha and
1.96 t/ha were obtained with IoN0 in the respective seasons.
Straw yield increased significantly with addition of irrigation attaining a maximum of 6.4
t/ha with four irrigations in 1999-2000 and 6.6 t/ha in 2000-2001 (Fig. 2). Application of one
and two irrigations, resulted in a percentage increase of the straw yield by, 17.6 and 30.9 in
the first season and 17.0 and 30.4 increase in the second season, over the control (lo),
respectively.
Table 3:Interactive effects of I and N on the grain yield (kg/ha) of late sown wheat at harvest
IxN effects on grain yield
To I1 12 13 T0 I1 12 13
No 2038.8 2649.9 3013.0 3301.2 1959.8 2407.8 288505 3271.7
N5o 2337.1 3199.9 3713.1 4108.7 2259.8 3011.0 3587.9 4042.1
Nloo 2637.1 3699.9 4359.7 4615.4 2535.1 3499.9 4186.9 4436.1
N within I I within N N within I I within N
SEm+ 144.3 149.52 130.49 136.2
CO(p=0.05) 298.1 302.2 268.9 275.2
Fig. 2. Straw yield increase with additional irrigation
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




27
12000.0
10454.2 10524.10000.0 9876.2 9553.1
8739.2 8354.4
9 8000.0 y
32
7061.0
. 6852.5
6000.0 6180.9 6445.8 6571.4 OS
4000.0- .0
4008.435.
2000.0 23375
0.0 LP r Y
10 I1 12 13 10 I1 12 13
Year 1999-2000 Irrigation levels Year 2000-2001
Consumptive Use of Water (CU) by Wheat
Increasing frequency of irrigation increased the CU progressively and markedly in both the seasons (Fig 3 and Table 4). The maximum consumptive use of 328.6 mm in 1999 2000 and 302.0 mm in 2000-2001 was attained with the four irrigation treatments, given at CR1,
jointing, flowering and dough stages of the crop growth. The minimum CU was obtained with no post sowing irrigation, i.e., control (191.5 mm and 178.2 mm) in corresponding
seasons, respectively. The percent increase in CU values due to 11, 12 and 13 over 10 treatment
was 21.5, 44.5, and 71.5, in 1999 2000 and 25.5, 39.8, and 69.3 in 2000 2001 seasons,
respectively. Moisture use rate of wheat increased with the increase in the number of
irrigation levels given. The increase in moisture use rate of wheat due to 11, 12 and 13 levels over the no irrigation treatment (lo), was in the order of 21.3%, 39.6%, and 68.5% in 20002001 season, respectively.
Water Use Efficiency by Wheat
The pooled average water use efficiency (kg grain/m3 water use) of I1 (1.35) and 12 (1.38) were higher than those of l0 (1.24) and 13 (1.27) treatments in both the seasons (Table 4).
Highest WUE for the first season crop was recorded with 11 (1.37), while it was with 12 (1.42) in the second season (Fig 3 and Table 4). Minimum water use efficiency was recorded with lo
(1.22 and 1.26 kg grain/m3 water use) in both the respective seasons.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




28
400
0 1999l-20oo
. 10 2000-2001
j 200C
Irrigation level Nitrogen level Zinc level
30
C 100 1999-2000
"002000-2001 .. ... .. .,
0 ,2,....,. ......
10 1 2 3 No NM N100 Z0 5
Irrigation level Nitrogen level Zinc level
1.6 191999-200
S2 2000-2001
00
S I1 2 13 No N Noo ZO Z5
Irrigation level Nitrogen level Zinc level
Figure 3. Influence of irrigation, nitrogen and zinc on the consumptive use of water, moisture use
13 2 0:0.2 0
0.6.
rate and water use efficiency by late sown wheat
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




29
Table 4:Influence of 1, N and z on seasonal consumptive water use, moisture use rate and WUE
of late sown wheat
CU (mm) WUE(kg grainlm3 water) Moisture use rate (mmi/day)
1999-00 2000-01 1999-2000 2000-2001 1999-2000 2000-2001
1o 191.52 178.20 1.22 1.26 1.60 1.49
11 232.67 223.66 1.37 1.33 1.94 1.86
12 276.10 249.10 1.34 1.42 2.30 2.08
13 1328.39 301.66 1.22 1.31 2.74 2.51
No 253.14 234.36 1.09 1.12 2.11 1.95
N50 257.08 237.95 1.30 1.35 2.14 1.98
Nioo 261.28 242.16 1.46 1.52 2.18 2.02
Zo 257.11 238.15 1.27 1.32 2.14 1.98
V5 .257.23 238.16 .1.30 1.34 .2.14 1.98
Nutrient Nitrogen and Zinc Effects on Water Use
Application of nitrogen did not enhance the CU appreciably as in case of irrigation
application. With application of nitrogen fertilizer at the rate of 5 0 and 100 kg /ha, the
percent increase in CU of wheat over no nitrogen (No) application was marginal i.e. 1.56 and
3.22 in the first season and 1.53 and 3.32 in the second season respectively (Table 4).
Similarly Zinc application did not have much influence on the consumptive use of wheat.
Application of nitrogen had very little effect on the daily moisture use rate of wheat, while
zinc application had practically no effect at all, in both the cropping seasons (Table 4).
Nitrogen application increased water use efficiency of the crop progressively up to 100
kg / ha (1.46 and 1.52 kg grain/in3 water use) during both the years of experimentation. Zinc application however, induced negligible influence in WUE of wheat in both the seasons (Fig.
3 and Table).
Nitrogen application up to 100 kg/ha did not bring any noticeable change in seasonal
consumptive use and daily moisture use rate of the crop (Table 4). Water use efficiency
increased progressively with N application and the mean maximum WUE recorded with 100
kg N/ha was 1.46 and 1.52 kg grain/n3 water use, in first and second seasons, respectively.
Moisture Extraction
The influence of irrigation and fertility (N and Z) interactive effects on the soil moisture
extraction patterns is depicted in Table 5. Maximum (58.6 %o to 66.8 %o) soil moisture was
extracted from within the 0- 30cm soil profile layer with the highest being under the four
irrigation regime (64.7 to 66.8 %o) and lowest under the no post-sowing irrigation treatment
(58.2 to 60.1 %). This trend of declining soil moisture extraction percentages with increasing irrigation frequency was evident at the 60-90 cm and 90-120 cm depth soil layer profiles also.
The ranges were 9.6 11.7 00 and 6.4 8.0 00 in the 60-90 cm and 90-120 cm soil profiles,
respectively.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




30
Application of nitrogen induced the plants to extract proportionately greater amount of
soil moisture from 60-120 cm layers as compared to the non- fertilized ones. Zinc application
on the other hand, caused negligible increase in moisture extraction pattern at different soil
profiles that followed the similar effect as in the case of irrigation effects.
Table 5: Influence of irrigation, nitrogen and zinc on soil moisture
extraction pattern (%)
1999-2000 2000-2001
Seasonal moisture extracton patem Seasonal moisture extracton patem
90-120
Treatment 0-30 cm 30-60 cm 60-90 cm 90-120 cm 0-30 cm 30-60 cm 60-90 cm cm
Irrigation
lo 59.35 21.90 11.65 7.10 58.10 21.62 12.53 7.75
I1 61.50 20.98 11.12 6.40 60.15 21.00 12.05 6.80
12 64.32 20.05 10.00 5.63 61.62 20.25 11.65 6.48
13 65.80 19.30 9.58 5.32 64.68 19.72 10.13 5.47
Nitrogen
No 63.56 20.70 10.58 5.16 61.43 20.54 11.78 6.26
N50 62.94 20.53 10.60 5.94 61.13 20.63 11.60 6.65
Nioo 61.73 20.45 10.59 7.24 60.86 20.78 11.40 6.96
Zinc
Zo 62.69 20.54 10.56 6.21 61.10 20.62 11.56 6.73
Z5 62.79 20.58 10.62 6.02 61.18 20.68 11.63 6.53
Production Function for Wheat
Water yield relationships depicted as production functions are considered as useful tools in
the management of water and nitrogen application for the purposes of optimizing crop
productivity. Crop production functions that relate water-yield relationships, are
mathematical equations relating crop response in terms of biomass or grain yield with water
availability and it's uptake by the crop.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




31
2.0
y = -0.0025x2 + 0.1288x 0.288. R2=0.108 c.~ 1.5 $
....... .... ...
= 1.0 "
0.5
10 20 30 40
Evapo-transpi ration (cm)
cc
x 4.0 Y = -0.0004x2+ 0.097x 2.71
-_ 3.0R2= 0.745e
3.0 I
ccc 2.0 (D 1.0
50 60 70 80 90 100 110 120
F Total water use up to anthesis (mm)
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




32
8000
S y= 5%;1"x + 726
q
, 6000
4000 = 20.8"x + 553
I
2 000 aA
2000y 6.7x +698
0
55 65 75 85 95 105 115 125
Seasonal water uptake (mm)
Water use efficiency is generally considered as a conservative term and is expressed as
the ratio of DMP or Y to water supply or water use, expressed in terms of evapotranspiration
(ET) or transpiration (Tr) on daily or seasonal basis (Sinclair et al, 1984). Water use efficiency generally ranged from 1.22 to 1.42 kg/m3 (Table 4), with the highest WUE
observed at one and two irrigation levels of the first and second season crop, respectively.
The values of WUE were higher than those reported from the Mediterranean region by Zhang and Oweis (1999) that ranged from 1.08 to 1.19 kg/m3 and close to those (1.23-1.46 kg/m3) of
the Loess plateau in China (Kang et al., 2002).
Regression analysis indicated a quadratic relationship between WUE and seasonal evapotranspiration (Fig. 5), which was reflected in a poor correlation of 11%. WUE reached its
maximum value at an evapo-transpiration of approximately 260-270 mm then started a
decrease with evapo-transpiration. Maximum WUE did not correspond to the highest grain
yield (Fig 2 and 3). When ET is relatively low, water availability is the limiting factor for
grain yield and an increase in ET results in significant increase in both grain yield and WUE.
However, the rate of increase in both grain yield and WUE starts to decrease as ET further
increases. Once WUE reaches its maximum value, an increase in total crop water use could still lead to a marginal increase in grain yield. Thus, WUE would decrease (Figure 5). This was in agreement with the findings of Singh (1977) and Kang et al. (2002). This means that
relatively high crop yield and WUE can be maintained while substantially reducing irrigation
volume under limited-irrigation management (Singh et al. 1991; Zhang and Oweis, 1999;
Kang et al. 2002).
When evaluating dry matter production, expansive growth deserves special attention, since it
is the means for developing leaf area for intercepting light and carrying out photosynthesis.
The sensitivity of expansive growth to small water deficits is marked by reduction in leaf area (Hsiao et al., 1976b; Vaux and Pruitt, 1983). Maximum LAI for wheat is generally noticed at
anthesis. LAI at flowering is mainly used for forecasting wheat yields, with the prime
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




33
assumption that there is no stress in the subsequent stages of crop growth. The LAI estimates
based on water use (Fig 8) accounted for 75% of the variations only.
It can be concluded that wheat grown under lower levels of irrigation water results in
lower consumptive use of water (Fig 3 and Table 4). This further influences the production of
less leaf area indices (Table 3 and Fig 6), that ultimately causes the decline in the total above
ground biomass and grain yield production (Fig 2).
Final biomass and grain yield of wheat depends upon the sensitivity of various stages
towards moisture availability that scheduling of irrigation is carried out. Dated production
functions take care of the crop stage sensitivity and thus gives better predictability for
biomass and yield estimates (Kalra, 1986; Singh et al., 1987). The highest rate of biomass
gains of 53.1 kg/ha-mm was obtained over the 60-90 day period, a period that falls within the maximum growth phase of wheat, followed by 28.3 and 6.7 kg/ha-mm during the 90-120 and
0-60 DAS periods (Fig 7). This is in agreement with the findings for semi and regions of
Hisar in India (Singh et al., 1987) and China (Zhang et al., 1999). Therefore, scheduling of
water application ought to consider providing less quantities of water during the 0-60 DAS
period as compared to the 90-120 and 60-90 DAS periods, respectively. This should be
aimed at providing just sufficient amounts of water and nitrogen for the necessary
developmental processes to take place. Further research aimed at optimizing growth and
production of wheat through the application of minimum quantities of water and nitrogen, at
various stages of wheat growth, is therefore recommended.
References
Dastane, N.G., Singh, M. Hukkeri, S.B. and Vasudevan, V.k. 1970. Review of work done on water
requirement of crops in India. New Bharat Prakashan, Poona, India.
Davis, G.J. 1994. Managing plant nutrients for optimum water use efficiency. Advances in Agronomy
53 :85-110.
English, M., Raja, S.N. (1997). Perspectives on deficit irrigation Agricultural Water Management, 32
(1): 1-14.
Hsiao, T.C., Feres, E., Acevedo, E. and Henderson, D.W. 1976b. Water stress and dynamics of growth
and yield of crop plants. Ecological Studies, 19: 281-305.
Jackson, M. L. 1958. Soil Chemical analysis. Prentice Hall of India. (Pvt.) Ltd. New Delhi 1973, India.
Kalra, N. 1986. Evakuation of soil water status, plant growt and canopy environment in relation to
variable water supply to wheat. P.hD. Theses. P.G. School, I.A.R.I., New Delhi: 103pp.
Kang, S.Z. and Dang, Y.H. 1987. Research on crop water production functions and optimal irrigation
scheduling. Water Resource Hydraulic Eng. 1: 1-12.
Khang, S., Zhang, L., Liang, Y., Hu, X., Cai, H. and Gu, B. 2002. Effects of limited irrigation on yield
and water use efficiency of winter wheat in the Loess plateau of China. Agric. Water Manage.
55 : 203-216.
Novoa, R and Loomis, S.S. 1981. Nitrogen in plant production. Plant and Soil. 58 (1) : 177-204.
Sinclair, T. R., Tanner, C. B. and Bennet, J. M. 1984. Water use efficiency in crop production.
Bioscience 34 : 46- 40.
Singh N.T., Vig A.C., Rachhpal Singh, Chaudhary M. R., Singh R., (1977). Influence of different
levels of irrigation and nitrogen on yield and nutrient uptake by wheat, Agronomy Journal, 71
(3) : 401-404.
Singh, A.K. 1991. Response of irrigated wheat (Triticum aestivum) to nitrogen and phosphorus in
farmers' fields in Ganga diara of Bihar, Indian J Agron., 41 (1) 157-159.
Singh, P., Wolkewitz, H. and Kumar, R. 1987. Comparative performance of different crop production
functions for wheat (Triticum aestivum). Irrigation Sci. 8 (4) : 273-290.
Stewart, J. I., Cuerca, R. H., Prutt, W.O., Hagon, R.M. and Tosso, J. 1977a. Determination and
utilizationof water production functions for principal California crops, W-67 Calif Contri.
Proj. Rep. University of California, Davis.
Tandon H.L.S. (Ed.) 1998. Methods of analysis of soils, plants, waters and fertilizers. Fertilizer
Development and Consultations Organization, New Delhi, India. Pp. 144+vi.
Vaux, H.J. and Pruitt, W.O. 1983. Crop water production functions. In: D. Hillel (Editor), Adv. in
Irrigation, 2. Academic Press, New York, pp 257-272.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




34
Watson, D.J. 1952. Adv. Agron. 4: 101-145.
Zhang H., Wang X., You M. and Liu C., (1999). Water-yield relations and water-use efficiency of
winter wheat in the North China Plain, Irrigation Science, 19 (1) : 37-45.
Zhang H.P, and Oweis T., (1999). Water-yield relations and water-use efficiency of winter wheat in the
North China Plain, Agricultural Water Management, 38 (3): 195-211.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




35
Agronomic and Economic Evaluation of Break
Crops and Management Practices on Grain
Yield of Wheat at Shambo, Western Oromiya, Ethiopia
Tolera Abera and Mathewos Belissa
Oromiya Agricultural Research Institute, Bako Agricultural Research Center P.O. Box 03, Email: akthirpha@yahoo.com Bako, Western Oromiya, Ethiopia
Abstract-Sustainable production of wheat (Triticum aestivum) may be limited by a number of crop management practices, including crop rotation and cultural practices.
A field experiment was conducted for three years at Shambo, Ethiopia, to evaluate the agronomic and economic effects of break crops and management practices on the grain yield of wheat. After a field pea break crop wheat grain yield was 32% higher than after a barley crop. Management practices gave significant effects on the mean grain yield of wheat. Wheat performed better after the field pea break crop with both farmer's management practices and the recommended practices. Both local and improved varieties gave greater yield with intensive management and fertilizer application. The highest economic benefits came from the use of field pea break crop with improved management practices and fertilizer application and the extension program under way in the area should consider this recommendation for sustainable
production of wheat.
Introduction
Wheat is one of the major cereal crops produced in the Ethiopian highlands (1,500 to 3,000 m.a.s.1) (Hailu et al., 1991), and in the highlands of Shambo, wheat ranks fourth among the cultivated crops in terms of production (Asfaw et al., 1997). Both biotic and abiotic factors
limit agricultural productivity in the country. Human degradation on the natural resource base
is generally very swift, but interventions to amend the decline are very slow. Lal (1989) indicated that conservation tillage and crop rotation are considered the major means of
sustaining agricultural productivity at the global scale. Soil fertility can, in many cases, be
maintained through the combined use of suitable legumes in a suitable crop rotation and
modem artificial fertilizers capable of correcting nutrient deficiencies (Whyte et al., 1969).
Practical cropping systems options with appropriate management practices for wheat in
western parts of Ethiopia have not been clearly identified. However, making use of their
indigenous knowledge, most farmers practice crop rotation based on short-term agronomic
benefits break crops, and most studies in the past focused on the short-term agronomic
benefits of break crops for wheat production (Hailu et al., 1989).
The application of chemical fertilizer has been recommended for wheat production in
eastern and southern Africa (ESA) without considering the sustainability of continuous
application (Tanner, 1997). Demonstration and extension of fertilizer and improved seeds
along with all the recommended practices were started in Ethiopia with the introduction of the
Minimum Package Program (MPP) (1967 78). In this Program, based on multi-locational trials and demonstration results, "blanket fertilizer recommendations" were made for major
cereal crops in the country. The "blanket" fertilizer rate recommendations of EPID during the
1970s are still used for wheat production (Asnakew et al., 1991), and there are no reports
available on the long-term effects of these recommended fertilizer rates on wheat production
in the Shambo highlands. However, increasing productivity and system sustainability through
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




36
crop rotation has been suggested to be a sound management options for small-scale farmers in
the Ethiopian highlands. Hailu et al. (1989) reported that faba bean as a break crop increased wheat grain yield by 1,100 kg ha' or 69 % cf. the yield of second-year continuously cropped
wheat, while the study of Asefa et al. (1992) indicated that a faba bean crop increased the
yield of the following wheat crop by 982 kg ha', or 44 %, compared to that of continuously
cropped wheat. However, studies have not been conducted in the highlands of Shambo in
relation to these agronomic interventions. Furthermore, the agronomic results of "multioptional and local input inclusive trials" are important to obtain useful information on break crops for wheat production in the area. Therefore, the objective of this study was to evaluate
the effects of pulse and cereal break crops along with management practices on wheat
productivity in the Shambo highlands.
Materials and Methods
The experiment was conducted during the 1997, 1998 and 1999 cropping seasons at Shambo:
9034'N latitude and 37006'E longitude at an altitude of 2400 meters above sea level. Mean
annual rainfall is 1,695 mm (NMSA, 2003). It has a cool humid climate with the mean
minimum, mean maximum, and average air temperatures of 8.150C, 15.720C, and 11.940C,
respectively. The soils are nitisols.
The experiment was laid out as split plot in a randomized complete block design with
precursor crops as main plots and management practices as sub-plots. The precursor crops were field pea (Pisum sativum) variety (G-22763-2c) and barley (Hordeum vulgar) variety (Shege) and were sown in the 1997 cropping season. Wheat was sown on these plots in six
management sub-treatments in the 1998 and 1999 seasons The wheat management practices
included wheat variety, cultural practices and fertilizer level, arranged as follows: Farmer's variety and farmer's traditional practices without fertilizer (FVFP-FE) Improved variety and farmer's traditional practices without fertilizer (IVFP-FE) Farmer's variety with improved practices without chemical fertilizer (FVIP-FE)
- Improved variety with improved agronomic practices without chemical fertilizer (IVIP-FE)
- Farmer's variety with all improved agronomic practices with chemical fertilizer (FVIP + FE) Improved variety with all improved agronomic practices with chemical fertilize (IVIP + FE).
The wheat varieties used were the local variety "Molgo" (awnless wheat) and HAR-604
(Galama), an improved variety. The improved cultural practices were the recommendation
practices as for wheat. For the improved cultural practices, the herbicide 2,4-D was applied 30
to 45 days after planting to control weeds while in plots with the farmer's cultural practices, hand weeding was done once at 25 days after sowing. The recommended fertilizer rates of 100 kg haI each for DAP and Urea were applied at planting n the plots receiving fertilizers.
The plot size was 10 m x 10 m. The seed rates were 150 kg ha for field pea, barley and
the improved wheat variety, whereas the normal farmer practice of 160 kg ha was used for
the local variety. Sowing dates were between mid June and early July.
The data were analyzed using MSTAT-C Computer Software Program. Selected
orthogonal contrasts were used to separate the effects of the two levels of the three factors:
management practices, fertilizer effects, and variety.
Economic evaluation was conducted using the partial budget, values to cost ratio (VCR)
and marginal analyses methods. Wheat grain yield was valued at the average open market price over the last five years of Ethiopian Bihr (EB) 143.00/100 kg. Plot grain yields were
reduced by 10 % to reflect actual production environments (CIMMYT, 1988). The wheat seed cost used was EB 2.40/kg for the improved variety and EB 1.43/kg for the local variety. Urea
and DAP were valued at the official prices of EB 192.00 and 256/100 kg respectively. The
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




37
cost utilized for labour for weeding was EB 3.50/day. The average rental price of a sprayer
was EB 10.00 ha', and the cost of 2,4-D herbicide was EB 0.42/liter.
Results and Discussion
The yields of the precursor crops sown in the 1997 cropping season are shown in Table 1.
Table 1. Mean grain yield of break crops, field pea and barley in 1997.
Crop Grain yield (kg ha-')*
Field pea 556
Barley 1549
Statistically significant t-value at P=0.0 1
With all management practices, the grain yield of wheat was significantly greater (P<0.05)
following field peas than following barly in both cropping seasons (Tables 2 and 3). The
mean grain yield of wheat was increased by 39% and 25% following field pea compared to
barley in the 1998 and 1999 cropping seasons respectively, and on average by 32%. This
benefit of the field pea break crop has also been shown by Tanner et al. (1991), who reported
that wheat after barley break crop was not significantly different from continuous wheat
production, and Tilahun et al. (2000). Tanner et al. (1999) showed that wheat grain yield was
consistently higher in rotation with dicotyledonous crops than with cereal crops, exhibiting
grain yield increments ranging from 22 to 54 %. In addition, in their studies, wheat grain yield
was higher in rotation with faba bean than with rapeseed, with grain yield increases ranging
from 8 to 31 %. This is likely due to the fixing of free atmospheric nitrogen by Rhizobium
bacteria in association with
Table 2. Effects of break crops and management practices on grain yield of wheat (kg
ha-i) at Shambo in 1998.
Treatment Management practices Mean
Break crop FVFP IVFP FVIP IVIP FVFP IVIP
-FE -FE -FE -FE +FE +FE
Barley 1093 1563 1300 1400 1590 1763 1451
Field pea 1510 2250 1650 2390 1970 2373 2024
Mean 1302 1906 1475 1895 1780 2068 1718
Yield change(%) 38 44 27 71 24 35 39
(Pea/barley) I I
Break crop Mgmt practices Break crop x Mgmt practices LSD (5%) 206 480 NS
CV % 8.25 22.94
Table 3. Effects of break crops and management practices on grain yield of wheat (kg ha-1) at
Shambo in 1999.
Treatment Management practices Mean
Break crop FVFP IVFP FVIP IVIP FVFP IVIP
-FE -FE -FE -FE +FE +FE
Barley 1087 1800 1400 1210 1767 1947 1535
Field pea 1567 2033 1633 2133 1933 2200 1917
Mean 1327 1917 1517 1672 1850 2073 1726
Yield change(%) 44 13 17 76 9 13 25
(Pea/barley) I I
Break crop Mgmt practices Break crop x Mgmt practices LSD (5%) 155 NS NS
CV% 6.26 25.40
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




38
Table 4. Effects of break crops and management practices on mean grain yield of wheat (kg ha-1)
at Shambo (combined over two years), 1998-1999.
Treatment Management practices Mean
Break crops FVFP IVFP FVIP IVIP FVFP IVIP
-FE -FE -FE -FE +FE +FE
Barley 1090 1681 1350 1305 1678 1855 1493
Field pea 1538 2142 1642 2262 1952 2287 1970
Mean 1314 1911 1496 1783 1815 2071
Yield change(%) 41 27 22 73 16 24 32
Break crop Mgmt practices Break crop x Mgmt practices LSD (5%) 83.02 345.7 NS
CV % 7.33 24.19
the legume break crops, as the amount of nitrogen left in the soil by a crop may influence the
yield of the crop that follows it (Martin et al., 1976).
The yield increment of local variety was lesser in field pea break as compared to barley
in two consecutive years as the management practices increased. This might be due to the
more nitrogen left in the soil from legume break crop. Tanner et al. (1991) reported that for
most of the characteristics, the effect of faba bean break crop was similar to that of Nfertilizer. The mean grain yield increment ranged from 24 of 71 % was observed from the
FVFP+FE and IVIP-FE following field pea as compared to barely in 1998 (Table 2). Mean
grain yield advantage of 9 to 76 % was recorded from FVFP+FE and IVIP-FE following field pea as compared to barely in 1998 (Table 3). The result argues that with improved agronomic
managements both fields gave similar yields.
Management practices had significant (P<0.05) effects for grain yield in 1998 and when
combined over the two years, but differences were not significant in 1999 (Tables 2, 3 and 4).
There was no significant interaction bewteen previous crop and management practices on
grain yield of wheat in either year, nor in the combined data (Tables 2, 3 and 4).
Orthogonal contrasts of all three factors (variety, cultural practices and fertilizer use)
showed highly significant mean grain yield difference (P< 0.01) (Table 5). The improved
variety gave a mean grain yield increases of 191 kg ha', the improved management practices a yield advantage of 78 kg ha', fertilizer application a mean wheat grain yield increase of 115
kg ha'. Thus improved agronomic practices are very advantageous for wheat production in
the Shambo highlands.
Table 5. Statistical significance of orthogonal comparisons in the wheat management trial at
Shambo, 1998-1999.
Contrasta Mean grain yield
Farmers vs. improved management practices **
Local vs. improved variety **
With fertilizer vs. without fertilizer application **
** Significant at 1% level of probability.
a See materials and methods for complete description of orthogonal contrasts used
Economic analysis of the management practices indicated that the highest net benefit of EB 1554.23 ha'I and a cost:benefit ratio of EB 1.26 profit per unit of investment were obtained
from the treatment with the improved variety, improved management practices and fertilizer
application (IVIP+FE)(Table 6). The second highest net benefit of EB 1537.83 ha', and a
cost:benefit ratio of EB 3.03 profit per unit investment, was achieved with the farmers'
variety, improved agronomic management practices without fertilizer application (FVIP-FE).
However, the marginal rate of return of IVIP+FE over FVIP-FE was only 2.2%, far too low to
warrant the recommendation of this treatment. The cost:benefit ratios of the other
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




39
management practices were EB 1.29, 1.19, 1.30 and 1.08 profit per unit investment, from
IVFP-FE, IVIP-FE, FVIP+FE and FVFP-FE respectively.
Table 6. Partial budget and marginal rate of return (MRR) analyses for the effects of
management practices on the grain yield of wheat at Shambo, 1998-1999.
Item Management practices
FVIP FVFP FVIP IVIP IVFP IVIP
-FE -FE +FE -FE -FE +FE
Average yield (kg ha-') of wheat 1496 1314 1815 1783 1911 2071
Adjusted yield (kg ha-') Wheat 1346 1183 1634 1605 1720 1864
Gross field benefit of Wheat grain 1925 1691 2336 2294 2459 2665
Average straw yield (kg ha-') 2400 1947 2850 2003 1870 2457
Gross field benefit of wheat straw 120 97 143 100 94 123
Total field benefit (EB ha-') 2045 1788 2478 2395 2553 2788
Costs that vary (EB ha-')
Wheat seed cost (EB ha-' 215 215 215 360 360 360
Urea 192 192
DAP 256 256
Rental price of sprayer (EB ha-') 10 10 10 10
Herbicide cost (EB ha-') 42 42 42 42
Fertilizer application cost (EB ha-') 11 10.5
Total labour cost (EB ha-') 241 645 350 683 756 364
Total costs that vary (EB ha-') 508 860 1075 1095 1117 1234
Net benefit 1538 929D 1403D 1300D 1436D 1554
Value to cost ratio 3.03 1.08 1.30 1.19 1.29 1.26
Marginal rate of return (MRR) I I 1 2.20%
Note: D= dominated treatment, Urea= EB 1.92/kg, DAP= EB 2.56/kg, Labour cost =EB
3.5/day, 2,4-D= EB 0.42/litter, Grain price= EB 1.37/kg, Local seed = EB 1.43/kg, Improved seed= EB 2.40/kg, Straw cost= EB 0.05/kg, Rental price of sprayer= EB 10.00 ha', Fertilizer
application cost= EB 10.50 ha', Yield was down adjusted with 10 % coefficient
It is noteworthy that the farmers' common practice (FVFP-FE) was dominated in the partial
budget analysis, because the best treatment (FVIP-FE) gave higher net benefits with lower
costs that vary. In the current study the low cost of seed of the farmers' variety resulted in this variety dominating the treatments with the improved variety. However, it should be noted that the full cost of buying new seed of the improved variety was used, whereas this can probably
be discounted as usually the farmer will use this seed for several years. The high cost of
commercial fertilizer and improved seed made the improved agronomic practices for wheat
production costly, and reduced prices of improved seed and commercial fertilizer will benefit
wheat producers of the area. However, with the average current prices of inputs, use of
farmers' variety with improved technology without fertilizer or the improved variety with
improved practices and fertilizer application were economically viable and profitable. Both
the local and improved varieties gave better grain yield under intensive management practices
with fertilizer application (Table 4). The least profitable management option of wheat is
farmer's variety with farmer's management practices and without fertilizer application.
However, the escalating price of chemical fertilizer and improved seed, together with the
reduced market price of grain drastically decrease the profitability of wheat production with
improved packages.
Conclusion
A previous field pea crop significantly increased the grain yield of wheat compared to a
previous barley crop. Both the local variety and the improved variety gave the greatest yields under improved agronomic management with chemical fertilizer application when following
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




40
after a field pea break crop. The economic analysis indicates that optimum market prices
should be determined by the policy makers for producers of wheat in order to get the potential
benefit of the recommended package. Alternative technologies should be developed on the
integration of inorganic and organic fertilizer use for profitable wheat production. Long term investigation on the economic benefit of rotation effects with intensive management practices and the extension of the results to the farmers is essential for sustainable wheat production in
the Shambo highlands.
Acknowledgment-The authors thank Oromiya Agricultural Research Institute for funding
the project. We are most grateful to our senior researchers Mr. Abdissa Gemeda and Mr.
Abubeker Mussa for their priceless contributions to us in executing this experiment. We are
also grateful to Mr. Asefa Mijena, Tesema Tesso and Yosef kenea for their assistance in
carrying out the experiment and efficiently collecting the data. Bako Research Center
Management is also thanked for facilitating the execution of the field trial.
References
Asefa Taa, Tanner, D G. and Amanuel Gorfu. 1992. The effects of tillage practices on bread wheat in
three different cropping sequences in Ethiopia. pp. 376-386. In: Tanner, D. G. and Mwangi,
W. (eds.). Seventh Regional Wheat Workshop for Eastern, Central and Southern Africa.
CIVMMYT, Nakuru, Kenya.
Asfaw Negassa, Abdissa Gemeda, Tesfaye Kumsa and Gemechu Gedeno. 1997. Agro ecological and
Socioeconomical Circumstances of farmers in East Wallagga Zone of Oromiya Region,
Ethiopia. Research Report No. 32. Institute of Agricultural Research, Addis Ababa, Ethiopia.
Asnakew Woldeab, Tekaligne Mamo, Mengesha Bekele and Tefera Ajema. 1991. Soil fertility
management studies on wheat in Ethiopia. pp.137-172. In: Hailu Gebre-Mariam, Tanner, D.G.
and Mengistu Hulluka (eds.). Wheat Research in Ethiopia: A Historical Perspective. Addis
Ababa, IAR/CIMMYT.
CIMMIYT. 1998. From Agronomic Data to Farmer Recommendations. An Economics Training
Manual. Completely Revised Edition. CIMMYT, Mexico, D.F., Mexico. 79 pp
Hailu Gebre, Amsal Tarekegne and Endale Asmare. 1989. Beneficial break crops for wheat production.
Ethiopian Journal of Agricultural Sciences Vol. 11: 15-24.
Hailu Gebre-Mariam, Tanner, D.G. and Mengistu Hulluka. 1991 Wheat Research in Ethiopia: A
Historical Perspective. Addis Ababa, IAR/CIMMYT.
Martin, J. H., Leonard, W. H., and Stamp, D. L. 1976. Principles of Field Crop Production (3rd ed.).
Collier Macmillan Publishing. London, UK.
Lal, R. 1989. Conservation tillage for sustainable agriculture: Tropics versus temperate environment.
Advances in Agronomy 42:85-197.
NMSA (National Meteorological Service Agency). 2003. Meteorological data of Shambo area for
1969-2002. NMSA, Addis Ababa, Ethiopia.
Tanner, D. G., Amanuel Gorfu and Kassahun Zewdie. 1991. Wheat agronomy research in Ethiopia. pp.
95-135. In: Hailu Gebre-Mariam, Tanner, D.G. and Mengistu Hulluka (eds.). Wheat Research
in Ethiopia: A Historical Perspective. Addis Ababa: IAR/CIMMYT.
Tanner, D.G. 1997. Sustainable wheat production: global perspectives and local initiatives. pp. 10-41
In: Woldeyesus Sinebo (ed.). Crop Management Research for Sustainable Production: Status
and Potentials. Proceedings of the Second Annual Conference of the Agronomy and Crop
Physiology Society of Ethiopia. ACPSE, Addis Ababa, Ethiopia.
Tanner, D.G., Verkuij, H., Asefa Taa and Regassa Ensermu. 1999. An agronomic and economic
analysis of long-term wheat based crop rotation trial in Ethiopia. pp. 213-248. In: The Tenth
Regional Wheat Workshop for Eastern, Centeral and Southern Africa. Addis Ababa, Ethiopia:
CIMMYT.
Tilahun Geleto, Kedir Nefo and Feyissa Tadesse. 2000. Crop rotation effects on grain yield and yield
components of bread wheat in the Bale highlands of southeastern Ethiopia. pp. 316-324. In:
The Eleventh Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis
Ababa, Ethiopia: CIMMYT.
Whyte, R.O., Neisser, G. N. and Trumble, H.C. 1969. Legumes in Agriculture. Food and Agricultural
Organization of the United Nations, Rome, Italy.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




41
Response of Bread Wheat to Nitrogen and Phosphorous Fertilizers at Different Agroecologies of Northwestern Ethiopia
Minale Liben, Alemayehu Assefa, Tilahun Tadesse and Abreham Mariye
Adet Agricultural Research Center, P. 0. Box 08, Bahir Dar, Ethiopia
Abstract-Wheat is one of the major cereal crops grown in northwestern Ethiopia, though its productivity in farmers' fields is very low. In northwestern Ethiopia, N and P deficiencies are among the major biophysical constraints. Wheat fertilizer response trials were conducted at different agro-ecologies on farmers' field during 1999, 2000 and 2001. The objective of the trial was to determine economically optimum rates of N and P fertilizer for the high-yielding semi-dwarf bread wheat cultivars HAR 1868 and HAR 604 at Farta and Laie-Gaient areas, respectively. Four rates of N (0, 46, 92 & 138 kg/ha for Farta where as 0, 41, 82 & 123 kg/ha for Laie-Gaient) and four rates of P205 (0, 23, 46 & 69 kg/ha for both localities) were laid out in a factorial arrangement in RCB design with three replications. The results indicated highly significant differences on grain yield, plant height, tiller and spike density and thousand-kernel weight amongst the fertilizer rates in the two locations. There was a linear increase in grain yield of bread wheat as the rates of N and P increased. The increase in grain yield is significant for nitrogen, P and N by P interaction. The highest grain yield, 3025 kg/ha at Farta and 2610 kg/ha at Laie-Gaient were obtained at the highest fertilizer rates of 138/69 kg N/ P205 ha-' and 123/69 kg N/ P205 ha-', respectively. Fertilizer rates of 138/46 kg N/P205 ha-' at Farta and 123/46 kg N/ P205 ha-' at Laie-Gaient were found economically feasible with net benefit (NB) of
Ethiopian Birr 3538 and 3062, respectively.
Introduction
Wheat ranks fifth in production and area and fourth in mean-yield among the principal cereal
crops grown in the country. The national mean yield is low ranging from about 1.1 t/ha for
peasant farmers to about 2 t/ha for state farms (Hailu et al., 1991). Soil degradation is a
serious problem in many parts of the country, largely as a result of mismanagement of natural
resources. Erosion and farmers' practice of continuous cultivation without fallow aggravate
the soil fertility problem. Northwestern Ethiopia is accompanied with diversified agro
ecologies. Poor soil fertility particularly N and P deficiencies are among the major
biophysical constraints in those agro ecologies (Hailu et al. 1991, UNDP 1996, Aleligne et
al. 1992, ANRS BOPED 2000, Aleligne & Regassa 1992). Previous experiments conducted at
the different agro ecologies of the region indicated variations in yield responses of the wheat crop to fertilizer application. Therefore, this wheat fertilizer response trial was conducted at the two agro ecologically different localities on farmers' field during 1999, 2000 and 2001.
The objective was to determine economically optimum rates of N and P fertilizer for the highyielding semi-dwarf bread wheat cultivars HAR 1868 and HAR 604 at Farta and Laie-Gaient
areas, respectively.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




42
Materials and Methods
The experiment was carried out on luvisols at Farta and Laie-Gaient region on farmers' fields for three consecutive years (1999-200 1) on a total of five sites at Farta and seven sites at LaieGaient. Laie-Gaient and Farta at an altitude of 2860 and 2630 m.a.s.l., respectively are
located on the same zone (south Gondar) of northwestern Ethiopia but represent different agro
ecologies. Laie-Gaient is characterized by erratic rainfall, short rainy season (late onset and
early cease) and low temperature compared to Farta area. The soil fertility status of Farta area
is total N 0.08%, available P 3.70%, OC 3.20, C/N 40 and pH of the soil 5.17.
The experiment consisted of 16 factorial combinations of four N rates 0,46,92 and 138
kg/ha for Farta, 0,41,82 and 123 kg/ha for Laie-Gaient and four P205 rates (0, 23, 46 & 69 kg/ha) at both sites. The design was a complete factorial arrangement within an RCB with
three replications. The improved varieties HAR-1868 (shina) at Farta and HAR-604 (Galema) at Laie-Gaient were used. Seeds were broadcasted at the recommended seed rate of 175 kg/ha.
DAP, urea and TSP were the sources of N and P205. All P and half of the N were applied at sowing, and the remaining N was top dressed at early to mid-tillering stage of the crop. The
gross and net plot sizes were 20m2 and 12m2, respectively. Data were collected on grain yield,
tiller and spike density at maturity, plant height and thousand-kernel weight. All data were subjected to analysis of variance (ANOVA) using the MSTATC microcomputer software.
The mean grain yield data over sites for each location was adjusted down by 10% and
subjected to partial budget and sensitivity analysis (CIMMYT, 1988). Total costs that varied
(fertilizer cost) for each treatments was calculated and treatments were ranked in order of
ascending total variable cost (TVC) and dominance analysis was used to eliminate those
treatments costing more but producing a lower net benefit than the next lowest cost treatment.
The marginal rate of return (MRR) was calculated for each non-dominated treatment and a
minimum acceptable MRR of 100% was assumed. Sensitivity analysis was made through the
assumption that cost of fertilizer and price grain increased by 10%, respectively.
Results and Discussion
Analysis on mean grain yield of the individual sites indicated significant responses for N
application in all sites. At Farta, all sites except one showed significant response to P
application where as at Laie-Gaient only one site showed significant responses to P (Table 1).
Fertilizer application increased grain yield ranging from 1554 to 2602 kgha' at Farta and
1208 to 2447 kgha-' at Laie-Gaient compared to the unfertilized control.
Results on combined analysis indicated significant response for N application in all traits
except thousand kernel weight at Laie-Gaient. The response for P application was significant
on grain yield, plant height and thousand kernel weight at Farta. At Laie-Gaient all traits showed significant response except thousand kernel weight. Most of the traits showed a
significant responses to NxP interaction (Table 2). Generally there was linear increase in all
the parameters as the N and P rates increased. This is in line with previous findings that at
most sites of northwestern Ethiopia, wheat grain yield responds positively to increasing
fertilizer rates (Minale et al. 1999; Ammanuel et al. 1990). The responses were relatively larger for N than P in all parameter considered. Biologically the highest grain yield 3052
kghaI at Farta and 2610 kgha at Laie-Gaient were obtained at the highest fertilizer rates in
both of the locations (Table 3 & 4). An increase in grain yield, 175.70 and 106.50%, over the
control was recorded at Farta and Laie-Gaient areas, respectively.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




Table 1: Results of ANOVA and mean grain yield of bread wheat on the individual sites of the two regions
CDFarta region Laie-Gaient region
Source
of variation Site-1 Site-2 Site-3 Site-4 Site-5 Site-1 Site-2 Site-3 Site-4 Site-5 Site-6 Site-7
N ** ** ** ** ** ** ** ** ** **
P205 ** ** NS ** NS NS NS NS NS NS
NX P05 NS NS NS NS NS NS NS NS NS NS NS
Meanyield (kgha-') of 495 1663 1510 490 1376 1689 1311 1134 889 1483 1381 961
0/0 N/ P205
Highest mean yield 2816 3453 3603 3092 2930 4136 3038 2342 2337 2925 2768 2218
(kgha-l)
N/ P205 comb. for 138/69 92/69 92/23 138/46 138/46 92/46 123/69 123/69 123/69 123/46 123/23 123/69
S highest yield
CV(%) 20.02 25.10 18.40 37.54 21.45 35.24 28.00 37.66 28.28 29.56 20.27 31.42
* ,** Indicate significance at the 5 &1% levels, respectively.
NS Indicate non-significance
Table 2: Results of combined analysis of variance over sites of the two locations
Farta area Laie-Gaient area
Source Grain plant Tiller /m2 Fertile 1000 Grain plant Tiller /m2 Fertile 1000
of variation yield height at spike/mn2 kernel yield height at spike/mn2 kernel
(kg/ha) (cm) maturity wt.(g) (kg/ha) (cm) maturity wt.(g)
Site (S) ** ** ** ** ** ** ** ** ** **
S N ** ** ** ** ** ** ** NS
S NxS NS ** ** ** NS NS NS **
P205 ** ** NS NS ** ** ** NS
P205 x S NS NS ** NS NS NS NS *
Nx P25 NS NS NS
Mean value of 0/0
N/P205 1107 63.1 319 304 26.8 1264 71.2 245 234 41.3
Highest mean
value 3052 82.3 391 381 32.8 2610 95.3 308 300 43.5
N/P205 comb. for
highest value 138/69 138/69 138/0 138/0 138/69 123/69 123/69 123/69 123/23 41/46
* ,** Indicate significance at the 5 &1% levels, respectively.
NS Indicate non-significance
cO




44
Table 3: Effect of nitrogen (N) and phosphorus (P205) on the grain yield (kg/ha) of bread wheat (Shina variety) at Farta area
N levels (kg/ha) P205 levels (kg/ha) Mean
0 23 46 69
0 1107 1429 1570 1802 1477
46 1535 1832 2115 1990 1868
92 2033 2359 2696 2426 2379
138 2116 2833 3025 3052 2757
Mean 1698 2113 2352 2318
C.V(%) 24.8
N P NxP
LSD 5Oo) 190 190 380
Table 4: Effect of N and P on the grain yield (kg/ha) of bread wheat (Galema variety) at Laie-Gaient area
P205 rates kg/ha Mean
N rates (kg/ha) 0 23 46 69
0 1264 1103 1280 1355 1251
41 1630 1413 1754 1605 1601
82 1817 2206 2237 2297 2139
123 1890 2457 2599 2610 2389
Mean 1650 1795 1967 1967
C.V(%) 31.2
N P NxP
LSD (5%) 174.8 174.8 349.6
The economic analysis indicated 138/46 kg N/P205 ha' was more profitable with net benefit (NB) ofbirr 3538.4 and an acceptable marginal rate of return (MRR) of 171.9% at Farta area (Table 5). The NB increased by birr 1958.6 over the unfertilized one through the application of 138/46 kg N/P205 ha'-1. The sensitivity analysis revealed the same recommendation as to the current situations.
Table 5: Results of the economic analysis of N and P205 fertilizers rate on grain yield of bread wheat at Farta area
At current cost and price Scenario (cost of fertilizer & price of grain
increased and decreased, respectively by
10%)
N/P205 TVC NB MRR TVC NB MRR
(kgha-1) (Eth. birr) (Eth. birr) (%) (Eth. birr) (Eth. birr) (%)
0/0 0 1579.8 0 1421.8
0/23 138.8 1900.5 231.0 152.7 1682.7 170.8
46/0 192.7 1998.7 182.4 211.9 1760.3 131.1
46/23 293.8 2321.5 319.2 323.2 2030.6 243.0
92/0 385.3 2516.0 212.5 423.9 2187.4 155.7
46/46 395.0 2624.5 1128.4 434.4 2283.0 905.3
92/23 486.5 2881.6 280.9 535.1 2496.1 211.6
92/46 587.6 3261.2 375.3 646.4 2817.5 288.9
138/23 679.1 3364.5 112.9 747.1 2892.2 74.2
138/46 780.3 3538.4 172.0 858.3 3028.5 122.5
Note: Cost of DAP = Birr 2.77/kg
Cost of Urea = Birr 1.92/kg
Price of wheat grain = Birr 1.58/kg
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




45
The highest NB of birr 3213.2 with acceptable MRR (122.51%) was obtained at fertilizer rates of 138/46 kg N/P205 ha-'. At Laie-Gaient, 123/46 kg N/P205 ha was more profitable with a NB ofbirr 3061.7 and MRR 117.5% (Table 6). The sensitivity analysis showed 123/46 kg N/P205 haI gave the highest NB, however, the MRR is below the acceptable level. Therefore, with the assumption that price of grain and fertilizers decreased and increased, respectively by 10%, fertilizer rate of 82/23 kg N/P205 ha-' will be the optimum rate to be recommended.
Table 6 Results of the economic analysis of N and P205 fertilizers rate on grain yield of bread wheat at LaieGaient area
At current cost and price Scenario (cost of fertilizer & price of grain
increased and decreased, respectively by
10%)
N/P205 TVC NB MRR TVC NB MRR
(kgha-1) (Eth. birr) (Eth. birr) (%) (Eth. birr) (Eth. birr) (%)
0/0 0 1846.7 0 1662
41/0 181.5 2200.5 194.9 199.7 1944.1 141.3
82/0 363.3 2290.7 49.6 399.6 1989.0 22.5
82/23 458.6 2764.0 496.5 504.4 2395.9 388.2
123/23 640.1 2949.4 102.2 704.1 2526.4 65.3
123/46 735.6 3061.7 117.5 809.2 2608.4 78.1
Note: Cost of DAP = Birr 2.70/kg
Cost of Urea = Birr 2.03/kg
Price of wheat grain = Birr 1.62/kg
References
Aleligne Kefyalew, and Regassa Ensermu. 1992. Bahir-Dar Mixed Farming Zone: Diagnostic Survey Report. Research
Report No 18. Institute of Agricultural Research, Addis Ababa, Ethiopia Amahara National Regional State
Bureau of Planning and Economic Development (ANRS BOPED). 2000. Annual Statistical Bulletin, Bahir
Dar Ethiopia.
CIVMMYT. 1988. From agronomic data to farmer recommendations: An economics training manual. Completely
revised edition. Mexico, D.F.
Hailu Gebre-Mariam, Tanner, D.G., and Mengistu Hulluka, (eds.). 1991. Wheat Research in Ethiopia: A historical
perspective. Addis Ababa: IAR/CIMMYT.
UNDP. 1996. Sustainable Agricultural and Environmental Rehabilitation Program (SAERP). House Hold Level SocioEconomic Survey of the Amhara Region. V.1 Produced by the Cooperative Endeavors of the Amhara
Regional Council. Bahir Dar, Ethiopia.
Ammanuel G, D.G. Tanner, Lemma Z, Tilahun G, Zewdu Y. and Eyasu E. 1990. Derivation of Economic fertilizer
recommendations for bread wheat in the Ethiopian highlands: On-Farm trial in the peasant sector. PP. 63-72.
In; Tanner, D.G., M. van Ginkel, and W. Mwangi, eds. Sixth Regional Wheat Workshop for Eastern, Central
and Southern Africa. Mexico, D.F; CIMMYT.
Minale Liben, Alemayehu Assefa, D.G. Tanner and Tilahun Tadesse. 1999. The response of bread wheat to N and P
application under improved drainage on Bichena Vertisols in northwestern Ethiopia. The Tenth Regional
Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




46
Grain Yield, Water Use and Water Use
Efficiency as Affected by Moisture Level Under Rain-Out Shelter
P. A. Ooro 1, M.G.Kinyua, and J.B.O. Ogola2
'K4RI-NPBRC, P.O. Njoro, Tel. 254-051-61070, Kenya, paooro@yiahoo corn
2Department ofplant production, University of Venda, Private Bag X5050, Thohoyandou, Republic of South Africa, *Corresponding author
Abstract-About one-third of the developing world's wheat area is located in environments that are regarded as marginal for wheat production, because of drought, heat, and soil problems. Nearly one-third of the area planted to bread wheat and about three-quarters of the area planted to durum wheat suffer from drought stress during the growing seasons. Despite these limitations, the world's dry and difficult cropping environments are increasingly crucial to food security in the developing world. Gains in wheat productivity in marginal environments are important because it is unlikely that increased productivity in the favourable environments will be sufficient to meet the projected growth in demand for wheat from the present to 2020. This study was aimed at evaluating wheat genotypes' water use efficiency (WUE) as affected by moisture regime. An experiment was conducted under min shelter for two seasons (2001-2002) with six cultivars (Duma, Ngamia, Chozi, Kwale, Mbuni and Pasa) tested under two moisture regimes (High and low moisture regimes). The experiment was a randomized complete block design (RCBD) with split arrangement of the treatments. Moisture regime was assigned as the main plot and cultivars as the sub-plot. An analysis of variance was carried out on combined season using SAS computer package. The genotypes tested were significantly different in their water use efficiency (WUE) under both low and high moisture conditions. The drought-tolerant cultivars (DTC)-Duma, Ngamia and Chozi-had significantly higher WUE under moisture stress than the drought-susceptible cultivars (DSC)-Kwale, Mbuni and Pasa. Under high moisture the WUE of the
DTC was decreased by 10-19% and in the DSC, it increased by 26-29.
Introduction
Cereals contribute significantly to food security in Kenya (Wanjama et al, 1993). Among
cereals, wheat has been ranked second to maize (KARI, 1991) and is grown on an estimated area of 105,000 ha (representing about 6% of the total area under cereal production in Kenya) giving an annual production of about 350,000 metric tones (Ekboir, 2002). The crop is mainly grown in the Rift Valley province with Nakuru, Uasin Gishu and Narok districts contributing
about 85% of the total national production (Wanjama et al, 1993). The national wheat
production has varied over time. For example, in the period 1988 2000, wheat production
declined annually by about 1.5% (Aquino et al., 2002).
Due to the reduction in wheat area, Kenya produces only 40% of the national demand of
approximately 720,000 tones, and importing about 60% of its wheat requirement (Aquino et
al., 2002). For example, between 1997-1999, wheat imports stood at 484,900 tones (5.4
million 90-kg bags), costing about 5.85 billion Kenya shillings (US$ 84 million). Kenya is
bound to remain a net importer of wheat and wheat products unless domestic wheat
production is significantly stepped up. Thus for wheat production to be increased to meet the
increasing demand in Kenya, among other technologies, wheat growing has to be expanded to
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




47
the marginal rainfall areas which consist of 83% of the total land area with unused arable
potential of approximately 200,000 hectares (World Bank, 1989).
Furthermore, it has been reported that 32% of the 99 million hectares of wheat grown in
developing countries experiences varying levels of drought stress (Rajaram et al., 1996). In
Kenya, for example, marginal rainfall areas have been able to achieve wheat yields of between one third and one half of those from high rainfall areas (Kinyua et al., 1989).
Promotion of wheat farming in the dry lands of Kenya can be done through several ways.
These include use of irrigation and/or drought resistant wheat varieties. Irrigation would require steady supply of water and due to scarcity of water in the marginal areas and large
capital required, irrigation is not easily achievable. Thus development of appropriate varieties
for marginal areas may be the most effective alternative for utilizing these areas for crop
production. Therefore, the main objective of this study was to evaluate WUE as an indirect
method for selecting drought tolerant wheat cultivars.
Materials and Methods
An experiment was conducted in the rain shelter for two seasons between August, 2001 and May, 2002 at the National Plant Breeding Research Centre (NPBRC), Kenya (00 20' S and
35056'E and altitude 2185 metres) [NPBRC Meteorological Station No. 9035021, 1999].
The site lies within the Agro-ecological zone LH3 (AEZ LH3) with a bi-modal rainfall pattern
(Jaetzold and Schmidt, 1983). The site receives an annual rainfall of about 960 mm with an
average maximum and minimum temperatures of 24 o C and 8 o C, respectively (NPBRC,
Njoro Meteorological station No.9035021, 1999).
The soil at the is site well drained, deep to very deep, dark reddish brown, friable and
smeary, silt clay, with humic topsoil classified as mollic Andosols (Jaetzold and Schmidt,
1983). The site was under fallow for two seasons before this study was carried out.
The experiment was carried out under a mobile rain-out shelter (which is an open ended structure measuring 30 m long and 7 m wide constructed at NPBRC) to exclude rain and consequently induce drought stress. The shelter consists of a roof mounted on wheels that
allows it to roll on two parallel-elevated concrete tracks. Translucent sheets (which allow up
to 90% photosynthetic photon flux density) covered the roof The sides had metal frame,
which was covered by foldable polythene sheets, with open-end 0.5 m above the ground. A
concrete barrier had also been constructed to a depth of 0.6 m below the ground level and
0.15 m above the ground level along the two longer sides of the shade from a rail. The length
of the rail was 30 m but the shelter rested on half the length. The barrier also helped to
prevent the rainwater from flooding the shade. The mobile rainout shelter used is similar to
that described by Legg et al. (1978) and Upchurch et al. (1983).
Six bread wheat genotypes Duma, Chozi, Ngamia, Kwale, Mbuni and Pasa were grown
under two watering regimes (High and low watering regimes) for two seasons. The
experiment was laid out in a randomized complete block design (RCBD) with split plot
arrangement with the watering regime being the main plot and cultivar as sub-plot.
Thirty-six (36) access tubes (PVC in nature) with 39.4 mm and 41.4 mm as internal and
outside diameters, respectively, were installed in the field before planting. In each sub-plot, an
access tube was installed in holes drilled by hand using a soil auger. The tubes were cut to a depth of 1.7 m out of which 1.1 m was driven into the holes and 0.6 m was left above the soil
surface. Moisture readings were taken by use a Troxler moisture meter (Neutron probe). In
the process of moisture reading a standard count was at the beginning of each moisture
regime.
Hydrogen ions, standard counts and moisture percentage by volume readings were taken
through the access tubes at an interval of 7 days at an interval of 10 cm in the upper 40 cm
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




48
and 20 cm below 40 cm. The hydrogen counts were divided by the standard counts to derive the count ratio that was later fitted in the calibration equations, for different soil levels within
the profile, to calculate the actual soil water content.
The seeds of the wheat varieties used in the trial were tested for germination before
planting and the seed-rate was adjusted according to the germination percentage. The
recommended seed-rate of 125 kg ha' was used.
The six cultivars were planted in plots of 0.8 metres length and 2 metres wide with ten
rows. The rows were 20 cm a part and planting was done by hand. The seeds were drilled
continuously on the furrows (5 cm deep) made by hand. Triple super-phosphate (0-46-0) was applied in the furrows at 60 kg P205 ha planting and calcium ammonium nitrate (CAN) at 40
kg N ha (recommended rate) was top-dressed at early tillering stage to supply Water was
supplied through a drip irrigation at a constant pressure as per the treatments. One drip lateral
was laid between two wheat rows.
Watering was done once weekly and a day before watering was done moisture reading
were taken by use of neutron probe. Watering was done when the moisture level had come to
50% field capacity and watering done according to treatments. The high moisture treatment
received 14.5 mm and the low moisture regime received 7.5 mm once weekly. The total
amount of water applied during each season was 217 mm and 113 mm for the high and low
moisture regimes, respectively.
Weed control was done by the use of Buctril Mc at the rate of 1.4 L ha' at the 3-4-leaf
stage of crop. The chemical was used to take care of broad-leaved weeds but for the grasses,
hand weeding was done. Six innermost rows were hand harvested for biomass and grain yield
determination.
Evapo-transpiration (ET) included both transpiration and direct evaporation of water
from the soil surface (Es,). Evapo-transpiration (ET) was determined by using the soil water
balance equation given below:
ET = -AS + I -D R (3.1)
where -AS is the change in storage (obtained by the difference in volumetric moisture content of the entire profile at the beginning and end of the season), I is irrigation water applied, D is the drainage and R is the run-off Drainage and run-off were assumed to be negligible hence
total crop water use for the whole season obtained by summing total irrigation where
applicable) and changes in storage for the entire profile (Ogola, 1999). Pilbeam et al., (1995) suggested that significant drainage is considered to have occurred where the soil water at the
base of the profile was increased above its initial value. Drip irrigation was used hence
drainage was assumed to be zero and thus not included in the water balance equation.
For the purposes of this study, water use efficiency (WUE) was defined as a ratio of grain
yield (GY) to evapo-transpiration (ET). Thus WUE was determined as follows:
WUE = GY/ET. (3.2)
where GY is the grain yield.
Grain yield (GY) was determined by harvesting one metre length of the two inner-most
rows. Harvesting was done by hand and threshed by the use of laboratory single head thresher
and grain yield was standardized at 12.5% moisture content.
Thousand-kernel weight (TKW) was determined after grain weight had been recorded from both experiments. A thousand seeds were picked randomly and weighed to determine TKW.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




49
Neutron probe calibration procedure
The soil profile of the site (NPBRC field 1) was described and the soil depth of the site
determined and the pumice was realized to start immediately after 60-cm depth. For
calibration, two access tubes were installed each at a depth of 110 cm. One of the tubes was
used to get wettest point of the calibration curve whereas the other one was for the drier
points of the curve (Ooro et al., 2001).
The place surrounding the tube for the wettest point was flooded with water and left to
settle for 24 hours and then soil samples for gravimetric moisture determination were taken at
10, 20, 30, 40, 60, 80 and 100 cm depths. At the corresponding depths, neutron probe
readings were taken for the hydrogen ion counts and hence volumetric water content. The
same approach was used around the access tube for the lower points of the calibration curve
(Manual of operation and Instruction, 1996).
The data generated from the neutron probe readings (from the wet and dry calibration
spots) were used to calculate count ratios (F/Standard count) plotted on the x-axis of the
calibration curves and two equations were then developed for the top soil (upper 20 cm) and below 20 cm depths. The two calibrations were developed due to the variations between the readings of depths up to 20 cm and those below 20 cm. Hydrogen ions readings up to 20 cm
depth were relatively lower due to scatter effect while this effect did affect readings below 20
cm. Plotting gravimetric moisture readings on the y-axis and the count ratios on the x-axis
developed the equations. The equations were as follows:
0 -20cm depth: Y = 0.0661n + 0.001615 (3.3)
30 100 cm depth: Y = 0.20960n + 0.001132 (3.4)
where Y is the actual moisture volumetricc moisture), n = count ratios (F/standard counts)
[H+ was read from the neutron probe].
The data read by neutron probe in the rain-shade experiment was fitted on the calibration
curve to give the actual moisture content.
Data from all the parameters, combined over two seasons (2001 and 2002), were
subjected to an analysis of variance (ANOVA) using SAS (SAS User's guide, 1985). Least significant difference procedure (Lsd) was used to carry out mean separations. The data was also subjected to correlation and path coefficient analysis was done to separate the direct and
indirect between the agronomic traits and water use efficiency.
Results
The interaction between genotype and moisture regime affected final grain yield. High
watering regime increased grain yield of both the drought tolerant and drought susceptible
varieties but the increase in yield due to high watering regime was greater in susceptible
(110%- 134%) than in tolerant (55-64%) varieties. The drought tolerant varieties produced
greater yield than the drought susceptible ones under low watering regime. Under high
moisture regime, in contrast, the drought sensitive varieties produced greater yields than the tolerant ones. Significant genotypic differences in grain yield were detected. Averaged over the low watering regimes, the drought tolerant cultivars had higher yield (1274 kg ha') while
the drought susceptible ones had relatively lower yield (972 kg ha'). Moreover, drought
tolerant varieties produced greater (by 39%) yield than the drought susceptible varieties under low watering regime. Averaged over all the genotypes, high watering regime increased grain
yield by 89% (from 1123.2 kg ha' to 2119.0 kg ha').
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




50
There was significant effect of the interaction between moisture regime and cultivar on
crop water use, such that the increase in water use due to high watering regime was greater for
drought sensitive (96%) than for drought tolerant (93%) varieties. Also, under low moisture
conditions Kwale used more water (by 10%) than all the other cultivars tested while under
high moisture, Chozi had significantly lower (by 3%) water use value than the rest of the
cultivars. Significant genotypic differences on water use were detected. Averaged over the
watering regimes, Kwale had the highest water use (139 mm) while Pasa had the lowest water
use (132 mm) as compared to the other varieties tested. High watering regime increased
water use (averaged over all varieties) by 95% (from 91 to 177 mm).
The interaction between variety and moisture regime affected water use efficiency
(WUE) significantly (Table 1). High moisture regime decreased (by 10%-19%) WUE of
drought tolerant varieties (Duma, Ngamia and Chozi) but increased (by 26%-29%) WUE of
drought sensitive varieties (Kwale, Mbuni and Pasa). Under low moisture regime, drought
tolerant varieties had higher WUE values than the drought sensitive varieties while the
converse was true under high moisture level (Table 1). No significant effects of cultivar on
WUE were detected. However, high watering regime increased WUE (averaged over all
wheat varieties) by 4% (from 9.6 to 9.9 kg ha' mm-') (Table 1).
Table la. Grain yield, water use and water use efficiency of six wheat genotypes under two
watering regimes
Treatment Grain Yield Water Use Water Use
Moisture regime Cultivar (kg ha-') (mm) Efficiency (kg ha-' mmn)
Low
Duma 1213.ab 90.b 10.7ab
Ngamia 1286.a 91.b 10.9a
Chozi 1324.a 91.b lI.la
Kwale 1152.abc98.a 9.0b
Mbuni 902.bc 88.b 7.9c
Pasa 862.c 87.0b 7.7c
High
Duma 1886.b 177.a 8.7b
Ngamia 2097.b 177. a 9.8ab
Chozi 2173.ab 171. b 9.6b
Kwale 2424.a 179. a 11.3a
Mbuni 2114.ab 177.a 10.lab
Pasa 2020. b 177.a 9.9ab
SE 11 3 0.58
F-test probabilities
Irrigation (I) P<0.001 P<0.001 P<0.05
Cultivar (C) P<0.05 P<0.05 ns
Interaction (CXI) P<0.05 P<0.001 P<0.001
Figures followed by the same letter within the same moisture regime in the same column are
not significantly different at Lsd 5%.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




51
Table lb. Varietal differences in yield components of wheat at two moisture levels
Treatment No. of Plant TKW No. of
Moisture regime Cultivar tillers/plant height (cm) (g/1000seeds) seeds/spike
Low
Duma 4.5ab 53.b 43.b 26.c
Ngamia 2.7c 52. b 39.c 40.a
Chozi 3.7bc 58.a 46.a 27.bc
Kwale 3.5bc 55.ab 45.ab 25.c
Mbuni 4.0b 55.ab 46.a 35.abc
Pasa 5.7a 58.a 43. b 36.ab
High
Duma 3.8bc 69.bcd 44.bc 33.b
Ngamia 3.7bc 66.cd 41.d 36.ab
Chozi 5.7a 81.a 49.a 28.b
Kwale 4.7ab 70. bc 46.b 34.ab
Mbuni 3.3c 73.bc 42.cd 35.ab
Pasa 4.5abc 65.d 40.d 43.a
SE 0.63 2. 1 9
F-test probabilities
Irrigation (I) P<0.001 P<0.05 ns P<0.001
Cultivar (C) P<0.001 ns P<0.05 P<0.05
Interaction (CXI) P<0.05 P<0.001 P<0.001 P<0.001
Figures followed by the same letter within the same moisture regime in the same column are
not significantly different at Lsd 5%.
Discussion
Improving drought tolerant of wheat has long been a major objective of most breeding
program because water deficits during some parts of growing period are common in most regions of the world where wheat is produced under rain-fed conditions. Therefore in this
study the cultivars tested were evaluated with an aim of understanding their reactions under
low and high moisture regimes. Significant interactions were observed between moisture level
and water use efficiency (WUE) of wheat cultivars. The WUE of drought tolerant wheat
cultivars (Duma, Ngamia and Chozi) was decreased (by 14%) under high moisture but increased (by 30%) WUE of drought sensitive varieties (Kwale, Mbuni and Pasa). This
behaviour of drought tolerant cultivars under low moisture was probably due to the fact they
had less tiller number per plant compared to the susceptible varieties. Fewer tillers of drought
tolerant wheat have been reported from studies by Ehlig and Lemert (1976) and Singh et
al., (1979). Low tiller number per plant under low moisture could be a mechanism that
drought tolerant used reducing the demand for water. In a study involving the evaluation of
the effects of moisture stress, reduced maximum tiller number of Triticum tauschii (wild
relative of wheat) were reported to be strongly reduced by low moisture. Those findings were
similar to those reported for cultivated wheat (Cone et al., 1995).
The significantly high WUE of a cultivar such as Chozi might have been due to
therelatively taller plants under low moisture which probably converted into higher biomass and relatively higher grain yield. Drought tolerant cultivars (Duma and Ngamia) maintained
shorter plant stature under low moisture, which is another mechanism of economizing on
moisture hence, increased WUE. The findings of this study are in disagreement with those reported by Rezgui et al., (1999) from their study involving evaluation of 61 durum wheat
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




52
genotypes in a moisture stressed environment at Mograne in Tunisia. From their study it came
out that taller plants were associated with increased WUE.
The thousand-kernel weight (TKW) for the drought tolerant cultivars (Chozi, Duma and
Ngamia) was higher under both low and high moisture. Less tillers per plant under low
moisture observed on the drought tolerant cultivars was boosted by the higher TKW value.
The high TKW value resulted in heavy individual seeds which in turn culminated in higher
grain yield which appeared to increas WUE mainly due to higher harvest index (HI). The fact
that the drought tolerant cultivars (Chozi, Duma and Ngamia) were shorter in stature HI was inversely related to plant height resulting into higher WUE under both low and high moisture
conditions. Ehdaie (1995) reported higher WUE of shorter wheat cultivars under both low and high moisture regimes in a study conducted on bread wheat to evaluate seasonal WUE
and its components in Spring wheat.
Lower seed number per spike was observed on drought tolerant cultivars (Duma and
Chozi) a part from Ngamia under low moisture. The low seed number per spike translated
into increased seed size and therefore increased TKW. Rebetzke et al., (2001) reported
increased kernel size due to die back of wasteful tillers in wheat under drought in Australia. A
study carried out to evaluate the stability of grain yield and its components under moisture
stress also revealed that the overall moisture deficit induced reduction in yield of specific
wheat cultivars were largely due to similar effects on kernels per spike (Guttieri et al., 2001).
Drought tolerant cultivars (Chozi, Duma and Ngamia) maintained high TKW even under
low moisture. Despite the fact that these cultivars had fewer tillers per plant under low
moisture, they had relatively higher WUE value. The high WUE value may be attributed to high TKW, which helped to maintain higher grain yield resulting into increased WUE. The drought tolerant cultivars (Duma and Chozi) a part from Ngamia had relatively lower seed
number per spike under low moisture. The high seed number per spike translated into
increased TKW which may be because the low seed number per spike resulted into less
competition giving rise to increased kernel size. Rebetzke et al., (2001) reported that wheat
cultivars that had tiller inhibition genes had increased kernel size. Correlation study
confirmed that tiller number per plant had a high negative (P<0.05) correlation with WUE of
Duma under low moisture. However, under the same moisture regime, Ngamia and Chozi were not significantly correlated to tiller number per plant. This showed that tiller number
controlled WUE positively on Duma but not Ngamia and Chozi under low moisture regime.
Conclusion and Recommendation
Moisture affected the grain yield and water use efficiency (WUE) of the genotypes tested significantly. Drought tolerant wheat cultivars (Duma, Ngamia and Chozi) under moisture stress used moisture more efficiently than the drought susceptible cultivars (Kwale, Mbuni
and Pasa). The increased WUE of drought tolerant wheat cultivars (Duma, Ngamia and Chozi) was the basis of their drought tolerance. This has an implication on the drought
tolerance selection in wheat research.
In view of the aforementioned results, the understanding of the traits that contributed to
the increased WUE of the drought tolerant wheat genotypes under moisture stressed
environment was addressed in the second part of this study. However, the results of that
section of the study have not been included in this paper.
References
Aquino, P., Carrion, F., and Calvo, R. 2002. Selected Wheat Statistics. In Ekboir J.(ed.) 2002.
CIMMYT 2000 2001 World Wheat over-view and outlook: Developing no till packages for
small-scale farmers. Mexico, D.F.:CIfl)T.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




53
Cone, A. E., Stafer, G.A., and Halloran, G.M. 1995. Effects of moisture stress on leaf appearance,
tillering and other aspects of development in Triticum tauschii.Euphytica. 86 (1): 56-64.
Ehdaie,B. 1995. Variation in water-use efficiency and its components in wheat: II. Well-watered pot
experiment. Crop Science. 1995. 35 (6). 1617-1626.
Ehlig, C.F., and LeMert, R.D. 1976. Water use and productivity of wheat under five irrigation
treatments. Soil Sci. Soc. Am. J 40:750-755.
Ekboir, J. (ed.) .2002. CIMMIYT 2000-2001World Wheat Overview and Outlook: Developing No-Till
Packages for Small-Scale Farmers. Mexico, DF:.CIMMYT.
Guttieri, M.J., Stark, J.C., O'Brien, K., and Souza, E. 2001. Relative sensitivity of Spring wheat grain
and quality parameters to moisture deficit. Crop Science. 41 (2): 327- 335.
Jaetzold, R., and Schmidt, H. 1983. Farm Management Handbook of Kenya, Vol IIB3, Central Kenya
(Rift Valley and Central Provinces). Ministry ofAgriculture in Cooperation with the German
Agric. Team (GAT) of the German Agencyfor Technical Cooperation (GTZ). Printedby
Typo-druck, Rossdorf, W. Germany. 381-416pp.
KARI. 1991. KARI agricultural research priorities to the year 2000. KARL, Nairobi, Kenya.
Kinyua, M.G., Wanjama, J.K., Kamwaga, J., and Migwi, S. 1989. The situation in wheat production in
Kenya. Kenya Agricultural Research Institute (K4RI) Report.
Legg, B.J., Day, W., Brown, N.J., and Smith, G.J. 1978. Small plots and automatic rain shelter. A field
appraisal. Journal ofAgricultural Science 91: 321 326.
Manual of Operation and Instruction, 1996. Calibration of Depth moisture gauge (Model 4300), pp.C24. Troxler Electronic Laboratories, Inc. and subsidiary Troxler International, Ltd, North
Carolina, USA.
National Plant Breeding Research Centre (NPBRC). 1999. Weather Data. Meteorological Station No.
9035021.
Ooro, P.A., Kinyua, M.G., Ogola, B.O., Owido, S.O., Kimurto, P.K., and Mwangi, S.K. 2001. Soil
Profile characterization and Moisture meter (Neutron Probe) calibration. Report, National
Plant Breeding Research Centre (NPBRC), Njoro.
Rajaram, S., Braun, H.J., and Ginkel, M. van. 1996. CIVMMYT's approach to breed for drought
tolerance. Eupytica. 92, (1/2), 147-153.
Rebetzke, G.J., Bonnett, D.G., Richards, R.A., Condon, A.G., and Herwaarden van, A. 2001. Breeding
for high water use efficiency increases grain yield. CSIRO, Plant Industry.
Rezgui, S., Yahyaoui, A., Amara, H., and Daaloul, A. 1999. Relationships between water
use efficiency and agronomic traits of selected durum wheat cultivars. Revue de l'INAT. 14
(1): 143-153.
SAS Institute Inc. 1985. SAS user's guide: Statistics, Version 5 Edition. SASInst. Inc. Carry, NC.
Singh, N.T., Singh, R., Mahajan, P.S. and Vig, A.C. 1979. Influence of supplemenatary irrigation and
pre-sowing soil water storage on wheat. Agron.J 71:483-486.
Upchurch, D.R., Ritchie, J.T., and Foale, M.A. 1983. Design of a large dual-structure rainout
shelter. Agronomy Journal 75: 845 -848.
Wanjama, J.K., Macharia, M., Karanja, D.D., Pinto, J.M., Maina, M.P.D., Kinyua, M.G., Wanyera, R.,
Faraj, A., and Gitari, J.N. 1993. Strategic Planning For Wheat Research To The Year 2013.
National Plant Breeding Research Centre, Njoro, pp.2-28.
World Bank. 1989. Kenya Agricultural growth and strategy options. Unpublished Sector Report.
Nairobi, Kenya.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




54
Part 2. BREEDING
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




55
Current Status of Stem Rust in Wheat Production in Kenya
R. Wanyera, M.G Kinyua, P. Njau, J.W. Kamundia and S.Kilonzo
Kenya Agricultural Research Institute- National Plant Breeding Research Centre
P.O. Private Bag, Njoro- 20107, Kenya. Email: cepnjr@wananchi.com, or rvkepawae@wananchi. com
Abstract-Stem rust Puccinia graminis f. sp tritici is among the three rusts currently threatening wheat (Triticum aestivum L) production in Kenya. A three-year survey (2002-2004) was carried out to determine the distribution of stem rust on the wheat varieties in the country, and to study the reaction of wheat varieties/lines to stem rust inocula from different sources. Fifty-three (53) Stem Rust Parental Collections (SRPC) were also screened in the experimental cage at Njoro whereas 16 commercial wheat varieties were screened in the field in Njoro, Mau-Narok and Eldoret arranged in a RCBD with 3 and 2 replications respectively. Fourteen (14), 9 and 11 sets of breeders F3, stem rust differentials from Mexico and Australia were also screened in the greenhouse. The survey results indicated that stem rust has reched the high altitude areas where maximum and minimum temperatures range from 22 to 260 C and 6 tol4C respectively. Out of 53 SRPC screened, 79.3% were susceptible, while 20.7% showed resistance. The 16 commercial varieties were susceptible. Varieties Ngamia, Duma, Yombi, Heroe were most susceptible with a mean score of 63%, 48%, 46%, and 43% respectively. Kenya Fahari, K. Paka and Njrbwl were least susceptible (5%, 4%, and 14%). The infection types of stem rust spores from different sources were different on the stem differentials from Australia. These preliminary data indicate the possibility of more than one stem rust race attacking the wheat crop. Joint effort is required to utilize the current biotechnology tools to
identify and breed for resistance to the new race(s).
Introduction
Stem rust (Puccinia graminisfp tritici) is among the cereal rusts that cause severe losses
throughout the world. The rust range can cause losses ranging from slight to complete
destruction of a wheat crop. Losses of 50-70-% have been reported in the field. Infected grain
is shriveled, light in weight and of low quality (Stubbs et all, 1986; Agrios, 1988; Wiese,
1991; Minzenmayer, 2000; Cereal Disease Laboratory, 2004). In addition to reducing grain
yield, rusts lower the crop's forage value and predispose plants to other diseases. Rusted
wheat plants are less palatable and are toxic to livestock (Wiese, 1991). Puccinia graminis commonly attacks wheat; other hosts include barley, rye, oats, wild barley and goat grass.
Wheat production in Kenya is highly affected by stem rust and other rusts like yellow
and leaf rusts. Under favorable environmental conditions, infection of the wheat crop with
stem rust disease reduces both the quantity and quality of the grain (KARI annual reports
2000, 2001, 2002, and 2003). At the beginning of the wheat breeding efforts in Kenya in 1927 up to the early 1980s, stem rust was the most serious of the three rusts. This was given a high
priority by the breeding team and due to wide resistance breeding, this problem disappeared for some time and subsequently many resistant commercial wheat varieties were developed
and released to farmers. Since 1992, severe epidemics have continued to occur on commercial
bread wheat and introductions (Danial et al, 1994; KARI annual reports 1999, 2002, and
2003). In 1996 the disease was recorded on some commercial wheat varieties in Mau-Narok
and Molo. By 2000, all the varieties had succumbed to the disease, and at present, they are
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




56
susceptible. Stem rust outbreak is now widespread in all the wheat growing areas in the country in low, medium and high altitude areas (KARl annual reports 2000-2003). The
resurgence of this disease even in the high altitude areas (2700-2900 masl) has necessitated
the need to revive work on this pathogen to incorporate genetic resistance to the prevalent
race(s) within the commercial varieties to work on this pathogen. Therefore, the objective of
this study was to determine the distribution of stem rust on wheat varieties around the
country and also to study the reaction of wheat varieties/ lines to rust inocula collected from
different sources.
Materials and Methods
1. Survey and collection of stem rust spore from the fields
The survey was carried out on farmers' commercial fields and also experimental plots in some
parts of Uasin-Gishu, Nakuru, Meru, Nyeri, Nyandarua and Narok districts between years
2002-2004. Farms were picked along routes at random, stopping at every 3- 5 kilometers or whenever the nearest farm was, especially in areas with sparsely distributed farms. In every
farm, plants were examined randomly by walking across the field. When sampling, a two meter distance from the edge was left to avoid border effect. Disease severities were taken
using modified Cobb's scale (Peterson, et al., 1948) and coefficient of infection calculated by
multiplying the disease severity and host response. Growth stage was entered in the field data
form. The stem rust (sr) spores (uredospores) were then collected using a suction pump and some were scrapped from the leaves and stems using sterile scalpel blades. In the laboratory the spores were dried using Silica gel in the dessicator, vacuumed, sealed in vials and stored
for further work.
2. Screening of Stem Rust Parental Collection (SRPC)
Fifty-three SRPC were planted by hand in the cage at Kari-Njoro and evaluated for stem rust infection. Variety Morocco, the universal suscept was used as the check. Each collection was planted in a hill at 2-3 seeds/hill. Diammonium phosphate (DAP) fertilizer was applied at the
recommended rate of 130kg/ha.The plants were sprayed with Metasystox insecticide at the rate of 0.5L/ha to control the Russian wheat aphid (RWA) and other aphid vectors for the
barley yellow dwarf disease. No fungicide was applied and the plants were monitored for the appearance of stem rust. Disease severity scores were taken twice at the heading stage using Cobb's modified scale (Peterson et al., 1948). The scores were then transformed using square
root (x+l) and the means transformed back to original scores. These were statistically
analysed using analysis of variance and least significance difference used to separate the
means.
3 .Screening of wheat commercial varieties against stem rust in the field
Sixteen wheat commercial varieties were planted in Njoro (2185 m asl) and Eldoret
Chepkoilel campus (2150 m asl ) while 14 varieties were planted in Mau-Narok-Purko ranch
(2900 m asl ) and evaluated for resistance against stem rust. In Njoro and Mau-Narok the
varieties were planted in a randomized complete block design with three replications. Plots
were 8 rows by 2 meters per variety. In Eldoret, two replicates of the same were planted.
Furrows were made using the seed drill and both seed and fertilizer placed by hand.
Immediately, after planting the experiment was sprayed with Stomp 500E a pre-emergent
herbicide at the rate of 3.OL/ha to control grass weeds. At the tillering stage, the experiment
was sprayed with Buctril MC at the rate of 1.25L/ha to control broad leaved weeds. The
Russian wheat aphid and other insect pests were controlled as above. Disease severity
readings were taken at the heading stage and data were analysed as above.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




57
4. Screening of breeders F3 lines and stem rust differential lines for resistance against
stem rust
Seventy-eight F3 lines and 50 and 40 stem rust differential lines from Cimmyt-Mexico and Australia respectively were planted in 14,8 and 11 sets respectively. Each set was planted in
4 inch plastic pots filled with soil and placed in the green-house. The pots were placed on
trays with gravel filled with water. In every pot there were four lines/varieties. The
uredospores collected from: Njbw2, ISR5RA(SR5), ISR9DRA, F1 material, Chozi, Ngamia
Duma, Heroe/Rwa8, Heroe/Rwa9, K7892/Rwa9, differentials Yalta, Anautka, Timboroa
(Matharu), Eldoret and National Dryland Wheat Preliminary Trial (NDWPT) The rust spores were tested for viability and suspended in distilled water. A drop of Tween 20 was added to
each collection and sprayed on 7 day old seedlings. The seedlings had been treated at
emergence with Maleic hydrazide 0.01 g in 30 mls of water to enhance spore production. The plants were incubated for 24 hrs in a polythene chamber kept moist by frequent spraying with
water under high relative humidity of 80-100%, and temperature of 18-220C. They were
removed and placed on the benches in the green house for 14 days and infection types
recorded using 0-4 scale proposed by Stakman et al. ( McIntosh et al, 1995) for stem rust: 0Immune, ;(fleck)-Very resistant,1-Resistant, 2-Resistant to moderately resistant,3-Moderately
resistant/moderately susceptible,4- Susceptible. Data on infection types was compiled and
analyzed using SAS system and means separated using least significance difference.
Results
1. Survey and collection of stem rust spore from the fields
The total number of commercial farms sampled between 2002 and 2004 were 61, from which 46 (75.4% ) of the farms had stem rust infection( Table 1). In some farms the wheat crop had
dried but the stem rust infection spores (basidiospores) were obvious. The farmers were in most cases not able to identify the varieties because they bought planting seed from fellow
farmers. Those who kept their own seed were growing old varieties.
2. Screening of Stem Rust Parental Collection (SRPC)
Disease evaluation was based on 29 lines out of the 53 SRPC and Morocco the susceptible
check. Twenty-three of these (79.3%) and Morocco were susceptible while six ( 20.7%) were
resistant. There were significance (P< 0.05) differences among the line in disease infection.
Table 2 shows that SRPC1(b), SRPC4(b), SRPC 25, SRPC 38(b) and SRPC 204 were highly
susceptible while SRPC10, SRPC 27(a), SRPC (b), SRPC 29, SRPC 166 and 245 were
resistant.
3. Screening of wheat commercial varieties in the field
There were significance (P< 0.05) differences among the varieties in stem rust infection
(Table3). Thel6 commercial wheat varieties were all infected with stem rust and the level of infection varied from variety to variety in all the three sites. In Njoro ( 2185 m asl), the most
infected varieties were: Duma and Ngamia (56.7%, K.Tembo (46.7%) Yombi and Chozi
(43.3%) Heroe and K. Kongoni (40%), Kwale (30%) and Chiriku, Mbega, Mbuni. K. Fahari, K.Paka. and Njrbwl were least infected. In Eldoret( 2150 m asl ), the most affected varieties were: Ngamia and Duma (70%), Chozi, Kwale, Heroe and K. Kongoni(50%), Yombi (45%), while Chiriku, K.Nyangumi, K. Fahari, K.Paka K. Tembo and Njrbwl were less infected. In
Mau-Narok (2900 m asl), the most infected were: Yombi (50%), K. Nyangumi (30%) and
Heroe (40%) wheras, Chiriku, Kwale, K. Nyangumi, Mbuni, K. Fahari and K. Kongoni were
less infected. Varieties Ngamia, Duma, Yombi, Heroe and Chozi were most susceptible
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




58
across the three sites with a mean disease score of 63.4%, 48.3%,46.1%, 43.3% and 40.0%
respectively.
Table 1: Stem rust scores and coefficient of infection on commecial wheat varieties year 20022004
District Place/Area Sr score/ Coefficient District Place/Area Sr score/ Coefficient
Year reactions of reactions of
infection infection
2002 UasinGishu Ainabkoi 20 MR 8 Nyeri Bagati 40S 40
Ainabkoi 40S 24 Bagati 60MS 36
Plateau 30MS 18 Naromoru 50MR 20
Plateau 60MS 36 Mweiga 5MS 3
Plateau 5MR 2 Mweiga 30MS 18
Plateau 70S 70 Kieni 40MS 24
Lembus 10S 10 Nairutia 20S 20
Lembus 5MR 2 Narok Olokurto 70S 70
Timboroa 5MR 2 Olokurto 20S 20
Timboroa O10S 10 Olokurto 50S 50
Timboroa 40S 24 Olokurto 30MS 30
Timboroa 20MR 8 Lopito 80S 80
Timboroa 5S 5 Elmasharani 80MS 32
2002 Nyeri Mbogo 5MR 2 Elmasharani 20S 20
Mbogo 20S 20 Ilipson 90S 90
Mbogo 10MR 4 Ntulele 60S 60
2003 Nakuru Lare 20S 20 Ntulele 5MR 1
2004 Nakuru M/Narok 80S 80 Navasha 40S 40
M/Narok 70S 70 Navasha 30MS 18
M/Narok 40S 40 Navasha 20S 20
M/Narok 15S 15
2004 Meru W/Embori 20S 20
Magara 30S 30
Kararu 10S 10
Mwireri 30S 30
M/Narok=Mau-Narok, W/Embori=Wangu Embori
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




59
Table2 :Mean Stem Rust Disease Infection score and relative ranking (Lsd) on the Stem Rust
Parental Collection ( SRPC) in the Cage Kai-Nloro, 2002
SRPC Pedigree Mean SRPC NO. Pedigree Mean
disease Disease
score score
SRPC 1(b) Africaxmayo 3529-1y-4m-3c-1t 45ab SRPC 36 E.Na48O(Pilotxmd, C1123 16)xGza 0.5 hij
13 92 ,1289-2 196
SRPC1( c) Africa/mayo 15 def SRPC 38(a) 17.5
cde
SRPC 2 Africa/mayo 3529-ly-4m-3y-Im 0.5 ii SRPC 38(b) 45 ab
SRPC 2(a) 30 bcd SRPC 41 (FederationxHope) xBolim, P1 124830 5 fghij
SRPC 4(b) Bza sibx(CJ 12633,W1S245) 35 bc SRPC 71 Gbx(fn-K58/N,JJ-50-17), 11-53-649 5 ghij
SRPC 8(b) Chiinca-A-ElongatumxRd,K.Se1.A 10 efgh SRPC 140 IVbxSR, LM-72-14-57 10 efghi
SRPC 10 Chiinca-A-ElongatumxRd)x(cl Oj SRPC 148 My54XL 1266-61,1448-4603 10 efgh
l2633xldaed
SRPC 12 C1 8154xfr2 JJJ-1009-2t-3b-lt-2b- 10 efgh SRPC 153 Mt-KxN-M,(fr-fn /Y2), 15224-5b-lt- loefgh
it lb-4t
SRPC 15 C1. 12632xCeres R6421198.A.2.1 3 hij SRPC 166 ND 463 j
SRPC 25 30 bcd SRPC 204 Sandos No63x C1l2633-daed 2) 40Oab
Gb 56xVeranopolis, 5134.3.32A
SRPC 27(a) (CI 1263 3,Wis245)2 Oj SRPC 238( 57.5a
-s5 lx(fr-fn /y)111-1 13-6-6b-3t- a)
2b.K. sel .1I
SRPC 27(b) (C012633, WIS245)/(for-fn/Y)2 jOj SRPC 245 -0
SRPC 28(b) 3 hij SRPC 250 5 Ihii
SRPC 29 Oj SRPC 263 2.5 hij
SRPC 32 Desc-C17800 /Bza3,14951-9b-lt 5 ghij Morocco 12.5 efg
Lsd P<0.05 1.7 Lsd P<0.05 1.7
Cv ______________ 26.4 CV 26.4
Means in the same column followed by the same letter do not differ significantly using Lsd
P<0.05
-Pedigree to be traced
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




60
Table 3: Mean % stem rust disease infection and relative ranking (Lsd) on commercial wheat
varieties in Njoro, Eldoret and Mau-Narok 2003
Variety Site/ Mean disease score % mean score
Njoro Reaction Mau-Narok Reaction Eldoret Reaction across the sites
type type type
Yombi 43.3 b S 50.0 a S 45.0 b MR 46.1
Duma 56.7 a S 18.3 abcd S 70.0a S 48.3
Chiriku 5.0 g MR 6.6 cd MR 5.0 f MR 5.5
Chozi 43.3 b S 26.6 bcd S 50.0b S 40.0
Ngamia 56.7 a S 70.0a MS 63.4
Njrbw2 20.0 def MR 11.6 bcd MR 25.0cd MS 18.9
Mbega 13.3 efg MS 23.2 abcd S 30.0c MS 22.2
Kwale 30.0 cd S 1.6 d MS 50.0b S 27.2
K.Nyangumi 21.7 de MS 30.3 abc MS 5.0 f MS 19.0
Mbuni 5.0 g MR 10.0 cd MS 20.0de S 11.7
K.Fahari 10.0 efg MR 5.0 cd MR 0.0f R 7.5
K.Paka 8.3 fg MR 5.0f MR 4.4
Heroe 40.0 bc S 40.0 ab S 50.0b S 43.3
K.Kongoni 40.0 bc MR 5.3 cd MR 50.0b MS 31.8
K.Tembo 46.7 ab S 17.0 bcd MR 5.0f MS 22.9
Njrbwl 11.7 efg MR 16.6 abcd MR 15.0f MR 14.4
P<0.05 12.3 3.092 9.482
Cv 26.16% 48.3% 14.38%
Means in the same column followed by the same letter do not differ significantly using Lsd
P<0.05
4. Screening of breeders F3 lines and stem rust differential lines for resistance against
stem rust
Significant differences were observed in sources of stem rust inocula on sr differential lines acquired from Australia (table 4a). Stem rust spores collected from sr differential Anautika,
ISR5RA(SR) and K 7892/Rwa9 had the highest mean infection (reaction) type while
Heroe/Rwa9, Timboroa (Matharu), Eldoret, and NDLWPT had the lowest mean infection (reaction) type. This was not significant on breeders F3 material and also on sr differential
lines acquired from Mexico. Test lines, the lines x source of inocula did not differ
significantly for all the breeders F3 and sr differential lines. The analysis of variance (Table
4b) showed significant differences among the sources of sr inocula for the differentials
acquired from Australia (F=0.001* *) were significant.
Table 4a :Mean disease score of stem rust inocula from different sources on sr differential
lines from Australia 2003
Source of inocula Mean disease score
Variety Chozi 0.82 def
ISR5RA(SR) 2.14 ab
Timboroa(Matharu) 0.31 f
ISRDRA 1.27 cde
Heroe/Rwa8 1.42 bcd
Heroe/Rwa9 0.30 f
K7892/Rwa9 2.01 abc
Yalta 0.52 ef
Anautika 2.69 a
Eldoret 0.36 f
NDLWPT 0.36 f
Lsd (P<0.05) 0.825
Means in the same column followed by the same letter do not differ significantly
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




61
using Lsd P<0.05
Table 4b: Analysis of variance for sources of stem rust inocula
Source DF SS F
Set 10 359.514 0.001**
Variety 39 183.086 0.4709
Set x Variety 288 1475.120 0.1133
Error 219 960.833
Discussion
Levels of stem rust infection varied with environment as well as varieties/lines. This could
have been due to the temperature and rain distribution which varies from place to place with some areas experiencing drought while others were wet. In Kenya, wheat is grown in many
agro- ecological zones that have different planting dates ( Ralph and Schmit, 1983). This
staggered planting provides a green crop most part of the year allowing stem rust spores to
move by wind from one area to another and that way spreads to different areas. It is therefore,
a problem to reduce the disease infection in susceptible varieties. Almost all the wheat fields
had stem rust infection and the problem is advancing and becoming more serious in high
altitude areas where maximum and minimum temperatures range from 22C- 26' C and 6C14'C respectively. The varied infection (reaction) types of stem rust inocula from different sources on differential lines from Australia probably showed that there could be more new
races attacking the wheat varieties. Although, Ngamia was not among the varieties screened
in Mau-Narok (2900m ASL), the records from a different study ( data not shown) showed
that the infection of stem rust was very high (Kari annual report, 2003).
Conclusion
The survey of stem rust revealed that the disease is on the increase in Kenya. Apparently all the wheat growing areas are prone to this disease. There is therefore, an urgent need to have more work done to determine the isolates so that the breeding programme can be effectively
sustained. Commercial varieties like K. Fahari, K.Paka, Chiriku and some of the old varieties
which have remained resistant/ moderately resistant to stem rust can be utilized in the
breeding programme. It is evident that stem rust is widely distributed in the wheat growing
areas and joint effort is required to utilize the current biotechnology tools to identify and
breed for resistance to the apparently new race(s).
Acknowledgement-The authors wish to acknowledge funds from CIMMYT for the
implementation of this study; Dr. Ravi Singh (CIMMYT, Mexico) for technical discussions and germplasm; Drs.C.R. Wellings and Robert Park (Australia) for the differentials; Mr. D.
Onyango for maintaining the stem rust spores; and colleagues in pathology section for
cooperation and technical support.
References
Agrios, G.N. 1988. Plant Pathology.3 rd edition. Academic Press. ISBN 0-12-044563-8 Cereal Disease Laboratory .(2004). Wheat Stem Rust (Puccinia graminis f.sp. tritici ) at
http://www.cdl.umn.edu
Danial, D.L., 1994. Aspects of Durable Resistance in Wheat to yellow rust. PhD Thesis. Wagenigen
Agricultural University, The Nertherlands, 144pp.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




62
Kenya Agricultural Research Institute. 1999-2003.Annual reports. Kenya Agricultural Research
Institute.2002-2003.Annual reports.
McIntosh,R. A.,C.R. Wellings and R.F.Park. 1995. Wheat Rusts: An Atlas of resistance Genes.
CSIRO, Australia.
Minzenmayer Richard (2000). Pest Management News at http://www.tpma.org
Peterson,R.F., A.B. Campell and A.E. Hannah. 1948. A diagrammatic Scale for estimating Rust
Severity on leaves and stems of cereals. Can.J. Res; 26:496-500.
Ralph Jaetzold and Helmut Schmidt. 1993. Farm Management Handbook Agriculture Vol.II, Part B.
Ministry of of Kenya in cooperation with German Agricultural Team (GAT) of Germany
Agency for Technical for Technical Cooperation ( GTZ).
Stubbs, R.W., J.M. Prescott, E.E. Saari and H.J.Dubin. Cereal Disease Methodology Manual. Centro
Internacional de Mejoramiento de Maiz y Trigo (CIMMYT), Mexico. 1986.
Wiese, M.V.1991. Compedium of Wheat Diseases 2nd Edition. APS Press.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




63
Genotype-by-Nutrient Interaction in Wheat Grown in a Marginal Environment in Kenya
J. Kamwaga, H. Okwaro, P. Njau, P. Kimurto, P. Ndungu. and E. Kimani
Abstract-Six wheat (Triticum aestivum L.) cultivars (Chozi, Ngamia, Duma, Njoro BW1, K.Heroe, and Chiriku) were grown in a marginal environment at two levels of fertilizer; 0:0:0, 30:60:0 kg ha-' (N:P205:K).and at a seed rate of 125 kg. ha- to assess their response to fertilizer application and nutrient use efficiencies. Grain yield, spike numbers, harvest index (HI), nitrogen (N) uptake, N concentration and nitrogen use efficiency (NUE) were recorded. There were no significant differences in grain yield or spikes m-2 among varieties when averaged across fertilizer levels. Differences in HI between varieties were significant, with cv. Duma, a variety recommended for marginal environments, having a higher HI than cv. K.Heroe. Fertilizer did not affect the HI. There were no significant differences among varieties in N acquisition, indicating similar nutrient uptake abilities, or in NUE, although this was reduced by fertilizer application, probably because the increased N uptake did not translate into a corresponding grain yield increase due to the poor growing conditions. Grain N content ranged between 2.2% in the fertilized plots and 1.8% in the unfertilized plots
(meaned across cultivars) and 2.2% in Njoro BW1 and 1.9% in Ngamia.
Introduction
Genotypic differences in nutrient utilization have been documented in wheat. These
differences have been attributed to both differences in the acquisition and the efficiency of use
of nutrients between cultivars (Inthapanya et al., 2000). One reason for differences in
efficiency has been attributed to the development of different root systems between wheat
cultivars. Nitrogen (N) and phosphorous (P) deficiencies limit the yield of wheat in most
wheat growing areas of Kenya. A number of approaches have been used to enhance
productivity of wheat in such environments: principally the development of site-specific fertilizer recommendations for various areas. Inorganic fertilizers are expensive and not
affordable to most wheat growers, while. Organic sources of nutrients are not easily available
in the quantities and qualities that would give adequate yields in farmers' fields. The use of
nitrogen fixing crops also has its limitations. Another approach would be to identify cultivars
that do well in low soil fertility conditions but that also respond well to applied fertilizer.
These would be the preferred cultivars because they would be useful to both farmers with fertilizer inputs and resource-poor farmers. This type of cultivar usually has a high harvest
index (El Bassam 1998)
The efficiency of utilization of both N and P determines grain and dry matter yields.
Variety by fertilizer level interactions will be very important: researchers need to find
varieties that have low yield interactions with fertilizer level, and that, therefore, respond in in
a similar manner in low and high fertility environments. This is especially important as
varieties widely adapted to diverse soil fertility status will be preferable.
Cultivars grown in the marginal environments of Kenya have not been evaluated for their
nutrient use efficiency, yet this is an important factor determining yield of wheat in such
areas. Identification of cultivars that do well under low nutrient and moisture conditions will
improve wheat production in such areas and the the breeding program will be enhanced by
incorporation of characters associated with nutrient use efficiency. Improved cultivar
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




64
response to nutrients will help reduce inputs and hence protect the environment (El Bassam
1998).
The objectives of this project were to:
I. Determine genotypic variation in nitrogen uptake between cultivars recommended for the
arid and semi arid areas (ASALs) of Kenya.
ii. Determine the efficiency of utilization of N in recommended cultivars.
Materials and Methods
The trial was established in Kajiado, Kenya, on a moderately fertile vertisol with P ststus
above the critical levels (Table 1). The treatments were arranged in a split plot design with
three replications, with fertilizer levels as the main treatments and varieties as sub-treatments.
The two fertilizer levels used were 0:0:0, and 30:60:0 kg ha' (N:P205:K), and the cultivars
assessed for nutrient use efficiency were: Ngamia, Chozi, Duma, Njoro BW1, Chiriku and K.
Heroe. Sub-plot size was 6m x 3m.
Seed was dribbled by hand into furrows spaced at 20 cm apart at a rate of 125 kg. ha-',
and fertilizer dribbled into the same furrows. At four weeks after seeding, Buctril MC. (a.j.
Bromoxynil +MCPA) was applied at 1.2 1 ha' to control weeds. At harvest, a 1 m2 quadrat
was harvested by cutting the plants at ground level. Parameters evaluated were total biomass (Total Dry Matter), grain yield and nitrogen (N) concentration in grain and straw, determined
using the kjeldahl method. N uptake, nitrogen use efficiency (NUE) and harvest index (HI)
were calculated from the evaluated parameters.
Table 1 Soil chemical and physical characteristics of the trial site
Depth (cm) pH % N P (ppm) % O.M.
0-15 5.8 0.15 33 2.4
15-30 6.5 0.09 30 2.0
30-60 6.9 0.07 31 1.7
60-90 6.9 0.08 31 0.8
Results and Discussion
There were no significant effects of genotype on grain yield, harvest index, nitrogen uptake, use efficiency or grain N concentration, nor were there fertizer by genotype interactions with
respect to these variables (Table 2), indicating that the varieties responded similarly to fertilizer application. However there were differences between varieties with respect to thousand kernel weight (TKW): Chozi had significantly higher TKW than all the other
varieties (Table 3). Fertilizer application significantly increased grain yield, total N uptake
and grain N concentration, but reduced nitrogen use efficiency, and had no significant effect
on HI nor TKW. (Table 2 and Table 4).
Table 2. Degrees of freedom F ratio and level of significance for grain yield, Harvest Index (HI),
Total N uptake, Grain N concentration and Nitrog n Use Efficiency.
Source of variation DF Grain HI Total N Grain N NUE TKW
yield Uptake (%)
Genotype 5 0.94 1.46 0.69 1.49 0.95 6.49**
Fertilizer 1 12.75** 0.68 65.6** 19.65** 18.51** 2.94
Genotype*Fertilizer 5 0.85 0.51 0.97 0.42 0.46 1.12
** Significant at (p 12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




65
Table 3. Variety yield parameters measured across fertilizer levels at Kajiado
Variety Spikes/m2 Grain yield kg Harvest TKW
ha-1 index (HI)
Duma 182 1853 0.40 A 38.9 B
Ngamia 179 1751 0.38 AB 35.6 BC
Njoro bwl 173 1491 0.37 AB 37.9 BC
Chiriku 170 1642 0.37 AB 34.6 C
Heroe 163 1417 0.31 B 39.0 B
Chozi 150 1522 0.37 AB 44.0 A
LSD 5% NS NS 0.06 3.8
C.V % 24 26.2 15.7 8.3
Table 4. Fertilizer effects on yield parameters measured across varieties at Kajiado
Fertilizer level Grain yield Harvest TKW NUE Total N Uptake Grain N
kg ha-1 index g grain g-1' gm2 Concentration
N (%)
O 1361 B 0.37 A 39.2 41.8 a 330 a 0.38 a
NXP 1864A 0.36 A 37.4 31.9b 600b 0.36b
LSD 5% 29.4 NS NS 0.05 69.6 0.04
There were no significant differences between genotypes in either nitrogen uptake
(acquisition) or utilization of the absorbed N: there were no significant differences in NUE
(grain production per unit of the total N ) (Table 5) However, cv. Njoro BW1 had a
significantly higher N concentration than c.v Ngamia. It has been shown that grain yield is a
function of NUE, total N uptake and HI as in the equation of Inthapanya et al. (2000).
Y=NUE X Total N uptake x HI
Where NUE= Nitrogen Use Efficiency
Hi=Harvest Index.
The results from this study show that of the three variables, the tested cultivars only
differed in HI, suggesting that the only potential for increasing yields in the wheat cultivars
would be in improving the HI of the cultivars.
Table 5. Total N uptake, grain nitrogen concentration and nitrogen use efficiency (NUE) meaned
over two fertilizer levels in wheat at Kajiado
Variety Total Nitrogen Uptake g m2 Grain N Concentration (%) Nitrogen Use
Efficiency (NUE) g
grain g-1 N
Duma 5.22 2.08 Ab 38
Njoro BW1 4.72 2.17 A 33
Chiriku 4.67 1.96 Ab 37
Ngamia 4.60 1.87 B 40
K. Heroe 4.50 1.93 Ab 34
Chozi 4.18 2.00 Ab 39
LSD NS 0.26 NS
C.V 2.15 10.8 18.8
Means followed by the same letter are not significantly different at p>0.05
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




66
References
Inthapanya P., Sipaseuth, Sihavong P., Sihathep V., Chanphengsay M., Fukai S., Basnayake J. 2000.
Genotype differences in nutrient uptake and utilization for grain yield production of rainfed lowland rice under fertilized and non-fertilized conditions. Field Crops Research 65 (2000)
57-68.
El Bassam N. 1998. A concept of selection of low input wheat varieties. Euphytica 100 (1/3) 95100.Improvement of Wheat
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




67
Improving Wheat Productivity for the
Drought Prone Areas of Kenya Using the Doubled Haploid Technique
Njau P. N1, Kimurto, P.K, Kinyua M. G1, Okwaro H. K1 and Ogolla J.B.02
1. National Plant Breeding Research Center, Njoro, P.O. Njoro, Kenya'] 2. Dept. ofAgronomy, Egerton University, P.O. Box 536, Njoro, Kenya
Abstract- Various breeding methods have been applied in an effort to develop superior wheat varieties for the marginal areas of Kenya. These include gernplasm introductions, mutation breeding and doubled haploid (DH) production. The application of DH in breeding for drought tolerance has proved to be very effective and efficient. This study was aimed at evaluating and validating DH lines that were developed in 2000 using the chromosome elimination technique. The resulting lines were compared with other lines introduced from CIMMN4YT and two mutants developed in Njoro. The lines were tested in multi-locational trials in 2002 and 2003. The sites included were Lanet, Mogotio, Naivasha, Mweiga, Kajiado and Katumam. The results show that the DHs were as good as the conventionally developed lines and yielded more than the check varieties Chozi and Duma. One of the DHs had average yielded 1.7tonsHa-', which was not significantly different from Chozi. The DH technique has proved to be applicable in saving time of breeding
without compromising the output.
Introduction
Wheat is the second most important cereal crop in Kenya (FAO, 2002) but Kenya's current
national wheat production (approximately 300,000 tons per annum) meets only about 50% of
the national demand. Moreover, increasing population and changing eating habits is
expected to have a bigger impact on wheat demand; estimated to reach 850,000 tons p.a. in
the year 2020 (FAO 2003). Land area currently suitable for wheat production in the high
potential areas is less than 2% and hence expansion of wheat production in these regions is limited. In recent years wheat production has expanded in the marginal rainfall areas (such
as lower Narok, Naivasha, Laikipia and Machakos) which were hitherto considered unsuitable for wheat growing. Introduction of wheat in the non-traditional areas began in 1992. Over this
period 4 varieties (Duma, Ngamia, Chozi and Njoro BW1) have been released (Kinyua et al,
2002). However, varieties with greater yields are still needed to meet the farmers' needs.
Future production increases must come largely from vertical expansion (i.e., greater
production per unit area), which will require more intensive research to further improve yield
potential and cultural technology. Wheat improvement in Kenya has been directed into developing broadly adapted, high yielding germplasm with high yield stability, durable
disease resistance and acceptable end use quality (CIMMYT 2002). Resistance to biotic stress
and tolerance to abiotic stress can be critically important in maintaining high yield and
adaptation over time and locations. Over the last decade wheat breeding has been enhanced
by the application of various biotechnological approaches; these have been used to accelerate
the breeding process and also complemented the conventional breeding methods. The DH
technique and mutation breeding have been used to improve stress tolerance in wheat. These
include breeding for drought tolerance, tolerance to soil acidity, lodging tolerance and
resistance to Russian wheat aphid.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




68
Drought is a multidimensional stress affecting plants at various levels of their organisation
(Blum, 1996). Improvement of yield under stress must combine a reasonably high yield
potential with specific plant factors which would buffer yield against severe reduction due to drought (Blum, 1983). Measurement of plant response to drought at the whole-plant level is
complex because it reflects the integration of stress effects and responses at all underlying
levels of organisation, over space and time (Blum, 1996). Conventional breeding (which involves multi-location testing and hence gauges spatial adaptation of genotypes) is used
extensively by the Kenya bread wheat program to identify temporally stable, drought tolerant
germplasm (Maarten, 1994 this does not appear in the reference list). However, practical
breeding programs for self-pollinated crops (such as wheat) must include a process of genetic fixation (for uniformity of agronomic traits) after genetic recombination to increase variability
(Inagaki, 1996). Repeated selection of heterozygous material can increase uniformity but
many generation cycles are required to reach homozygosity in loci associated with agronomic
traits. Haploid production followed by chromosome doubling offers a quick method for
developing homozygous breeding lines (Bentolila et al, 1992, Murignaux et al. 1993,
Beanziger et al., 1989a, Baenziger, 1996). The DH lines derived from hybrid progenies can be used as recombinant inbred lines with favourable gene combinations (Inagaki, 1996). This
technique could thus complement the conventional breeding programs to accelerate the
release of new varieties.
Since the double haploids are completely homozygous, all stocks are identical and no
purification process is required. In contrast, in the conventional system stocks are usually
derived from a single plant of an advanced generation (Baenziger, 1996) hence several
generations are required to build up sufficient quantities of seeds for release. Compared to
selection during the early generations a DH system increases the efficiency of selection for
both qualitative and quantitative characters. This study aimed to evaluate and validate DH lines and compare them with conventionally bred materials. It was hypothesised that the
DH technique is superior to conventional breeding when dealing with complex factors like
drought.
Materials and Methods
Introduction
Two experiments were carried out. Experiment I involved development of the double
haploids (including embryo rescue, regeneration of haploid plantlets and chromosome doubling of haploids) and preliminary evaluation of the DHs. Experiment II was a field
evaluation of the performance of the DHs as compared with the conventionally developed
lines
Development of doubled haploids
Six F1 hybrids were produced by crossing three drought tolerant lines (Duma, K. Mbweha and
Ngamia) as females with two high yielding commercial varieties (Kwale and Kenya Chiriku) as males. The F1 plants were emasculated at anthesis. The glumes and awns were not clipped, as is the case with normal-crossing emasculation. The emasculated spikes were then covered
with polyethylene bags to maintain high humidity. One day before (predicted) anthesis, emasculated spikes were pollinated with freshly collected maize pollen. The pollen was
collected by picking mature anthers and placing them on petri-dishes. Once the anthers burst to release the pollen, a soft brush was used to brush the emasculated spikelets with the pollen.
Extreme care was taken not to damage the stigma. After pollination, the uppermost intemodes
of wheat culms (with pollinated spikes) were injected with a 100mg 1-1 2, 4-D solution daily
(for two consecutive days) to increase the rate of fertilisation and embryo formation.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




69
After about 14 days of spike growth, wheat caryopses were removed from spikes and
sterilized (in 2% sodium hypochlorite) for 15 minutes. After rinsing with sterilized water
(under laminar flow bench), embryos were aseptically excised and transferred onto half
strength Murashige and Skoog medium supplemented with 20g L-' sucrose and 8g L-' agarose
in petri dishes (Inagaki 1996). These were then placed in the fridge (at 4 C) for three days after which they were incubated (at 25 C) in the dark until embryos germinated (5-7 days).
The germinated embryos were transferred to a lighted growth room with controlled
temperature of 24 C and 16 hours day length. The light intensity was Ca 5000 Lax. Cultured
plants with fully developed roots and leaves were transplanted to potted soil for further
growth. The temperature was maintained at 21 C -25 C using 24 hours day length. The size
of the stomata was used to determine the ploidy level of the plants. This is because haploid
plants have stomata half the size of diploid plants. At the third tillering stage, a leaf from each
plant was cut and coated with clear nail varnish. The varnish was pealed off (upon drying) and mounted on high power objective of light microscope; the size of the stomata was then
compared with that of known diploid plants.
In addition, at the third tillering stage the plants were watered and then removed from the
pots. The roots were trimmed to about 2 cm below the crown (to increase solution
absorption). The plants were then immersed in colchicine solution (0.2% colchicine, 2%
dimethyl sulfoxide and 15 drops of tween-20) for 3 hours at room temperature. The plants
were then washed thoroughly with running tap water for 3 hours after which they were
planted in pots (using forest soil enriched with copper dust) at a temperature of 200 C-25 o C
under high humidity (90-100%).
Preliminary evaluation of the doubled haploids
Twenty DH lines were selected (based on the amount of seed available). These selections and
their five parents were initially screened in the rain-shelter at the National Plant Breeding
Research Centre (NPBRC) Njoro in 2000 and later evaluated in the field (in Njoro and
Katumani) through observation trials in 2001. They were planted in a randomized complete block design (RCBD) with three replications. Each plot was 1 metre long with two rows 20
cm apart. Five seeds were planted per row. DAP (18-46-0) fertilizer was applied at planting at
the rate of 150 kg ha'. Drip irrigation was used to water the plots (Chapin Watermatics, 1999). Water was applied at the rate of 20 mm every fortnight (up to grain filling stage),
giving a total of 200 mm water during the crop season. The amount and frequency of water
application simulates the amount and nature of rainfall pattern usually received in most
marginal areas during the cropping season (Mugo et al., 1998; Jaetzold and Smith, 1983).
The following parameters were measured (both in the rain shelter and in the field): Ear length
(measured from the base of the spike to the tip of the apical spikelet, excluding the awns);
number of spikes per plant, sterile spikelets per head; number of grains per head counted on 10 spikes selected randomly in each experimental at maturity; and weight of 10-kernels (g).
All the data was subjected to analysis of variance using the general linear model (SAS, 1996).
Field performance of the double haploids
Eight DH lines (DH4, DH5, DH6, DH7, DH9, DH12, DH15, DH16) and 2 mutants (BM1 and
BM3) were selected in a preliminary yield trial in 2001. The selection was on the basis of
drought tolerance. These lines, together with Chozi and Njoro BW1 (checks), were entered in
the National Performance Trial (NPT) in 2002 and consequently planted at three sites (i.e.,
Naivasha, Katumani, and Lanet). The design was a RCBD replicated 3 times. The seeds were
drilled in plots of four rows that were 6 metres long and 20 cm apart.
Seven DHs (DH4, DH5, DH6, DH7, DH9, DH12 and DH16) were selected form the first year
of NPT and entered into the second year of National Performance Trial (in 2003) alongside 7
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




70
conventionally developed lines (R840, R960, R962, R963, R965, R966 and K7872) and the
check varieties (Njoro BWl and Chozi). These were evaluated at Katumani, Naivasha,
Mweiga, Lanet and Ravine. The Design was a RCBD replicated three times. The plots were 4
rows, 20 cm apart and 6 m long.
The following parameters were measured in both the first and second year NPT: grain yield,
Kernel weight, plant height and reaction to rust diseases.
ANOVA was used for data analysis and means were separated using LSD
Results and Discussion
Development of Doubled Haploids
A total of 1800 florets were cross pollinated out of which 890 Fl seeds were harvested and
the embryos excised and planted in-vitro. This shortened the F1 seed production by over four months compared to the in-vivo method where the seed is left to dry in the field. The in-vitro F1 plants grew faster due to the conducive environment and reached heading stage one month
earlier than in the conventional method.
Over 2,880 florets of F, plants were cross pollinated with maize. 413 of these developed
seeds from which 57 embryos were rescued (Table 2). Out of the 57 embryos rescued 46
were haploid. When treated with colchicine 24 of the 40 survived. The time taken for
pollination to colchicine treatment was 8 weeks and the DHs were ready for harvesting in 20
weeks.
Performance of doubled haploids
There was high variability among the DHs in the rainout shelter for the yield components measured (Table 1). These results are comparable to those reported by Baenziger (1996);
whilst there was variation among doubled haploid lines, there was little variation within
individual haploid lines. This increased the efficiency of selecting lines with superior characteristics (Njau et al, 2000) because the response to selection is indicated by the
variability between the treatments. Such variation can be used to select lines with the required
characteristics (Njau 2001).
There were significant differences in yield and other growth parameters between the
DHs and their respective parents (Table 1) and some DHs showed heterosis for drought
tolerance over their better parent or the mean of the two parents (Table 4). For example, one
of the DHs (DH17) developed from the cross between Kenya Mbweha and Kenya Chiriku had
greater number of spikes (6.1) than the parents (2.0 and 5.0, respectively for Kenya Chiriku and Kenya Mbweha). DH17 also had more grains per head (43) than the mean of the parents (27.5). DH1 12 (developed from Duma/Kenya Chiriku cross) had more grains (%) than either of the parents. The expression of heterosis in DH lines (Njau 2002) makes the DH technique
superior to that of conventional breeding. This is because heterosis is lost in conventional
breeding due to repeated selfing of the hybrids. Similar results have been reported elsewhere
(Ba-Bong and Swaminathan, 1995).
Over 26% of the cost of developing a fixed line using conventional selection was saved
in the production of the DHs (data not provided). This saving was attributed to the fewer
number of generations required to produce homozygous lines when using haploids. Similar
findings have been reported in rice (Sunint, 1993).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




71
Table 1. Mean performance for various parameters of 20 DHs as compared to their parents
tested for drought tolerance under the rainshelter at Njoro. Kenya.
Parents DH lines No. of tillers Effective Spikelets/He Sterile Grains/ weight of
at booting Heads ad Spikelets head 10 grains
K. Mbweha 5.00 5.0 15.333 2.000 24.333 0.367
Kwale 4.21 2.16 14.533 3.994 27.936 0.306
(K.mbweha x DH11 5.33 4.00* 16.000** 2.00* 34.00 0.301
Kwale) DH18 5.71 3.67* 17.533** 2.494* 35.936 0.317**
DHI19 3.71" 3.66 15.533 1.494* 32.936 0.27*
K. Chiriku 4.33 2.00 15.333 3.000 30.667 0.312
(K. Chiriku x DH17 7.39** 6.11** 19.301 2.009 43.489*** 0.282
KMbeha) DHI14 2.67* 2.33 17.333 1.333* 36.667** 0.343
Ngamia 3.67 3.00 10.00 1.667* 41.667 0.349
(Ngamia x DH12 4.39 4.11 19.301** 1.009* 41.489** 0.336
Kwale) DH13 5.21 3.66 16.033** 1.995* 24.94' 0.21*
DHIl0 4.33 2.00* 13.333 1.006* 27.67* 0.346
DHIl13 4.30 3.95 15.349** 1.996* 41.755** 0.304
DHIl15 3.33* 3.33 15.333** 0.67*** 30.00* 0.367
DHIl16 8.39** 6.1*** 19.301** 0.09*** 28.489 0.304
(Ngamia DHIl17 0.803* 0.45** 13.765 4.980** 4.33** 0.12**
xK.Chiriku) DH120 5.00 4.33 14.000** 1.667** 28.67* 0.19*
Duma 6.33 3.00 15.000 2.000 28.333 0.467
DH14 2.80* 2.45 13.349 2.996 19.255 0.305*
(Duma x DH15 3.39 3.11 19.301 2.009** 35.489 0.310*
Kwale) DH16 7.39 2.89 14.617 1.510** 33.308 0.367*
DH19 7.00 4.00 11.333 2.333** 29.333 0.266*
DHI18 4.30 2.45 12.349 1.996** 32.255 0.326*
(Duma x DHlll 3.71* 2.16 14.533 0.995 40.936* 0.307*
K.Chiriku) DHIl12 5.67 3.67 17.333 2.000 44.00** 0.421**
* Significantly better at 5% level than the poorer parent.
** Significantly better than the mean of the parent at P<0.5
*** Significantly better than the better parent at 5 % level
Significantly poorer than the better parent at 5 % level.
* Significantly poorer than the poor parents at 5 % level.
Preliminary performance of the double haploids
Significant differences in yield Among genotypes were noted in Katumani and Naivasha
while no differences were noted in Lanet. Two lines DH6 and DH4 performed quite well at all the sites. DH6 was better than all the other entries in Naivasha and it was better than the
checks at all the sites. DH7 and DH9 also performed well at all the sites
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




72
Table 2. Grain yield of the test lines (DHs, mutants and check varieties) at 3 different sites (t ha-)
LINE YIELD (t ha-')
KATUMANI NAIVASHA LANET
DH4 .9AB 1.0BC 1.4AB
DH5 .3E .3D 1.I1B
DH6 .9AB 1.5A 1.5AB
DH7 1.0A 1.3AB 1.I1B
DH9 .7BCD 1.4AB 1.6A
DH12 .7ABCD .9C 1.2AB
DH15 .7CD .7C 1.3AB
DH16 .8ABC .8C 1.2AB
BM1 .6CD .8C O1.0B
BM3 .7BCD .7C 1.3AB
NJORO BW1 .8ABC .9C 1.3AB
CHOZI .5DE. .9C 1.I1B
LSD .27 .37 0.56
SE 0.22 0.11
P(F-ratio) <0.05 <0.05 <0.05
Values followed by the same letter are not significantly different at p = 0.05
Genotype did not affect 1000-kernel weight in Katumani and Naivasha but did affect 1000kernel weight in Lanet (Table 3). The double haploids had greater 1000-kernel weight (16%
on average) than the check varieties and the mutant BM3 in Lanet (Table 3). Also, the mutant
BM1 had 9% 13% greater 1000-kernel weight than the double haploids in Lanet (Table 3).
Table 3. 1000 kernel weight of the entries in the three sites during the year 2002
Line 1000-Kernel weight (g)
Katumani Naivasha Lanet
DH4 76.6 73.4 83.1A
DH5 72.0 78.8 81.0AB
DH6 72.4 76.7 78.7AB
DH7 72.4 72.4 78.4AB
DH9 70.7 71.2 78.0AB
DH12 77.4 78.3 78.0AB
DH15 75.9 75.8 77.9AB
DH16 76.6 76.4 77.4AB
BM1 73.9 72.7 75.6AB
BM3 72.9 68.2 69.7CD
NJORO BW1 74.2 76.1 69.4CD
CHOZI 75.3 78.6 66.9D
Lsd 10.85
P (F-ratio) ns ns <0.05
SED 4.1 3.2 4.04
Values followed by the same letter are not significantly different at p = 0.05
Genotype affected the average coefficient of infection (ACI) by stem rust in 2002 (Fig 1).
Chozi (check variety) was more susceptible to stem rust than the double haploids (except
DH15 and DH16) and the mutants (Figure 1). NjoroBWl (check variety) had lower stem rust
infection than the mutants and most double haploids (DH5, DH9, DH12, DH15, DH16)
(Figure 1).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




73
120 Fig 1. Stem rust average coefficiet of infection
100
80
60
40
20
0
o*
LinesO
Performance of the DHs as compared to the conventionally developed lines
There was a significant effect of site on variety in terms of grain yield and hectolitre weight
in 2003 (Table 4). Katumani had lower yields than Lanet (by 80%), Mogotio (by 70%),
Naivasha (by 70%), Mweiga (by 30%) and Kajiado (by 30%). Kajiado and Mweiga also had lower yields compared with Lanet (38%), Mogotio (31%) and Naivasha (31%) (Table 4). The
lower yield in Katumani (compared with other sites) could be due to the low rainfall (100
mm) recorded in Katumani during the growing period of the crop.
Lanet had lower hectolitre weight (by 20%, 19%, 10%, 13% and 7% for Mogotio, Naivasha,
Mweiga, Kajiado and Katumani, respectively) compared with the other sites (Table 4).
Katumani had 12%, 11% and 5% lower hectolitre weight than Mogotio, Naivasha and
Kajiado, respectively. Also, Mweiga had lower hectolitre weight compared with Mogotio
(9%), Naivasha (8%) and Kajiado (2%). Moreover, Kajiado had 6% lower hectolitre weight
compared with Lanet and Mogotio (Table 4).
Table 4. Average yield in tons per hectare and hectolitre weight in the six sites
Site Yield (t ha-') Hectolitre weight (g)
Lanet 1.8a 67.3d
Mogotio 1.7a 80.9a
Naivasha 1.7a 80.2a
Mweiga 1.3b 74.3c
Kajiado 1.3b 76.0b
Katumani 1.0Loc 72. lc
Mean 1.5 75.1
LSD 0.18 1.70
SED 0.21 4.21
P (F-ratio) <0.01 <0.01
CV (%) 29.7 5.6
The yields and hectolitre weight for the lines averaged across the sites are shown in Table 5.
Genotype affected grain yield and hectolitre weight (Table 5). Chozi (check variety) had 55%
greater yield than DH6 and DH16 and 113% greater yield compared with DH12.
NJOROBW1 (check variety) also had greater yield (by 88%) than DH12 (Table 5).
Chozi had greater hectolitre weight compared with R960, R962 and R966 (9%), K7872
(12%), R965 (10%), which are lines developed through conventional breeding (Table 5).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




74
Chozi also had greater hectolitre weight (by 8% 13%; average 11%) compared with a
number of double haploid lines (DH4, DH9, DH5, DH7, DH16, DH12) (Table 5). The check
variety NJOROBW1 had 7% greater hectolitre weight than DH12 (Table 5). These results
show that in addition to being reducing the breeding time, lines developed using the DH
technique, compete well with conventionally developed lines which have taken over 13 years
to develop.
Table 5. Yield and hectolitre weight of the sixteen lines averaged over the sites
Line Yield (t ha-1) Hectolitre weight (g)
R960 1.9a 74.5bcd
DH4 1.7a 73.0cd
R840 1.7a 78. lab
Chozi 1.7a 81.3a
R966 1.6ab 74.3bcd
R963 1.6ab 76.6abcd
K7872 1.5abc 72.3cd
DH9 1.5abc 75.0bcd
NJBW1 1.5abc 77.0abc
DH5 1.5abc 73.5bcd
R965 1.5abc 73.9bcd
DH7 1.4abc 74.7bcd
R962 1.4abc 74.6bcd
DH16 1. lbcd 72.9cd
DH6 1.lcd 78.0ab
DH12 0.8d 72.0d
Mean 1.5 75.1
LSD .29 2.77
SED 0.09 17.7
P (F-ratio) <0.01 0.01
Conclusion
The DH technique proved to be quite useful in breeding for complex characters such as
drought and compares well with the conventional methods. The time saved in DH
development makes the methodsuperior to the other methods. It is important to note that the
multiloctional testing increases the efficiency of measuring yield stability. The two lines
(DH4 and R960) are recommended for release as they both out yielded the check varieties at
almost all sites.
References
Ba Bong B. and M.S. Swaminathan 1995. Magnitude of hybrid vigour retained in double haploid lines
of some heterotic rice hybrids. Theo. Appl. Genet. 90:253-257.
Baenziger P.S. C.J Peterson, M.R Morris and P.J Mattern 1989a. Quantifying gametoclonal variation in
wheat doubled haploids. M. Muluszynski (Ed). Current Options for cereal improvement pp.
1-9. Kluwer Academic Publishers. Boston.
Baenziger S.P. 1996. Reflections on doubled haploids in plant breeding. S.M. Jains, S.K Sopory and
R.E Veilleus (Ed). In vitro Haploid production in Higher plants 1: 35-48. Klumer Academic
Publishers. Netherlands.
Bentolila S, T. Hardy, C. Guitton and G. Freyssient 1992. Comparative genetic analysis of F2 plants
and anther culture derived plants of maize. Genome 35: 575-582.
Bitch C., G. Sabine and J. Lelley 1998. Effect of parental genotypes on haploid
embryo and plantlet formation in wheat x maize crosses. Euphytica 103: 319323.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




75
Blum A. 1996. Crop responses to drought and the interpretation of adaptation. Plant Growth
Regulation 20:135-148.
FAO, 1986. Production yearbook. Vol. 40. FAO ROME.
Inagaki M.N. (1990). Wheat Haploid through Bulbosum technique. Biotechnology in Agriculture and
Forestry 13:448-58.
Inagaki M.N. (1996). Technical advances in wheat haploid production using ultra-wide crosses.
IIRCAS Journal 4:51-62.
Jaetzold R. and H. Schmidt, 1983. Farm management Handbook of Kenya. MOA and GTZ.
Kinyua M.G. 1997. Transfer of genes of resistance to yellow rust (Puccinia stiifarmis L.) from wild
emmer Triticum dicoccoides into Kenyan wheat commercial varieties. PhD Thesis. Nairobi
University.
Kinyua, M.G,., Karanja, L., and Njau, P.N. 2002. Drought tolerant wheat varieties developed through
mutation breeding technique. Workshop on Dryland Farming January 2002. Paper submitted
and accepted.
Kinyua, M.G., J.K. Wanjama, J. Kamwaga, M. Migui 1989. The situation of wheat production in
Kenya KARI Report.
Lefebure D. and P. Devaux 1996. Doubled Haploid of wheat from wheat x maize crosses: genotypic
influence, fertility and inheritance of the IBL-IRS chromosomes. Theor. Appl. Genet. 93:
1267-1273.
Murignaux A., D. Barloy, P, Leray and M. Beckont 1993. Molecular and morphological evaluation of
doubled haploid lines in maize, Homogeneity within DH lines. Theor. Appl. Genet. 86:837842.
Njau P. N. 2001. Development of drought tolerant wheat lines through chromosome elimination
technique. Msc. Thesis. Egerton University.
Njau P.N., M.G. Kinyua and R.S. Pathak.2000. Drought Tolerant Doubled Haploid (DH) of Bread
Wheat for Kenya. Proceedings of the 7h Biennial KARI Scientific Conference Paper
No.40.KARI Headquarters. Nairobi.
Sunint L.R., C.P. Martinez, A. Ramirez and Z. Lentini, 1993. Rice anther culture versus conventional
breeding: A cost/benefit analysis. Trends in CIAT commodities working document NO.128.
Contro International De Agricultura Tropical, Call.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




76
Grain Yield Stability of Bread Wheat
Genotypes in Favorable and Stressed Environments
Desalegn Debelo, Solomon Gelalcha, Balcha Yaie, Bedada Girma, Berhanu Mamo, and Debebe Masresha
EARO, Kulumsa R. C., P. 0. Box 489, Asella, Ethiopia E-mail: desalegnd@freemail. et
Abstract-A multilocational trial consisting of 18 bread wheat genotypes including standard and local checks was conducted in 39 locations between 1999 and 2001 across the diverse wheat agro-ecologies of Ethiopia. Yield stability of the genotypes was assessed in high potential, moisture stress and waterlogged growing environments using the AMMI (Additive Main Effects and Multiplicative Interaction) model. The results showed highly significant genotypic and G x E interaction affects in all three environmental categories. Genotypes FH 8-2, HAR 3354, ETBWC 037 showed low positive interaction while HAR 3116, and FH 6-1-7 showed low negative IPCA axis 1 scores indicating stable performance in the high potential wheat growing areas. In the low moisture areas FH 8-1, ETBWC026, HAR 3224 and HAR 2870 showed low positive IPCA axis 1 scores. Among these genotypes, HAR 2818 gave the highest mean grain yield and is better suited to moisture stressed areas. The analysis of waterlogged areas showed that HAR 3354, FH 4-2-11 and ETBWC037 had high mean grain yield performance and low G x E interaction. From the results we concluded that FH 8-2, HAR 3354, ETBWC 037, HAR 3116 and FH 6-1-7 are widely adaptable; FH 8-1, ETBWC026, HAR 3224 and HAR 2870 perform better in the moisture stressed areas while FH 4-2-11 and
ETBWC037 are tolerant to waterlogged vertisols.
Introduction
In developing crop varieties with high and stable yields, data from multi-environments,
representing spatial and temporal target domains, play an important role in estimating and
predicting yield and yield stability. The pattern of response of genotypes to different
environments is used by the plant breeder to select the best varieties for any particular region.
Yield response of genotypes tested in different environments is almost always subject to
genotype x environment interaction (GxE). To estimate the level of interaction of genotypes
to environments and eliminate as much as possible the unexplainable and extraneous variability contained in the data, several statistical techniques have been developed to
describe G x E and measure the stability of genotypes. These techniques include conventional analysis of variance (Yates and Cochran, 1938; Cochran and Cox, 1957); the joint regression method proposed by Yates and Cochran (1938),modified by Finlay and Wilkinson (1963) and further improved by Eberhart and Russell (1966); combined analysis of variance for each pair
of genotypes (Plaised and Peterson, 1959); ecovalence (Wrickle, 1962;1964); stability
variance (Shuka, 1972); cluster analysis (Lin et al., 1986); and cultivar superiority measure
(Lin and Binns, 1988).
Furthermore, the additive main effect and multiplicative interaction method, widely
known as AMMI model, has recently been developed and used to analyze multi12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




77
environmental trials (Gauch and Zobel, 1988; Zobel et al., 1988; Crossa et al., 1990). It is a
unified approach that fits the additive main effects of genotypes and environments by the
usual analysis of variance and then describes the non-additive part by principal components
analysis fitted to the AMMI model according to the following equation:
Yge -p + ag +6e + Z An ygn &n + ege,
n-1
where Yge is the yield of genotype g in environment e; Pi is the grand mean;
ag are the genotype mean deviations (the genotype means minus the grand mean);
3e are the environment mean deviations; Xn is the eigenvalue of principal components
analysis (PCA) axis n; ygn and ygn are the genotype and environment PCA scores for PCA
axis n; i is the number of PCA axes retained in the model; Oge is the residual.
Wheat is a widely grown crop in the world, produced under a wide range of environments (Hanson et al., 1982). In Ethiopia, wheat is grown between 6 and 140N latitudes; and between 35 and 420E longitude at altitudes from 1500 m to 3200 m. The current total area of production of both durum (Triticum turgidum var. durum) and bread wheat (Triticum aestivum) is estimated to be between 1.2 to 1.5 million hectares. The environmental variation within the wheat growing range is noticeable over short distances. This type of environment obviously affects the adaptation of a particular crop variety and complicates the process of plant breeding.
To counterbalance this problem, categorizing environments and conducting multienvironment variety evaluation and stability analysis are imperative. In the last two decades, a considerable number of bread wheat varieties were released in Ethiopia; all were assessed for adaptation and stability over different environmental categories before release (Desalegn et al., 1996; Debebe et al., 2000; Bedada et al., 1999; Desalegn et al., 2001). The purpose of this paper is to report on the stability of yield performance and yield potential of advanced bread wheat genotypes in different environments in Ethiopia.
Materials and Methods
A national variety trial consisting of 18 bread wheat genotypes was conducted from 1999 to 2001 across the wheat growing agro-ecologies in Ethiopia. The genotypes included advanced materials from a breeding program and selections from introductions. The entries were established on six rows of 2.5m plots with a distance of 20cm between rows. The seeds were drilled into the row at the rate of 150kg ha'. A randomized complete block design with four replications was used at all locations. Site-specific agronomic practices were applied. The central 4 rows were harvested to determine the yield potential of each genotype.
The testing sites included nine high potential locations, two waterlogged sites (Ginchi and Arsi-Robe) and three moisture stress sites (Alemaya, Asasa and Dhera) as described in Table 1.
For the three environmental categories, analysis of variance was performed on yield data of individual trials and AMMI analysis was carried out separately on pooled data using Agrobase 99 (Agronomix Software, Inc., 1999).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




78
Table 1. Environmental characteristics of national bread wheat testing sites in Ethiopia.
Location Altitude (masl) Long-term Major production constraint(s)
Annual
Rainfall (mm)
Kulumsa 2200 830 SR
Bekoji 2780 1098 YR, STB
Adet 2240 1250 SR
Holetta 2400 1078 STB
Arsi-Negele 1800 763 SR
Sinana 2400 1023 YR, SR, STB
Hossana 2450 YR, SR
Kokate 2150 SR, YR
Alemaya 1900 638 SR, drought
Ginchi 2100 1093 Waterlogging
Arsi-Robe 2400 873 Waterlogging
Asasa 2360 644 Terminal drought
Dhera 1680 617 Intermittent drought
SR= Stem rust; YR=yellow rust; STB=Septoria blotch
Results and Discussion
In high potential wheat growing areas, the AMMI analysis showed that environments,
genotypes and G x E interaction were highly significant. In this set of environment, IPCA 1
explained 33 per cent of the G x E interaction sum of squares and IPCA2 explained 23 per
cent of the interaction. The first principal component explained more of the interaction
variation and this could be interpreted in terms of morphological, phenological, agronomic or
disease factors that may affect the performance the genotypes (Table 2).
In the drought stressed sites, the environments, genotypes and G x E were highly
significantly different. About 41 per cent of the interaction sum of squares was explained by IPCA1. IPCA2 and IPCA3 explained 17% and 14% of the interaction, respectively. The first principal component factor has a high contribution to the interaction sum of squares while the
IPCA2 and IPCA3 are small. This indicates that one primary factor is influencing the G x E
interaction in this set of environment (Table 2).
Likewise, in the waterlogged areas the environments, genotypes and G x E interaction
were significantly different (Table 2). IPCA1 explained 42% of sum of squares of the
interaction, while IPCA2 and IPCA3 were responsible for 24% and 17 % of the interaction
sum of squares, respectively. This indicates that one fundamental factor affects G x E
interaction; this could be either genotypic or environmental in nature.
In AMMI analysis, IPCA scores of genotypes are stability indicators. IPCA scores closer
to zero (0), either from a positive or negative direction, indicate stable performance of
genotypes over sampled environments for the particular trait under consideration.
Accordingly, in high potential wheat growing environments, FH 8-2, which had IPCA
score nearest to zero (Table 3), was the most stable variety although it was low yielding. FH
8-2 was developed from a cross between Ethiopian and Germany materials for multiple
diseases resistance. Moreover, genotypes HAR 3316, HAR 3354, FH 6-1-7 and ETBWC 037 showed stable and predictable yield performance across the different environments (Figure 1).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




79
TABLE 2. Analysis of variance for multi-environment bread wheat trial over the period of 1999
to 2001.
Source Favorable Environment Low moisture Water logged environment
environment
df MS Prob df MS Prob df MS Prob
Total 1799 57 43
5 1
Environments 24 43510329.0 ** 7 98028937.3 ** 5 160558072.6 **
Genotype 17 5002302.5 ** 17 2974331.8 ** 17 6143106.4 **
GxE. 408 1279672.6 ** 11 585980.1 ** 85 1158614.0 **
9
IPCA 1 40 4251121.2 ** 23 1234844.4 ** 21 1969767.6 **
IPCA 2 38 3222205.7 ** 21 565213.7 ** 19 1258203.6 **
IPCA 3 36 1396068.3 ** 19 499430.6 17 1023044.6 *
IPCA 4 34 1234018.4 **
IPCA 5 32 969418.0 **
IPCA 6 30 979453.2 **
IPCA 7 28 779923.7 **
Residual 1275 414285.4 40 284442.936 30 433747.8
8 6
** =significant at 0.05 and 0.01 levels respectively
TABLE 3. Genotype IPCA Axis 1 Scores for High potential environment
No Genotype Score Mean No Genotype Score
1 a HAR 1522 20.5423 3688.481 11 11 HAR 2870 29.0941
2 b HAR 3224 16.4640 3908.441 10 10 HAR 2818 26.0322
3 c HAR 3116 -2.0432 4091.361 8 8 HAR 2812 20.8600
4 d HAR 3354 2.1311 4036.38 1 1 1 HAR 1522 20.5423
5 e ETBWC026 -10.4608 3621.871 2 2 HAR 3224 16.4640
6 f ETBWC028 -10.5972 3872.901 14 14 FH 9-3-4 10.7835
7 g ETBWC037 2.7556 3560.85 1 9 9 HAR 2814 7.7190
8 h HAR 2812 20.8600 3691.621 7 7 ETBWC037 2.7556
9 i HAR 2814 7.7190 3832.50I 4 4 HAR 3354 2.1311
10 j HAR 2818 26.0322 3963.501 17 17 FH 8-2 1.1397
11 k HAR 2870 29.0941 3746.581 3 3 HAR 3116 -2.0432
12 1 FH 4-2-11 -17.3445 3787.621 16 16 FH 6-1-7 -2.5637
13 m FH 81 -27.9384 4039.081 5 5 ETBWC026 -10.4608
14 n FH 9-3-4 10.7835 4117.901 6 6 ETBWC028 -10.5972
15 o FH 7-1-5 -24.6687 4059.521 12 12 FH 4-2-11 -17.3445
16 p FH 6-1-7 -2.5637 3459.06 15 15 FH 7-1-5 -24.6687
17 q FH 8-2 1.1397 3372.471 13 13 FH 81 -27.9384
18 r L.CHECK -41.9050 3986.091 18 18 L.CHECK -41.9050
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




80
29.094
k
h
I D
I 6
I b
.1n
W R I S
a P
I N g d-- -- -- -- -- --- -- -- -- -- q .. ... .. .. .. .. .. .................. .. .. .. .. .. ....
I ---I
* c 01
V pA
IH s
e C
TI
I- m II
-41.90 i
253$.8 3824.2 5109.3
I I
I N9 4I
key for the letters in the AMMI biplot
A= A-NEGELLE 01, B=ADET 01, C=BEKOJI 01, D=DEBRE-ZEIT 01, E=HOLETTA 01, F=KOKATEO1, G=-- KULUMSA 01, H=--SINANAO1, I=A-NEGELLE 00, J= ADET 00, K=
BEKOJI 00, L=D-ZEIT 00, M= HOLETTA 00, N=HOSSAENA 00, O=KOKATE 00, P=KULUMSA 00, Q=SINANA 00, R=A-NEGELLE 99, S=ADET 99, T=BEKOJI 99,
U=HOLETTA 99, V=HOSSAENA 99, W= KOKATE 99, X=KULUMSA 99, Y= SINANA
99; a=HAR 1522, b=HAR 3224, c=HAR 3116, d=HAR 3354,e=ETBWC026, f=ETBWC028,
g=ETBWC037, h=HAR 2812, I=HAR 2814, j=HAR 2818, k=HAR 2870, I=FH 4-2-11,
m=FH 81, n=FH 9-3-4, o=FH 7-1-5, p=FH 6-1-7, q=FH 8-2, r=L.CHECK
Figure 1. AMMI biplot for bread wheat variety trial consisting 18 genotypes and 25
environments (high potential) over the period of 1999 to 2001. (Note: 1 genotype and 3
environments in place of others with similar means and not shown)
On the other hand, FH 8-1 and FH 7-1-5 showed high negative IPCA1 scores, which is evidence of their specific adaptability to favorable environments. It was noted that the
genotypes have considerably high variation around the mean grain yield of 3824 kg ha'. The test sites also showed year-to-year variation in mean grain yield and were evenly distributed
in all four quadrants indicating the importance of seasonal climatic variation. Clearly, two or
more seasons of testing are better than a single year.
High seasonal variation was observed among the drought stressed environments (Figure
2). This is probably attributable to seasonal rainfall variation both in amount and distribution.
The genotypes tended to vary widely in grain yield around the grand mean (3576.8 kg ha-).
Among these genotypes, HAR 2818 gave the highest mean grain yield indicating that it could
be an ideal variety for low moisture stress areas; while ETBWC026 was the most stable
genotype in this environment (Table 4).
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




81
For water logged environment, genotypes like ETBWCO37 and FH 4-2-11, which had IPCA score close to zero, were found to perform stably and may be used in such areas as Arsi Robe
and Ginchi (Table 5)
39.3
h
H
P
.r
A
G
i.
0 D q
.... .. .. .... .. .. .... .. ... .. .. . . . . . . .
k m
B d
ic
C
-29.4 1 o
1998. 3576.8 5219.8
C
F
key for the letters in the AMMI biplot
A=AS01, B=DH01, D=AS00, E=DHOO, F=AS99, G=DH99, H=AL99, I=AL01; a=HAR
1522, b=HAR 3224, c=HAR 3116, d=HAR 3354,e=ETBWCO26, f=ETBWCO28,
g=ETBWCO37, h=HAR 2812, I=HAR 2814, j=HAR 2818, k=HAR 2870, l=FH 4-2-11,
m=FH 81, n=FH 9-3-4, o=FH 7-1-5, p=FH 6-1-7, q=FH 8-2, r=L.CHECK
Figure 2. AMMI biplot for bread wheat variety trial consisting 18 genotypes and 9 environments
(moisture stress areas) over the period of 1999 to 2001.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




82
TABLE 4. Genotype IPCA Axis 1 Scores for Moisture stress environment
No Genotype Score Mean No Genotype Score
1 a HAR 1522 -8.4258 3571.911 8 HAR 2812 30.0680
2 b HAR 3224 -0.9773 3829.221 16 FH 6-1-7 18.7863
3 c HAR 3116 -19.9266 4065.53 1 18 L.CHECK 14.0816
4 d HAR 3354 -4.8353 3701.381 9 HAR 2814 8.0520
5 e ETBWC026 0.2994 3561.941 17 FH 8-2 5.5885
6 f ETBWC028 4.4483 3638.341 6 ETBWC028 4.4483
7 g ETBWC037 3.9407 3348.191 7 ETBWC037 3.9407
8 h HAR 2812 30.0680 3445.281 5 ETBWC026 0.2994
9 i HAR 2814 8.0520 3545.061 2 HAR 3224 -0.9773
10 j HAR 2818 -5.8004 4134.161 13 FH 81 -2.1974
11 k HAR 2870 -2.9071 3390.591 11 HAR 2870 -2.9071
12 1 FH 4-2-11 -13.4219 3373.721 4 HAR 3354 -4.8353
13 m FH 81 -2.1974 3680.751 10 HAR 2818 -5.8004
14 n FH 9-3-4 -13.3237 3891.501 1 HAR 1522 -8.4258
15 o FH 7-1-5 -13.4493 3535.34 14 FH 9-3-4 -13.3237
16 p FH 6-1-7 18.7863 2938.75 12 FH 4-2-11 -13.4219
17 q FH 8-2 5.5885 3054.971 15 FH 7-1-5 -13.4493
18 r L.CHECK 14.0816 3676.221 3 HAR 3116 -19.9266
TABLE 5. Genotype IPCA Axis 1 Scores for water-logged environment
No Genotype Score Mean No Genotype Score
1 a HAR 1522 1.0963 4994.58 1 6 ETBWC028 21.1901
2 b HAR 3224 -3.0194 5094.63 17 FH 8-2 18.9806
3 c HAR 3116 -12.8362 5441.63 16 FH 6-1-7 12.4562
4 d HAR 3354 -1.9025 5489.251 15 FH 7-1-5 11.3533
5 e ETBWC026 -15.6519 4746.421 11 HAR 2870 9.9798
6 f ETBWC028 21.1901 4486.791 18 L.CHECK 9.9606
7 g ETBWC037 1.8652 4576.63 10 HAR 2818 6.9080
8 h HAR 2812 -9.4631 4733.50 7 ETBWC037 1.8652
9 i HAR 2814 -3.0989 4962.171 12 FH 4-2-11 1.6042
10 j HAR 2818 6.9080 5120.921 1 HAR 1522 1.0963
11 k HAR 2870 9.9798 4760.33 1 4 HAR 3354 -1.9025
12 1 FH 4-2-11 1.6042 4129.791 2 HAR 3224 -3.0194
13 m FH 81 -16.7825 4533.42 9 HAR 2814 -3.0989
14 n FH 9-3-4 -32.6399 5225.58I 8 HAR 2812 -9.4631
15 o FH 7-1-5 11.3533 4313.711 3 HAR 3116 -12.8362
16 p FH 6-1-7 12.4562 4298.671 5 ETBWC026 -15.6519
17 q FH 8-2 18.9806 3399.08 1 13 FH 81 -16.7825
18 r L.CHECK 9.9606 4744.171 14 FH 9-3-4 -32.6399
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




83
33.7
-32.6h
Key for the letters in the AMMI biplot
A=Arsi-Robe 01, B=Ginchi 01, C=Arsi-Robe 00, D=Ginchi 00, E=Arsi-Robe 99, F=Ginchi 99; a-=H4R15211741AR3224c-H4R3116, d HAR3354e-EIBWCT126,f EIBW~l2g-EIBWCT)373rflAR281ZI-R4R2814j44R2818,k-R4R2870,1-FH4-2-11,nr =FH81,n-fFH9L40 +H7-1-5, civ
ppH6-1-7,qH8-1LCHELK
Figure 3. AMMI biplot for bread wheat variety trial consisting 18 genotypes and 6 environments (water logging stress) over the period of 1999 to 2001. (Note: one environments in place of other
with similar means and not shown)
The genotypes were ranked for their yield stability in the favorable environments
according to each genotype by environment interaction statistical analysis procedure
used.(Table 6). Spearman's rank correlation coefficient was computed for each of the possible pairs of the G x E statistics (Table 7). Student's t test for the Spearman's rank correlation coefficients showed no significant correlation with AMMI, but was highly
significantly associated with the other procedures. A similar trend was noted for the two
extreme moisture stress environments (data not presented). Clearly, the stability analysis
procedures do not conform to AMMI analysis.
From the results, we can conclude that HAR 3116 and HAR 3354 are high yielding and
stable across high potential environments; HAR 3224 and HAR 2818 perform better in the drought stress environment while FH 4-2-11 and ETBWCO37 are tolerant to waterlogged
vertisols.
12th Recional Wheat Workshop for Eastern. Central, and Southern Africa. Nakuru. Kenva. 22-26 November 2004
0
1- g a
A b -d
-32.6 Ii
e
3277 4-725_1 739Ogl4
Key for the tllters in the AMMI biplot
A=Arsi-Robe 01, B=Ginchi 01, C=Arsi-Robe 00, D=Ginchi 00, E=Arsi-Robe 99, F=Ginchi 99; a=HAR152Zb=HAR3224c-HAR3116, d=HAR 3354e=EiBWO)26,f-EiBWI28g=EiBWG37,h=HAR2812IJAR281zj=AR2818, k-HAR2S70,1l=F-I4-2-11,m=FH 81, n=FH9-3-4oc=FH 7-1-5,
p=FH6-1-7, q=FH 8-Zr LCHECK
Figure 3. AMMI biplot for bread wheat variety trial consisting 18 genotypes and 6 environments (water logging stress) over the period of 1999 to 2001. (Note: one environments in place of other
with similar means and not shown)
The genotypes were ranked for their yield stability in the favorable environments
according to each genotype by environment interaction statistical analysis procedure
used.(Table 6). Spearman's rank correlation coefficient was computed for each of the possible pairs of the G x F statistics (Table 7). Student's t test for the Spearman's rank correlation coefficients showed no significant correlation with AMMI, but was highly
significantly associated with the other procedures. A similar trend was noted for the two
extreme moisture stress environments (data not presented). Clearly, the stability analysis
procedures do not conform to AMMI analysis.
From the results, we can conclude that HAR 3116 and HAR 3354 are high yielding and
stable across high potential environments; HAR 3224 and HAR 2818 perform belier in the drought stress environment while FH 4-2-11 and ETBWCO37 are tolerant to waterlogged
vertisols.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




84
TABLE 6. Estimates of various G x E statistics and stability parameters for 18 bread wheat genotypes
grown at 25 environments during 1999 to 2001.
No5 Genotypes Cultivar Superiority Ecovalence, Stability Variance Beta~ Devition Measure, Lin and Wricke ShklaF1 ay and from
Binns Wilkinson Linearity
F iFT52T 996717 58176 T110788 10 1MOT- T97T- f2T39
/-TAR 3 515 9 -44546 73T532r-- 79468 TD75 7UP2Tz-F AP 3TT 525681 -5871 TO2379TTF 7-))3~ T1T fTT92T6
z -F AR 5~35 30726 929 TOMT= V Fr73 -- I.T79W T7T75 E7DWCU27- 973059 -50279 TT4Tg Y- ffrgTnF ETBWCU2W9 7 16908 -900414~ T613OYW IT6103~ ETT 7702
F ETWCU77T 1024086 231 7T67- T3 -- 9 'Yg5W-9-376-gT
-F F iAF2W1 1086634 -70015 TTM02T 17087 TT f97TgTV -F ARfF21T 767383 -63142 T03UMT1137356 TTIN fT2BT9fT i2RTW 828490 -777 TT44TT 12 TT7 ~?Tff TTTF iAR W7 1084103 2560 T656UT724171 T7Y59- 7975
T --FT42T-T- 743175 -7039 TT2TT212976 OT fgT71T-T FHW1 463092 87227 T33TO-FTI3'939 T10- 5 033-9
14 FH 9-3-4 555071 7247264 1278882 1336333 0.995 193433
T51 T 7- F-- 422436 -63745 TMTT56- ir4053~ T1TY5- f2198W5T6-S FHTT--- 1279747 fO17829 T7T7TT 2?Y27 TDN9W i3T73TF
fT HTF- 1288069 261173 qT9~7T-47482 TfTO0T ---9ff9
T9I 1W LTHECK 7 0 84 48 f8476 7T7TTrT-2F 3810~~ TT 7T76-TTABLE 7. Spearman's rank correlation for various stability parameters
Cultivar Ecovalence Stability Stability Deviation AMIVM
Superiority (Wricke) Variance Variance from
Measure (Shukla) (Finlay & Linearity
Cultivar 0.88* 0.88*Wikinson) .6 TTSuperiority
Measure
Ecovalence 1.00* T1T9 0.9U0 7T
(Wricke)
Stability 1T9 U. 90* 01T3
Variance Shukla)
Stability 0.9U0TFT*
Variance Filay
& Wikinson)
Deviation from1F
Linearity
*Student's t test is sigmuficant at 0.0 Flevel ofsigmificance
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




85
References
Bedada Girma, Desalegn Debelo and Bekele Geleta. 1999. Evaluation and characterization of some
bread wheat genotypes for waterlogging tolerance in Ethiopia. African Crop Science
Proceedings, vol.4.pp.113-116, Uganda.
Crossa, J. 1990. Statistical analysis of multilocation trials. Advances in Agronomy 44: 55-86.
Cochran, W. G. and Cox, G. M. 1957. Experimental designs. John Wiley &Sons Inc., New york, 2nd
Ed.
Debebe Masresha, Desalegn Debelo, Bedada Girma, Solomon Gelalcha and Balcha Yaie.2000. Bread
wheat yield stability and environmental clustering of major wheat growing zones in Ethiopia.
pp. 78-86. In: CIVMMYT. 2000. The Eleventh Regional Wheat Workshop for Eastern, Central
and Southern Africa. Addis Ababa, Ethiopia: CIMMYT.
Desalegn Debelo, Bekele Geleta, Balcha Yaie and Zewdie Alemayehu. Grain Yield Response of Some
Bread Wheat Cultivars in Diverse Environments of Ethiopia. pp.382-86. In: Tanner, D. G.,
Payne, T. S., and Abdalla, O. S. Eds. 1996. The Ninth Regional Wheat Workshop for Eastern,
Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT.
Desalegn Debelo, Bedada Girma, Zewdie Alemayehu and Solomon Gelalcha. 2001. Drought tolerance
of some bread wheat genotypes in Ethiopia. African Crop Sci. J. 9(2): 385-392.
Eberhart, S. A. and W. A. Russell. 1966. Stability parameters for comparing varieties. Crop sci.6: 3640.
Finlay, K. W. and G. N. Wilkinson. 1963. The analysis of adaptation in plant breeding programs.
Austral. J Agri. Res. 14: 742-754.
Guach, H. G. 1988. Model selection and validation for yield trials with interaction. Biometrics 88: 705715.
Guach, H. G. and R. W. Zobel. 1988. Predictive and postdictive successes of statistical analysis of yield
trials. Theor. Appl. Genet.76: 1-10.
Hanson, H., Borlaug, N. E. and Anderson, R. G. 1982. Wheat in the third world. Boulder, CO, USA,
Westview Press.
Lin, C. S. and Binns, M. R. 1988. A superiority measure of cultivar performance for cultivar x location
data. Can. J. Plant Sci. 68: 193-198.
Plaised, R. L. and Peterson, L. C. 1959. Atechnique for evaluating the ability of selection to yield
consistently in different locations and seasons. Amer. Potato J. 36: 381-385.
Shukla, G. K. 1972. Some statistical aspects of partitioning genotype-environmental components of
variability. Heredity 28: 237-245.
Wright, A. J. 1971. The analysis and prediction of some two factors interactions in grass breeding. J
Agric. Sci. 76: 301-306.
Yates, E. and Cochran. 1938. The analysis of grouped experiments. J. Agric. Sci. 28: 556-580.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




86
Seedling and Adult Plant Resistance in
Ethiopian Wheat Varieties to Local Puccinia graminis Isolates
1 2
Emebet Fekadul, Belayneh Admassu and Zerihun Kassaye
Ethiopian Agricultural Research Organization, Plant Protection Research Center,
P. 0. Box 3 7, Ambo, Ethiopa: E-mail: 'f emebet5ivahoo.com 2 belay]20@yaho corn
Abstract-Wheat entries showing resistance to stem rust at the seedling stage do not necessarily possess adult plant resistance, and vice-versa. This investigation was therefore undertaken to evaluate some wheat varieties from Ethiopia for their resistance to four virulent stem rust isolates collected from Ambo and Debre Zeit (two isolates from each locality) at seedling as well as adult plant stages. Out of 15 durum (Triticum turgidum) and 14 bread wheat (T. aestivum) varieties tested, 10 durum and 7 bread wheat varieties were found to be resistant both at the seedling and adult stages to the first Ambo isolate. Whereas 11 durum and 13 bread wheat varieties were resistant to the second isolate at both growth stages. Eight durum and 8 bread wheat varieties exhibited a resistant infection type to the two Debre Zeit isolates at both growth stages. Some materials, for instance durum wheat varieties Kilinto and Gerardo were resistant at the seedling stage but were susceptible at the adult stage, whereas variety DZ 169505 exhibited a susceptible and resistant reaction at the seedling and adult stages respectively to the first Debre Zeit isolate. Our results indicate that selection for resistance must involve rigorous field evaluation with a
range of isolates to achieve effective resistance.
Introduction
Stem rust, caused by Puccinia graminis f sp. tritici, occurs every year in the wheat growing
regions of Ethiopia. For over 20 years, stem rust has been controlled through the use of
resistant cultivars. Appearance of new races and conditions favorable for diffusion of the pathogen enhance the likelihood of epidemics that may significantly reduce wheat yield.
Although host resistance is the best method to avoid epidemics and minimize yield losses, changes in pathogen virulence and the attendant loss of resistance complicates this control
strategy. Goludan et al. (1928) reported that adult plant resistance was inherited
independently of seedling resistance. Mutkekar et al. (1985) also reported similar results
stating that entries showing resistance at the seedling stage do not necessarily posses adult
plant resistance and vice-versa. The present investigation was undertaken to evaluate selected
wheat varieties for their resistance to stem rust at seedling as well as adult plant stages. The
resistant varieties are recommended as the best sources for breeding programs.
Materials and Methods
Stem rust field populations from Ambo and Debre Zeit were multiplied on the adult plants of
the susceptible variety Morocco in a greenhouse under conditions favourable for the
development of the pathogen. Two most virulent isolates from each location were selected for
the present study.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




87
Four sets of the seedlings and adult plants of 29 bread and durum wheat cultivars were grown
in 4*4 cm size earthen pots in a glass house where they were maintained disease free. After
inoculating seedlings at 2-leaf stage and adult plants at 5-leaf stage with individual races, each
set were placed in a moist chamber for 24 hours for incubation and then transferred to glass
house benches for the development of uredinia (pustules). Rust reactions were recorded about
14 days after inoculation by following the 0-4 scale (Stakman and Levine, 1922) where 0 -;
= immune, 1-2 = resistant, and 3 and 4 = susceptible.
Results
Seedling and adult plant reactions of the cultivars are given in Tables 1 and 2 for isolates from
Ambo and Debre Zeit, respectively. Variety 2 exhibited highly resistant reactions to both
isolates at both growth stages with Ambo isolates whereas varieties 5, 6, 7, 9, 10, 11, and 14
exhibited moderately resistant reactions and hence are suggested to be used as alternatives
(Table 1).
Bread wheat varieties exhibited highly resistant reaction compared to durum wheats.
Some varieties were immune to the pathogen isolates at either of the two growth stages.
Varieties 8 and 9 were highly resistant with immune or resistant reactions to the Ambo
isolates at both growth stages. Varieties 3, 7, 10, 13, 14 had a reaction type ranging from immune to moderately resistant (Table 1). Hence, these varieties could provide additional
resistance sources. With Debre Zeit isolates varieties 8, 9, 13 had immune to highly resistant reaction type. As alternatives 2, 3, 10, 12 and 14 could be used as their reactions ranged from
immune to moderately resistant to the two pathogen isolates (Table 2).
Some materials, for instance durum wheat varieties Kilinto and Gerardo were resistant at
the seedling stage but were susceptible at the adult stage, whereas variety DZ 169505
exhibited a susceptible and resistant reaction at the seedling and adult stages, respectively to the first Debre Zeit isolate (Table 2). Resistance that could be to all isolates can be generated
by carefully choosing the right parents based on the data we have presented here.
Table 1. Reaction of seedling and adult wheat plants to p.graminis isolate from Ambo
Isolate 1 Isolate 2
Plot no. Variety Seedling Adult infection Seedling infection Adult
infection type (0-4scale) type(O-4scale) infection
type(0- Type (04scale) 4scale)
Durum
1 Assassa 2 3 2 2
2 Tob-66 1 1 1 1
3 Kilinto 1 3 2 2
4 Foka 3 2 1 1
5 Gerardo 1 1 2 2
6 DZ1640 2 1 2 2
7 DY1050 2 2 2 2
8 Cocorit-71 1 3 1 3
9 Yielma 2 1 1 1
10 DZ1928-2 1 1 1 2
11 DZ 1691 2 2 2 2
12 DZ 1695-5 2 2 3 2
13 DZ1543 1 2 1 3
14 Boohai 2 2 2 2
15 Bichena 3 2 1 3
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




88
Local check 4 4 4 4
Bread Wheat
1 Kubsa 2 4 2 3
2 Wabe 3 3 0 1
3 Galama 2 2 2 2
4 Tuse 3 3 0 1
5 Katar 3 2 2
6 Shina 3 3 0 2
7 Hawi 0 0 0
8 Simba 0 0 0
9 Wetera 0 0 0
10 HAR 2192 1 2 0 2
11 HAR 2508 3 1 1
12 F-H-6-1-7 1 3 1
13 Abola 2 0 0
14 Tura 2 1 0
Local check 4 4 4 4
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




89
Table 2. Seedling and adult plant reactions (0-4 scale) of wheat varieties when tested with to P.
graminis isolates from DebreZeit
Isolate 1 Isolate 2
No. Variety Seedling Adult plant Seedling reaction Adult plant
reaction reaction reaction
Durum wheat
1 Assassa 2 2 2 3
2 Tob-66 1 2 1
3 Kilinto 2 3 1
4 Foka 1 2 1 1
5 Gerardo 2 3 1 1
6 DZ1640 1 2 1 2
7 DY1050 2 2 2 2
8 Cocorit-71 1 1 2
9 Yielma 3 4 1 2
10 DZ1928-2 2 2 1 2
11 DZ 1691 1 2 2 2
12 DZ 1695-5 3 2 1 2
13 DZ1543 2 2 1 2
14 Boohai 3 3 3 1
15 Bichena 3 4 2 2
Susceptible check 4 4 4 4
Bread Wheat
1 Kubsa 1 3 1 0
2 Wabe 2 2 1 0
3 Galama 2 2 2
4 Tuse 2 1 3
5 Katar 3 3 1 0
6 Shina 2 4 1 2
7 Hawi 3 0 0
8 Simba 0 0 0
9 Wetera 0 0
10 HAR 2192 1 2 1 0
11 HAR 2508 2 3 1 3
12 F-H-6-1-7 1 2 1
13 Abola 1 0
14 Tura 2 0
Susceptible check 3 4 3 4
REFERENCES
Goluden, C.H., Neatby, K.W. and Welsh, J.N. (1928). The inheritance of resistance of Puccinia
graminis tritici in a cross between two varieties of Triticum vulgare. Phytopathology 18:631658.
Mutkekar, M.L., Bhangale, G.T., Patil, J.V. and Kalekar, A.R. (1985). Source of resistance to stem rust
of wheat. Cereal Rust Bulletin 13:19-22.
Stakeman, E.C. and Levine, M.N. (1922). The determination of physiologic forms of Puccinia
graminis on Triticum spp. Tech. Bull. 10 Univ. Minn. Agric. Exp. Stn. 1-10pp.
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




90
Evaluation of Kenyan Wheat (Triticum
aestivum L.) Lines for Bread Making Quality
Kimani E.N., J. Ndung'u, M.G. Kinyua and J. Owuoche
K4R- Njoro P.O. Private Bag- 20107 Njoro
Abstract-The grain composition and baking quality of wheat (Triticum aestivum L.) are important in determining end-use product acceptability, although some of these factors are influenced by environmental conditions. This study investigates the bread making quality and protein content of Kenyan wheat advanced lines. Eleven lines and one wheat variety (control) were planted in four locations in 2001 and 2002, and evaluated for flour extraction rate; protein content; farinographic dough water absorption and dough development time (DDT); and bread loaf volume, using official methods of the American Association of Cereal Chemists. There were significant (p<0.0 1) effects associated with year of planting for protein, DDT; water absorption and loaf volume. Effects due to genotype were significant (p<0.01) for flour yield and DDT. No significant (P>0.05) genotype x year interaction was observed. Genotype K7972-1 had 5.94% higher mean FY than the control K. Heroe.
Lines 92B20 and R932 showed 4.5 and 2.67mins, respectively, longer DDT than K.
Heroe. R899 had the highest mean flour Water Absorption (WA), 2.18% more than the control. R946 had the highest Loaf Volume (LV) of 619.38cm3 (20.1% higher than K. Heroe). K. Heroe recorded the lowest loaf volume of 515.63 cm3. 92B19, 92B24 and R946 had higher protein content, 19.2%, and 17.1% and 15.9%, respectively, than K. Heroe. Protein content and loaf volume correlated significantly (r = 0.657** and 0.707**, respectively) with DDT. Protein content and loaf volume correlated significantly (r= 0.642**) while flour yield correlated negatively (r= 0.462**) with flour water absorption. From this study, we suggest that 92B9, 92B19
and R946 can be used in the breeding program to improve the quality of wheat.
Introduction
Grain and baking quality of wheat (Triticum aestivum L.) refers to its suitability for the type of processing or utilization for which the raw material is destined (Morris and Rose, 1996).
Grain composition and baking quality are variable factors that depend on both genotype and
the growing conditions. Environmental conditions such as weather-related factors, soil
fertility, temperature, and soil moisture regimes have a major influence on grain and end use quality of wheat (Peterson et al, 1998). Therefore, there is genetic variation in wheat baking quality although some of the quality factors are influenced by the environment (Busch et al.,
1969; Baezinger et al., 1985; Bassett et al., 1989; Peterson et al., 1992). Wheat strength is
usually associated with flour protein. Both protein quantity and quality are considered to be
primary factors in measuring the quality of flour in relation to bread making. The quantitative
expression of crude protein is related to total organic nitrogen in the flour, whereas quality
evaluations relate specially to physicochemical characteristics of the gluten-forming
component. Protein quality criteria are related primarily to the gluten portion of the flour
protein. Glutenins and gliadins are the major components of the storage protein in wheat and make a significant contribution to dough theology and baking quality (Payne, 1987; Weegels
et al. 1996). Quality is appraised largely by subjecting the flour to several physical testing
devices, which measure various theological characteristics. The tests are performed usually on flour-water dough. They characterize the dough as related to the properties of the gluten
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




91
portion of the protein. Resistance to extension of dough at rest, hydration time, maximum development time and tolerance or resistance to breakdown at a predetermined consistency
during mechanical mixing, are some of the most common dough parameters evaluated.
Mixing characteristics are usually related to gluten quality, and can be reasonably well
defined by the use of recording dough mixers such as the farinograph. Mixing requirements of
bread dough correlate with some of the measurements obtained with recording dough mixers.
Performance of flour is evaluated experimentally under conditions that are similar to
those applied by the millers and bakers. The evaluation of flour is carried out by test baking
depending on intended use. Tests are performed under rigidly controlled and standardized
condition (AACC methods).
Flour water absorption is an important factor in production of all types of baked goods.
Usually high water absorption values are desirable, since they tend to increase unit yields of
baked products. It is measured as the amount of water required to yield dough of
predetermined consistency (Pratt, 1971).
In Kenya wheat is mainly used for chapatti and bread making but rarely used for pasta
production. The increased demand for these products has prompted the breeding of more new varieties with good qualities. Although yield and resistance to diseases and pest have been the
main objective of breeding work, recently, there has been an increasing emphasis on
improvement of quality of the grain for milling and baking. The objective of this study was to
investigate the bread making quality of advanced wheat lines, evaluated at four locations in
Kenya.
Materials and Methods
Two wheat genotypes developed in Kenya (K7972-1 and K7872), five CIMMYT lines (R946,
R891, R892, R932, R899) and four from Kenya Seed Company (92B9, 92B 19, 92B20 and
92B24) were used in this experiment. K. Heroe, a variety was used as a control. The
genotypes are described in Table 4.
The experiment was conducted at Lanet (04'S; 363'E), Naivasha (06'S; 365'E), Mau
Narok (06'S; 358'E) in Rift Valley Province and at Katumani (16'S; 374'E) in the Eastern
Province of Kenya. The cultural practices were followed according to the recommendations in
all sites in order to sustain plant growth and production of grain. The seeding rate was
maintained at 138 Kg ha'. Diammonium phosphate fertilizer was applied at the rate of 231
Kg ha' in order to supply 42 Kg N and 1Og of DAP (18% N + 20% P) per plot during
planting time. Pre-emergence herbicide STOMPR (Pendimethalin 500g/litre) was applied at
sowing time to restrict early weed growth. The experiment was laid out in a partially balanced
lattice design, replicated three times and in each replicate, there were 4 blocks with 4 entries.
At physiological maturity, the crop was harvested by Hege combine Model 140, seed cleaned
and composited. Due to inadequate seed for entries, the sites were considered as replicates.
About 800-1000g of wheat samples were sieved and cleaned, tempered to 15% moisture content (MC) and left overnight at room temperature about 20-21C. These samples were
then milled using a Buhler mill (Buhler Brothers Ltd., Switzerland) to determine flour
extraction. The milled products were withdrawn and weights of all six flour streams (blended)
and of shorts and bran were determined. Flour yield was calculated using equation:
Flour
FlourYield = F 1O0
total recovered milleqbroducts
Chemical, rheological, and baking tests were carried out to evaluate the baking quality of the wheat. The flour water absorption (WA) and dough development time (DDT) were analysed
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




92
using a small bowl (50g) Brabender Farinograph (brabender OHG. Duisburg Germany) as
described in the AACC method 54-21 (AACC, 1983). The amount of flour used was 50g
(14% moisture basis). Flour water absorption is the amount of water required to centre the
farinograph curve on the 500 brabender unit line.
Dough development time (DDT) is the duration required to develop dough visco elastic
properties. It is the interval from first addition of water to that point of maximum consistency
range immediately before first indication of weakening. This value has also been referred to
as "peak" or "peak time".
Bread making was performed using a modified basic straight dough method. The baking
formula used included 1Og flour at 14% moisture basis (mb), 1.2g yeast, 2.5g sugar, Ig salt,
3g dry milk powder, 3g fat and lml of bromate phosphate solution. Dough ingredients were
mixed until optimum mixing time and then the dough was fermented (30'C and 80% relative humidity) for about four hours, with intervals of moulding the dough and finally proofing for 55minutes. This was then baked for 25 minutes in the oven at 220'C. Loaf volume (LV) was
measured using the Rapeseed displacement method 10 minutes after the loaf was removed
from the oven, using a Loaf Volumeter (National Mfg. Co., Lincohl, Nebraska.)
Crude flour protein (PC) analysis was done using the Kjeldahl method (N X 5.7) using a
Tecator Kjeltec system.
SAS system was used to do statistical analysis using the General Linear Model. Analysis
of variance for all the parameters, involving sites and years was done for each quality trait.
Pearson correlation was done between some of the baking quality parameters.
Results
The mean year effects were significantly different (p< 0.01) for all the traits (Table 1). The
effects due to genotype were also significant (p<0.05) for FY and DDT. No significant
(P>0.05) effects due to year x genotype interaction were observed.
A comparison of the genotypes for the two years showed much difference across the
traits (Table 2). The means of the second year were lower for flour yield and higher for DDT,
LV and PC than those of the first year.
Among the tested genotypes, K7972-1 produced 5.94% more FY than the control K.
Heroe (Table 4). 92B20 and R932 showed longer differences of DDT of 4.5 and 2.67min,
respectively, as compared to K. Heroe. R899, 92B 19 and 92B24 showed higher water
absorption than the control (Table 4).
The loaf volumes were generally low for all the genotypes (lower than the desirable
700cm3 and above). The highest loaf volume was shown by R946 with 619.38 cm3. This was higher than K. Heroe by 20.10%. Genotype 92B 19, 92B24 and R946 exhibited protein ranging
from 118.30 -115.01 g/kg. This was 19.2%, 17.1 % and 15.9% than K. Heroe (Table 4).
Table 1: Mean Squares for flour yield, Dough Development, Flour Water absorption, loaf volume
and protein content of the 12 wheat genotypes (Triticum aestivum L.)
Source DF FY DDT WA LV PC
(%) (min) (cm3) g.Kg-1
Year 1 343.01** 0.90** 32.67** 112951.40** 1889.08**
Genotype 11 29.68* 0.37** 3.92 9422.79 237.93
Rep 3 107.78 7.22 12.23 314602.39 8021.20
Gen*Year 11 29.19 0.13 9.14 7482.89 138.81
Error 64k 12.86 0.10 4.14 5492.40 167.14
C.V 5.54 13.72 3.36 13.26 11.75
*, ** P<0.05, P<0.01
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




93
portion of the protein. Resistance to extension of dough at rest, hydration time, maximum development time and tolerance or resistance to breakdown at a predetermined consistency
during mechanical mixing, are some of the most common dough parameters evaluated.
Mixing characteristics are usually related to gluten quality, and can be reasonably well
defined by the use of recording dough mixers such as the farinograph. Mixing requirements of
bread dough correlate with some of the measurements obtained with recording dough mixers.
Performance of flour is evaluated experimentally under conditions that are similar to
those applied by the millers and bakers. The evaluation of flour is carried out by test baking
depending on intended use. Tests are performed under rigidly controlled and standardized
condition (AACC methods).
Flour water absorption is an important factor in production of all types of baked goods.
Usually high water absorption values are desirable, since they tend to increase unit yields of
baked products. It is measured as the amount of water required to yield dough of
predetermined consistency (Pratt, 1971).
In Kenya wheat is mainly used for chapatti and bread making but rarely used for pasta
production. The increased demand for these products has prompted the breeding of more new varieties with good qualities. Although yield and resistance to diseases and pest have been the
main objective of breeding work, recently, there has been an increasing emphasis on
improvement of quality of the grain for milling and baking. The objective of this study was to
investigate the bread making quality of advanced wheat lines, evaluated at four locations in
Kenya.
Materials and Methods
Two wheat genotypes developed in Kenya (K7972-1 and K7872), five CIMMYT lines (R946,
R891, R892, R932, R899) and four from Kenya Seed Company (92B9, 92B 19, 92B20 and
92B24) were used in this experiment. K. Heroe, a variety was used as a control. The
genotypes are described in Table 4.
The experiment was conducted at Lanet (04'S; 363'E), Naivasha (06'S; 365'E), Mau
Narok (06'S; 358'E) in Rift Valley Province and at Katumani (16'S; 374'E) in the Eastern
Province of Kenya. The cultural practices were followed according to the recommendations in
all sites in order to sustain plant growth and production of grain. The seeding rate was
maintained at 138 Kg ha'. Diammonium phosphate fertilizer was applied at the rate of 231
Kg ha' in order to supply 42 Kg N and 1Og of DAP (18% N + 20% P) per plot during
planting time. Pre-emergence herbicide STOMPR (Pendimethalin 500g/litre) was applied at
sowing time to restrict early weed growth. The experiment was laid out in a partially balanced
lattice design, replicated three times and in each replicate, there were 4 blocks with 4 entries.
At physiological maturity, the crop was harvested by Hege combine Model 140, seed cleaned
and composited. Due to inadequate seed for entries, the sites were considered as replicates.
About 800-1000g of wheat samples were sieved and cleaned, tempered to 15% moisture content (MC) and left overnight at room temperature about 20-21C. These samples were
then milled using a Buhler mill (Buhler Brothers Ltd., Switzerland) to determine flour
extraction. The milled products were withdrawn and weights of all six flour streams (blended)
and of shorts and bran were determined. Flour yield was calculated using equation:
Flour
FlourYield = F 1O0
total recov eredmilledproducts
Chemical, rheological, and baking tests were carried out to evaluate the baking quality of the wheat. The flour water absorption (WA) and dough development time (DDT) were analysed
12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22-26 November 2004




Full Text

PAGE 3

Proceedings of the 12th Regional Wheat Workshop for Eastern, Central and Southern Africa Nakuru, Kenya, 22 November 2004

PAGE 4

CIMMYT (http://www.cimmyt.org/) is an internationally funded, not-for-profit organization that conducts research and training related to maize and wheat throughout the developing world. Drawing on strong science and effective partnerships, CIMMYT works to create, share, and use knowledge and technology to increase food security, improve the productivity and profitability of farming systems, and sustain natural resources. Financial support for CIMMYT's work comes from many sources, including the members of the Consultative Group on International Agricultural Research (CGIAR) (http://www.cgiar.org/), national governments, foundations, development banks, and other public and private agencies. The Kenya Agricultural Research Institute (KARI) ( http://www.kari.org/ ) was established in 1979 with the express mission of increasing sustainable agricultural produ ction by generating appropriate technologies through research, and disseminating these to the farming community. Inherent to this mission is the protection, conservation, and improvement of the basic resources, both natural and human. Such resources are critical for Kenyas agricultural development and expansion of the nations scientific and tec hnological capacity. KARI has an extensive history of productive collaborators with nationa l and international institutes and universities, as well as with the private sector. KARI/CIMMYT 2006. All rights reserved. The designations employed in the presentation of materials in this publication do not imply the expression of any opinion whatsoever on the part of KARI, CIMMYT or their contributory organizations concerning the le gal status of any country, territory, city, or area, or of its authorities, or concerning the delimitation of its frontie rs or boundaries. KARI and CIMMYT enc ourage fair use of this material. Proper citation is requested. Correct citation: M.G. Kinyua, J. Kamwaga, J.O. Owuoche, A. C.Ndiema, P.N. Nj au, D. K. Friesen, D. Ouya (Editors) 2006. Proceedings of the 12th Regional Wheat Workshop for Eastern, Central, and Southern Africa. Nakuru, Kenya, 22 November 2004. Mexico, D.F.; CIMMYT and KARI ISBN: 970-648-135-4 AGROVOC Descriptors Triticum durum; Triticum aestivum; Varieties; Plant breeding; Breeding methods; Plant production; Production increase; Plant diseases; Food production; Food security; Production factors; Economic policies; Pests of plants; Pest resistance; Yield increases; Kenya; Ethiopia; Tanzania; Sudan; Africa Additional Keywords CIMMYT; KARI AGRIS Category Codes F30 Plant Genetics and Breeding E14 Development Economics and Policies H20 Plant Diseases Dewey Decimal Classification 633.116 ISBN

PAGE 5

Introduction Like its predecessors, the 12th Regional Wheat Workshop for Eastern, Central and Southern Africa, held at Merica Ho tel in Nakuru, Kenya 22 November 2004 provided a forum where wheat scientists fr om the region could come together to exchange their results, views, and ideas on the way forward in securing this important food crop. Ethiopia has an annual production of 1,600,000 me tric tons, of both bread and durum wheats, produced mainly on small scale fa rms. A large gap exists between wheat production and demand in Ethiopia, and the country relies heavily on imports of nearly 800,000 metric tons each year. Kenyas domestic bread wheat requirement, which currently stands at 720,000 tons p.a., is projected to reach 1 million tons by 2010. Like Ethiopia, Kenya pr oduces less wheat than she needs, importing some 400,000 tons each year. Clearly therefore, increased wheat production in the region through approaches such as breeding, agronomy, crop protection and improved economic policies, is not just desirable, but essential for the food security of millions of people. This proceedings presents valuable new info rmation from research conducted in Kenya (11 articles), Ethiopia (9), The Sudan (2), and Tanzania (1), and is organized into four sections, namely Agronomy Breeding, Protection, And Economics. The Agronomy section contains 6 papers, ra nging in subject mater from water and fertilizer use efficiency in wheat varieties, to a comparison of field pea and barley as break crops in wheat farming. Breeding is the largest section, with 14 articles, including results from a three-year survey of the wheat stem rust situation in Kenya. The Protection section, with 3 articles, reports on the chemical management of weeds, and the Russian wheat aphid Diuraphis noxia a new and damaging pest in the region. Results from a survey of the natural enemies of D. noxia are also presented, as the starting point for designi ng a biological control strategy for the aphid. In his economics article, Professor Abbas Elsir M. Elamin deals with the financial and economic profitability of wheat producti on in The Sudan, and its international competitiveness. The comprehensive analysis discusses government reform policies in the 1990s, and their role in wheat production, consumpti on and input use from that time to the present. We believe that this volume represents an important contribution to the regional and global knowledge bases on wheat producti on and economics, and that a better understanding of the subject matter will cont ribute to the ultimate goal of improved livelihoods for the people of Africa. Romano Kiome, PhD Permanent Secretary, Ministry of Agriculture

PAGE 6

Acknowledgments The Workshop organizers would like to acknowledge the contributions of several institutions and individual s towards the success of the workshop, and the publication of this proceedings. We thank Monsanto, Bayer Crop science, and Osho chemicals, for funding the workshop; KARI and CIMMYT for technical and administrative support; and all the workshop participants, for their cooperat ion during all stages, starting with the preparation and presentation of the papers, to their review/revision for publication. CIMMYT reviewers Karim Ammar, Hans-J oachim Braun, Jose Crossa, Hugo De Groote, Etienne Duveiller, Augustine Langyint uo, Ivan Monasterio, Wilfred Mwangi, Tom Payne, Javier Pena, Matthew Reynol ds, Ken Sayre, Ravi Singh, Richard Trethowan, and Pat Wall, provided thorough and constructive comments on the manuscripts, including suggestions for impr ovement. This was a valuable contribution towards seeing the proceedings to comple tion, for which we are very grateful. Dr. Miriam Kinyua Chair, Organizing Committee

PAGE 7

Contents Part 1. AGRONOMY Effect of Nitrogen Fertilizer Levels and Vari eties on Gluten Content and Some Rheological Characteristics of Durum Wheat Flour Bemnet Gashawbeza, Solomon Assefa, Ameh a Yaekob, Alemayehu Zemede, Jemanesh Kifetew and Bekele Mekuria ...............................................................................................2 Impact of Irrigation Frequency and Farmyard Manure on Wheat Productivity on a SalineSodic Soil in Dongola, Sudan Elmoiez M. Fadul and Mukhtar A. Mustafa .....................................................................9 Consumptive Use of Water and Water Use Efficiency by Wheat ( Triticum aestivum ) in Relation to Irrigation and Nitrogen Antony M. Kibe Subedar Singh and Naveen Kalra ......................................................21 Agronomic and Economic Evaluation of Break Crops and Management Practices on Grain Yield of Wheat at Shambo, Western Oromiya, Ethiopia Tolera Abera and Mathewos Belissa.................................................................................35 Response of Bread Wheat to Nitrogen and Phosphorous Fertilizers at Different Agroecologies of Northwestern Ethiopia Minale Liben, Alemayehu Assefa, Tilahun Tadesse and Abreham Mariye....................41 Grain Yield, Water Use and Water Use Efficiency as Affected by Moisture Level Under Rain-Out Shelter P. A. Ooro M.G.Kinyua and J.B.O. Ogola ...................................................................46 Part 2. BREEDING Current Status of Stem Rust in Wheat Production in Kenya R. Wanyera, M.G Kinyua, P. Njau, J.W. Kamundia and S.Kilonzo...............................55 Genotype-by-Nutrient Interaction in Wheat Gr own in a Marginal Environment in Kenya J. Kamwaga, H. Okwaro, P. Njau, P. Kimurto, P. Ndungu. and E. Kimani...................63 Improving Wheat Productivity for the Drought Prone Areas of Kenya Using the Doubled Haploid Technique Njau P. N Kimurto, P.K Kinyua M. G Okwaro H. K and Ogolla J.B.O ...................67 Grain Yield Stability of Bread Wheat Genotyp es in Favorable and Stressed Environments Desalegn Debelo, Solomon Gelalcha, Balc ha Yaie, Bedada Girma, Berhanu Mamo, and Debebe Masresha...............................................................................................................7 6 Seedling and Adult Plant Resistance in Ethiopian Wheat Varieties to Local Puccinia graminis Isolates Emebet Fekadu Belayneh Admassu and Zerihun Kassaye............................................86 Evaluation of Kenyan Wheat ( Triticum aestivum L.) Lines for Bread Making Quality Kimani E.N., J. Ndungu, M.G. Kinyua and J. Owuoche................................................90

PAGE 8

Physiological Traits Associated with Drought Tolerance in Bread Wheat ( Triticum Aestivum L.) under Tropical Conditions Kimurto, P.K. M.G. Kinyua J.B.O. Ogola J.M. Macharia and P.N. Njau ...............96 Allelism of Resistance Genes to Phaeosphaeria nodorum in Wheat C. A. Kuwite and G. R. Hughes ......................................................................................109 Evaluation of Kenyan Breadwheat ( Triticum aetivum L.) Varieties for Resistance to Russian Wheat Aphid in Multi-location Trials J. Malinga M. G. Kinyua A. Kamau J. K. Wanjama P. Njau and J. Kamundia ....................................................................................................................117 On-Farm Evaluation and Comparison of New and Old Wheat Varieties R. V. Ndondi C .A. Kuwite and R.Shekibula ................................................................125 Evaluation of Improved Wheat Varieties Under Different Management Practices in Eastern Wallagga Highlands Tolera Abera Daba Feyisa Girma W. Tsadik Hasan Yusuf and Gemechu Keneni 134 Cell Membrane Stability (CMS) as Screening Tec hnique for Drought Tolerance in Bread and Durum Wheat Genotypes Alemayehu Zemede H. Martens and M.T. Labuschagne ...........................................140 Physiological Races and Virulence Diversity of Puccinia graminis f. sp. tritici on Wheat in Ethiopia Belayneh Admassu Emebet Fekadu and Zerihun Kassaye..........................................145 Participatory Evaluation of Bread Wheat Vari eties in the Central Highlands of Ethiopia Kassa Getu, Kassahun Zewdie, Yeshimebet Gebrehiwot and Addisu Alemayehu.........151 Part 3. PROTECTION Survey of Natural Enemies of the Russian Wheat Aphid, Diuraphis Noxia (Kurdijimov) in Kenya M.Macharia, M. Njuguna, and I. Koros.........................................................................161 Evaluation of the Herbicide Monitor, Alone a nd in Tank-Mixes for Weed Control in Wheat ( Triticum aestivum L.) D.O.K. Amadi, J.N. Kamwaga and G.O. Mbanda..........................................................166 Control of the Russian Wheat Aphid, Diuraphis Noxia (Kurdijumov) in Wheat Using Systemic Insecticides in Kenya M. Macharia, M. Njuguna and I. Koros.........................................................................175 Part 4. ECONOMICS Analysis of Marketing and Pricing Polici es on Technology, Input Use and Production of Wheat in the Sudan Abbas Elsir M. Elamin.....................................................................................................181 Appendixes Opening Speech.....................................................................................................................195trn Closing Speech......................................................................................................................197rn Question and Answer Sessions..............................................................................................199trn Participants E-mail Addresses..............................................................................................231trn

PAGE 9

Part 1. AGRONOMY

PAGE 10

Effect of Nitrogen Fertilizer Levels and Varieties on Gluten Content and Some Rheological Characteristics of Durum Wheat Flour Bemnet Gashawbeza1, Solomon Assefa1, Ameha Yaekob1, Alemayehu Zemede1, Jemanesh Kifetew1and Bekele Mekuria2 1 Ethiopian Agricultural Research Organization, Debre Zeit Agricultural Research Center, P.O.Box 32, Debre Zeit, Ethiopia; 2 Kality Food Share Company, P.O.Box 1819, Addis Ababa, Ethiopia Abstract Improved gluten and rheological characteristics of durum wheat ( Triticum durum Desf) have recently received increased attention due to their positive effect on the quality of pasta products. An experiment was conducted at Debre Zeit and Akaki in the Central Highlands of Ethiopia, during the 2000/01 cropping season, to study the influence of nitrogen fertilization levels (0, 30, 60, 90 and 120 KgNha -1 ) on gluten content and some rheological ch aracteristics (dough resistance, expansion and extensibility) of five durum wheat varieties at Akaki and Debre Zeit. At Debre Zeit, gluten content significantly varied (P<0.05) depending on wheat genotype (variety) and nitrogen fertilizer levels, but at Akaki only varietal differences were significant. Nitrogen level increment consistently increased wet gluten content of varieties at Debre Ziet and; the highest nitrogen level, 120 Kgha-1, gave significantly higher wet gluten percent 29.5 % than all other treatment levels, except 90 Kgha-1 (27.3%). Tob 66 (30.8%) had the highest wet gluten content followed by Foka (27.5%). At Akaki, Tob 66 was the highest (24.9%), followed by Boohai (22.4%). The studied rheological characteristics at Debre Zeit showed that nitrogen fertilizer application had no effect on either of the traits (dough resistance and extensibility), but varietal differences were observed in dough resistance. Therefore, durum wheat quality improvement should concentrate both on selection of appropriate varieties and management practices such as nitrogen fertilizer application. Further study on input-responsive semi dwarf varieties is recommended. Introduction Durum wheat is used for pasta production because of its hard grain texture, amber color, and other grain quality traits rela ted to endosperm protein. Increasing grain protein concentration and improving other grain quality traits have been the major objectives in most varietal development processes (Metho, Hammes and Beer, 1997). This is because the processing quality of wheat is largely dependent on the am ount and quality of endosperm protein. Grain protein content affects milling and other industr ial qualities of wheat. As a result, premiums are commonly paid for protein levels above base line (Woolfolk et al. 2002). Gluten is an important endosperm protein th at affects pasta quality. It is a visco-elastic component of wheat dough responsible for physical dough properties (Curic et al. 2001). Due to its strong relation with greater c ooked firmness and increased tolerance to overcooking, strong gluten varieties are preferred (Josephides et al. 1987). Besides gluten content, rheological characteristics that measur e gluten quality are also important criteria of

PAGE 11

selection for quality durum wheat. D'Egidio et al. (1990) have shown that Alveogrphic parameters are excellent indicator of durum wheat gluten properties. Quality durum wheat grain used for production of pasta is achieved by selection of appropriate variety, environment a nd management practices (Rharrabti et al. 2001). Under high rainfall and low soil fertility conditions gr ain protein production is limited (Simmonds, 1989). The effect of environment and manage ment methods, especially availability of nitrogen in the soil during ripening, on vitr eousness of durum wheat has been reported (Mosconi and Bozzini, 1973; Ottman et al. 2000). Numerous authors have obtained improved grain protein upon intensiv e nitrogen fertilization (Geleto et al. 1996; Gashawbebeza et al. 2002; Virga et al. 2003). The work of Hadjichristodoulou (1979) showed significant effects of nitrogen fertili zation, genotype and loca tion on vitreousness, a physical grain quality correlated with protein content, of durum wheat. Dexter et al. (1982) also found significant difference between varieties for several quality traits at different nitrogen fertilizer levels. Durum wheat is a traditional crop for the Et hiopian farmers. Its production however has been limited to locally made foods. The current increasing of food processing industries and importance of pasta products by the urban people of the country has increased the importance of quality durum wheat production. The objective of this experiment was, theref ore, to evaluate the effect of nitrogen fertilizer and varieties on gluten content and some rheological properties of durum wheat flour. Materials and Methods A field experiment was conducted during the 2000/01 cropping season under rain-fed conditions at two locations in the Central Hi ghlands of Ethiopia (Debre Zeit and Akaki). Debre Zeit (8o 44 N, 39 o 02E) is mid-highland (1900 m.a.s.l.) characterized by moderate rainfall (851mm average annual rainfall); 17.9oC average mean temperature and Pellic Vertisol soil; and Akaki (8 o52N, 38 o 47E) mid altitude area (2100masl) characterized by average annual rainfall of 1086mm and 15.6oC average mean temperature. Five medium tall to tall durum wheat varieties viz (Kilinto, To b66, Foka, Assasa and Boohai) that were selected for their industrial quality by the local p asta industries were planted at five different nitrogen levels (0, 30, 60, 90 and 120 kgha-1 N) with uniform basal application of 10kgha-1 phosphorus in the form of Triple Super Phos phate (TSP). The experiment was laid out in Randomized Complete Block (RCB) in factorial a rrangement with three replications. The plot size was 3m x 4m (12m2) and data was recorded from 10.4m2. Seeding rate was 150 kgha-1. Nitrogen was split applied half at planting and the remaining half at full tillering. Wet gluten content was determined from flour and was determ ined by gluten wash method (ICC standard No. 106/2) while rheological characteristics vi z. extensibility (L in mm) and dough resistance (P=height X 1.1 in mm) were measured usi ng Chopin Alveograph (ICC No.121) (ICC, 2000). Rheological data was taken for Debre Zeit only. Data was analyzed by ANOVA using MSTATC. Results Wet gluten. ANOVA showed that wet gluen content significantly varied depending on varieties, locations and; location X nitrogen level interaction was significant (P<0.05) but no marked difference between the nitrogen levels in the combined analysis. However, the difference between varieties and fertilizer levels were significant (P<0.05) at Debre Zeit. At Debre Zeit, the highest nitrogen level (120 Kgha-1) had significantly higher wet gluten percent

PAGE 12

than all other treatment levels except 90 Kgha-1. At Akaki, on the other hand, only varietal differences were significant (Table 1). Varietal differences were evident at both locations. In the combined analysis Tob 66 gave the highest wet gluten content (27.9%) while the lowest was Kilinto (21.18%). The response to applied nitrogen was higher at Debr e Zeit and Tob 66 also gave the highest wet gluten percent (30.8) while the least was K ilinto (22.55%). But no significant difference was observed between Boohai, Foka and Assassa. Dough resistance. Varietal difference was significant for Dough resistance. But no marked difference between nitrogen levels was observed. Boohai (113.9) gave the highest dough resistance followed by Foka ( 101.5) while the least was Assassa. Dough extensibility. This trait was not aff ected by the applied nitrogen fertilizer levels (Table 2). Besides, there was no statistically significant difference between varieties. Table 1. Nitrogen fertilizer level effects on we t gluten (%) of durum wheat varieties at two locations (2000/01) Wet gluten Treatment Debre Zeit Akaki Mean Nitrogen (Kg ha-1) 0 24.4c 22.2 23.3 30 26.5bc 21.8 24.2 60 26.5bc 22.8 24.6 90 27.3ab 21.29 24.3 120 29.5a 20.95 25.2 LSD (0.05) 2.543 NS NS CV (%) 12.91 15.57 14.1 Variety Boohai 27.02b 22.4ab 24.73b Foka 27.5b 20.65b 24.11b Kilinto 22.5c 19.8b 21.18c Tob 66 30.8a 24.93a 27.9a Assassa 26.2b 21.29b 21.29b Means within a column with similar letters are not significantly different at P<0.05 by DMRT. NS = no statistically si gnificant difference Table 2. Effect of nitrogen levels on rheological characteristics of durum wheat flour at Debre Zeit (2000/01) Treatment Dough resistance (P) Expansion (G) Extensibility (L) P/L Nitrogen 0 95.6 19.35 78.8 1.1 30 97.8 18.09 68.6 1.4 60 93.7 19.1 76.9 1.2 90 88.8 18.5 73.4 1.2 120 101.7 19.27 70.1 1.4 Variety Boohai 113.9a 20.3 78.8 1.4 Foka 101.5ab 18.7 73.3 1.4 Kilinto 92.5bc 18.8 74.6 1.2 Tob66 91.8bc 18.0 68.7 1.3 Assassa 78.1c 18.6 72.1 1.1 LSD(0.05) 15.8 NS NS CV(%) 22.59 11.6 18.64 Means within a column with similar letters are not significantly different at P<0.05 by DMRT. NS = no statistically significant difference

PAGE 13

Generally, nitrogen level increment increased wet gluten contents of durum wheat varieties at Debre Zeit but not on rheological characteris tics (dough resistance and extensibility). The response of gluten contents was linear with the respective applications of 0, 30, 60, 90, and 120 kgha-1. In similar studies, the positive effect of nitrogen increment on grain protein (Prima et al. 1982 and Geleto et al. 1996) and also on gluten content of wheat (Prima et al. 1982 and Varga et al. 2003) was reported. At Akaki, on the other hand, the response to the applied nitrogen was not consistent. The work of Campbell et al. (1993) also showed significant contribution of environmental factors on grain nitrogen content of wheat in addition to nitrogen application. As in most grain quality traits in wheat, protein content is known to be affected by genetic and environment mainly location, fertilizer app lication and other management practices. The effect of nitrogen fertilizer on whole meal protein content of durum wheat was reported (Gashawbeza et al. 2002). Thus this difference could be ascribed to the difference between the two locations in air temperature and degree of water-logging. Ak aki is more waterlogged than Debre Zeit and could have inhibited efficient utilizati on of the applied nitrogen (Gashawbeza et al. 2002). According to Spiertz and Devos (1983), under high temperature especially during grain filling period protein synthesis is favored. Therefore, the environment at Debre Zeit could favor protein synthesis due to its relatively higher mean temperature and less water-logging stress. Besides, no correlation between gluten content and grain yield at Debre Zeit (r=0.15) suggesting improving gluten content couldn't risk grain yield. From the result of the study, for high rainfall and waterlogged environments like Ak aki, it appears that other crop management practices such as drainage of excess water and adjustment of planting dates should be studied, in addition to nitrogen appl ication, to increase gluten content of durum wheat. Gluten content of varieties was significant at both locations and Tob 66 had the highest while kilinto had the lowest (Table 1). As show n in Fig.1, at Debre Zeit, there is a general linear increasing trend of gluten content of va rieties with increasing nitrogen level. However, the increase was sharp beyond 90 N kg/ha for Tob 66, which gave 37 % gluten content at the highest nitrogen fertilizer level. Besides, Tob 66 had consistently high wet gluten starting from the nil nitrogen treatment to the highest leve l and, this could suggest the efficiency of the variety in converting assimilates to protein. At Akaki on the other hand, there is no definite trend for the applied nitrogen levels and gluten content (Fig.2).

PAGE 14

0 5 10 15 20 25 30 35 40 0306090120 Nitrogen Level(kg/ha)Wet gluten (%) Assassa Boohai Foka Kilinto Tob 66 Fig. 1. Gluten Content of duru m wheat varieties at different nitrogen levels at Debre Zeit 0 5 10 15 20 25 30 0306090120 Nitrogen level (kg/ha)Gluten content (5) Kilinto Tob 66 Foka Boohai Assassa Fig. 2. Gluten Content of du rum wheat varieties at differen t nitrogen levels at Akaki According to ICC (2001), the rheological characteristics of a dough are expressed as the resistance of the dough to stretching and its ex tensibility until it begins to rupture. (Curic et al. 2001) measured the physical properties of the dough of wheat flour, which primarily

PAGE 15

depend on the gluten quality. In this study, the observed non-significant difference in rheological properties due to increasing applicati on of nitrogen fertilizer could suggest that the traits are affected more by genotype rath er than by environment in which varieties are grown. Unlike this, however, Virga et al. (2003) reported the use of the intensive production system in bread wheat resulted in a significant improvement in dough resistance. Curic et al. (2001) reported gluten quality of wheat to be dependent on location and region of cultivation. In this study, varietal differences were observe d in dough resistance. Assassa showed the least performance in dough resistance but it was within the acceptable range for pasta industry use. From the alveograph curve, the resistance to stretching is taken from the maximum height reached by the curve (P) while extensibility is ta ken from the total length (L) in the horizontal direction of the curve. Alveograph strength is taken from the area under the alveogram (W) X 10-3 ergs and the greater the reference area shows the higher the value of W. Alveograph strength (W X 10-3 ergs) and the ratio of dough resistance to extensibility (P/L) are the two commonly used indicators of quality of durum wheat rheological characteristics. When W is greater or equal to 250 and P/L ratio is greater th an 0.8, the flour is considered to be obtained from strong grains (Professional pasta website, 1999). Considering P/L ratio, all varieties fall under strong gluten category. Discussion Increasing grain protein, especially that of glut en content of durum wh eat and its strength has recently received higher attention due to the pr emium product produced out of it. In Ethiopia, growing selected varieties like Boohai, Foka and Tob 66 with high gluten content and acceptable dough characteristics under appropriate growing environment a nd; could be a good option for durum wheat growers to produce high quality durum wheat. Considering the varied agro-climatic conditions of the country, which va ries within a short distance, the durum wheat growing environments should be characterized for suitability to produce industrial quality durum wheat. Regarding the studied rheological properties, the observed non-significant difference between nitrogen fertilizer levels c ould show selection of appropriate varieties could be more important factor than the grow ing environment. Although early to conclude, the observed significant differences in gluten content between 60kgha-1, recommended for grain yield at Debre Zeit, and 120Kg ha-1 N levels in this study could suggest the need to consider economically important quality traits besides grain yield in fertilizer rate recommendations. The durum wheat breeding program should also concentrate on development of varieties with desirable chemical grain quality as well as physical dough properties required by Pasta industries. Such type of study on input responsive semi dwarf durum wheat varieties is recommended. AcknowledgmentsThe authors are thankful to th e technical staff of Durum Wheat Research Project of the Debre Zeit Agricultural Research Center. The authors would like to acknowledge Kality Food Share Co mpany for quality analysis. References Campbell, C.A., Selles. F., Zentner, R.P. and MacConkey B.G. 1993. Nitrogen management for zero till spring wheat: disposition in plant and utilization efficiency. Commun. Soil Sci. Plant Anal 24:2223-2239 Curic, D., Karlovic, D., Tusak, D., Petrovic,B.and D ugum, J.2001. Gluten as a standard of wheat flour quality. Biotechnol.39: 353-361

PAGE 16

D'Egidio,M.G., Mariani, B.M., Nardi,S., Novaro,P and Cubadda,R.(1990). Chemical and technological variables and their relationships: A predictive equation for pasta cooking quality. Cereal Chem., 67,275-281 Dexter, J.E., Crowle, W.C. Matsuo, R.R. and Kosmul ax, F.G. 1982. Effect of N fertilization on quality characteristics of five North Am erican amber durum wheat vari etys. 1. Can.J. Plant Sci. (Canada), 62: 90-902 Gashawbeza, B., Mekuria, B., Yaekob, A., Mitiku, D., Tadesse, T., and Kifetew, J. 2002. Effect of nitrogen fertilization on grain protein conten t of durum wheat varietys. Proceedings of 6th Annual Conference of Ethiopian Soil Science Society. (in press) Geleto, T.,Tanner, D.G., Mamo, T. and Gebeyehu, G.,. 1996. Response of rainfed bread and durum wheat to source, level and timing of nitrogen fertilizer on two Ethiopian Vertisols: II. N uptake, recovery and efficiency. Hadjichristodoulou, A. 1979. Genetic and environmental effects on vitreousness of durum wheat. Euphytica 28:711-716 International Association for Cereal Science and Technology. 2000. ICC Standards 106/2,Viena, Austria Josephides, C.M.; Joppa, L.R. and Youngs, V.L. 1987. Effect of chromosome 1 B on gluten strength and other characteristics of duru m wheat. Crop Sci. 27:212-216 Metho, A.L, Hammes, P.S. and De Beer,J.M. 1997. Ef fect of cultivar and soil fertility on grain yield, yield components and grain nitrogen content of wheat. African Crop Science Proceedings, Vol. 3: 695-709 Mosconi, G and Bozzini, A. 1973. Effects of appli cation of late nitrogen fertilizer to durum wheat. Revista di Agronomia:75 (Cited in Field Crop Abstr. (1975) 4226) Ottman, M.J., Doerge, T.A. and Edward, E.M. 2000 Durum grain quality as affected by nitrogen fertilization near anthesis and irrigation duri ng grain fill. Agron. J. 92:1035-1041. Prima, G.Di., Sarrino, R. and Stringi, L. 1982. N itrogen, its role in controlling yield and quality of durum wheat in the warm arid zone of Sicily. Inst. Agron. Gen. Erbace, Italy (Cited in An Annoted Bibliography in Durum Wheat 1972-1984, ICARDA) Rharrabti,Y., Villegas,D., Moral,L. F.G., Aparicio, N., Elhani, S. and Royo, C. 2001. Environmental and genetic determination of protein content and grain yield in durum wheat under Mediterranean conditions. Plant Breeding 120,381-388 Simmonds, D.H. 1989. Inherent qu ality factors in wheat. Wheat and wheat quality in Australia. CSIRO, Australia, P31-61 Spiertz, J.H.J. and Devos, N.M. 1983. Agronomical and physiological aspects of the role of nitrogen in yield formation of cereal s. Plant and soil 75:379-391 Woolfolk, C.W., Raun, W.R. Johnson. G.V. Thom ason, W.E., Mullen, R. W., Wynn, K. J. and K.W.freeman.2002. Influence of late-season foliar nitrogen application on yield and grain nitrogen in winter wheat. Agron.J. Agron.J. 94:429-434

PAGE 17

Impact of Irrigation Frequency and Farmyard Manure on Wheat Productivity on a SalineSodic Soil in Dongola, Sudan Elmoiez M. Fadul1 and Mukhtar A. Mustafa2 1Dongola Agric. Research Station (ARC), Sudan. 2Faculty of Agriculture, University of Khartoum, Sudan Abstract A field experiment was conducted in January 2001 and December 2002, at Dongola University Farm to investigat e the effects of irrigation frequency and farmyard manure application on salt leaching and on wheat ( Triticum aestivum L.) growth on a salinesodic sandy loam soil classified as fine loam, mixed, hyperthermic, sodic haplocalcids. Each experiment consisted of three irrigation frequencies: 7, 14, and 21 days, and three levels of farm yard manure (F.Y.M): 0, 4.8 and 9.7 ton/fed. The quantity of water applied was proportional to the irrigation frequency and estimated from knowledge of reference evapotranspiration as predicted by Jensen and Haise equation, a crop factor and an irrigation efficiency value. Each treatment was replicated thrice in a split-plot design. Irrigation treatments accommodated main plots (7 18m) and manure application sub-plots (7 x 6m). Data, collected at harvest, showed that all irrigation treatments caused salt leaching, which decreased with increase of soil depth. The data did not reflect the salt distribution between irrigation intervals. Th e irrigated weekly (F1) was the superior treatment; all forage and grain yields and their components incr eased with irrigation frequency. INTRODUCTION Northern Sudan is dominated by hyper-arid, arid and semiarid ecological zones that favor the formation of salt-affected soils (Nachtergael e, 1976; Mustafa, 1986). Dongola, in the Northern State, has two main soil orders: Entis ols in the first terrace and Aridisols in the upper second and third terraces. Entisols, at the close proximity of the Nile, are fertile, nonsaline, non-sodic and highly productive soils. Ho wever, they are endangered by gully erosion at the riverside and sand encroachment from th e adjacent desert. Furthermore, the land is intensively cultivated and frac tionated due to land tenure laws. The accessibility of good quality Nile water (Mustafa, 1973) and Nubian aquifer water prompted horizontal expansion of agriculture in salt-affected soils of the uppe r terraces (Aridisols). The productivity of these soils is constrained by osmotic and specifi c ion effects and nutritional imbalance. Wheat ( Triticum aestivum ) is a strategic field crop in Sudan, since it constitutes the main staple food for most of the urban population. It is grown in Northern and Central Sudan. However, the new trend is to restrict its production to the northern State because of its favorable climatic environment and consequent higher yield as compared to Central Sudan. The national need for more wheat production necessitated the expansion of wheat production to salt-affected areas. The proper use of these soils requires appropriate soil and water management. The share of the Nile water of Sudan is limited because of the large expansion of irrigated agriculture in the arid and semi-ari d regions. Thus, there is need to economize on water use by increasing its efficiency. The present research was undertaken to inves tigate the effect of irrigation frequency and farm yard manure on salt leaching and wheat grow th on a salinesodic Aridisol in Dongola.

PAGE 18

Materials and Methods A field experiment was conducted in two succes sive seasons (Jan. 2001 April 2001 and Dec. 2001April 2002) on an old alluvium saline-s odic sandy loam soil classified as fine loam, mixed, hyperthermic, sodic Haplocalcids (Soil Survey Staff 1996) at Dongola (19 N 29 30 E), 228 above sea level. The characteristics of this soil are presented in Table 1. The treatments consisted of three irrigation fre quencies:7 (F1), 14 (F2) and 21 (F3) days, and three levels of farm yard manure: 0 (M0), 4.8 (M1) and 9.7 (M3) ton/feddan (1 feddan = 0.42 ha).Each treatment was replicated three times in a split-plot design. Irrigation frequencies were designated to main plots and farmyard manure levels were accommodated in subplots. The main plots were arranged in co mpletely randomized block design. The land was disc ploughed to 20-cm depth, and leveled using a long-span blade leveler. Nine main plots (7 18 m), each subdivided into three sub-plot s (7 x 6 m) were constructed using earth embankments. The main plots were 1-meter ap art to check lateral water movement. In the second season, the experiment was repeated in th e same plots, which were harrowed before seeding. Wheat (Wadi El Niel variety ) was sown on 1st January 2001 in the first season and on 27th December 2001 in the second season. Seeds were hand dibbled in rows 20 cm apart in each subplot with a seed rate of 60-kg/ feddan. The quantity of water applied per irrigation (Q, mm) was estimated by the following relationship: Q = kc x ETp x F x Ei Where: kc = a crop coefficient, ETp = poten tial evapotranspiration (mm/day), F = irrigation frequency (days) and Ei = efficiency of irriga tion water application (0.7). kc was estimated using the FAO (1984) procedure, ETp was estimated using the following Jensen and Haise (1963) equations: ETp = CT (T Tx) Rs CT = 1/ (C1+7.6 CH) C1 = 38 2E/ 305 CH= 50 mb/ (e2-e1) Tx = 2.5 0.14(e2-e1) E/550 Where: T= mean air temperature ( C), E = the site elevation (m), e1 and e2 = the saturation vapor pressure of water (mb) at the mean monthly minimum and maximum air temperature of the warmest month in the year, respectively, Rs = short wave incoming solar radiation. Thus, the water applied at each irrigation event w as proportional to the irrigation frequency and varied with the month. However all treatments received the same total amount (seasonal) of water by the time of harvest. These amount s were 856.6 and 774.7 mm in the two successive seasons. The estimated irrigation water requirement (IWR) in each month expressed in mm/day is presented in Table 2. The quantity of water applied per irriga tion, Q is the product of IWR and F. Thus using Table 2, a predet ermined quantity of water was applied to each subplot using Parshall flume. Three irrigations at 7-day frequency were applied to all subplots for establishing the crop before subjecting it to different irrigation regimes. The FYM was

PAGE 19

heaped, mixed with water, and composted for on e month before application. The average EC of the composted FYM was 20.7 dS/m. It was a pplied at a predetermined rate at the third irrigation. In all treatments, P fertilizer (Tri ple super phosphate) was a pplied before sowing in the order of 80 kg/feddan. Twenty-one and 63 da ys after sowing, urea at the rate of 20 kg N/ feddan was applied to all treatments. Fertilizers were applied at the same rate and stage in both seasons. Manual weeding was done whenever needed. Soil samples were collected, before sowing a nd at harvest, by an auger from each subplot at an increment of 0.2 m from the surface to a depth of one meter. Gravimetric moisture content, electrical conductivity (ECe) and sodium adsorption ratio (SAR) of the saturation extract were determined using standard procedures (Richards, 1954, Black, 1962). At harvest, ten plants were selected at ra ndom from each subplot and their height, head length and leaf area index were determined. On e-meter square was ta ken at random and the number of heads was counted. Ten heads were selected at random and threshed collectively and the number of seeds per head was counted Thousand grains were taken randomly from each treatment and weighed. The total biomass enclosed in an area of 5 x 6 m2 in each subplot was cut at ground level and the weight was measured and expressed in ton/feddan. The harvested crop was left to dry for a week thresh ed and the total gain yield was determined in ton/feddan. The crop water use (CWU) was de termined by the following water balance equation: CWU = I + P + M where: I = Amount of irrigation water (mm), P = Amount of rainfall (mm) and M = The difference between cumulative water content befo re sowing minus that at harvest in one meter depth (mm). The crop water use efficiency (CWUE) was calculated by dividing grain yield by crop water use (kg/m3)

PAGE 20

Table 1. Seedling and adult plant reactions (0-4 scale) of wheat varietie s when tested with to P. graminis isolates from Ambo Isolate 1 Isolate 2 No. Variety Seedling reaction Adult plant reaction Seedling reaction Adult plant reaction Durum wheat 1 Assassa 2 3 2 2 2 Tob-66 1 1 1 1 3 Kilinto 1 3 2 2 4 Foka 3 2 1 1 5 Gerardo 1 1 2 2 6 DZ1640 2 1 2 2 7 DY1050 2 2 2 2 8 Cocorit-71 1 3 1 3 9 Yielma 2 1 1 1 10 DZ1928-2 1 1 1 2 11 DZ 1691 2 2 2 2 12 DZ 1695-5 2 2 3 2 13 DZ1543 1 2 1 3 14 Boohai 2 2 2 2 15 Bichena 3 2 1 3 Susceptible check 4 4 4 4 Bread wheat 1 Kubsa 2 4 2 3 2 Wabe 3 3 0 1 3 Galama 2 2 2 2 4 Tuse 3 3 0 1 5 Katar 3 2 2 ; 6 Shina 3 3 0 2 7 Hawi 0 ; 0 0 8 Simba 0 ; 0 0 9 Wetera 0 0 0 ; 10 HAR 2192 1 2 0 2 11 HAR 2508 3 1 ; 1 12 F-H-6-1-7 1 3 ; 1 13 Abola ; 2 0 0 14 Tura 2 1 0 ; Susceptible check 4 4 4 4

PAGE 21

Table 2. Estimated irrigation water requiremen ts (IWR) of wheat for the two seasons in Dongola.* Month ETp (mm/day) kc ETcrop (mm/day) IWR (mm/day) First season (2001) January 3.50 0.70 2.45 3.50 February 4.48 1.2 5.38 7.70 March 6.70 1.2 8.04 11.4 April 8.69 0.70 6.08 8.70 Second season (2001-2002) December 3.64 0.60 2.18 3.10 January 3.50 0.70 2.45 3.50 February 4.48 1.10 4.93 7.00 March 6.70 1.10 7.37 10.5 ETp = Potential evapotranspiration estimated by Jensen and Haise equation, kc = crop factor, ETcrop = Crop (actual) evapotranspiration Results and Discussion The results of the two seasons showed that irrigation per se caused significant salt leaching. Furthermore, salt leaching was not significantly affected by irrigation frequency (IF), farmyard manure (FYM) or by their interaction. Thus, the ECe data of each season were averaged over the three levels of FYM and plo tted to reflect the main effects of irrigation frequency (Fig.1a and Fig.1c), and averaged over the three levels of irrigation frequency and plotted to reflect the main effects of FYM (Fig.1 b and Fig.1d). In general, the effectiveness of salt leaching decreased with increase in the depth of the soil layer. It may be cautioned that this data was collected at harvest. It indicates the overall effect of treatments at harvest, but it does not reflect the salt distributio n between irrigation intervals. For the first season, the ECe redistribution profiles depicted a top desalinized zone (0 60 cm) and a salt accumulation zone (60-100 cm). The main effect data of IF showed that F1, F2 and F3 reduced the initial ECe of the 0 20 cm by 89, 90 and 79 %, respectively (Fig. 1a). The reductions of the initial ECe (20-40 cm) by the same treatments in sequence were 56, 50 and 59 %, respectively. The reductions of the initial ECe (40-60 cm) by the same treatments in sequence were 22, 36 and 19 %, respectiv ely. Irrigation with good quality Nile water (Mustafa, 1973) caused desalinization of these layers It is evident that the efficiency of salt leaching decreased with increasing soil depth. This may be attributed partly to the gradual decrease of water available for salt leaching and consequent decrease in downward water movement with increase in soil depth, and partly due to the rise of salt by capillarity during the drying cycle. The initial ECe (0-20 cm) was reduced by 85 84 and 88% due to the application of M0, M1 and M3, respectively (Fig.1 b). The same treatm ents in sequence reduced the initial ECe (20-40 cm) by 53, 58 and 54 %, respectively. At the 40 cm depth, the initial ECe was reduced by 28, 26 and 23, respectively. The reductions of the initial ECe in the successive depths due to FYM were of nearly similar order of magnitude as that due to IF. The effect is conceivably an indirect effect of irrigation. FYM is expected to promote water movement, enhance root development and ramification, a nd thereby enhance salt leaching. However, its effect was not significant. Although, the FYM was washed before its application, its residual salinity limited its effectiveness in salt le aching. Furthermore, vaporization during composting reduced its potential as a nitrogen source.

PAGE 22

For the second season, irrigation reduced the sa linity throughout the profile (0 100 cm) (Fig.1c and Fig.1d). Treatments F1, F2 and F3 reduced the initial ECe (0-20 cm) by 59, 80 or 84 %, respectively. M0, M1 and M3 reduced the initial ECe (0-20 cm) by59, 81 and 84 %, respectively. It seems that the reduction of salin ity was due mainly to irrigation and not to the application of farmyard manure. This is becau se the top layer is permeable and did not need FYM to promote water movement. It is evident that salt leaching was greater in the first season because of the initial higher salt level in th e virgin uncultivated soil. However, trends of salt leaching were similar. Fig. 1a indicates that F2 was relatively more efficient in salt leaching than F1 or F3, since it desalinized the top 70 cm whereas F1 and F3 desalinized the top 60 cm. However, the effect was not significant. The second season data showed that irrigati on reduced the initial ECe throughout the soil profile (0-100 cm). This was a cumulative effect since the experimental plots of the first season were irrigated in the second season. Treatments F1, F2 and F3 reduced the initial ECe (0-20 cm) by 59, 80 and 84 %, respectively. M0, M1 and M3 reduced the initial ECe at the same depth by 59, 81 and 84%, respectively. The results indicate that the effect of FYM was an indirect effect of IF. FYM did not enhan ce salt leaching by water because the inherent salt content of the manure was not reduced greatly and it was not incorporated in the soil. Furthermore, salt leaching indicated, as percentage decrease of the initial ECe, was greater in the first season. This is because the initial ECe in the first season was greater than in the second season. SAR redistribution The main effects of treatments on SAR are presented on Figs. 2a, b, c, d. Irrigation per se significantly reduced the initial SAR in the top 50 cm in the first season and through out the whole profile in the second season. In the first season, F1 and F2 reduced the initial SAR throughout the profile; however, the percenta ge decrease in SAR was minimal below 50 cm depth. F3 increased the initial SAR in the third laye r. In general, the effectiveness of dealkalization decreased with increase in soil dept h. This trend was similar to desalinization and could be explained on the same manner. Dealkalization was due to dissolution of cabearing compounds, replacement of Na+ by Ca++ ions on the exchange complex and leaching of the more mobile Na+ ions (Mustafa and Abdelmagid, 1981). It is evident that the initial SAR of the top layer was rapidly and markedly reduced merely by leaching with water, this may attributed to the low SAR of the irri gation water and decrease in soil solution SAR values during leaching because of dilution and faster leaching of sodium compared with calcium and magnesium. Similar results were obtained in India (Leffellar and Sharma, 1977; Dahiya et al ., 1981,1982). For the first season, the initial SAR of the 0-20 cm was reduced by 72, 80 and 64 % by F1, F2 and F3, respectively. The reductions in the second layer caused by the same frequencies in sequence were 35, 27 and 39 %, respectively. The initial SAR values of the third layer were reduced by 6 and 7 and increased by 12 % by the same frequencies in sequence. The differences between treatments were not significant. The effect of FYM on SAR redistribution profile was qualitatively similar to that of irrigation frequency. The trend of FYM effect on SAR profile was similar to that of salt leaching and it may be explained in the same manner.

PAGE 23

Fig. (1a) Mean electrical conductivity (dS/m) profile as affected by irrigation frequency (days) at the end of the first season (Dec. 2000April 2001) 20 40 60 80 100 05101520253035404550 Soil depth, cm ECe, dS/m 0510152025303540Soil depth, cmECe, dS/m20 40 60 80 100 Fig. (1b) Mean electrical conductivity (dS/m) profile as affected by farm yard manure (ton) at the end of the first season (Dec. 2000April 2001) 7 21 14 Initial Initial M0 M1 M2

PAGE 24

051015202530Soil depth, cm ECe, dS/m20 40 60 80 100 F1 Fig. (1c) Mean electrical conductivity (dS/m) profile as affected by irrigation frequency (days) at the end of the second se ason (Dec. 2001April 2002) 0510152025303540 Soil depth, cm ECe, dS/m20 40 60 80 Fig. (1d) Mean electrical conductivity (dS/m) profile as affected by farm yard manure (ton) at the end of the second season (Dec. 2001April 2002) 100 Initial F3 F2 Initial M0 M1 M2

PAGE 25

010203040506070 Soil depth, cm SAR20 40 60 80 100 Fig. (2a) Mean sodium adsorption ratio (SAR) profile as affected by irrigation frequency (days) at the end of the first season (Dec. 2000April 2001) 010203040506070 Soil depth, cm SAR20 40 60 80 Fig. (2b) Mean sodium adsorption ratio (SAR) profile as affected by farm yard manure (ton) at the end of the first season (Dec. 2000April 2001) Initial M0 M1 M2 100 Initial 7 21 14

PAGE 26

20 0102030405060 Soil depth, cm SAR40 60 80 100 Initial F3 F1 F2 Fig. (2c) Mean sodium adsorption ratio (SAR) profile as affected by irrigation frequency (days) at the end of the second season (Dec. 2001April 2002) 0102030405060 Soil depth, cm SAR20 40 60 80 100 Initial M0 M1 M2 Fig. (2d) Mean sodium adsorption ratio (SAR) profile as affected by farm yard manure (ton) at the end of the second se ason (Dec. 2001April 2002)

PAGE 27

Yield components and total grain yield Since irrigation had the predominant impact on growth and yield of wheat, only the main effects of irrigation frequency are reported in Tabl e 3. In general, all variables increased with increase in FYM, but the effect in many cases wa s not significant. In the first seasons, LAI, head length, number of grains per head, and to tal biomass increased with increase of irrigation frequency; but the difference was only significant between F1 and F3;. Plant height, 1000 grains weight total grain yield and water u se efficiency increased si gnificantly with increase of irrigation frequency. The number of heads/m2 increased with increase of irrigation frequency but the effect was not significant. In all cases, F1 was the superior treatment. In the second season, all forage and grain yields and their components increased with increase in irrigation frequency but the effects were only significant in the case of LAI and number of grains per head. For the first season, the impact of irriga tion frequency and FYM on grain yield, and consumptive water use efficiency, which were significant, is presented in Table 4. It is evident that treatment F1M2 was the superior treatment and treatment F3M0 was the inferior treatment. In comparison to the inferior treatment, the total grain yield was significantly increased by 170 % and the CWUE was increased by 100 %. Increasing the irrigation interval would prolong both osmotic and water str ess, reduce cell elongation and decrease plant growth and grain yield (Heyn, 1940; Mustaf a and Abdelmagid, 1981). Irrigating weekly will increase the water potential between irrigation and alleviate both stresses. Treatment F1M2 was also superior in the second season, but the e ffect was not significant. In comparison to the inferior treatment, the total grain yield was significantly increased by 46 % and the CWUE was increased by 31 %. The beneficial impact of the superior treatment was reduced in the second season because was relatively ame liorated in the first season. It is recommended that in these higher terrace saline-sodic sandy loam Aridisols, wheat should irrigated weekly at a rate equivalent to the potential evapotranspiration. Farm Yard manure may be washed first properly composted a nd incorporated into the plough layer after the first month. Table 3. Main irrigation effect (means at harvest for the three levels of F.Y.M) on the growth and yield of wheat in two successive season (2000/012001/02) Irrigation frequency /day Type of data F1 F2 F3 LSD First season (April 2001) LAI 1.6 1.2 1.0 LSD 0.0108 = 0.30 Plant height (cm) 65.7 56.1 49.5 LSD 0.0022 = 5.06 Head length (cm) 6.8 6.2 6.0 LSD 0.0042 = 0.41 No. of heads /m2 378.6 339.7 307.1 NS No. of grains / head 30.6 25.6 20.3 LSD 0.0185 = 5.63 1000 grains weight (gm) 33.3 30.7 29.7 LSD 0.0002 = 0.64 Total biomass ton/fed 2.9392 2.2050 1.5563 LSD 0.0225 = 0.8073 Total grain yield ton/fed 1.0834 0.8220 0.5220 LSD 0.0045 = 0.20963 Water use efficiency kg/m3 0.301 0.228 0.145 LSD 0.0046 = 0.06 Second season (April 2002) LAI 1.5 1.4 1.1 LSD 0.041 = 0.29 Plant height (cm) 61.7 58.8 56.4 NS Head length (cm) 6.30 6.00 5.70 NS No. of heads /m2 422.9 398.1 391.6 NS No. of grains / head 31.3 26.0 24.1 LSD 0.0076 = 3.2 1000 grains weight (gm) 34.0 33.1 31.9 NS Total biomass ton/fed 2.3092 2.0549 1.9795 NS Total grain yield ton/fed 0.9303 0.7966 0.7060 NS Water use efficiency Kg/m3 0.285 0.244 0.217 NS F1, F2 and F3 mean 7, 14 and 21 days respectively.

PAGE 28

Table 4.Mean grain yield (ton/feddan) and cons umptive water use efficiency of wheat (kg/m3) as affected by irrigation frequency and farm yard manure.* Frequency (day) M0 M1 M2 Mean Grain yield 7 0.828 1.121 1.302 1.083 a 14 0.622 0.929 0.915 0.822 b 21 0.482 0.511 0.573 0.522 c Mean 0.644 a 0.854 b 0.930 b Water use efficiency 7 0.230 0.311 0.361 0.301 a 14 0.172 0.258 0.254 0.228 b 21 0.134 0.141 0.159 0.145 c Mean 0.179 a 0.237 b 0.258 b M0 = Zero farm yard manure, M1= 4.8 ton / feddan farmyard manure, M2 = 9.7 ton/feddan. Main frequency effect (F) LSD 0.0045 = 0.20963 Main farmyard manure effect (M) LSD 0.0003 = 0.11100 Main frequency effect (F) LSD 0.0046 = 0.06 Main farmyard manure (M) LSD 0.0003 = 0.03 References Black, C. A. (1962). Methods of Soil Analysis. Part I. Agronomy Monograph No. 9. American Society of Agronomy, Inc., Publisher, Madison, Wisconsin, USA. Dahiya, I.S., Malik, R.S. and Singh, M. (1981). Field studies on leach ing behaviour of a highly saline sodic soil under two methods of water application in the presence of crops. J. of Agric. Sci. Cambridge. 97: 383-389. Dahiya, I.S., Malik, R.S. and Singh, M. (1982). Reclaiming a saline sodic, sandy loam soil under rice production. Agricultural water management. 5: 61-72. FAO (1984). Irrigation and drainage in crop water requirements. Paper No. 24. Rome, P. 35-44. Heyn, A.N.F. (1940). The physiology of cell elongation. Bot. Rev. 6: 515. Jensen, M. E. and Haise, H. R. (1963). Estimating Evapotranspiration from Solar Radiation. Proc. Am. Soc. CN. Engr., J. 89:5-41. Leffellaar, P.A. and Sharma, R.P. (1977). Leaching of a highly saline sodic soil J. of hydrology. 32: 203-219. Mustafa, M. A. (1973). Appraisal of the water qua lity of the Blue and the White Niles for irrigation use. African Soils. 18: 113-124. Mustafa, M.A. (1986). Sa lt affected soils in the Sudan their distribution, properties and management. Reclamation and revegetation research. 5 : 115 124. Mustafa, M.A. and Abdelmagid, E.A. (1981). The effect of irrigation interval, urea N, and gypsum on salt redistribution on a highly saline-sodic montmorillonitic clay soil under forage sorghum. Soil Sci. 132 4th: 308-315. Nachtergaele, F.O.F. (1976). Studies on saline and sodic soils in Sudan. FAO/ UNDP project for strengthening the soil survey administration: SUD/71/553. Technical Bulletin N. 24. May 1976. Wad Medani, Sudan. Richards, L. A. (1954). Diagnosis and improvement of saline and alkali soils. Hand book 60., USDA. Soil Survey Staff (1996). Soil Taxonomy (key). Seventh Edition.

PAGE 29

Consumptive Use of Water and Water Use Efficiency by Wheat ( Triticum aestivum ) in Relation to Irrigation and Nitrogen Antony M. Kibe1, Subedar Singh2, and Naveen Kalra3 1Egerton University, Agronomy Department, P.O. Box 536 Njoro, Kenya ( akmwangi@yahoo.com ); 2Indian Agricultural Research Institute, Water Technology Center, 110012 N. Delhi, India; 3Indian Agricultural Research Institute, Center for Agriculture Systems Simulation, 110012 N. Delhi, India Abstract A field experiment was carried out on a sandy loam soil to study the use relationships to yields of a late sown wheat cultivar, HD-2285 under limited and adequate water and nitrogen regimes. The experiment was laid out in a split-plot design with four irrigation levels (I0, no post-sowing irrigation; I1, one irrigation at CRI; I2, two irrigations, each at CRI and flowering; I3, four irrigations each given at CRI, jointing, flowering and dough stages) in main plots and a combination of three N levels, viz. N0, N50 and N100 and two zinc levels, Z0 and Z5 in sub-plots, (subscript numbers signifying N and Zn quantities in kg/ha), in three replicates. The consumptive use of water (CU) by wheat increased with every additional irrigation level to a maximum of 328.4 mm and 301.7mm. Water use efficiency (WUE) was maximum (1.38 kg grain/m3 water use) with two irrigation treatments given at the crown root initiation (CRI) and flowering stages. This did not correspond with the above ground biomass and grain yield production, which were highest under four irrigation treatments (I3). The moisture use rate increas ed with increase in irrigation water to a maximum of 2.74 and 2.51 mm/day in the first and second seasons, respectively. Moisture extraction was ma ximum (59.4% 65.8%) from the 0-30 cm soil layer. Water use efficiency increas ed markedly with in crease in nitrogen application attaining a maximum (1.42 and 1.52 kg grain/m3 water use) under 100 kg N/ha application. Maximum WUE did not correspond to the highest grain yield. The rate of increase in both grain yield and WUE started to decline as ET further increased beyond 270 mm. Introduction The importance of irrigation water to wheat is to provide the water for enhanced transpiration. This permits higher leaf area indices and stomatal apertures and therefore, a greater rate of biomass production which often lead to higher yields (Davis, 1994). Improving water use efficiency (WUE) in crop production and sustaina ble use of water resources are the need of the day. Fully satisfying crop water requirement s may be prohibitive in terms of sustainable utilization of limited water. The solution theref ore, is to limit water application to specific stages thereby minimizing loss of yield from water stress. Thus, it is imperative to study and know which critical growth stages of wheat; a small amount of water application would results in optimal WUE and minimum loss of yield. Apart from water, other factors such as essential nutrients play important roles in optimizing growth and yield of wheat. Major wh eat growing regions of the world are poor in available nitrogen. Nitrogen fertilization is essential for obtaining reasonable yields. Improvement of plant nutrition often leads to increased leaf area index (LAI). As the LAI increases, transpiration increases and evapora tion from the soil surface declines because of shading and canopy closure. Nitrogen fertiliza tion can enhance new leaf growth (increased LAI and CGR) and delay plant senescence (increased leaf area duration), resulting in

PAGE 30

increased transpiration demand. Nutrient defi ciencies however often lead to more rapid senescence (Davis, 1994). Thus, a better unders tanding of the response and relationship of wheat growth (i.e., dry mater accumulation and LAI) to nutrient and water inputs is necessary in our assessment of a cultivar. This will also enable us make appropriate decisions for purposes of optimizing the productivity of wheat under various environments. Seasonal consumptive use of water by wh eat is universally acknowledged to increase with increasing available soil moisture. This is commensurate with the increase in growth and yield of wheat. Water use efficiency is however, known to decrease with increase in irrigation water applied beyond an optimum (Zhang, et al 1999). The quantities of moisture used and the efficiency of its utilization diffe rs however, as influenced by factors such as the environment, cultivars, duration of the cultivar s on the land and the rooting pattern of the crop. The effects of irrigation on crop producti on are usually quantified using crop production functions that relate crop yield to the amount of water applied (English and Raja, 1996). These functions are used to optimize on-farm irrigation and economic evaluation of irrigation water application. Water production functions are mathematical equations relating crop response in terms of biomass or grain yield with water availability or its uptake by the crop. These functions can be used in managing wate r resource for achieving maximum returns with minimum amount of water application as irrigation. The use of water-yield relationships has been employed in the development of models th at are used for prediction purposes. This type of work has mostly been done successfully in th e developed countries. There is a need for more work in the developing countries, where water is limited, for purposes of determining the optimum amounts of water required in produc ing appreciable and sustainable crop yields. In light of the considerations mentioned above, the present investigation was undertaken at the Water Technology Center, Indian Agricu ltural Research Institute, New Delhi during the winter seasons of 1999-2000 and 2000-2001 with the following objectives: 1. To determine the influence of irrigation, nitrogen and zinc on the growth of wheat 2. To determine the seasonal consumptive water use, water use efficiency and moisture use rate of wheat under varying irrigation and nutrient levels 3. To compute production functions for wheat under varying water supply conditions. Materials and Methods Four irrigation levels (I0, no post-sowing irrigation; I1, one irrigation at CRI; I2, two irrigations, each at CRI and flowering; I3, four irrigations each given at CRI, jointing, flowering and dough stages) were allotted to main plots and six fertilizer levels (N0Z0, no nitrogen and zinc; N0Z5, no nitrogen and 5 kg Zn/ha; N50Z0, 50 kg N/ha and no Zinc; N50Z5, 50 kg N/ha and 5 kg Zn/ha; N100Z0, 100 kg N/ha and no zinc; N100Z5, 100 kg N/ha and 5 kg Zn/ha) were allotted to the sub-plots of a split plot design replicated thrice. Dry matter production (DMP), Leaf Area Index (LAI), grain and straw yield data was recorded. Periodic soil samples from 0-120 cm soil depth were collected layer-wise to compute water use by crop as described by Dastan e (1972). Crop evapo-transpiration to yield relationships was determined through regression analysis. Soil samples were taken from plots at dept h intervals of 0-15, 15-30, 30-60, 60-90,and 90-120 cm soil profiles and dried to, respectivel y, to determine the soil moisture. Samples were taken at sowing time; 48 hours before; and af ter irrigating; and at harvest, with the help of a tube auger. Samples were dried in the oven at 1050 C for over 24hr till constant dry weight. The moisture content was measured thermo-gravimetrically. Seasonal consumptive water use (CU) and the water use efficiency (WUE) of the crop, daily moisture use rate

PAGE 31

(MUR) and soil moisture extraction pattern of the crop were computed from the periodic soil moisture content data. The seasonal consumptive water use by the crop under different treatments was worked out using the formula described by Dastane (1972): N n CU = (Ep X 0.6) + (M 1I M 2I ) x Asi x Di + ER + GWC 1 i=1 100 where CU = Consumptive use of water (mm) Ep= Pan evaporation value from the USWB cl ass A open pan evaporimeter for the period from the date of irrigation to the date of soil sampling after each irrigation (mm) A = constant used for obtaining ET value from Ep value for the given period of time M1i = Per cent soil moisture (w/w) of the ith layer of the soil at the time of sampling after each irrigation M2i = Per cent soil moisture (w/w) of the ith layer of the soil at the time of sampling before each irrigation Asi = Apparent specific gravity of the ith layer of the soil Di = Depth of ith layer of soil (mm) ER = Effective rainfall, if any during the period under consideration, mm GWC =Ground water contribution to the root z one moisture during the given period of time (mm) n = Number of soil layers N = Number of days between pre-irrigation and post-irrigation soil moisture samplings The ground water contribution (GWC) was consid ered negligible as the ground water table during crop growth periods remained below 4.0 from soil surface during in both the seasons. Crop water use efficiency was computed using the following formula and was expressed as kg grain /m3 water used. Grain yield (Economic yield) WUE = _________________________________ Consumptive Water Use (Seasonal) Leaf area index was measured using the leaf area to dry leaf weight relationship. At each dry mater sampling date, a sub sample of fresh leaves from the main sample (1m row length) was taken from one replication and leaf area was re corded with the help of LI-3100 Area Meter, LI-COR, Inc. The leaves were then dried at 60 C for about 24 hr until a constant weight was obtained. Specific leaf area (SLA) was calculated as the ratio of leaf area (cm2) to leaf weight (g). This SLA value for each treatment at each sampling time was multiplied by the corresponding dry leaf weight values, to get leaf area index. Leaf area index is expressed as total leaf area of the crop (one side only) per unit ground area occupied by the crop. The area occupied by the plants under consideration in the field was taken as 0.25 m x 1.00 m. The experimental data pertaining to each character was analyzed statistically by using Analyses of Variance technique (ANOVA) for split plot. Standard error of mean difference and significant difference (LSD) at 5% leve l of significance were worked out for each character. Pooled analysis of the two years data was done only for grain and straw yield. The

PAGE 32

computed parameters such as data on soil moisture studies were explained on the basis of comparative performance. Results and Discussion Growth of Wheat (Dry matter and Leaf area index) The four irrigation treatment applied at CRI, jointing, flowering and dough stages along with one pre-sowing irrigation increased the dry matte r production of wheat at harvest significantly over other irrigation treatments in both the 1999-00 and 2000-01 seasons. Total above ground dry matter (yields) produc tion (TDMP) at 30, 60 and 90 days after sowing was significantly more with added irrigation, compared to the no post-sowing irrigated TDMP production in second season (Tab le 1). In the first season however, only the 90 days after sowing (DAS) above ground DM showed significant differences due to irrigation treatments. This was attributed to the 68.2mm rainfall (Fig. 1) during the initial part of the first season that mitigated irrigation effects on DMP up to the 60 DAS. Similarly, application of irrigation increased the leaf ar ea index (LAI) of wheat significantly over the non-irrigated control treatment at 60 and 90 DAS, in the second year (Table 2). Table 1: Influence of I, N and Z on total dry ma tter production (g/m2), at different growth stages Treatment 30 DAS 60 DAS 90 DAS Harvest 1999-00 2000-01 1999-00 2000-01 1999-00 2000-01 1999-00 2000-01 I0 304.9 222.5 1299.2 999.4 5711.8 5192.3 7061.0 6852.5 I1 316.5 218.9 1306.5 1082.9 6966.1 6181.1 8739.2 8354.4 I2 316.6 224.4 1309.6 1139.9 7518.3 6695.2 9876.2 9553.1 I3 317.8 236.0 1308.0 1238.7 7915.8 7300.5 10454.2 10524.1 SEm+ 4.34 1.77 10.37 7.57 54.56 42.77 89.60 75.26 CD(p=0.05) NS 5.3 NS 22.8 164.3 128.9 269.9 226.8 N0 292.2 202.9 1086.3 929.5 5423.3 5174.9 7998.4 7744.8 N50 306.6 224.2 1295.2 1085.9 7098.1 6228.7 9029.8 8889.2 N100 343.1 246.4 1536.0 1330.3 8562.6 7623.2 10069.7 9829.6 SEm+ 2.11 2.2 12.34 8.64 65.24 52.89 97.7 87.2 CD(p=0.05) 15.4 6.6 37.2 26.1 196.6 159.4 300.5 262.2 Z0 313.5 225.2 1303.9 1114.6 7018.6 6331.7 8982.2 8800.3 Z5 314.5 225.7 1307.9 1115.9 7037.4 6352.8 9083.1 8842.0 SEm+ 4.17 1.79 10.05 53.26 43.18 85.4 81.4 71.2 CD(p=0.05) NS NS NS NS NS NS NS NS

PAGE 33

Fig 1:Pattern of rainfall, evaporation, average temperature and average relative humidity during Late sown wheat growth periods of 1999-00 and 2000-01 0 10 20 30 40 50 6049 50 51 52 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19Standard meteorological weeksRainfall and pan evaporation (mm)0 10 20 30 40 50 60 70 80 90Average temperature (oC) and average relative humidity (%) Rainfall Evaporation Average temperature Average humidity. 0 10 20 30 40 50 6049 50 51 52 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19Stardard meteorological weeksRainfall and pan evaporation (mm)0 10 20 30 40 50 60 70 80 90Average temperature (oC) and average relative humidity (%) Rainfall Evaporation Average temperature Average humidity.

PAGE 34

Table 2 :Influence of irrigation, Ni trogen and Zinc on the LAI of late sown wheat at 30, 60 and 90 DAS Treatment 30 DAS 60 DAS 90 DAS 1999-00 2000-01 1999-00 2000-01 1999-00 2000-01 I0 0.164 0.291 0.595 0.561 2.349 2.007 I1 0.163 0.289 0.614 0.564 3.178 2.697 I2 0.163 0.293 0.607 0.566 3.467 2.922 I3 0.162 0.292 0.613 0.576 3.603 3.11 SEm+ 0.01 0.05 0.007 0.008 0.036 0.028 CD(p=0.05) NS NS NS NS 0.01 0.09 N0 0.14 0.226 0.519 0.493 2.696 2.314 N50 0.159 0.278 0.616 0.525 3.272 2.77 N100 0.19 0.368 0.687 0.682 3.48 2.963 SEm+ 0.001 0.005 0.008 0.008 0.047 0.034 CD(p=0.05) 0.003 0.014 0.024 0.025 0.141 0.103 Z0 0.163 0.291 0.606 0.566 3.142 2.676 Z5 0.163 0.291 0.608 0.568 3.157 2.693 SEm+ 0.001 0.004 0.006 0.007 0.038 0.028 CD(p=0.05) NS NS NS NS NS NS Grain and Straw yield of Wheat The maximum average grain yield of 4.0 tons in first season and 3.9 tons per hectare (t/ha) in the second season was obtained with the four irrigation treatment (I3) given at CRI, jointing, flowering and dough stages of the crop (Fig. 2). Minimum yields of 2.3 t/ha in both the seasons were obtained with no post sowing irrigation (I0). The per cent increase in grain yield due to one (I1), two (I2) and four irrigation levels over the no irrigation treatment was 36.2, 58.0 and 71.5 in 1999-2000, and the corresponding per cent increase in 2000-2001 was 32.0, 57.8 and 73.9, respectively. Each and every ad dition of either irrigation or nitrogen resulted in a progressive and significant increase in grain yield up to I3N100 level in first and second year of experimentation, respectively (Table 3). The minimum grain yield of 2.00 t/ha and 1.96 t/ha were obtained with I0N0 in the respective seasons. Straw yield increased significantly with add ition of irrigation attaining a maximum of 6.4 t/ha with four irrigations in 1999-2000 and 6.6 t/ha in 2000-2001 (Fig. 2). Application of one and two irrigations, resulted in a percentage in crease of the straw yield by, 17.6 and 30.9 in the first season and 17.0 and 30.4 increase in the second season, over the control (I0), respectively. Table 3:Interactive effects of I and N on the grain yield (kg/ha) of late sown wheat at harvest IxN effects on grain yield I0 I1 I2 I3 I0 I1 I2 I3 N0 2038.8 2649.9 3013.0 3301.2 1959.8 2407.8 288505 3271.7 N50 2337.1 3199.9 3713.1 4108.7 2259.8 3011.0 3587.9 4042.1 N100 2637.1 3699.9 4359.7 4615.4 2535.1 3499.9 4186.9 4436.1 N within I I within N N within I I within N SEm+ 144.3 149.52 130.49 136.2 CD(p=0.05) 298.1 302.2 268.9 275.2 Fig. 2. Straw yield increase with additional irrigation

PAGE 35

Consumptive Use of Water (CU) by Wheat Increasing frequency of irrigation increased the CU progressively and markedly in both the seasons (Fig 3 and Table 4). The maximum consumptive use of 328.6 mm in 1999 2000 and 302.0 mm in 2000001 was attained with th e four irrigation treatments, given at CRI, jointing, flowering and dough stages of the crop growth. The minimum CU was obtained with no post sowing irrigation, i.e., control (191.5 mm and 178.2 mm) in corresponding seasons, respectively. The percent increase in CU values due to I1, I2 and I3 over I0 treatment was 21.5, 44.5, and 71.5, in 1999 2000 a nd 25.5, 39.8, and 69.3 in 2000 2001 seasons, respectively. Moisture use rate of wheat increased with the increase in the number of irrigation levels given. The increase in moisture use rate of wheat due to I1, I2 and I3 levels over the no irrigation treatment (I0), was in the order of 21.3%, 39.6%, and 68.5% in 20002001 season, respectively. Water Use Efficiency by Wheat The pooled average water use efficiency (kg grain/m3 water use) of I1 (1.35) and I2 (1.38) were higher than those of I0 (1.24) and I3 (1.27) treatments in both the seasons (Table 4). Highest WUE for the first season crop was recorded with I1 (1.37), while it was with I2 (1.42) in the second season (Fig 3 and Table 4). Minimu m water use efficiency was recorded with I0 (1.22 and 1.26 kg grain/m3 water use) in both the respective seasons. I0I1I2I3I0I1I2I3 Y 7061.0 8739.2 9876.2 10524.7 9553.1 8354.4 6852.5 10454.2 4723.4 5556.0 6180.9 6571.4 6000.0 5381.5 4601.0 6445.8 2337.6 3183.2 3695.3 3953.3 3553.1 2972.9 2251.5 4008.4 0.0 2000.0 4000.0 6000.0 8000.0 10000.0 12000.0 Irrigation levelsYield (Kg/ha) Y S T Year 2000-2001 Year 1999-2000

PAGE 36

Figure 3. Influence of irri gation, nitrogen and zinc on the consumptive use of water, moisture use rate and water use efficiency by late sown wheat

PAGE 37

Table 4:Influence of I, N and z on seasonal consumptive water use, moisture use rate and WUE of late sown wheat CU (mm) WUE(kg grain/m3 water) Moisture use rate (mm/day) 1999-00 2000-01 1999-2000 2000-2001 1999-2000 2000-2001 I0 191.52 178.20 1.22 1.26 1.60 1.49 I1 232.67 223.66 1.37 1.33 1.94 1.86 I2 276.10 249.10 1.34 1.42 2.30 2.08 I3 328.39 301.66 1.22 1.31 2.74 2.51 N0 253.14 234.36 1.09 1.12 2.11 1.95 N50 257.08 237.95 1.30 1.35 2.14 1.98 N100 261.28 242.16 1.46 1.52 2.18 2.02 Z0 257.11 238.15 1.27 1.32 2.14 1.98 Z5 257.23 238.16 1.30 1.34 2.14 1.98 Nutrient Nitrogen and Zinc Effects on Water Use Application of nitrogen did not enhance the CU appreciably as in case of irrigation application. With application of nitrogen fe rtilizer at the rate of 50 and 100 kg /ha, the percent increase in CU of wheat over no nitrogen (N0) application was marginal i.e. 1.56 and 3.22 in the first season and 1.53 and 3.32 in the second season respectively (Table 4). Similarly Zinc application did not have much influence on the consumptive use of wheat. Application of nitrogen had very little effect on the daily moisture use rate of wheat, while zinc application had practically no effect at all, in both the cropping seasons (Table 4). Nitrogen application increased water use effi ciency of the crop progressively up to 100 kg / ha (1.46 and 1.52 kg grain/m3 water use) during both the years of experimentation. Zinc application however, induced negligible influe nce in WUE of wheat in both the seasons (Fig. 3 and Table). Nitrogen application up to 100 kg/ha did not bring any noticeable change in seasonal consumptive use and daily moisture use rate of the crop (Table 4). Water use efficiency increased progressively with N application and the mean maximum WUE recorded with 100 kg N/ha was 1.46 and 1.52 kg grain/m3 water use, in first and second seasons, respectively. Moisture Extraction The influence of irrigation and fertility (N a nd Z) interactive effects on the soil moisture extraction patterns is depicted in Table 5. Maximum (58.6 % to 66.8 %) soil moisture was extracted from within the 030cm soil profile layer with the highest being under the four irrigation regime (64.7 to 66.8 %) and low est under the no post-sowing irrigation treatment (58.2 to 60.1 %). This trend of declining so il moisture extraction percentages with increasing irrigation frequency was evident at the 60-90 cm and 9020 cm depth soil layer profiles also. The ranges were 9.6 11.7 % and 6.4 8.0 % in the 60 cm and 9020 cm soil profiles, respectively.

PAGE 38

Application of nitrogen induced the plants to extract proportionately greater amount of soil moisture from 60-120 cm layers as compared to the nonfertilized ones. Zinc application on the other hand, caused negligible increase in moisture extraction pattern at different soil profiles that followed the similar effect as in the case of irrigation effects. Production Function for Wheat Water yield relationships depicted as production functions are considered as useful tools in the management of water and nitrogen app lication for the purposes of optimizing crop productivity. Crop production functions that relate water-yield relationships, are mathematical equations relating crop response in terms of biomass or grain yield with water availability and its uptake by the crop. Table 5: Influence of irrigation, nitrogen and zinc on soil moisture extraction pattern (%) 1999-2000 2000-2001 Seasonal moisture extracton patern Seasonal moisture extracton patern Treatment 0-30 cm 30-60 cm 60-90 cm 90-120 cm 030 cm 30-60 cm 60-90 cm 90-120 cm Irrrigation I0 59.35 21.90 11.65 7.10 58.10 21.62 12.53 7.75 I1 61.50 20.98 11.12 6.40 60.15 21.00 12.05 6.80 I2 64.32 20.05 10.00 5.63 61.62 20.25 11.65 6.48 I3 65.80 19.30 9.58 5.32 64.68 19.72 10.13 5.47 Nitrogen N0 63.56 20.70 10.58 5.16 61.43 20.54 11.78 6.26 N50 62.94 20.53 10.60 5.94 61.13 20.63 11.60 6.65 N100 61.73 20.45 10.59 7.24 60.86 20.78 11.40 6.96 Zinc Z0 62.69 20.54 10.56 6.21 61.10 20.62 11.56 6.73 Z5 62.79 20.58 10.62 6.02 61.18 20.68 11.63 6.53

PAGE 39

Fig. 5 : Relationship of water use efficiency to evapo-transpiration Fig 6: Relationship of maximum leaf area index and total water use at anthesis in late sown wheat y = -0.0025x2 + 0.1288x 0.288. R2=0.108 0.5 1.0 1.5 2.0 10203040 Evapo-transpiration (cm)Water use efficiency (q/ha-cm) y = -0.0004x2 + 0.097x 2.71 R2 = 0.745 1.0 2.0 3.0 4.0 5060708090100110120 Total water use up to anthesis (mm)Leaf area index at anthesis

PAGE 40

Figure 7: Above ground biomass gains in various growth stages of wheat as related to water use Water use efficiency is generally consider ed as a conservative term and is expressed as the ratio of DMP or Y to water supply or wate r use, expressed in terms of evapotranspiration (ET) or transpiration (Tr) on daily or seasonal basis (Sinclair et al, 1984). Water use efficiency generally ranged from 1.22 to 1.42 kg/m3 (Table 4), with the highest WUE observed at one and two irrigation levels of the first and second season crop, respectively. The values of WUE were higher than those re ported from the Mediterranean region by Zhang and Oweis (1999) that ranged from 1.08 to 1.19 kg/m3 and close to those (1.23-1.46 kg/m3) of the Loess plateau in China (Kang et al. 2002). Regression analysis indicated a quadratic relationship between WUE and seasonal evapotranspiration (Fig. 5), which was reflected in a poor correlation of 11%. WUE reached its maximum value at an evapo-transpiration of approximately 260-270 mm then started a decrease with evapo-transpiration. Maximum WUE did not correspond to the highest grain yield (Fig 2 and 3). When ET is relatively low, water availability is the limiting factor for grain yield and an increase in ET results in significant increase in both grain yield and WUE. However, the rate of increase in both grain yield and WUE starts to decrease as ET further increases. Once WUE reaches its maximum valu e, an increase in total crop water use could still lead to a marginal increase in grain yield. Thus, WUE would decrease (Figure 5). This was in agreement with the findings of Singh (1977) and Kang et al. (2002). This means that relatively high crop yield and WUE can be main tained while substantia lly reducing irrigation volume under limited-irrigation management (Singh et al. 1991; Zhang and Oweis, 1999; Kang et al. 2002). When evaluating dry matter production, expansive growth deserves special attention, since it is the means for developing leaf area for intercepting light and carrying out photosynthesis. The sensitivity of expansive growth to small wate r deficits is marked by reduction in leaf area (Hsiao et al. 1976b; Vaux and Pruitt, 1983). Maximum LAI for wheat is generally noticed at anthesis. LAI at flowering is mainly used for forecasting wheat yields, with the prime y = 53.1*x + 726 y = 6.7*x + 698 y = 20.8*x + 553 0 2000 4000 6000 8000 5565758595105115125 Seasonal water uptake (mm)

PAGE 41

assumption that there is no stress in the subsequent stages of crop growth. The LAI estimates based on water use (Fig 8) accounted for 75% of the variations only. It can be concluded that wheat grown under lower levels of irrigation water results in lower consumptive use of water (Fig 3 and Tabl e 4). This further influences the production of less leaf area indices (Table 3 and Fig 6), that ultimately causes the decline in the total above ground biomass and grain yield production (Fig 2). Final biomass and grain yield of wheat de pends upon the sensitivity of various stages towards moisture availability th at scheduling of irrigation is carried out. Dated production functions take care of the crop stage sensitivit y and thus gives better predictability for biomass and yield estimates (Kalra, 1986; Singh et al. 1987). The highest rate of biomass gains of 53.1 kg/ha-mm was obtained over the 60 0 day period, a period that falls within the maximum growth phase of wheat, followed by 28.3 and 6.7 kg/ha-mm during the 90-120 and 0-60 DAS periods (Fig 7). This is in agreement with the findings for semi arid regions of Hisar in India (Singh et al. 1987) and China (Zhang et al. 1999). Therefore, scheduling of water application ought to c onsider providing less quantities of water during the 0-60 DAS period as compared to the 90-120 and 60-90 DAS periods, respectively. This should be aimed at providing just sufficient amounts of water and nitrogen for the necessary developmental processes to take place. Further research aimed at optimizing growth and production of wheat through the application of minimum quantities of water and nitrogen, at various stages of wheat growth, is therefore recommended. References Dastane, N.G., Singh, M. Hukkeri, S.B. and Vasude van, V.k. 1970. Review of work done on water requirement of crops in India. New Bharat Prakashan Poona, India. Davis, G.J. 1994. Managing plant nutrients for optimum water use efficiency. Advances in Agronomy 53 : 85-110. English, M., Raja, S.N. (1997). Perspectives on deficit irrigation Agricultural Water Management 32 (1) : 1-14. Hsiao, T.C., Feres, E., Acevedo, E. and Henderson, D.W. 1976b. Water stress and dynamics of growth and yield of crop plants. Ecological Studies, 19: 281-305. Jackson, M. L. 1958. Soil Chemical analysis. Prentice Hall of India. (Pvt.) Ltd. New Delhi 1973, India. Kalra, N. 1986. Evakuation of soil water status, plant growt and canopy environment in relation to variable water supply to wheat. P.hD. Thes es. P.G. School, I.A.R.I., New Delhi: 103pp. Kang, S.Z. and Dang, Y.H. 1987. Research on crop water production functions and optimal irrigation scheduling. Water Resource Hydraulic Eng 1: 1-12. Khang, S., Zhang, L., Liang, Y., Hu, X., Cai, H. an d Gu, B. 2002. Effects of limited irrigation on yield and water use efficiency of winter wh eat in the Loess plateau of China. Agric. Water Manage 55 : 203-216. Novoa, R and Loomis, S.S. 1981. Nitrogen in plant production. Plant and Soil 58 (1) : 177-204. Sinclair, T. R., Tanner, C. B. and Bennet, J. M. 1984. Water use efficiency in crop production. Bioscience 34 : 4640. Singh N.T., Vig A.C., Rachhpal Singh, Chaudhary M. R., Singh R., (1977). Influence of different levels of irrigation and nitrogen on yield and nutrient uptake by wheat, Agronomy Journal 71 (3) : 401-404. Singh, A.K. 1991. Response of irrigated wheat ( Triticum aestivum ) to nitrogen and phosphorus in farmers fields in Ganga diara of Bihar, Indian J. Agron ., 41 (1) 157-159. Singh, P., Wolkewitz, H. and Kumar, R. 1987. Comparative performance of different crop production functions for wheat ( Triticum aestivum ). Irrigation Sci 8 (4) : 273-290. Stewart, J. I., Cuerca, R. H., Prutt, W.O., Hagon, R.M. and Tosso, J. 1977a. Determination and utilizationof water production functions for principal California crops, W-67 Calif Contri. Proj. Rep University of California, Davis. Tandon H.L.S. (Ed.) 1998. Methods of analysis of soils, plants, waters and fertilizers. Fertilizer Development and Consultations Organization, New Delhi, India. Pp. 144+vi. Vaux, H.J. and Pruitt, W.O. 1983. Crop wate r production functions. In: D. Hillel (Editor), Adv. in Irrigation 2. Academic Press, New York, pp 257-272.

PAGE 42

Watson, D.J. 1952. Adv. Agron. 4 : 101-145. Zhang H., Wang X., You M. and Liu C., (1999). Water-yield relations and water-use efficiency of winter wheat in the North China Plain, Irrigation Science 19 (1) : 37-45. Zhang H.P, and Oweis T., (1999). Water-yield relations and water-use efficiency of winter wheat in the North China Plain, Agricultural Water Management 38 (3) : 195-211.

PAGE 43

Agronomic and Economic Evaluation of Break Crops and Management Practices on Grain Yield of Wheat at Shambo, Western Oromiya, Ethiopia Tolera Abera and Mathewos Belissa Oromiya Agricultural Research Institute, Bako Agricultural Research Center P.O. Box 03, Email: akthirpha@yahoo.com Bako, Western Oromiya,Ethiopia Abstrac tSustainable production of wheat ( Triticum aestivum ) may be limited by a number of crop management practices, including crop rotation and cultural practices. A field experiment was conducted for three years at Shambo, Ethiopia, to evaluate the agronomic and economic effects of break crops and management practices on the grain yield of wheat. After a field pea br eak crop wheat grain yield was 32% higher than after a barley crop. Management practi ces gave significant effects on the mean grain yield of wheat. Wheat performed better after the field pea break crop with both farmers management practices and th e recommended practices. Both local and improved varieties gave greater yield with intensive management and fertilizer application. The highest economic benefits came from the use of field pea break crop with improved management practices and fertilizer application and the extension program under way in the area should consider this recommendation for sustainable production of wheat. Introduction Wheat is one of the major cereal crops produced in the Ethiopian highlands (1,500 to 3,000 m.a.s.l) (Hailu et al. 1991), and in the highlands of Shambo, wheat ranks fourth among the cultivated crops in terms of production (Asfaw et al., 1997). Both biotic and abiotic factors limit agricultural productivity in the country. Hu man degradation on the natural resource base is generally very swift, but interventions to amend the decline are very slow. Lal (1989) indicated that conservation tillage and crop ro tation are considered the major means of sustaining agricultural productivity at the globa l scale. Soil fertility can, in many cases, be maintained through the combined use of su itable legumes in a suitable crop rotation and modern artificial fertilizers capable of correcting nutrient deficiencies (Whyte et al., 1969). Practical cropping systems options with appr opriate management practices for wheat in western parts of Ethiopia have not been clearly identified. However, making use of their indigenous knowledge, most farmers practice crop rotation based on short-term agronomic benefits break crops, and most studies in the past focused on the short-term agronomic benefits of break crops for wheat production (Hailu et al. 1989). The application of chemical fertilizer has been recommended for wheat production in eastern and southern Africa (ESA) without c onsidering the sustainability of continuous application (Tanner 1997). Demonstration and extensi on of fertilizer and improved seeds along with all the recommended practices were started in Ethiopia with the introduction of the Minimum Package Program (MPP) (1967 78). In this Program, based on multi-locational trials and demonstration resu lts, "blanket fertilizer recomme ndations" were made for major cereal crops in the country. The "blanket" fer tilizer rate recommendations of EPID during the 1970s are still used for wheat production (Asnakew et al. 1991), and there are no reports available on the long-term effects of these recommended fertilizer rates on wheat production in the Shambo highlands. However, increasing pr oductivity and system sustainability through

PAGE 44

crop rotation has been suggested to be a sound ma nagement options for small-scale farmers in the Ethiopian highlands. Hailu et al. (1989) reported that faba bean as a break crop increased wheat grain yield by 1,100 kg ha-1 or 69 % cf. the yield of second-year continuously cropped wheat, while the study of Asefa et al. (1992) indicated that a faba bean crop increased the yield of the following wheat crop by 982 kg ha-1, or 44 %, compared to that of continuously cropped wheat. However, studies have not been conducted in the highlands of Shambo in relation to these agronomic interventions. Furthermore, the agronomic results of "multioptional and local input inclusive trials" are important to obtain useful information on break crops for wheat production in the area. Therefore, the objective of this study was to evaluate the effects of pulse and cereal break crops along with management practices on wheat productivity in the Shambo highlands. Materials and Methods The experiment was conducted during the1997, 1998 and 1999 cropping seasons at Shambo: 9o34'N latitude and 37o06'E longitude at an altitude of 2400 meters above sea level. Mean annual rainfall is 1,695 mm (NMSA, 2003). It has a cool humid climate with the mean minimum, mean maximum, and average air temperatures of 8.15oC, 15.72oC, and 11.94oC, respectively. The soils are nitisols. The experiment was laid out as split plot in a randomized complete block design with precursor crops as main plots and management practices as sub-plots. The precursor crops were field pea ( Pisum sativum) variety (G-22763-2c) and barley ( Hordeum vulgar) variety (Shege) and were sown in the 1997 cropping season. Wheat was sown on these plots in six management sub-treatments in the 1998 and 1999 seasons The wheat management practices included wheat variety, cultural practices and fertilizer level, arranged as follows: Farmer's variety and farmer's traditiona l practices without fertilizer (FVFP-FE) Improved variety and farmer's traditi onal practices without fertilizer (IVFP-FE) Farmer's variety with improved practices without chemical fertilizer (FVIP-FE) Improved variety with improved agronomic prac tices without chemical fertilizer (IVIP-FE) Farmer's variety with all impr oved agronomic practices with chemical fertilizer (FVIP + FE) Improved variety with all impr oved agronomic practices with chemical fertilize (IVIP + FE). The wheat varieties used were the local variety Molgo" (awnless wheat) and HAR-604 (Galama), an improved variety. The improved cultural practices were the recommendation practices as for wheat. For the improved cultural practices, the herbicide 2,4-D was applied 30 to 45 days after planting to control weeds while in plots with the farmer's cultural practices, hand weeding was done once at 25 days after sowing. The recommended fertilizer rates of 100 kg ha-1 each for DAP and Urea were applied at planting n the plots receiving fertilizers. The plot size was 10 m x 10 m. The seed rates were 150 kg ha-1 for field pea, barley and the improved wheat variety, whereas the normal farmer practice of 160 kg ha-1 was used for the local variety. Sowing dates were between mid June and early July. The data were analyzed using MSTAT-C Computer Software Program. Selected orthogonal contrasts were used to separate the e ffects of the two levels of the three factors: management practices, fertilizer effects, and variety. Economic evaluation was conducted using the partial budget, values to cost ratio (VCR) and marginal analyses methods. Wheat grain yi eld was valued at the average open market price over the last five years of Ethiopian Bihr (EB) 143.00/100 kg. Plot grain yields were reduced by 10 % to reflect actual production environments (CIMMYT, 1988). The wheat seed cost used was EB 2.40/kg for the improved variet y and EB 1.43/kg for the local variety. Urea and DAP were valued at the official prices of EB 192.00 and 256/100 kg respectively. The

PAGE 45

cost utilized for labour for weeding was EB 3.50/ day. The average rental price of a sprayer was EB 10.00 ha-1, and the cost of 2,4-D herbicide was EB 0.42/liter. Results and Discussion The yields of the precursor crops sown in th e 1997 cropping season are shown in Table 1. Table 1. Mean grain yield of break cr ops, field pea and barley in 1997. Crop Grain yield (kg ha-1)* Field pea 556 Barley 1549 Statistically significant t-value at P=0.01 With all management practices, the grain yield of wheat was significantly greater (P<0.05) following field peas than following barly in both cropping seasons (Tables 2 and 3). The mean grain yield of wheat was increased by 39% and 25% following field pea compared to barley in the 1998 and 1999 cropping seasons respectively, and on average by 32%. This benefit of the field pea break crop has also been shown by Tanner et al. (1991), who reported that wheat after barley break crop was not significantly different from continuous wheat production, and Tilahun et al. (2000). Tanner et al. (1999) showed that wheat grain yield was consistently higher in rotation with dicotyled onous crops than with cereal crops, exhibiting grain yield increments ranging from 22 to 54 %. In addition, in their studies, wheat grain yield was higher in rotation with faba bean than with rapeseed, with grain yield increases ranging from 8 to 31 %. This is likely due to the fi xing of free atmospheric nitrogen by Rhizobium bacteria in association with Table 2. Effects of break crops and management practices on grain yield of wheat (kg ha-1) at Shambo in 1998 Treatment Management practices Break crop FVFP -FE IVFP -FE FVIP -FE IVIP -FE FVFP +FE IVIP +FE Mean Barley 1093 1563 1300 1400 1590 1763 1451 Field pea 1510 2250 1650 2390 1970 2373 2024 Mean 1302 1906 1475 1895 1780 2068 1718 Yield change (%) (Pea/barley) 38 44 27 71 24 35 39 Break crop Mgmt practices Break crop x Mgmt practices LSD (5%) 206 480 NS CV % 8.25 22.94 Table 3. Effects of break crops and managemen t practices on grain yield of wheat (kg ha-1) at Shambo in 1999. Treatment Management practices Mean Break crop FVFP -FE IVFP -FE FVIP -FE IVIP -FE FVFP +FE IVIP +FE Barley 1087 1800 1400 1210 1767 1947 1535 Field pea 1567 2033 1633 2133 1933 2200 1917 Mean 1327 1917 1517 1672 1850 2073 1726 Yield change (%) (Pea/barley) 44 13 17 76 9 13 25 Break crop Mgmt practices Break crop x Mgmt practices LSD (5%) 155 NS NS CV % 6.26 25.40

PAGE 46

Table 4. Effects of break crops and management practices on mean grain yield of wheat (kg ha-1) at Shambo (combined over two years), 1998-1999. Treatment Management practices Mean Break crops FVFP -FE IVFP -FE FVIP -FE IVIP -FE FVFP +FE IVIP +FE Barley 1090 1681 1350 1305 1678 1855 1493 Field pea 1538 2142 1642 2262 1952 2287 1970 Mean 1314 1911 1496 1783 1815 2071 Yield change (%) 41 27 22 73 16 24 32 Break crop Mgmt practices Break crop x Mgmt practices LSD (5%) 83.02 345.7 NS CV % 7.33 24.19 the legume break crops, as the amount of nitroge n left in the soil by a crop may influence the yield of the crop that follows it (Martin et al., 1976). The yield increment of local variety was lesser in field pea break as compared to barley in two consecutive years as the management pr actices increased. This might be due to the more nitrogen left in the soil from legume break crop. Tanner et al (1991) reported that for most of the characteristics, the effect of fa ba bean break crop was similar to that of Nfertilizer. The mean grain yield increment ranged from 24 of 71 % was observed from the FVFP+FE and IVIP-FE following field pea as compared to barely in 1998 (Table 2). Mean grain yield advantage of 9 to 76 % was recorded from FVFP+FE and IVIP-FE following field pea as compared to barely in 1998 (Table 3). The result argues that with improved agronomic managements both fields gave similar yields. Management practices had significant (P<0.05) effects for grain yield in 1998 and when combined over the two years, but differences we re not significant in 1999 (Tables 2, 3 and 4). There was no significant interaction bewteen previous crop and management practices on grain yield of wheat in either year, nor in the combined data (Tables 2, 3 and 4). Orthogonal contrasts of all three factors (variety, cultural practices and fertilizer use) showed highly significant mean grain yield di fference (P< 0.01) (Table 5). The improved variety gave a mean grain yield increases of 191 kg ha-1, the improved management practices a yield advantage of 78 kg ha-1, fertilizer application a mean wheat grain yield increase of 115 kg ha-1. Thus improved agronomic practices are very advantageous for wheat production in the Shambo highlands. Table 5. Statistical significance of orthogonal comparisons in the wheat management trial at Shambo, 1998-1999. Contrasta Mean grain yield Farmers vs. improved management practices ** Local vs. improved variety ** With fertilizer vs. without fertilizer application ** ** Significant at 1% level of probability. a See materials and methods for complete description of orthogonal contrasts used Economic analysis of the management practices indicated that the highest net benefit of EB 1554.23 ha-1 and a cost:benefit ratio of EB 1.26 prof it per unit of investment were obtained from the treatment with the improved variety, improved management practices and fertilizer application (IVIP+FE)(Table 6). The second highest net benefit of EB 1537.83 ha-1, and a cost:benefit ratio of EB 3.03 profit per unit investment, was achieved with the farmers' variety, improved agronomic management practic es without fertilizer application (FVIP-FE). However, the marginal rate of return of IV IP+FE over FVIP-FE was only 2.2%, far too low to warrant the recommendation of this treatme nt. The cost:benefit ratios of the other

PAGE 47

management practices were EB 1.29, 1.19, 1.30 and 1.08 profit per unit investment, from IVFP-FE, IVIP-FE, FVIP+FE and FVFP-FE respectively. Table 6. Partial budget and marginal rate of return (MRR) analyses for the effects of management practices on the grain yi eld of wheat at Shambo, 1998-1999. Management practices Item FVIP -FE FVFP -FE FVIP +FE IVIP -FE IVFP -FE IVIP +FE Average yield (kg ha-1) of wheat 1496 1314 1815 1783 1911 2071 Adjusted yield (kg ha-1) Wheat 1346 1183 1634 1605 1720 1864 Gross field benefit of Wheat grain 1925 1691 2336 2294 2459 2665 Average straw yield (kg ha-1) Gross field benefit of wheat straw Total field benefit (EB ha-1) 2400 120 2045 1947 97 1788 2850 143 2478 2003 100 2395 1870 94 2553 2457 123 2788 Costs that vary (EB ha-1) Wheat seed cost (EB ha-1 Urea DAP Rental price of sprayer (EB ha-1) Herbicide cost (EB ha-1) Fertilizer application cost (EB ha-1) Total labour cost (EB ha-1) 215 10 42 241 215 645 215 192 256 10 42 11 350 360 10 42 683 360 756 360 192 256 10 42 10.5 364 Total costs that vary (EB ha-1) 508 860 1075 1095 1117 1234 Net benefit 1538 929D 1403D 1300 D 1436D 1554 Value to cost ratio 3.03 1.08 1.30 1.19 1.29 1.26 Marginal rate of return (MRR) 2.20% Note: D= dominated treatment, Urea= EB 1.92/kg, DAP= EB 2.56/kg, Labour cost =EB 3.5/day, 2,4-D= EB 0.42/litter, Grain price= EB 1.37/kg, Local seed = EB 1.43/kg, Improved seed= EB 2.40/kg, Straw cost= EB 0.05/kg, Rental price of sprayer= EB 10.00 ha-1, Fertilizer application cost= EB 10.50 ha-1, Yield was down adjusted with 10 % coefficient It is noteworthy that the farmers common practice (FVFP-FE) was dominated in the partial budget analysis, because the best treatment (FVI P-FE) gave higher net benefits with lower costs that vary. In the current st udy the low cost of seed of the fa rmers variety resulted in this variety dominating the treatments with the improve d variety. However, it should be noted that the full cost of buying new seed of the improve d variety was used, whereas this can probably be discounted as usually the farmer will use this seed for several years. The high cost of commercial fertilizer and improved seed made the improved agronomic practices for wheat production costly, and reduced prices of improve d seed and commercial fertilizer will benefit wheat producers of the area. However, with th e average current prices of inputs, use of farmers' variety with improved technology with out fertilizer or the improved variety with improved practices and fertilizer application were economically viable and profitable. Both the local and improved varieties gave better grai n yield under intensive management practices with fertilizer application (Table 4). The l east profitable management option of wheat is farmer's variety with farmer's management practices and without fertilizer application. However, the escalating price of chemical fe rtilizer and improved seed, together with the reduced market price of grain drastically decr ease the profitability of wheat production with improved packages. Conclusion A previous field pea crop significantly increased the grain yield of wheat compared to a previous barley crop. Both the local variety an d the improved variety gave the greatest yields under improved agronomic management with chem ical fertilizer application when following

PAGE 48

after a field pea break crop. The economic analysis indicates that optimum market prices should be determined by the policy makers for producers of wheat in order to get the potential benefit of the recommended package. Altern ative technologies should be developed on the integration of inorganic and organic fertilizer use for profitable wheat production. Long term investigation on the economic benefit of rotation effects with intensive management practices and the extension of the results to the farmer s is essential for sustainable wheat production in the Shambo highlands. AcknowledgmentThe authors thank Oromiya Ag ricultural Research Institute for funding the project. We are most grateful to our senior researchers Mr. Abdissa Gemeda and Mr. Abubeker Mussa for their priceless contributions to us in executing this experiment. We are also grateful to Mr. Asefa Mijena, Tesema Tesso and Yosef kenea for their assistance in carrying out the experiment and efficiently collecting the data. Bako Research Center Management is also thanked for facilitating the execution of the field trial. References Asefa Taa, Tanner, D G. and Amanuel Gorfu. 1992. Th e effects of tillage practices on bread wheat in three different cropping sequences in Ethiopia. pp. 376-386. In: Tanner, D. G. and Mwangi, W. (eds.). Seventh Regional Wheat Workshop for Eastern, Central and Southern Africa. CIMMYT, Nakuru, Kenya. Asfaw Negassa, Abdissa Gemeda, Tesfaye Kumsa an d Gemechu Gedeno. 1997. Agro ecological and Socioeconomical Circumstances of farmers in East Wallagga Zone of Oromiya Region, Ethiopia. Research Report No. 32. Institute of Agricultural Research, Addis Ababa, Ethiopia. Asnakew Woldeab, Tekaligne Mamo, Mengesha Be kele and Tefera Ajem a. 1991. Soil fertility management studies on wheat in Ethiopia. pp.137-172. In: Hailu Gebre-Mariam, Tanner, D.G. and Mengistu Hulluka (eds.). Wheat Research in Ethiopia: A Historical Perspective. Addis Ababa, IAR/CIMMYT. CIMMYT. 1998. From Agronomic Data to Farmer Recommendations. An Economics Training Manual. Completely Revised Edition. CIMMYT, Mexico, D.F., Mexico. 79 pp Hailu Gebre, Amsal Tarekegne and Endale Asmare. 1989. Beneficial break crops for wheat production. Ethiopian Journal of Agricultural Sciences Vol. 11: 15-24. Hailu Gebre-Mariam, Tanner, D.G. and Mengistu Hulluka. 1991 Wheat Research in Ethiopia: A Historical Perspective. A ddis Ababa, IAR/CIMMYT. Martin, J. H., Leonard, W. H., and Stamp, D. L. 1976. Principles of Field Crop Production (3rd ed.). Collier Macmillan Publishing. London, UK. Lal, R. 1989. Conservation tillage for sustainable agriculture: Tropics versus temperate environment. Advances in Agronomy 42:85-197. NMSA (National Meteorological Service Agency). 2003. Meteorological data of Shambo area for 1969-2002. NMSA, Addis Ababa, Ethiopia. Tanner, D. G., Amanuel Gorfu and Kassahun Zewdie. 1991. Wheat agronomy research in Ethiopia. pp. 95-135. In : Hailu Gebre-Mariam, Tanner, D.G. and Mengistu Hulluka (eds.). Wheat Research in Ethiopia: A Historical Perspectiv e. Addis Ababa: IAR/CIMMYT. Tanner, D.G. 1997. Sustainable wheat production: gl obal perspectives and local initiatives. pp. 10-41 In: Woldeyesus Sinebo (ed.). Crop Management Research for Sustainable Production: Status and Potentials. Proceedings of the Second An nual Conference of the Agronomy and Crop Physiology Society of Ethiopia. ACPSE, Addis Ababa, Ethiopia. Tanner, D.G., Verkuij, H., Asef a Taa and Regassa Ensermu. 1999. An agronomic and economic analysis of long-term wheat based crop rotation trial in Ethiopia. pp. 213-248. In: The Tenth Regional Wheat Workshop for Eastern, Centeral and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. Tilahun Geleto, Kedir Nefo and Feyissa Tadesse. 2000. Crop rotation effects on grain yield and yield components of bread wheat in the Bale highlands of southeastern Ethiopia. pp. 316-324. In: The Eleventh Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. Whyte, R.O., Neisser, G. N. and Trumble, H.C. 1969. Legumes in Agriculture. Food and Agricultural Organization of the United Nations, Rome, Italy.

PAGE 49

Response of Bread Wheat to Nitrogen and Phosphorous Fertilizers at Different Agroecologies of Northwestern Ethiopia Minale Liben, Alemayehu Assefa, Tilahun Tadesse and Abreham Mariye Adet Agricultural Research Center, P. O. Box 08, Bahir Dar, Ethiopia Abstract Wheat is one of the major cereal cr ops grown in northwestern Ethiopia, though its productivity in farmers fields is very low. In northwestern Ethiopia, N and P deficiencies are among the major biophysical constraints. Wheat fertilizer response trials were conducted at different agro-ecologies on farmers field during 1999, 2000 and 2001. The objective of the trial was to determine economically optimum rates of N and P fertilizer for the high-yielding semi-dwarf bread wheat cultivars HAR 1868 and HAR 604 at Farta and Laie-Gaient areas, respectively. Four rates of N (0, 46, 92 & 138 kg/ha for Farta where as 0, 41, 82 & 123 kg/ha for Laie-Gaient) and four rates of P2O5 (0, 23, 46 & 69 kg/ha for both localities) were laid out in a factorial arrangement in RCB design with three replications. The results indicated highly significant differences on grain yield, plant height, tiller and spike density and thousand-kernel weight amongst the fertilizer rates in the two locations. There was a linear increase in grain yield of bread whea t as the rates of N and P increased. The increase in grain yield is significant for nitrogen, P and N by P interaction. The highest grain yield, 3025 kg/ha at Farta a nd 2610 kg/ha at Laie-Gaient were obtained at the highest fertilizer rates of 138/69 kg N/ P2O5 ha-1 and 123/69 kg N/ P2O5 ha-1, respectively. Fertilizer rates of 138/46 kg N/P2O5 ha-1 at Farta and 123/46 kg N/ P2O5 ha-1 at Laie-Gaient were found economically feasible with net benefit (NB) of Ethiopian Birr 3538 and 3062, respectively. Introduction Wheat ranks fifth in production and area and f ourth in mean-yield among the principal cereal crops grown in the country. The national mean yield is low ranging from about 1.1 t/ha for peasant farmers to about 2 t/ha for state farms (Hailu et al ., 1991). Soil degradation is a serious problem in many parts of the country, largely as a result of mismanagement of natural resources. Erosion and farmers practice of continuous cultivation without fallow aggravate the soil fertility problem. Northwestern Ethiopia is accompanied with diversified agro ecologies. Poor soil fertility particularly N and P deficiencies are among the major biophysical constraints in those agro ecologies (Hailu et al. 1991, UNDP 1996, Aleligne et al. 1992, ANRS BOPED 2000, Aleligne & Regas sa 1992). Previous experiments conducted at the different agro ecologies of the region indicat ed variations in yield responses of the wheat crop to fertilizer application. Therefore, this wheat fertilizer response trial was conducted at the two agro ecologically different localities on farmers field during 1999, 2000 and 2001. The objective was to determine economically optimum rates of N and P fertilizer for the highyielding semi-dwarf bread wheat cultivars HAR 1868 and HAR 604 at Farta and Laie-Gaient areas, respectively.

PAGE 50

Materials and Methods The experiment was carried out on luvisols at Farta and Laie-Gaient region on farmers fields for three consecutive years (1999-2001) on a total of five sites at Farta and seven sites at LaieGaient. Laie-Gaient and Farta at an altitude of 2860 and 2630 m.a.s.l., respectively are located on the same zone (south Gondar) of nor thwestern Ethiopia but represent different agro ecologies. Laie-Gaient is characterized by erra tic rainfall, short rainy season (late onset and early cease) and low temperature compared to Farta area. The soil fertility status of Farta area is total N 0.08%, available P 3.70%, OC 3.20, C/N 40 and pH of the soil 5.17. The experiment consisted of 16 factorial co mbinations of four N rates 0,46,92 and 138 kg/ha for Farta, 0,41,82 and 123 kg/ha for Laie-Gaient and four P2O5 rates (0, 23, 46 & 69 kg/ha) at both sites. The design was a complete factorial arrangement within an RCB with three replications. The improved varieties HAR -1868 (shina) at Farta and HAR-604 (Galema) at Laie-Gaient were used. Seeds were broad casted at the recommended seed rate of 175 kg/ha. DAP, urea and TSP were the sources of N and P2O5. All P and half of the N were applied at sowing, and the remaining N was top dressed at early to mid-tillering stage of the crop. The gross and net plot sizes were 20m2 and 12m2, respectively. Data were collected on grain yield, tiller and spike density at maturity, plant height and thousand-kernel weight. All data were subjected to analysis of variance (ANOVA) using the MSTATC microcomputer software. The mean grain yield data over sites for each location was adjusted down by 10% and subjected to partial budget and sensitivity analys is (CIMMYT, 1988). Total costs that varied (fertilizer cost) for each treatments was calculated and treatments were ranked in order of ascending total variable cost (TVC) and dominan ce analysis was used to eliminate those treatments costing more but producing a lower net benefit than the next lowest cost treatment. The marginal rate of return (MRR) was cal culated for each non-dominated treatment and a minimum acceptable MRR of 100% was assumed. Sensitivity analysis was made through the assumption that cost of fertilizer and price grain increased by 10%, respectively. Results and Discussion Analysis on mean grain yield of the indivi dual sites indicated significant responses for N application in all sites. At Farta, all s ites except one showed significant response to P application where as at Laie-Gaient only one site showed significant responses to P (Table 1). Fertilizer application increased grai n yield ranging from 1554 to 2602 kgha-1 at Farta and 1208 to 2447 kgha-1 at Laie-Gaient compared to the unfertilized control. Results on combined analysis indicated signifi cant response for N application in all traits except thousand kernel weight at Laie-Gaient. The response for P application was significant on grain yield, plant height and thousand kernel weight at Farta. At Laie-Gaient all traits showed significant response except thousand kern el weight. Most of the traits showed a significant responses to NxP interaction (Table 2). Generally there was linear increase in all the parameters as the N and P rates increased. This is in line with previous findings that at most sites of northwestern Ethiopia, wheat grain yield responds positively to increasing fertilizer rates (Minale et al. 1999; Ammanue l et al. 1990). The responses were relatively larger for N than P in all parameter consider ed. Biologically the highest grain yield 3052 kgha-1 at Farta and 2610 kgha-1 at Laie-Gaient were obtained at the highest fertilizer rates in both of the locations (Table 3 & 4). An increase in grain yield, 175.70 and 106.50%, over the control was recorded at Farta and Laie-Gaient areas, respectively.

PAGE 51

Table 1: Results of ANOVA and mean grain yield of bread wheat on the individual sites of the two regions Farta region Laie-Gaient region Source of variation Site-1 Site-2 Site-3 Site-4 Site-5 Site-1 Site-2 Site-3 Site-4 Site-5 Site-6 Site-7 N ** ** ** ** ** ** ** ** ** ** P2O5 ** ** NS ** NS NS NS NS NS NS N X P2O5 NS NS NS NS NS NS NS NS NS NS NS Mean yield (kgha-1) of 0/0 N/ P2O5 495 1663 1510 490 1376 1689 1311 1134 889 1483 1381 961 Highest mean yield (kgha-1) 2816 3453 3603 3092 2930 4136 3038 2342 2337 2925 2768 2218 N/ P2O5 comb. for highest yield 138/69 92/69 92/23 138/46 138/46 92/46 123/69 123/69 123/69 123/46 123/23 123/69 CV(%) 20.02 25.10 18.40 37.54 21.45 35.24 28.00 37.66 28.28 29.56 20.27 31.42 Indicate significance at the 5 &1% levels, respectively. NS Indicate non-significance Table 2: Results of combined analysis of variance over sites of the two locations Farta area Laie-Gaient area Source of variation Grain yield (kg/ha) plant height (cm) Tiller /m2 at maturity Fertile spike/m2 1000 kernel wt.(g) Grain yield (kg/ha) plant height (cm) Tiller /m2 at maturity Fertile spike/m2 1000 kernel wt.(g) Site (S) ** ** ** ** ** ** ** ** ** ** N ** ** ** ** ** ** ** NS NxS NS ** ** ** NS NS NS ** P2O5 ** ** NS NS ** ** ** NS P2O5 x S NS NS ** NS NS NS NS N x P2O5 NS * NS NS Mean value of 0/0 N/P2O5 1107 63.1 319 304 26.8 1264 71.2 245 234 41.3 Highest mean value 3052 82.3 391 381 32.8 2610 95.3 308 300 43.5 N/P2O5 comb. for highest value 138/69 138/69 138/0 138/0 138/69 123/69 123/69 123/69 123/23 41/46 Indicate significance at the 5 &1% levels, respectively. NS Indicate non-significance

PAGE 52

Table 3: Effect of nitrogen (N) and phosphorus (P2O5) on the grain yield (kg/ha) of bread wheat (Shina variety) at Farta area P2O5 levels ( kg/ha) N levels (kg/ha) 0 23 46 69 Mean 0 1107 1429 1570 1802 1477 46 1535 1832 2115 1990 1868 92 2033 2359 2696 2426 2379 138 2116 2833 3025 3052 2757 Mean 1698 2113 2352 2318 C.V(%) 24.8 N P N x P LSD (5%) 190 190 380 Table 4: Effect of N and P on the grain yield (kg/ha) of bread wheat (Galema variety) at Laie-Gaient area P2O5 rates kg/ha N rates (kg/ha) 0 23 46 69 Mean 0 1264 1103 1280 1355 1251 41 1630 1413 1754 1605 1601 82 1817 2206 2237 2297 2139 123 1890 2457 2599 2610 2389 Mean 1650 1795 1967 1967 C.V(%) 31.2 N P N x P LSD (5%) 174.8 174.8 349.6 The economic analysis indicated 138/46 kg N/P2O5 ha-1 was more profitable with net benefit (NB) of birr 3538.4 and an acceptable marginal rate of return (M RR) of 171.9% at Farta area (Table 5). The NB increased by birr 1958.6 over the unfertilized one through the application of 138/46 kg N/P2O5 ha-1. The sensitivity analysis revealed the same re commendation as to the current situations. Table 5: Results of the economic analysis of N and P2O5 fertilizers rate on grain yi eld of bread wheat at Farta area At current cost and price Scenario (cost of fertilizer & price of grain increased and decreased, respectively by 10%) N/P2O5 (kgha-1) TVC (Eth. birr) NB (Eth. birr) MRR (%) TVC (Eth. birr) NB (Eth. birr) MRR (%) 0/0 0 1579.8 0 1421.8 0/23 138.8 1900.5 231.0 152.7 1682.7 170.8 46/0 192.7 1998.7 182.4 211.9 1760.3 131.1 46/23 293.8 2321.5 319.2 323.2 2030.6 243.0 92/0 385.3 2516.0 212.5 423.9 2187.4 155.7 46/46 395.0 2624.5 1128.4 434.4 2283.0 905.3 92/23 486.5 2881.6 280.9 535.1 2496.1 211.6 92/46 587.6 3261.2 375.3 646.4 2817.5 288.9 138/23 679.1 3364.5 112.9 747.1 2892.2 74.2 138/46 780.3 3538.4 172.0 858.3 3028.5 122.5 Note : Cost of DAP = Birr 2.77/kg Cost of Urea = Birr 1.92/kg Price of wheat grain = Birr 1.58/kg

PAGE 53

The highest NB of birr 3213.2 with acceptable MRR ( 122.51%) was obtained at fertilizer rates of 138/46 kg N/P2O5 ha-1. At Laie-Gaient, 123/46 kg N/P2O5 ha-1 was more profitable with a NB of birr 3061.7 and MRR 117.5% (Table 6). The sensitivity analysis showed 123/46 kg N/P2O5 ha-1 gave the highest NB, however, the MRR is below the acceptable level. Therefore, with the assumption that price of grain and fertilizers decreased and increased, respectively by 10%, fertilizer rate of 82/23 kg N/P2O5 ha-1 will be the optimum rate to be recommended. Table 6 Results of the economic analysis of N and P2O5 fertilizers rate on grain yield of bread wheat at LaieGaient area At current cost and price Scenario (cost of fertilizer & price of grain increased and decreased, respectively by 10%) N/P2O5 (kgha-1) TVC (Eth. birr) NB (Eth. birr) MRR (%) TVC (Eth. birr) NB (Eth. birr) MRR (%) 0/0 0 1846.7 0 1662 41/0 181.5 2200.5 194.9 199.7 1944.1 141.3 82/0 363.3 2290.7 49.6 399.6 1989.0 22.5 82/23 458.6 2764.0 496.5 504.4 2395.9 388.2 123/23 640.1 2949.4 102.2 704.1 2526.4 65.3 123/46 735.6 3061.7 117.5 809.2 2608.4 78.1 Note : Cost of DAP = Birr 2.70/kg Cost of Urea = Birr 2.03/kg Price of wheat grain = Birr 1.62/kg References Aleligne Kefyalew, and Regassa Ensermu. 1992. Bahir-Dar Mixed Farming Zone: Diagnostic Survey Report. Research Report No 18. Institute of Agricultural Research, Ad dis Ababa, Ethiopia Amahara National Regional State Bureau of Planning and Economic Development (ANRS BOPED). 2000. Annual Statistical Bulletin Bahir Dar Ethiopia. CIMMYT. 1988. From agronomic data to farmer recommendations: An economics training manual. Completely revised edition. Mexico, D.F. Hailu Gebre-Mariam, Tanner, D.G., and Mengistu Hulluka, (eds.). 1991. Wheat Research in Ethiopia: A historical perspective Addis Ababa: IAR/CIMMYT. UNDP. 1996. Sustainable Agricultural and Environmental Rehabilitation Program (SAERP). House Hold Level SocioEconomic Survey of the Amhara Region. V.l Produ ced by the Cooperative Endeavors of the Amhara Regional Council. Bahir Dar, Ethiopia. Ammanuel G, D.G. Tanner, Lemma Z, Tilahun G, Zewdu Y. and Eyasu E. 1990. Derivation of Economic fertilizer recommendations for bread wheat in the Ethiopian highlands: On-Farm trial in the peasant sector. PP. 63-72. In; Tanner, D.G., M. van Ginkel, and W. Mwangi, eds. Sixth Regional Wheat Workshop for Eastern, Central and Southern Africa. Mexico, D.F; CIMMYT. Minale Liben, Alemayehu Assefa, D.G. Tanner and Tilahun Tadesse. 1999. The response of bread wheat to N and P application under improved drainage on Bichena Vertisols in northwestern Ethiopia. The Tenth Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT.

PAGE 54

Grain Yield, Water Use and Water Use Efficiency as Affected by Moisture Level Under Rain-Out Shelter P. A. Ooro 1, M.G.Kinyua1, and J.B.O. Ogola2 1KARI-NPBRC, P.O. Njoro, Tel. 254-051-61070, Kenya, paooro@yahoo.com 2Department of plant production, University of Venda, Private Bag X5050, Thohoyandou, Republic of South Africa, *Corresponding author Abstract About one-third of the developing worlds wheat area is located in environments that are regarded as marginal for wheat production, because of drought, heat, and soil problems. Nearly one-third of the area planted to bread wheat and about three-quarters of the area planted to durum wheat suffer from drought stress during the growing seasons. Despite these limitations, the worlds dry and difficult cropping environments are increasingly crucial to food security in the developing world. Gains in wheat productivity in margin al environments are important because it is unlikely that increased productivity in the favourable environments will be sufficient to meet the projected growth in demand for wheat from the present to 2020. This study was aimed at evaluating wheat genotypes water use efficiency (WUE) as affected by moisture regime. An experiment was conducted under rain shelter for two seasons (2001-2002) with six cultivars (Duma, Ngamia, Chozi, Kwale, Mbuni and Pasa) tested under two moisture regimes (High and low moisture regimes). The experiment was a randomized complete block design (RCBD) with split arrangement of the treatments. Moisture regime was assigned as the main plot and cultivars as the sub-plot. An analysis of variance was carried out on combined season using SAS computer package. The genotypes tested were significantly different in their water use efficiency (WUE) under both low and high moisture conditions. The drought-tolerant cultivars (DTC)Duma, Ngamia and Chozihad significantly higher WUE under moisture stress than the drought-susceptible cultivars (DSC)Kwale, Mbuni and Pasa. Under high moisture the WUE of the DTC was decreased by 10-19% and in the DSC, it increased by 26-29. Introduction Cereals contribute significantly to food security in Kenya (Wanjama et al, 1993). Among cereals, wheat has been ranked second to maize (KARI, 1991) and is grown on an estimated area of 105,000 ha (representing about 6% of th e total area under cereal production in Kenya) giving an annual production of about 350,000 metric tones (Ekboir, 2002). The crop is mainly grown in the Rift Valley province with Nakuru, Uasin Gishu and Narok districts contributing about 85% of the total national production (Wanjama et al, 1993). The national wheat production has varied over time. For example, in the period 1988 2000, wheat production declined annually by about 1.5% (Aquino et al. 2002). Due to the reduction in wheat area, Kenya produces only 40% of the national demand of approximately 720,000 tonnes, and importing about 60% of its wheat requirement (Aquino et al. 2002). For example, between 1997-1999, wheat imports stood at 484,900 tonnes (5.4 million 90-kg bags), costing about 5.85 billi on Kenya shillings (US$ 84 million). Kenya is bound to remain a net importer of wheat and wheat products unless domestic wheat production is significantly stepped up. Thus for wheat production to be increased to meet the increasing demand in Kenya, among other technolog ies, wheat growing has to be expanded to

PAGE 55

the marginal rainfall areas which consist of 83% of the total land area with unused arable potential of approximately 200,000 hectares (World Bank, 1989). Furthermore, it has been reported that 32% of the 99 million hectares of wheat grown in developing countries experiences varyi ng levels of drought stress (Rajaram et al., 1996). In Kenya, for example, marginal rainfall areas ha ve been able to achieve wheat yields of between one third and one half of t hose from high rainfall areas (Kinyua et al. 1989). Promotion of wheat farming in the dry lands of Kenya can be done through several ways. These include use of irrigation and/or drought resistant wheat varieties. Irrigation would require steady supply of water and due to scarcity of water in the marginal areas and large capital required, irrigation is not easily achievab le. Thus development of appropriate varieties for marginal areas may be the most effectiv e alternative for utilizing these areas for crop production. Therefore, the main objective of this study was to evaluate WUE as an indirect method for selecting drought tolerant wheat cultivars. Materials and Methods An experiment was conducted in the rain sh elter for two seasons be tween August, 2001 and May, 2002 at the National Plant Breeding Research Centre (NPBRC), Kenya (0o 20' S and 35o56'E and altitude 2185 metres) [NPBRC Mete orological Station No. 9035021, 1999]. The site lies within the Agro-ecological zone LH3 (AEZ LH3) with a bi-modal rainfall pattern (Jaetzold and Schmidt, 1983). The site receiv es an annual rainfall of about 960 mm with an average maximum and minimum temperatures of 24 o C and 8 o C, respectively (NPBRC, Njoro Meteorological station No.9035021, 1999). The soil at the is site well drained, deep to very deep, dark reddish brown, friable and smeary, silt clay, with humic topsoil classified as mollic Andosols (Jaetzold and Schmidt, 1983). The site was under fallow for two seasons before this study was carried out. The experiment was carried out under a mobile rain-out shelter (which is an open ended structure measuring 30 m long and 7 m wide c onstructed at NPBRC) to exclude rain and consequently induce drought stress. The shelte r consists of a roof mounted on wheels that allows it to roll on two parallel-elevated concre te tracks. Translucent sheets (which allow up to 90% photosynthetic photon flux density) covered the roof. The sides had metal frame, which was covered by foldable polythene sheet s, with open-end 0.5 m above the ground. A concrete barrier had also been constructed to a depth of 0. 6 m below the ground level and 0.15 m above the ground level along the two longer sides of the shade from a rail. The length of the rail was 30 m but the shelter rested on half the length. The barrier also helped to prevent the rainwater from flooding the shade. Th e mobile rainout shelter used is similar to that described by Legg et al. (1978) and Upchurch et al (1983). Six bread wheat genotypes Duma, Chozi, Ng amia, Kwale, Mbuni and Pasa were grown under two watering regimes (High and low watering regimes) for two seasons. The experiment was laid out in a randomized complete block design (RCBD) with split plot arrangement with the watering regime being the main plot and cultivar as sub-plot. Thirty-six (36) access tubes (PVC in nature) with 39.4 mm and 41.4 mm as internal and outside diameters, respectively, were installed in the field before planting. In each sub-plot, an access tube was installed in holes drilled by hand using a soil auger. The tubes were cut to a depth of 1.7 m out of which 1.1 m was driven into the holes and 0.6 m was left above the soil surface. Moisture readings were taken by use a Troxler moisture meter (Neutron probe). In the process of moisture reading a standard count was at the beginning of each moisture regime. Hydrogen ions, standard counts and moisture percentage by volume readings were taken through the access tubes at an interval of 7 days at an interval of 10 cm in the upper 40 cm

PAGE 56

and 20 cm below 40 cm. The hydrogen counts were divided by the standard counts to derive the count ratio that was later fitted in the calib ration equations, for differe nt soil levels within the profile, to calculate the actual soil water content. The seeds of the wheat varieties used in th e trial were tested for germination before planting and the seed-rate was adjusted accord ing to the germination percentage. The recommended seed-rate of 125 kg ha-1 was used. The six cultivars were planted in plots of 0.8 metres length and 2 metres wide with ten rows. The rows were 20 cm a part and plan ting was done by hand. The seeds were drilled continuously on the furrows (5 cm deep) made by hand. Triple super-phosphate (0-46-0) was applied in the furrows at 60 kg P2O5 ha-1 planting and calcium ammonium nitrate (CAN) at 40 kg N ha-1 (recommended rate) was top-dressed at ea rly tillering stage to supply Water was supplied through a drip irrigation at a constant pressure as per the treatments. One drip lateral was laid between two wheat rows. Watering was done once weekly and a day be fore watering was done moisture reading were taken by use of neutron probe. Watering was done when the moisture level had come to 50% field capacity and watering done according to treatments. The high moisture treatment received 14.5 mm and the low moisture regime received 7.5 mm once weekly. The total amount of water applied during each seas on was 217 mm and 113 mm for the high and low moisture regimes, respectively. Weed control was done by the use of Buctril Mc at the rate of 1.4 L ha-1 at the 3-4-leaf stage of crop. The chemical was used to take care of broad-leaved weeds but for the grasses, hand weeding was done. Six innermost rows were hand harvested for biomass and grain yield determination. Evapo-transpiration (ET) included both tran spiration and direct evaporation of water from the soil surface (Esc). Evapo-transpiration (ET) was de termined by using the soil water balance equation given below: ET = S + I D R (3.1) where S is the change in storage (obtained by th e difference in volumetric moisture content of the entire profile at the beginning and end of the season), I is irrigation water applied, D is the drainage and R is the run-off. Drainage a nd run-off were assumed to be negligible hence total crop water use for the whole season obtained by summing total irrigation where applicable) and changes in storage for the entire profile (Ogola, 1999). Pilbeam et al ., (1995) suggested that significant drainage is considered to have occurred where the soil water at the base of the profile was increased above its in itial value. Drip irrigation was used hence drainage was assumed to be zero and thus not included in the water balance equation. For the purposes of this study, water use effi ciency (WUE) was defined as a ratio of grain yield (GY) to evapo-transpiration (ET). Thus WUE was determined as follows: WUE = GY/ET. (3.2) where GY is the grain yield. Grain yield (GY) was determined by harvesting one metre length of the two inner-most rows. Harvesting was done by hand and threshed by the use of laboratory single head thresher and grain yield was standardized at 12.5% moisture content. Thousand-kernel weight (TKW) was determined after grain weight had been recorded from both experiments. A thousand seeds were picked randomly and weighed to determine TKW.

PAGE 57

Neutron probe calibration procedure The soil profile of the site (NPBRC field 1) was described and the soil depth of the site determined and the pumice was realized to start immediately after 60-cm depth. For calibration, two access tubes were installed each at a depth of 110 cm. One of the tubes was used to get wettest point of the calibration curve whereas the other one was for the drier points of the curve (Ooro et al., 2001). The place surrounding the tube for the wettest point was flooded with water and left to settle for 24 hours and then soil samples for grav imetric moisture determination were taken at 10, 20, 30, 40, 60, 80 and 100 cm depths. At the corresponding depths, neutron probe readings were taken for the hydrogen ion c ounts and hence volumetric water content. The same approach was used around the access tube for the lower points of the calibration curve (Manual of operation and Instruction, 1996). The data generated from the neutron probe readings (from the wet and dry calibration spots) were used to calculate count ratios (H+/Standard count) plotted on the x-axis of the calibration curves and two equations were then developed for the top soil (upper 20 cm) and below 20 cm depths. The two calibrations were developed due to the variations between the readings of depths up to 20 cm and those below 20 cm. Hydrogen ions readings up to 20 cm depth were relatively lower due to scatter effect while this effect did affect readings below 20 cm. Plotting gravimetric moisture readings on the y-axis and the count ratios on the x-axis developed the equations. The equations were as follows: 0 20 cm depth: Y = 0.0661n + 0.001615 (3.3) 30 100 cm depth: Y = 0.20960n + 0.001132 (3.4) where Y is the actual moisture (volumetric moisture), n = count ratios (H+/standard counts) [H+ was read from the neutron probe]. The data read by neutron probe in the rain -shade experiment was fitted on the calibration curve to give the actual moisture content. Data from all the parameters, combined over two seasons (2001 and 2002), were subjected to an analysis of variance (ANOV A) using SAS (SAS Users guide, 1985). Least significant difference procedure (Lsd) was used to carry out mean separations. The data was also subjected to correlation and path coefficien t analysis was done to separate the direct and indirect between the agronomic traits and water use efficiency. Results The interaction between genotype and moisture regime affected final grain yield. High watering regime increased grain yield of both the drought tolerant and drought susceptible varieties but the increase in yield due to high watering regime was greater in susceptible (110%-134%) than in tolerant (55-64%) varieti es. The drought tolerant varieties produced greater yield than the drought susceptible ones under low watering regime. Under high moisture regime, in contrast, the drought sens itive varieties produced greater yields than the tolerant ones. Significant genotypic differences in grain yield were detected. Averaged over the low watering regimes, the drought tole rant cultivars had higher yield (1274 kg ha-1) while the drought susceptible ones had re latively lower yield (972 kg ha-1). Moreover, drought tolerant varieties produced greater (by 39%) yi eld than the drought susceptible varieties under low watering regime. Averaged over all the ge notypes, high watering regime increased grain yield by 89% (from 1123.2 kg ha-1 to 2119.0 kg ha-1).

PAGE 58

There was significant effect of the interaction between moisture regime and cultivar on crop water use, such that the increase in water use due to high watering regime was greater for drought sensitive (96%) than for drought tolerant (93%) varietie s. Also, under low moisture conditions Kwale used more water (by 10%) th an all the other cultivars tested while under high moisture, Chozi had significantly lower (by 3%) water use value than the rest of the cultivars. Significant genotypic differences on wa ter use were detected. Averaged over the watering regimes, Kwale had the highest water use (139 mm) while Pasa had the lowest water use (132 mm) as compared to the other varieties tested. High watering regime increased water use (averaged over all vari eties) by 95% (from 91 to 177 mm). The interaction between variety and moisture regime affected water use efficiency (WUE) significantly (Table 1). High moisture regime decreased (by 10%-19%) WUE of drought tolerant varieties (Duma, Ngamia a nd Chozi) but increased (by 26%-29%) WUE of drought sensitive varieties (Kwale, Mbuni and Pasa). Under low moisture regime, drought tolerant varieties had higher WUE values than the drought sensitive varieties while the converse was true under high moistu re level (Table 1). No significant effects of cultivar on WUE were detected. However, high wateri ng regime increased WUE (averaged over all wheat varieties) by 4% (from 9.6 to 9.9 kg ha-1 mm-1) (Table 1). Table 1a. Grain yield, water use and water use efficiency of six wheat genotypes under two watering regimes Treatment Grain Yield Water Use Water Use Moisture regime Cultivar (kg ha-1) (mm) Efficiency (kg ha-1 mm-1) Low Duma 1213.ab 90.b 10.7ab Ngamia 1286.a 91.b 10.9a Chozi 1324.a 91.b 11.1a Kwale 1152.abc98.a 9.0b Mbuni 902.bc 88.b 7.9c Pasa 862.c 87.0b 7.7c High Duma 1886.b 177.a 8.7b Ngamia 2097.b 177. a 9.8ab Chozi 2173.ab 171. b 9.6b Kwale 2424.a 179. a 11.3a Mbuni 2114.ab 177.a 10.1ab Pasa 2020. b 177.a 9.9ab SE 11 3 0.58 F-test probabilities Irrigation (I) P<0.001 P<0.001 P<0.05 Cultivar (C) P<0.05 P<0.05 ns Interaction (CXI) P<0.05 P<0.001 P<0.001 Figures followed by the same letter within the same moisture regime in the same column are not significantly different at Lsd 5%.

PAGE 59

Table 1b. Varietal differences in yield co mponents of wheat at two moisture levels Treatment No. of Plant TKW No. of Moisture regime Cultivar tillers/plant height (cm) (g/1000seeds) seeds/spike Low Duma 4.5ab 53.b 43.b 26.c Ngamia 2.7c 52. b 39.c 40.a Chozi 3.7bc 58.a 46.a 27.bc Kwale 3.5bc 55.ab 45.ab 25.c Mbuni 4.0b 55.ab 46.a 35.abc Pasa 5.7a 58.a 43. b 36.ab High Duma 3.8bc 69.bcd 44.bc 33.b Ngamia 3.7bc 66.cd 41.d 36.ab Chozi 5.7a 81.a 49.a 28.b Kwale 4.7ab 70. bc 46.b 34.ab Mbuni 3.3c 73.bc 42.cd 35.ab Pasa 4.5abc 65.d 40.d 43.a SE 0.63 2. 1 9 F-test probabilities Irrigation (I) P<0.001 P<0.05 ns P<0.001 Cultivar (C) P<0.001 ns P<0.05 P<0.05 Interaction (CXI) P<0.05 P<0.001 P<0.001 P<0.001 Figures followed by the same letter within the same moisture regime in the same column are not significantly different at Lsd 5%. Discussion Improving drought tolerant of wheat has long been a major objective of most breeding program because water deficits during some pa rts of growing period are common in most regions of the world where wheat is produced under rain-fed conditions. Therefore in this study the cultivars tested were evaluated with an aim of understanding their reactions under low and high moisture regimes. Significant inte ractions were observed between moisture level and water use efficiency (WUE) of wheat cultivars. The WUE of drought tolerant wheat cultivars (Duma, Ngamia and Chozi) was d ecreased (by 14%) under high moisture but increased (by 30%) WUE of drought sensitive varieties (Kwale, Mbuni and Pasa). This behaviour of drought tolerant cultivars under low moisture was probably due to the fact they had less tiller number per plant compared to the susceptible varieties. Fewer tillers of drought tolerant wheat have been reported from st udies by Ehlig and Lemert (1976) and Singh et al., (1979). Low tiller number per plant under low moisture could be a mechanism that drought tolerant used reducing the demand for water. In a study involving the evaluation of the effects of moisture stress, reduced maximum tiller number of Triticum tauschii (wild relative of wheat) were reported to be strongl y reduced by low moisture. Those findings were similar to those reported for cultivated wheat (Cone et al ., 1995). The significantly high WUE of a cultivar such as Chozi might have been due to therelatively taller plants under low moisture which probably converted into higher biomass and relatively higher grain yield. Drought tole rant cultivars (Duma and Ngamia) maintained shorter plant stature under low moisture, wh ich is another mechanism of economizing on moisture hence, increased WUE. The findings of this study are in disagreement with those reported by Rezgui et al. (1999) from their study involving evaluation of 61 durum wheat

PAGE 60

genotypes in a moisture stressed environment at Mograne in Tunisia. From their study it came out that taller plants were associated with increased WUE. The thousand-kernel weight (TKW) for the drought tolerant cultivars (Chozi, Duma and Ngamia) was higher under both low and high moisture. Less tillers per plant under low moisture observed on the drought tolerant cu ltivars was boosted by the higher TKW value. The high TKW value resulted in heavy individual seeds which in turn culminated in higher grain yield which appeared to increas WUE mainly due to higher harvest index (HI). The fact that the drought tolerant cultivars (Chozi, Duma and Ngamia) were shorter in stature HI was inversely related to plant height resulting into higher WUE under both low and high moisture conditions. Ehdaie (1995) re ported higher WUE of shorter wheat cultivars under both low and high moisture regimes in a study conducte d on bread wheat to evaluate seasonal WUE and its components in Spring wheat. Lower seed number per spike was observed on drought tolerant cultivars (Duma and Chozi) a part from Ngamia under low moisture The low seed number per spike translated into increased seed size and therefore increased TKW. Rebetzke et al., (2001) reported increased kernel size due to die back of wasteful tillers in wheat under drought in Australia. A study carried out to evaluate the stability of grain yield and its components under moisture stress also revealed that the overall moisture deficit induced reduction in yield of specific wheat cultivars were largely due to simila r effects on kernels per spike (Guttieri et al., 2001). Drought tolerant cultivars (Chozi, Duma a nd Ngamia) maintained high TKW even under low moisture. Despite the fact that these cultivars had fewer tillers per plant under low moisture, they had relatively higher WUE valu e. The high WUE value may be attributed to high TKW, which helped to maintain higher grain yield resulting into increased WUE. The drought tolerant cultivars (Duma and Chozi) a part from Ngamia had relatively lower seed number per spike under low moisture. The hi gh seed number per spike translated into increased TKW which may be because the low seed number per spike resulted into less competition giving rise to increased kernel size. Rebetzke et al., (2001) reported that wheat cultivars that had tiller inhibition genes had increased kernel size. Correlation study confirmed that tiller number per plant had a high negative (P<0.05) correlation with WUE of Duma under low moisture. However, under the same moisture regime Ngamia and Chozi were not significantly correlated to tiller number per plant. This showed that tiller number controlled WUE positively on Duma but not Ngam ia and Chozi under low moisture regime. Conclusion and Recommendation Moisture affected the grain yield and water use efficiency (WUE) of the genotypes tested significantly. Drought tolerant wheat cultivars (Duma, Ngamia and Chozi) under moisture stress used moisture more efficiently than the drought susceptible cultivars (Kwale, Mbuni and Pasa). The increased WUE of drought to lerant wheat cultivars (Duma, Ngamia and Chozi) was the basis of their drought toleran ce. This has an implication on the drought tolerance selection in wheat research. In view of the aforementioned results, the unde rstanding of the traits that contributed to the increased WUE of the drought tolerant wheat genotypes under moisture stressed environment was addressed in the second part of this study. However, the results of that section of the study have not been included in this paper. References Aquino, P., Carrion, F., and Calvo, R. 2002. Selected Wheat Statistics. In Ekboir J.(ed.) 2002. CIMMYT 2000 2001 World Wheat over-view and outlook: Developing no till packages for small-scale farmers Mexico, D.F.:CIMMYT.

PAGE 61

Cone, A. E., Stafer, G.A., and Halloran, G.M. 1995. Effects of moisture stress on leaf appearance, tillering and other aspects of development in Triticum tauschii .Euphytica. 86 (1): 56-64. Ehdaie,B. 1995. Variation in water-use efficiency and its components in wheat: II. Well-watered pot experiment. Crop Science. 1995. 35 (6). 1617-1626. Ehlig, C.F., and LeMert, R.D. 1976. Water use and productivity of wheat under five irrigation treatments. Soil Sci. Soc. Am. J. 40:750-755. Ekboir, J. (ed.) .2002. CIMMYT 2000-2001World Wheat Overview and Outlook: Developing No-Till Packages for Small-Scale Farmers Mexico, DF:CIMMYT Guttieri, M.J., Stark, J.C., O'Brien, K., and Souza, E. 2001. Relative sensitivity of Spring wheat grain and quality parameters to moisture deficit. Crop Science. 41 (2): 327335. Jaetzold, R., and Schmidt, H. 1983. Farm Management Handbook of Kenya, Vol IIB, Central Kenya (Rift Valley and Central Provinces) Ministry of Agriculture in Cooperation with the German Agric. Team (GAT) of the German Agency for Technical Cooperation (GTZ). Printed by Typo-druck, Rossdorf, W. Germany. 381-416pp. KARI. 1991. KARI agricultural resear ch priorities to the year 2000. KARI, Nairobi, Kenya Kinyua, M.G., Wanjama, J.K., Kamwag a, J., and Migwi, S. 1989. The situation in wheat production in Kenya. Kenya Agricultural Research Institute (KARI) Report. Legg, B.J., Day, W., Brown, N.J., and Smith, G.J. 1 978. Small plots and automatic rain shelter. A field appraisal. Journal of Agricultural Science 91: 321 326. Manual of Operation and Instruction, 1996. Calibration of Depth moisture gauge (Model 4300), pp.C24. Troxler Electronic Laboratories, Inc. and su bsidiary Troxler International, Ltd, North Carolina, USA National Plant Breeding Research Centre (NPBRC). 1999. Weather Data. Meteorological Station No. 9035021. Ooro, P.A., Kinyua, M.G., Ogola, B.O., Owido, S.O., Kimurto, P.K., and Mwangi, S.K. 2001. Soil Profile characterization and Moisture meter (Neutron Probe) calibration. Report, National Plant Breeding Research Centre (NPBRC), Njoro. Rajaram, S., Braun, H.J., and Ginkel, M. van. 1996. CIMMYT's approach to breed for drought tolerance. Eupytica 92, (1/2), 147-153. Rebetzke, G.J., Bonnett, D.G., Richards, R.A., Condon, A.G., and Herwaarden van, A. 2001. Breeding for high water use efficiency increases grain yield. CSIRO, Plant Industry. Rezgui, S., Yahyaoui, A., Amara, H., and Daaloul, A. 1999. Relationships between water use efficiency and agronomic traits of selected durum wheat cultivars. Revue de l'INAT. 14 (1): 143-153. SAS Institute Inc. 1985. SAS users guide: Statistics, Version 5 Edition. SAS Inst. Inc. Carry, NC. Singh, N.T., Singh, R., Mahajan, P.S. and Vig, A.C. 1979. Influence of supplemenatary irrigation and pre-sowing soil water storage on wheat. Agron.J. 71:483-486. Upchurch, D.R., Ritchie, J.T., and Foale, M.A. 1983. Design of a large dual-structure rainout shelter. Agronomy Journal 75: 845 848. Wanjama, J.K., Macharia, M., Karanja, D.D., Pinto, J.M., Maina, M.P.D., Kinyua, M.G., Wanyera, R., Faraj, A., and Gitari, J.N. 1993. Strategic Pl anning For Wheat Research To The Year 2013. National Plant Breeding Research Centre, Njoro, pp.2-28. World Bank. 1989. Kenya Agricultural growth and strategy options. Unpublished Sector Report. Nairobi, Kenya.

PAGE 62

Part 2. BREEDING

PAGE 63

Current Status of Stem Rust in Wheat Production in Kenya R. Wanyera, M.G Kinyua, P. Njau, J.W. Kamundia and S.Kilonzo Kenya Agricultural Research InstituteNational Plant Breeding Research Centre P.O. Private Bag, Njoro20107, Kenya. Ema il:cepnjr@wananchi.com, or rvkepawae@wananchi.com Abstract Stem rust Puccinia graminis f. sp tritici is among the three rusts currently threatening wheat ( Triticum aestivum L) production in Kenya. A three-year survey (2002-2004) was carried out to determine the distribution of stem rust on the wheat varieties in the country, and to study the re action of wheat varieties/lines to stem rust inocula from different sources. Fifty-three (53) Stem Rust Parental Collections (SRPC) were also screened in the experi mental cage at Njoro whereas 16 commercial wheat varieties were screened in the fiel d in Njoro, Mau-Narok and Eldoret arranged in a RCBD with 3 and 2 replications respec tively. Fourteen (14), 9 and 11 sets of breeders F3, stem rust differentials from Mexico and Australia were also screened in the greenhouse. The survey results indica ted that stem rust has reched the high altitude areas where maximum and minimum temperatures range from 22 to 26o C and 6 to14oC respectively. Out of 53 SRPC screened, 79.3% were susceptible, while 20.7% showed resistance. The 16 commercial varieties were susceptible. Varieties Ngamia, Duma, Yombi, Heroe were most su sceptible with a mean score of 63%, 48%, 46%, and 43% respectively. Kenya Fa hari, K. Paka and Njrbw1 were least susceptible (5%, 4%, and 14%). The infection types of stem rust spores from different sources were different on the stem differentials from Australia. These preliminary data indicate the possibility of mo re than one stem rust race attacking the wheat crop. Joint effort is required to utilize the current biotechnology tools to identify and breed for resi stance to the new race(s). Introduction Stem rust ( Puccinia graminis f p.tritici ) is among the cereal rusts that cause severe losses throughout the world. The rust range can cause losses ranging from slight to complete destruction of a wheat crop. Losses of 50-70-% have been reported in the field. Infected grain is shriveled, light in weight and of low quality ( Stubbs et all 1986; Agrios, 1988; Wiese, 1991; Minzenmayer, 2000; Cereal Disease Labor atory, 2004). In addition to reducing grain yield, rusts lower the crops forage value a nd predispose plants to other diseases. Rusted wheat plants are less palatable and are toxic to livestock (Wiese, 1991). Puccinia graminis commonly attacks wheat; other hosts include barley rye, oats, wild barley and goat grass. Wheat production in Kenya is highly affected by stem rust and other rusts like yellow and leaf rusts. Under favorable environmenta l conditions, infection of the wheat crop with stem rust disease reduces both the quantity a nd quality of the grain (KARI annual reports 2000, 2001, 2002, and 2003). At the beginning of the wheat breeding efforts in Kenya in 1927 up to the early 1980s, stem rust was the most ser ious of the three rusts. This was given a high priority by the breeding team and due to wide resistance breeding, this problem disappeared for some time and subsequently many resistan t commercial wheat varieties were developed and released to farmers. Since 1992, severe epid emics have continued to occur on commercial bread wheat and introductions (Danial et al 1994; KARI annual reports 1999, 2002, and 2003). In 1996 the disease was reco rded on some commercial wheat varieties in Mau-Narok and Molo. By 2000, all the varieties had succu mbed to the disease, and at present, they are

PAGE 64

susceptible. Stem rust outbreak is now widesp read in all the wheat growing areas in the country in low, medium and high altitude areas (KARI annual reports 2000-2003). The resurgence of this disease even in the high altitude areas (2700-2900 masl) has necessitated the need to revive work on this pathogen to incorporate genetic resi stance to the prevalent race(s) within the commercial varieties to work on this pathogen. Therefore, the objective of this study was to determine the distribution of stem rust on wheat varieties around the country and also to study the reaction of wheat varieties/ lines to rust inocula collected from different sources. Materials and Methods 1. Survey and collection of stem rust spore from the fields The survey was carried out on farmers commercial fields and also experimental plots in some parts of Uasin-Gishu, Nakuru, Meru, Nyeri, Nyandarua and Narok districts between years 2002-2004. Farms were picked along routes at random, stopping at every 35 kilometers or whenever the nearest farm was, especially in areas with sparsely distributed farms. In every farm, plants were examined randomly by walk ing across the field. When sampling, a two meter distance from the edge was left to avoi d border effect. Disease severities were taken using modified Cobbs scale (Peterson, et al .,1948) and coefficient of infection calculated by multiplying the disease severity and host response. Growth stage was entered in the field data form. The stem rust (sr) spores (uredospores) were then collected using a suction pump and some were scrapped from the leaves and stems using sterile scalpel blades. In the laboratory the spores were dried using Silica gel in the d essicator, vacuumed sealed in vials and stored for further work. 2. Screening of Stem Rust Parental Collection (SRPC) Fifty-three SRPC were planted by hand in the cag e at Kari-Njoro and evaluated for stem rust infection. Variety Morocco, the universal suscep t was used as the check. Each collection was planted in a hill at 2-3 seeds/hill. Diammonium phosphate (DAP) fertilizer was applied at the recommended rate of 130kg/ha.The plants were sprayed with Metasystox insecticide at the rate of 0.5L/ha to control the Russian wheat aphid (RWA) and other aphid vectors for the barley yellow dwarf disease. No fungicide was a pplied and the plants were monitored for the appearance of stem rust. Disease severity scores were taken twice at the heading stage using Cobbs modified scale ( Peterson et al .,1948 ). The scores were then transformed using square root (x+1) and the means transformed back to original scores. These were statistically analysed using analysis of variance and least significance difference used to separate the means. 3 .Screening of wheat commercial varieties against stem rust in the field Sixteen wheat commercial varieties were planted in Njoro (2185 m asl) and Eldoret Chepkoilel campus (2150 m asl ) while 14 varieti es were planted in Mau-Narok-Purko ranch ( 2900 m asl ) and evaluated for resistance ag ainst stem rust. In Njoro and Mau-Narok the varieties were planted in a randomized comple te block design with three replications. Plots were 8 rows by 2 meters per variety. In Eldoret two replicates of the same were planted. Furrows were made using the seed drill a nd both seed and fertilizer placed by hand. Immediately, after planting the experiment was sprayed with Stomp 500E a pre-emergent herbicide at the rate of 3.0L/ha to control grass weeds. At the tillering stage the experiment was sprayed with Buctril MC at the rate of 1.25L/ha to control broad leaved weeds. The Russian wheat aphid and other insect pests were controlled as above. Disease severity readings were taken at the heading st age and data were analysed as above.

PAGE 65

4. Screening of breeders F3 lines and stem rust differential lines for resistance against stem rust Seventy-eight F3 lines and 50 and 40 stem rust differential lines from Cimmyt-Mexico and Australia respectively were planted in 14,8 an d 11 sets respectively. Each set was planted in 4 inch plastic pots filled with soil and placed in the green-house. The pots were placed on trays with gravel filled with water. In ever y pot there were four lines/varieties. The uredospores collected from : Njbw2, ISR5RA(SR5), ISR9DRA, F1 material Chozi, Ngamia Duma, Heroe/Rwa8 Heroe/Rwa9, K7892/Rwa9 differentials Yalta, Anautka, Timboroa (Matharu), Eldoret and National Dryland Wheat Preliminary Trial (NDWPT) The rust spores were tested for viability and suspended in dis tilled water. A drop of Tween 20 was added to each collection and sprayed on 7 day old seedlings. The seedlings had been treated at emergence with Maleic hydrazide 0.01 g in 30 mls of water to enhance spore production. The plants were incubated for 24 hrs in a polythene chamber kept moist by frequent spraying with water under high relative humidity of 80-100%, and temperature of 18-22oC. They were removed and placed on the benches in the green house for 14 days and infection types recorded using 0-4 scale proposed by Stakman et al. ( McIntosh et al, 1995) for stem rust: 0 Immune, ;(fleck)-Very resistant, 1 -Resistant, 2 -Resistant to moderately resistant, 3 -Moderately resistant/moderately susceptible, 4 Susceptible. Data on infection types was compiled and analyzed using SAS system and means sep arated using least significance difference. Results 1. Survey and collection of stem rust spore from the fields The total number of commercial farms samp led between 2002 and 2004 were 61, from which 46 (75.4% ) of the farms had stem rust infection( Table 1). In some farms the wheat crop had dried but the stem rust infection spores (basidi ospores) were obvious. The farmers were in most cases not able to identify the varieti es because they bought planting seed from fellow farmers. Those who kept their own seed were growing old varieties. 2. Screening of Stem Rust Parental Collection (SRPC) Disease evaluation was based on 29 lines out of the 53 SRPC and Morocco the susceptible check. Twenty-three of these (79.3%) and Morocco were susceptible while six ( 20.7%) were resistant. There were significance (P< 0.05) differences among the line in disease infection. Table 2 shows that SRPC1(b), SRPC4(b), SRPC 25, SRPC 38(b) and SRPC 204 were highly susceptible while SRPC10, SRPC 27(a), SRPC (b), SRPC 29, SRPC 166 and 245 were resistant. 3. Screening of wheat commercial varieties in the field There were significance (P< 0.05) differences among the varieties in stem rust infection (Table3). The16 commercial wheat varieties were all infected with stem rust and the level of infection varied from variety to variety in all the three sites. In Njoro ( 2185 m asl), the most infected varieties were: Duma and Ngamia (56.7%, K.Tembo (46.7%) Yombi and Chozi (43.3%) Heroe and K. Kongoni (40%), Kwale (30% ) and Chiriku, Mbega, Mbuni. K. Fahari, K.Paka. and Njrbw1 were least infected. In Eldor et( 2150 m asl ), the most affected varieties were: Ngamia and Duma (70%), Chozi, Kwa le, Heroe and K. Kongoni(50%), Yombi (45%), while Chiriku, K.Nyangumi, K. Fahari, K.Paka K. Tembo and Njrbw1 were less infected. In Mau-Narok (2900 m asl), the most infected were: Yombi (50%), K. Nyangumi (30%) and Heroe (40%) wheras, Chiriku, Kwale, K. Nyangumi, Mbuni, K. Fahari and K. Kongoni were less infected. Varieties Ngamia, Duma, Yombi, Heroe and Chozi were most susceptible

PAGE 66

across the three sites with a mean disease sco re of 63.4%, 48.3%,46. 1%, 43.3% and 40.0% respectively. Table 1: Stem rust scores and coefficient of infection on commeci al wheat varieties year 20022004 Year District Place/Area Sr score/ reactions Coefficient of infection District Place/Area Sr score/ reactions Coefficient of infection 2002 UasinGishu Ainabkoi 20 MR 8 Nyeri Bagati 40S 40 Ainabkoi 40S 24 Bagati 60MS 36 Plateau 30MS 18 Naromoru 50MR 20 Plateau 60MS 36 Mweiga 5MS 3 Plateau 5MR 2 Mweiga 30MS 18 Plateau 70S 70 Kieni 40MS 24 Lembus 10S 10 Nairutia 20S 20 Lembus 5MR 2 Narok Olokurto 70S 70 Timboroa 5MR 2 Olokurto 20S 20 Timboroa 10S 10 Olokurto 50S 50 Timboroa 40S 24 Olokurto 30MS 30 Timboroa 20MR 8 Lopito 80S 80 Timboroa 5S 5 Elmasharani 80MS 32 2002 Nyeri Mbogo 5MR 2 Elmasharani 20S 20 Mbogo 20S 20 Ilipson 90S 90 Mbogo 10MR 4 Ntulele 60S 60 2003 Nakuru Lare 20S 20 Ntulele 5MR 1 2004 Nakuru M/Narok 80S 80 Navasha 40S 40 M/Narok 70S 70 Navasha 30MS 18 M/Narok 40S 40 Navasha 20S 20 M/Narok 15S 15 2004 Meru W/Embori 20S 20 Magara 30S 30 Kararu 10S 10 Mwireri 30S 30 M/Narok=Mau-Narok, W/Embori=Wangu Embori

PAGE 67

Table2 : Mean Stem Rust Disease Infection sco re and relative ranking (L sd) on the Stem Rust Parental Collection ( SRPC) in the Cage Kari-Njoro, 2002 SRPC Pedigree Mean disease score SRPC N0. Pedigree Mean Disease score SRPC 1(b) Africaxmayo 3529-1y-4m-3c-1t 45ab SRPC 36 E.Na480(Pilotxmd, CII2316)xGza 1392 ,1289-2196 0.5 hij SRPC1( c) Africa/mayo 15 def SRPC 38(a) 17.5 cde SRPC 2 Africa/mayo 3529-1y-4m-3y-1m 0.5 ij SRPC 38(b) 45 ab SRPC 2(a) 30 bcd SRPC 41 (FederationxHope) xBolim, PI124830 5 fghij SRPC 4(b) Bza sibx(CI 12633,WIS245) 35 bc SRPC 71 Gbx(fn-K58/N,II-50-17), II-53-649 5 ghij SRPC 8(b) Chinca-A-ElongatumxRd,K.Sel.A 10 efgh SRPC 140 MbxSR, LM-72-14-57 10 efghi SRPC 10 Chinca-A-ElongatumxRd)x(cl 12633xIdaed 0 j SRPC 148 My54XL1266-61,1448-4603 10 efgh SRPC 12 CI 8154xfr2 III-1009-2t-3b-1t-2b1t 10 efgh SRPC 153 Mt-KxN-M,(fr-fn /y2),15224-5b-1t1b-4t 10efgh SRPC 15 C1.12632xCeres R6421198.A.2.1 3 hij SRPC 166 ND 463 0 j SRPC 25 30 bcd SRPC 204 Sandos No63x CI12633-Idaed2) Gb 56xVeranopolis, 5134.B.3.B2.A 40 ab SRPC 27(a) (CI 12633,Wis245)2 s51x(fr-fn /y)III-113-6-6b-3t2b.K.sel .1 0 j SRPC 238( a) 57.5a SRPC 27(b) (CII2633, WIS245)/(for-fn/y)2 0 j SRPC 245 0 j SRPC 28(b) 3 hij SRPC 250 5 ghij SRPC 29 0 j SRPC 263 2.5 hij SRPC 32 Desc-CI7800 /Bza3,14951-9b-1t 5 ghij Morocco 12.5 efg Lsd P< 0.05 1.7 Lsd P< 0.05 1.7 Cv 26.4 Cv 26.4 Means in the same column followed by the same letter do not differ significantly using Lsd P< 0.05 -Pedigree to be traced

PAGE 68

Table 3: Mean % stem rust disease infection an d relative ranking (Lsd) on commercial wheat varieties in Njoro, Eldoret and Mau-Narok 2003 Variety Site/ Mean disease score Njoro Reaction type Mau-Narok Reaction type Eldoret Reaction type % mean score across the sites Yombi 43.3 b S 50.0 a S 45.0 b MR 46.1 Duma 56.7 a S 18.3 abcd S 70.0a S 48.3 Chiriku 5.0 g MR 6.6 cd MR 5.0 f MR 5.5 Chozi 43.3 b S 26.6 bcd S 50.0b S 40.0 Ngamia 56.7 a S 70.0a MS 63.4 Njrbw2 20.0 def MR 11.6 bcd MR 25.0cd MS 18.9 Mbega 13.3 efg MS 23.2 abcd S 30.0c MS 22.2 Kwale 30.0 cd S 1.6 d MS 50.0b S 27.2 K.Nyangumi 21.7 de MS 30.3 abc MS 5.0 f MS 19.0 Mbuni 5.0 g MR 10.0 cd MS 20.0de S 11.7 K.Fahari 10.0 efg MR 5.0 cd MR 0.0f R 7.5 K.Paka 8.3 fg MR 5.0f MR 4.4 Heroe 40.0 bc S 40.0 ab S 50.0b S 43.3 K.Kongoni 40.0 bc MR 5.3 cd MR 50.0b MS 31.8 K.Tembo 46.7 ab S 17.0 bcd MR 5.0f MS 22.9 Njrbw1 11.7 efg MR 16.6 abcd MR 15.0f MR 14.4 P< 0.05 12.3 3.092 9.482 Cv 26.16% 48.3% 14.38% Means in the same column followed by the same letter do not differ significantly using Lsd P< 0.05 4. Screening of breeders F3 lines and stem rust differential lines for resistance against stem rust Significant differences were observed in sources of stem rust inocula on sr differential lines acquired from Australia (table 4a). Stem rust spores collected from sr differential Anautika, ISR5RA(SR) and K 7892/Rwa9 had the highest mean infection (reaction) type while Heroe/Rwa9, Timboroa (Matharu), Eldoret, and NDLWPT had the lowest mean infection (reaction) type. This was not significant on breeders F3 material and also on sr differential lines acquired from Mexico. Test lines, the lines x source of inocula did not differ significantly for all the breeders F3 and sr differential lines. The analysis of variance (Table 4b) showed significant differences among the sources of sr inocula for the differentials acquired from Australia (F=0.001**) were significant. Table 4a :Mean disease score of stem rust inocula from diff erent sources on sr differential lines from Australia 2003 Source of inocula Mean disease score Variety Chozi 0.82 def ISR5RA(SR) 2.14 ab Timboroa( Matharu) 0.31 f ISRDRA 1.27 cde Heroe/Rwa8 1.42 bcd Heroe/Rwa9 0.30 f K7892/Rwa9 2.01 abc Yalta 0.52 ef Anautika 2.69 a Eldoret 0.36 f NDLWPT 0.36 f Lsd (P< 0.05) 0.825 Means in the same column followed by the same letter do not differ significantly

PAGE 69

using Lsd P< 0.05 Table 4b: Analysis of varian ce for sources of stem rust inocula Source DF SS F Set 10 359.514 0.001** Variety 39 183.086 0.4709 Set x Variety 288 1475.120 0.1133 Error 219 960.833 Discussion Levels of stem rust infection varied with environment as well as varieties/lines. This could have been due to the temperature and rain distribution which varies from place to place with some areas experiencing drought while others were wet. In Kenya, wheat is grown in many agroecological zones that have different planting dates ( Ralph and Schmit,1983). This staggered planting provides a green crop most part of the year allowing stem rust spores to move by wind from one area to another and that wa y spreads to different areas. It is therefore, a problem to reduce the disease infection in susceptib le varieties. Almost all the wheat fields had stem rust infection and the problem is a dvancing and becoming more serious in high altitude areas where maximum and minimum temperatures range from 22oC26o C and 6oC14oC respectively. The varied infection ( reaction) types of stem rust inocula from different sources on differential lines from Australia probabl y showed that there could be more new races attacking the wheat varieties. Although, Ngamia was not among the varieties screened in Mau-Narok (2900m ASL), the records from a different study ( data not shown) showed that the infection of stem rust was very high (Kari annual report, 2003). Conclusion The survey of stem rust revealed that the disease is on the increase in Kenya. Apparently all the wheat growing areas are prone to this disease. There is therefore, an urgent need to have more work done to determine the isolates so th at the breeding programme can be effectively sustained. Commercial varieties like K. Fahari, K.Paka, Chiriku and some of the old varieties which have remained resistant/ moderately resi stant to stem rust can be utilized in the breeding programme. It is evident that stem rust is widely distributed in the wheat growing areas and joint effort is required to utilize the current biotechnology tools to identify and breed for resistance to the apparently new race(s). Acknowledgement-The authors wish to acknowledge funds from CIMMYT for the implementation of this study; Dr. Ravi Singh (CIMMYT, Mexico) for technical discussions and germplasm; Drs.C.R. Wellings and Robert Pa rk (Australia) for the differentials; Mr. D. Onyango for maintaining the stem rust spores; and colleagues in pathology section for cooperation and technical support. References Agrios, G.N. 1988. Plant Pathology.3 rd ed ition. Academic Press. ISBN 0-12-044563-8 Cereal Disease Laboratory .(2004). Wheat Stem Rust ( Puccinia graminis f.sp. tritici ) at http://www.cdl.umn.edu Danial, D.L., 1994. Aspects of Durable Resistance in Wheat to yellow rust. PhD Thesis. Wagenigen Agricultural University, The Nertherlands, 144pp.

PAGE 70

Kenya Agricultural Research Institute.1999-2003.A nnual reports. Kenya Agricultural Research Institute.2002-2003.Annual reports. McIntosh,R. A.,C.R. Wellings and R.F.Park. 1995. Wheat Rusts: An Atlas of resistance Genes. CSIRO, Australia. Minzenmayer Richard (2000). Pest Management News at http:// www.tpma.org Peterson,R.F., A.B. Campell and A.E. Hannah. 1948. A diagrammatic Scale for estimating Rust Severity on leaves and stems of cereals. Can.J. Res; 26:496-500. Ralph Jaetzold and Helmut Schmidt. 1993. Farm Management Handbook Agriculture Vol.II Part B. Ministry of of Kenya in cooperation with German Agricultural Team (GAT) of Germany Agency for Technical for Te chnical Cooperation ( GTZ). Stubbs, R.W., J.M. Prescott, E.E. Saari and H.J.Dubin. Cereal Disease Methodology Manual. Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT), Mexico.1986. Wiese, M.V.1991. Compedium of Wheat Diseases 2nd Edition. APS Press.

PAGE 71

Genotype-by-Nutrient Interaction in Wheat Grown in a Marginal Environment in Kenya J. Kamwaga, H. Okwaro, P. Njau, P. Kimurto, P. Ndungu. and E. Kimani Abstract Six wheat ( Triticum aestivum L.) cultivars (Chozi, Ngamia, Duma, Njoro BW1, K.Heroe, and Chiriku) were grown in a marginal environment at two levels of fertilizer; 0:0:0, 30:60:0 kg ha-1 (N:P2O5:K).and at a seed rate of 125 kg. ha-1 to assess their response to fertilizer application and nu trient use efficiencies. Grain yield, spike numbers, harvest index (HI), nitrogen (N) uptake, N concentration and nitrogen use efficiency (NUE) were record ed. There were no significant differences in grain yield or spikes m-2 among varieties when averaged across fertilizer levels. Differences in HI between varieties were significant, w ith cv. Duma, a variety recommended for marginal environments, having a higher HI than cv. K.Heroe. Fertilizer did not affect the HI. There were no significant diff erences among varieties in N acquisition, indicating similar nutrient uptake abilities, or in NUE, although this was reduced by fertilizer application, probab ly because the increased N uptake did not translate into a corresponding grain yield increase due to the poor growing conditions. Grain N content ranged between 2.2% in the fertilized plots and 1.8% in the unfertilized plots (meaned across cultivars) and 2.2% in Njoro BW1 and 1.9% in Ngamia. Introduction Genotypic differences in nutrient utilizati on have been documented in wheat. These differences have been attributed to both differen ces in the acquisition and the efficiency of use of nutrients between cultivars (Inthapanya et al., 2000). One reason for differences in efficiency has been attributed to the devel opment of different root systems between wheat cultivars. Nitrogen (N) and phosphorous (P) defi ciencies limit the yield of wheat in most wheat growing areas of Kenya. A number of approaches have been used to enhance productivity of wheat in such environments: pr incipally the development of site-specific fertilizer recommendations for various areas. Inorganic fertilizers are expensive and not affordable to most wheat growers, while. Orga nic sources of nutrients are not easily available in the quantities and qualities that would give ade quate yields in farmers fields. The use of nitrogen fixing crops also has its limitations. Another approach would be to identify cultivars that do well in low soil fertility conditions but that also respond well to applied fertilizer. These would be the preferred cultivars because they would be useful to both farmers with fertilizer inputs and resource-poor farmers. This type of cultivar usually has a high harvest index (El Bassam 1998) The efficiency of utilization of both N and P determines grain and dry matter yields. Variety by fertilizer level interactions will be very important: researchers need to find varieties that have low yield interactions with fe rtilizer level, and that, therefore, respond in in a similar manner in low and high fertility e nvironments. This is especially important as varieties widely adapted to diverse soil fertility status will be preferable. Cultivars grown in the marginal environments of Kenya have not been evaluated for their nutrient use efficiency, yet this is an importan t factor determining yield of wheat in such areas. Identification of cultivars that do well under low nutrient and moisture conditions will improve wheat production in such areas and th e the breeding program will be enhanced by incorporation of characters associated with nutrient use efficiency. Improved cultivar

PAGE 72

response to nutrients will help reduce inputs and hence protect the environment (El Bassam 1998). The objectives of this project were to: I. Determine genotypic variation in nitrogen uptake between cultivars recommended for the arid and semi arid areas (ASALs) of Kenya. ii. Determine the efficiency of utiliza tion of N in recommended cultivars. Materials and Methods The trial was established in Kajiado, Kenya, on a moderately fertile vertisol with P ststus above the critical levels (Table 1). The treatme nts were arranged in a split plot design with three replications, with fertilizer levels as th e main treatments and varieties as sub-treatments. The two fertilizer levels used were 0:0:0, and 30:60:0 kg ha-1 (N:P2O5:K), and the cultivars assessed for nutrient use efficiency were: Ngamia, Chozi, Duma, Njoro BW1, Chiriku and K. Heroe. Sub-plot size was 6m x 3m. Seed was dribbled by hand into furrows spaced at 20 cm apart at a rate of 125 kg. ha-1, and fertilizer dribbled into the same furrows. At four weeks after seeding, Buctril MC. ( a.i. Bromoxynil +MCPA ) was applied at 1.2 l ha-1 to control weeds. At harvest, a 1 m2 quadrat was harvested by cutting the plants at ground le vel. Parameters evaluated were total biomass (Total Dry Matter), grain yield and nitrogen (N) concentration in grain and straw, determined using the kjeldahl method. N uptake, nitrogen use efficiency (NUE) and harvest index (HI) were calculated from the evaluated parameters. Table 1 Soil chemical and physical characteristics of the trial site Depth (cm) pH % N P (ppm) % O.M. 0-15 5.8 0.15 33 2.4 15-30 6.5 0.09 30 2.0 30-60 6.9 0.07 31 1.7 60-90 6.9 0.08 31 0.8 Results and Discussion There were no significant effects of genotype on grain yield, harvest index, nitrogen uptake, use efficiency or grain N concentration, nor were there fertizer by genotype interactions with respect to these variables (Table 2), indicati ng that the varieties responded similarly to fertilizer application. However there were di fferences between varieties with respect to thousand kernel weight (TKW): Chozi had significantly higher TKW than all the other varieties (Table 3). Fertilizer application significantly increased grain yield, total N uptake and grain N concentration, but reduced nitrogen use efficiency, and had no significant effect on HI nor TKW. (Table 2 and Table 4). Table 2. Degrees of freedom F ratio and level of significance for grain yield, Harvest Index (HI), Total N uptake, Grain N concentration and Nitrogen Use Efficiency. Source of variation DF Grain yield HI Total N Uptake Grain N (%) NUE TKW Genotype 5 0.94 1.46 0.69 1.49 0.95 6.49** Fertilizer 1 12.75** 0.68 65.6** 19.65** 18.51** 2.94 Genotype*Fertilizer 5 0.85 0.51 0.97 0.42 0.46 1.12 ** Significant at (p<0.01)

PAGE 73

Table 3. Variety yield parameters measu red across fertilizer levels at Kajiado Variety Spikes/m 2 Grain yield kg ha-1 Harvest index (HI) TKW Duma 182 1853 0.40 A 38.9 B Ngamia 179 1751 0.38 AB 35.6 BC Njoro bw1 173 1491 0.37 AB 37.9 BC Chiriku 170 1642 0.37 AB 34.6 C Heroe 163 1417 0.31 B 39.0 B Chozi 150 1522 0.37 AB 44.0 A LSD 5% NS NS 0.06 3.8 C.V % 24 26.2 15.7 8.3 Table 4. Fertilizer effects on yield paramet ers measured across varieties at Kajiado Fertilizer level Grain yield kg ha-1 Harvest index TKW NUE g grain g-1 N Total N Uptake g m-2 Grain N Concentration (%) O 1361 B 0.37 A 39.2 41.8 a 330 a 0.38 a N X P 1864 A 0.36 A 37.4 31.9 b 600 b 0.36 b LSD 5% 29.4 NS NS 0.05 69.6 0.04 There were no significant differences between genotypes in either nitrogen uptake (acquisition) or utilization of the absorbed N: there were no significant differences in NUE (grain production per unit of the total N ) (Table 5) However, cv. Njoro BW1 had a significantly higher N concentration than c.v Ngam ia. It has been shown that grain yield is a function of NUE, total N uptake and HI as in the equation of Inthapanya et al (2000). Y=NUE X Total N uptake x HI Where NUE= Nitrogen Use Efficiency Hi=Harvest Index. The results from this study show that of th e three variables, the tested cultivars only differed in HI, suggesting that the only poten tial for increasing yields in the wheat cultivars would be in improving the HI of the cultivars. Table 5. Total N uptake, grain nitrogen concentrat ion and nitrogen use ef ficiency (NUE) meaned over two fertilizer levels in wheat at Kajiado Variety Total Nitrogen Uptake g m-2 Grain N Concentration (%) Nitrogen Use Efficiency (NUE) g grain g-1 N Duma 5.22 2.08 Ab 38 Njoro BW1 4.72 2.17 A 33 Chiriku 4.67 1.96 Ab 37 Ngamia 4.60 1.87 B 40 K. Heroe 4.50 1.93 Ab 34 Chozi 4.18 2.00 Ab 39 LSD NS 0.26 NS C.V 2.15 10.8 18.8 Means followed by the same letter are not significantly different at p>0.05

PAGE 74

References Inthapanya P., Sipaseuth, Sihavong P., Sihathep V., Chanphengsay M., Fuka i S., Basnayake J. 2000. Genotype differences in nutrient uptake and utilization for grain yield production of rainfed lowland rice under fertilized and non-fertilized conditions. Field Crops Research 65 (2000) 57-68. El Bassam N. 1998. A concept of selection of low input wheat varieties. Euphytica 100 (1/3) 95100.Improvement of Wheat

PAGE 75

Improving Wheat Productivity for the Drought Prone Areas of Kenya Using the Doubled Haploid Technique Njau P. N1, Kimurto, P.K2, Kinyua M. G1, Okwaro H. K1 and Ogolla J.B.O2 1. National Plant Breeding Research Center, Njoro, P.O. Njoro, Kenya`1 2. Dept. of Agronomy, Egerton University, P.O. Box 536, Njoro, Kenya AbstractVarious breeding methods have been applied in an effort to develop superior wheat varieties for the marginal areas of Kenya. These include germplasm introductions, mutation breeding and doubled haploid (DH) production. .The application of DH in breeding for drought tolerance has proved to be very effective and efficient. This study was aimed at evaluating and validating DH lines that were developed in 2000 using the chromosome elimination technique. The resulting lines were compared with other lines introduced from CIMMYT and two mutants developed in Njoro. The lines were tested in multi-locational trials in 2002 and 2003. The sites included were Lanet, Mogotio, Naivasha, Mweiga, Kajiado and Katumani. The results show that the DHs were as good as the conventionally developed lines and yielded more than the check varieties Chozi and Duma. One of the DHs had average yielded 1.7tonsHa-1, which was not significantly different from Chozi. The DH technique has proved to be applicable in saving time of breeding without compromising the output. Introduction Wheat is the second most important cereal crop in Kenya (FAO, 2002) but Kenyas current national wheat production (approximately 300,00 0 tons per annum) meets only about 50% of the national demand. Moreover, increasing population and changi ng eating habits is expected to have a bigger impact on wheat demand; estimated to reach 850,000 tons p.a. in the year 2020 (FAO 2003). Land area currently suitable for wheat production in the high potential areas is less than 2% and hence expans ion of wheat production in these regions is limited. In recent years wheat production has expa nded in the marginal rainfall areas (such as lower Narok, Naivasha, Laikipia and Machakos ) which were hitherto considered unsuitable for wheat growing. Introduction of wheat in th e non-traditional areas began in 1992. Over this period 4 varieties (Duma, Ngamia, Chozi and Njoro BW1) have been released (Kinyua et al 2002). However, varieties with greater yields are still needed to meet the farmers needs. Future production increases must come largely from vertical expansion (i.e., greater production per unit area), which will require more in tensive research to further improve yield potential and cultural technology. Wheat impr ovement in Kenya has been directed into developing broadly adapted, high yielding germplasm with high yield stability, durable disease resistance and acceptable end use quality (CIMMYT 2002). Resistance to biotic stress and tolerance to abiotic stress can be critica lly important in maintaining high yield and adaptation over time and locations. Over the l ast decade wheat breeding has been enhanced by the application of various biotechnological approaches; these have b een used to accelerate the breeding process and also complemented the conventional breeding methods. The DH technique and mutation breeding have been used to improve stress tolerance in wheat. These include breeding for drought tolerance, tole rance to soil acidity, lodging tolerance and resistance to Russian wheat aphid.

PAGE 76

Drought is a multidimensional stress affecting plan ts at various levels of their organisation (Blum, 1996). Improvement of yield under stress must combine a reasonably high yield potential with specific plant factors which would buffer yield against severe reduction due to drought (Blum, 1983). Measurement of plant res ponse to drought at the whole-plant level is complex because it reflects the integration of stress effects and responses at all underlying levels of organisation, over space and tim e (Blum, 1996). Conventional breeding (which involves multi-location testing and hence gauges spatial adaptation of genotypes) is used extensively by the Kenya bread wheat program to identify temporally stable, drought tolerant germplasm (Maarten, 1994 this does not appear in the reference list). However, practical breeding programs for self-pollinated crops (suc h as wheat) must include a process of genetic fixation (for uniformity of agronomic traits) after genetic recombination to increase variability (Inagaki, 1996). Repeated selection of hetero zygous material can increase uniformity but many generation cycles are required to reach ho mozygosity in loci associated with agronomic traits. Haploid production followed by chro mosome doubling offers a quick method for developing homozygous breeding lines (Bentolila et al, 1992, Murignaux et al. 1993, Beanziger et al., 1989a, Baenziger, 1996). The DH lines derived from hybrid progenies can be used as recombinant inbred lines with favourab le gene combinations (Inagaki, 1996). This technique could thus complement the conve ntional breeding programs to accelerate the release of new varieties. Since the double haploids are completely homozygous, all stocks are identical and no purification process is required. In contrast, in the conventional system stocks are usually derived from a single plant of an advanced generation (Baenziger, 1996) hence several generations are required to build up sufficient quantities of seeds for release. Compared to selection during the early generations a DH system increases the efficiency of selection for both qualitative and quantitative characters. This study aimed to evaluate and validate DH lines and compare them with conventionally bred materials. It was hypothesised that the DH technique is superior to conventional breed ing when dealing with complex factors like drought. Materials and Methods Introduction Two experiments were carried out. Experi ment I involved development of the double haploids (including embryo rescue, regene ration of haploid plantlets and chromosome doubling of haploids) and preliminary evalua tion of the DHs. Experiment II was a field evaluation of the performance of the DHs as co mpared with the conventionally developed lines Development of doubled haploids Six F1 hybrids were produced by crossing three drought tolerant lines (Duma, K. Mbweha and Ngamia) as females with two high yielding comme rcial varieties (Kwale and Kenya Chiriku) as males. The F1 plants were emasculated at anthesis. The glumes and awns were not clipped, as is the case with normal-crossing emasculation. The emasculated spikes were then covered with polyethylene bags to maintain high humidity. One day before (predicted) anthesis, emasculated spikes were pollinated with fresh ly collected maize pollen. The pollen was collected by picking mature anthers and placi ng them on petri-dishes. Once the anthers burst to release the pollen, a soft brush was used to brush the emasculated spikelets with the pollen. Extreme care was taken not to damage the stig ma. After pollination, the uppermost internodes of wheat culms (with pollinated spik es) were injected with a 100mg l-1 2, 4-D solution daily (for two consecutive days) to increase the rate of fertilisation and embryo formation.

PAGE 77

After about 14 days of spike growth, wheat caryopses were removed from spikes and sterilized (in 2% sodium hypochlorite) for 15 minutes. After rinsing with sterilized water (under laminar flow bench), embryos were asep tically excised and transferred onto half strength Murashige and Skoog medium supplemented with 20g L-1 sucrose and 8g L-1 agarose in petri dishes (Inagaki 1996). These we re then placed in the fridge (at 4 oC) for three days after which they were incubated (at 25 oC) in the dark until embryos germinated (5-7 days). The germinated embryos were transferred to a lighted growth room with controlled temperature of 24 oC and 16 hours day length. The light intensity was Ca 5000 Lax. Cultured plants with fully developed roots and leaves were transplanted to potted soil for further growth. The temperature was maintained at 21 oC -25 oC using 24 hours day length. The size of the stomata was used to determine the ploidy level of the plants. This is because haploid plants have stomata half the size of diploid plan ts. At the third tillering stage, a leaf from each plant was cut and coated with clear nail varnis h. The varnish was pealed off (upon drying) and mounted on high power objective of light microscope; the size of the stomata was then compared with that of known diploid plants. In addition, at the third tillering stage the plants were watered and then removed from the pots. The roots were trimmed to about 2 cm below the crown ( to increase solution absorption). The plants were then immersed in colchicine solution (0.2% colchicine, 2% dimethyl sulfoxide and 15 drops of tween-20) fo r 3 hours at room temperature. The plants were then washed thoroughly with running tap water for 3 hours after which they were planted in pots (using forest soil enriched with copper dust) at a temperature of 20o C-25 o C under high humidity (90-100%). Preliminary evaluation of the doubled haploids Twenty DH lines were selected (based on the am ount of seed available). These selections and their five parents were initially screened in the rain-shelter at the National Plant Breeding Research Centre (NPBRC) Njoro in 2000 and later evaluated in the field (in Njoro and Katumani) through observation trials in 2001. Th ey were planted in a randomized complete block design (RCBD) with three replications. Each plot was 1 metre long with two rows 20 cm apart. Five seeds were planted per row. DAP (18-46-0) fertilizer was applied at planting at the rate of 150 kg ha-1. Drip irrigation was used to water the plots (Chapin Watermatics, 1999). Water was applied at the rate of 20 mm every fortnight (up to grain filling stage), giving a total of 200 mm water during the crop season. The amount and frequency of water application simulates the amount and nature of rainfall pattern usually received in most marginal areas during the cropping season (Mugo et al. 1998; Jaetzold and Smith, 1983). The following parameters were measured (both in the rain shelter and in the field): Ear length (measured from the base of the spike to the tip of the apical spikelet, excluding the awns); number of spikes per plant, sterile spikelets pe r head; number of grains per head counted on 10 spikes selected randomly in each experimental at maturity; and weight of 10-kernels (g). All the data was subjected to analysis of va riance using the general linear model (SAS, 1996). Field performance of the double haploids Eight DH lines (DH4, DH5, DH6, DH7, DH9, DH12, DH15, DH16) and 2 mutants (BM1 and BM3) were selected in a preliminary yield trial in 2001. The selection was on the basis of drought tolerance. These lines, together with Chozi and Njoro BW1 (ch ecks), were entered in the National Performance Trial (NPT) in 2002 a nd consequently planted at three sites (i.e., Naivasha, Katumani, and Lanet). The design w as a RCBD replicated 3 times. The seeds were drilled in plots of four rows that were 6 metres long and 20 cm apart. Seven DHs (DH4, DH5, DH6, DH7, DH9, DH12 and DH16) were selected form the first year of NPT and entered into the second year of Na tional Performance Trial (in 2003) alongside 7

PAGE 78

conventionally developed lines (R840, R960, R962, R963, R965, R966 and K7872) and the check varieties (Njoro BW1 and Chozi). These were evaluated at Katumani, Naivasha, Mweiga, Lanet and Ravine. The Design was a RCBD replicated three times. The plots were 4 rows, 20 cm apart and 6 m long. The following parameters were measured in bo th the first and second year NPT: grain yield, Kernel weight, plant height and reaction to rust diseases. ANOVA was used for data analysis a nd means were separated using LSD Results and Discussion Development of Doubled Haploids A total of 1800 florets were cross pollinated out of which 890 F1 seeds were harvested and the embryos excised and planted in-vitro This shortened the F1 seed production by over four months compared to the in-vivo method where the seed is left to dry in the field. The in-vitro F1 plants grew faster due to the conducive e nvironment and reached heading stage one month earlier than in the conventional method. Over 2,880 florets of F1 plants were cross pollinated with maize. 413 of these developed seeds from which 57 embryos were rescued (T able 2). Out of the 57 embryos rescued 46 were haploid. When treated with colchici ne 24 of the 40 survived. The time taken for pollination to colchicine treatment was 8 week s and the DHs were ready for harvesting in 20 weeks. Performance of doubled haploids There was high variability among the DHs in th e rainout shelter for the yield components measured (Table 1). These results are comparable to those reported by Baenziger (1996); whilst there was variation among doubled haploi d lines, there was little variation within individual haploid lines. This increased the efficiency of selecting lines with superior characteristics (Njau et al, 2000) because the response to selection is indicated by the variability between the treatments. Such variati on can be used to select lines with the required characteristics (Njau 2001). There were significant differences in yiel d and other growth pa rameters between the DHs and their respective parents (Table 1) a nd some DHs showed heterosis for drought tolerance over their better parent or the mean of the two parents (Table 4). For example, one of the DHs (DH17) developed from the cross between Kenya Mbweha and Kenya Chiriku had greater number of spikes (6.1) than the parent s (2.0 and 5.0, respectively for Kenya Chiriku and Kenya Mbweha). DH17 also had more grains per head (43) than the mean of the parents (27.5). DH112 (developed from Duma/Kenya Chiriku cross) had more grains (%) than either of the parents. The expression of heterosis in DH lines (Njau 2002) makes the DH technique superior to that of conventional breeding. This is because heterosis is lost in conventional breeding due to repeated selfing of the hybrids. Similar results have been reported elsewhere (Ba-Bong and Swaminathan, 1995). Over 26% of the cost of developing a fi xed line using conventional selection was saved in the production of the DHs (data not provided). This saving was attributed to the fewer number of generations required to produce hom ozygous lines when using haploids. Similar findings have been reported in rice (Sunint, 1993).

PAGE 79

Table 1. Mean performance for various parameter s of 20 DHs as compared to their parents tested for drought tolerance under the rainshelter at Njoro. Kenya. Parents DH lines No. of tillers at booting Effective Heads Spikelets/He ad Sterile Spikelets Grains/ head weight of 10 grains K. Mbweha Kwale (K.mbweha x Kwale) DH11 DH18 DH119 5.00 4.21 5.33 5.71 3.71 5.0 2.16 4.00 3.67 3.66 15.333 14.533 16.000** 17.533** 15.533 2.000 3.994 2.00* 2.494* 1.494* 24.333 27.936 34.00 35.936 32.936 0.367 0.306 0.301 0.317** 0.27 K. Chiriku (K. Chiriku x K Mbeha) DH17 DH114 4.33 7.39** 2.67 2.00 6.11** 2.33 15.333 19.301 17.333 3.000 2.009 1.333* 30.667 43.489*** 36.667** 0.312 0.282 0.343 Ngamia (Ngamia x Kwale) DH12 DH13 DH110 DH113 DH115 DH116 3.67 4.39 5.21 4.33 4.30 3.33 8.39** 3.00 4.11 3.66 2.00 3.95 3.33 6.1*** 10.00 19.301** 16.033** 13.333 15.349** 15.333** 19.301** 1.667* 1.009* 1.995* 1.006* 1.996* 0.67*** 0.09*** 41.667 41.489** 24.94 27.67 41.755** 30.00* 28.489 0.349 0.336 0.21 0.346 0.304 0.367 0.304 (Ngamia xK.Chiriku) DH117 DH120 0.803 5.00 0.45 4.33 13.765 14.000** 4.980 1.667** 4.33 28.67 0.12 0.19 Duma (Duma x Kwale) DH14 DH15 DH16 DH19 DH118 6.33 2.80 3.39 7.39 7.00 4.30 3.00 2.45 3.11 2.89 4.00 2.45 15.000 13.349 19.301 14.617 11.333 12.349 2.000 2.996 2.009** 1.510** 2.333** 1.996** 28.333 19.255 35.489 33.308 29.333 32.255 0.467 0.305 0.310 0.367 0.266 0.326 (Duma x K.Chiriku) DH111 DH112 3.71 5.67 2.16 3.67 14.533 17.333 0.995 2.000 40.936* 44.00** 0.307* 0.421** Significantly better at 5% level than the poorer parent ** Significantly better than the mean of the parent at P< 0.5 *** Significantly better than the better parent at 5 % level Significantly poorer than the better parent at 5 % level. Significantly poorer than the poor parents at 5 % level. Preliminary performance of the double haploids Significant differences in yield Among genotypes were noted in Katumani and Naivasha while no differences were noted in Lanet. Two lines DH6 and DH4 performed quite well at all the sites. DH6 was better than all the othe r entries in Naivasha and it was better than the checks at all the sites. DH7 and DH9 also performed well at all the sites

PAGE 80

Table 2. Grain yield of the test lines (DHs, muta nts and check varieties) at 3 different sites (t ha-1) YIELD (t ha-1) LINE KATUMANI NAIVASHA LANET DH4 .9AB 1.0BC 1.4AB DH5 .3E .3D 1.1B DH6 .9AB 1.5A 1.5AB DH7 1.0A 1.3AB 1.1B DH9 .7BCD 1.4AB 1.6A DH12 .7ABCD .9C 1.2AB DH15 .7CD .7C 1.3AB DH16 .8ABC .8C 1.2AB BM1 .6CD .8C 1.0B BM3 .7BCD .7C 1.3AB NJORO BW1 .8ABC .9C 1.3AB CHOZI .5DE. .9C 1.1B LSD .27 .37 0.56 SE 0.22 0.11 P(F-ratio) <0.05 <0.05 <0.05 Values followed by the same letter are not significantly different at p = 0.05 Genotype did not affect 1000-kernel weight in Katumani and Naivasha but did affect 1000kernel weight in Lanet (Table 3). The double haploids had greater 1000-kernel weight (16% on average) than the check varieties and the muta nt BM3 in Lanet (Table 3). Also, the mutant BM1 had 9% 13% greater 1000-kernel weight th an the double haploids in Lanet (Table 3). Table 3. 1000 kernel weight of the entries in th e three sites during the year 2002 Line 1000-Kernel weight (g) Katumani Naivasha Lanet DH4 76.6 73.4 83.1A DH5 72.0 78.8 81.0AB DH6 72.4 76.7 78.7AB DH7 72.4 72.4 78.4AB DH9 70.7 71.2 78.0AB DH12 77.4 78.3 78.0AB DH15 75.9 75.8 77.9AB DH16 76.6 76.4 77.4AB BM1 73.9 72.7 75.6AB BM3 72.9 68.2 69.7CD NJORO BW1 74.2 76.1 69.4CD CHOZI 75.3 78.6 66.9D Lsd 10.85 P (F-ratio) ns ns <0.05 SED 4.1 3.2 4.04 Values followed by the same letter are not significantly different at p = 0.05 Genotype affected the average coefficient of in fection (ACI) by stem rust in 2002 (Fig 1). Chozi (check variety) was more susceptible to stem rust than the double haploids (except DH15 and DH16) and the mutants (Figure 1). Njor oBW1 (check variety) had lower stem rust infection than the mutants and most double haploids (DH5, DH9, DH12, DH15, DH16) (Figure 1).

PAGE 81

Performance of the DHs as compared to the conventionally developed lines There was a significant effect of site on variety in terms of grain yield and hectolitre weight in 2003 (Table 4). Katumani had lower yields than Lanet (by 80%), Mogotio (by 70%), Naivasha (by 70%), Mweiga (by 30%) and Kajia do (by 30%). Kajiado and Mweiga also had lower yields compared with Lanet (38%), Mogotio (31%) and Naivasha (31%) (Table 4). The lower yield in Katumani (compared with other sites) could be due to the low rainfall (100 mm) recorded in Katumani during the growing period of the crop. Lanet had lower hectolitre weight (by 20%, 19%, 10%, 13% and 7% for Mogotio, Naivasha, Mweiga, Kajiado and Katumani, respectively) compared with the other sites (Table 4). Katumani had 12%, 11% and 5% lower hect olitre weight than Mogotio, Naivasha and Kajiado, respectively. Also, Mweiga had lowe r hectolitre weight compared with Mogotio (9%), Naivasha (8%) and Kajiado (2%). Moreove r, Kajiado had 6% lower hectolitre weight compared with Lanet and Mogotio (Table 4). Table 4. Average yield in tons per hectar e and hectolitre weight in the six sites Site Yield (t ha-1) Hectolitre weight (g) Lanet 1.8a 67.3d Mogotio 1.7a 80.9a Naivasha 1.7a 80.2a Mweiga 1.3b 74.3c Kajiado 1.3b 76.0b Katumani 1.0c 72.1c Mean 1.5 75.1 LSD 0.18 1.70 SED 0.21 4.21 P (F-ratio) <0.01 <0.01 CV (%) 29.7 5.6 The yields and hectolitre weight for the lines av eraged across the sites are shown in Table 5. Genotype affected grain yield and hectolitre wei ght (Table 5). Chozi (check variety) had 55% greater yield than DH6 and DH16 and 113% greater yield compared with DH12. NJOROBW1 (check variety) also had greater yield (by 88%) than DH12 (Table 5). Chozi had greater hectolitre weight comp ared with R960, R962 and R966 (9%), K7872 (12%), R965 (10%), which are lines develope d through conventional breeding (Table 5). Fig 1. Stem rust average coefficiet of infection -20 0 20 40 60 80 100 120D H 4 DH5 D H 6 DH7 D H9 DH12 DH15 DH1 6 BM1 BM 3 NJOR O BW 1 CHOZILinesACI

PAGE 82

Chozi also had greater hectolitre weight (b y 8% 13%; average 11%) compared with a number of double haploid lines (DH4, DH9, DH5, DH7, DH16, DH12) (Table 5). The check variety NJOROBW1 had 7% greater hectolitre weight than DH12 (Table 5). These results show that in addition to being reducing th e breeding time, lines developed using the DH technique, compete well with conventionally de veloped lines which have taken over 13 years to develop. Table 5. Yield and hectolitre weight of the sixteen lines averaged over the sites Line Yield (t ha-1) Hectolitre weight (g) R960 1.9a 74.5bcd DH4 1.7a 73.0cd R840 1.7a 78.1ab Chozi 1.7a 81.3a R966 1.6ab 74.3bcd R963 1.6ab 76.6abcd K7872 1.5abc 72.3cd DH9 1.5abc 75.0bcd NJBW1 1.5abc 77.0abc DH5 1.5abc 73.5bcd R965 1.5abc 73.9bcd DH7 1.4abc 74.7bcd R962 1.4abc 74.6bcd DH16 1.1bcd 72.9cd DH6 1.1cd 78.0ab DH12 0.8d 72.0d Mean 1.5 75.1 LSD .29 2.77 SED 0.09 17.7 P (F-ratio) <0.01 0.01 Conclusion The DH technique proved to be quite useful in breeding for complex characters such as drought and compares well with the co nventional methods. The time saved in DH development makes the methodsuperior to the othe r methods. It is important to note that the multiloctional testing increases the efficiency of measuring yield stability. The two lines (DH4 and R960) are recommended for release as they both out yielded the check varieties at almost all sites. References Ba Bong B. and M.S. Swaminathan 1995. Magnitude of hybrid vigour retained in double haploid lines of some heterotic rice hybrids. Theo. Appl. Genet. 90 :253-257. Baenziger P.S. C.J Peterson, M.R Morris and P.J Ma ttern 1989a. Quantifying gametoclonal variation in wheat doubled haploids. M. Muluszynski (Ed). Current Options for cereal improvement pp. 1-9. Kluwer Academic Publishers. Boston. Baenziger S.P. 1996. Reflections on doubled haploids in plant breeding. S.M. Jains, S.K Sopory and R.E Veilleus (Ed). In vitro Haploid production in Higher plants 1: 35-48. Klumer Academic Publishers. Netherlands. Bentolila S, T. Hardy, C. Guitton and G. Freyssient 1992. Comparative genetic analysis of F2 plants and anther culture derived plants of maize. Genome 35: 575-582. Bitch C., G. Sabine and J. Lelley 1998. E ffect of parental genotypes on haploid embryo and plantlet formation in wheat x maize crosses. Euphytica 103: 319323.

PAGE 83

Blum A. 1996. Crop responses to drought and the interpretation of adaptation. Plant Growth Regulation 20:135-148. FAO, 1986. Production yearbook. Vol. 40. FAO ROME. Inagaki M.N. (1990). Wheat Haploid through Bulbosum technique. Biotechnology in Agriculture and Forestry 13:448-58. Inagaki M.N. (1996). Technical advances in wheat haploid production using ultra-wide crosses. IIRCAS Journal 4:51-62. Jaetzold R. and H. Schmidt, 1983. Farm management Handbook of Kenya. MOA and GTZ. Kinyua M.G.1997. Transfer of genes of resistance to yellow rust ( Puccinia stiifarmis L.) from wild emmer Triticum dicoccoides into Kenyan wheat commercial varieties. PhD Thesis. Nairobi University. Kinyua, M.G,., Karanja, L., and Njau, P.N. 2002. Drought tolerant wheat varieties developed through mutation breeding technique. Workshop on Dryland Farming January 2002. Paper submitted and accepted. Kinyua, M.G., J.K. Wanjama, J. Kamwaga, M. Mi gui 1989. The situation of wheat production in Kenya KARI Report. Lefebure D. and P. Devaux 1996. Doubled Haploid of wheat from wheat x maize crosses: genotypic influence, fertility and inheritance of the IB L-IRS chromosomes. Theor. Appl. Genet. 93: 1267-1273. Murignaux A., D. Barloy, P, Leray and M. Beckon t 1993. Molecular and morphological evaluation of doubled haploid lines in maize, Homogeneity within DH lines. Theor. Appl. Genet. 86:837842. Njau P. N. 2001. Development of drought tolerant wheat lines through chromosome elimination technique. Msc. Thesis Egerton University. Njau P.N., M.G. Kinyua and R.S. Pathak.2000. Drought Tolerant Doubled Haploid (DH) of Bread Wheat for Kenya. Proceedings of the 7th Biennial KARI Scientific Conference Paper No.40.KARI Headqu arters. Nairobi. Sunint L.R., C.P. Martinez, A. Ramirez and Z. Len tini, 1993. Rice anther culture versus conventional breeding: A cost/benefit analysis. Trends in CIAT commodities working document NO.128. Contro International De Agricultura Tropical, Call.

PAGE 84

Grain Yield Stability of Bread Wheat Genotypes in Favorable and Stressed Environments Desalegn Debelo, Solomon Gelalcha, Balcha Yaie, Bedada Girma, Berhanu Mamo, and Debebe Masresha EARO, Kulumsa R. C., P. O. Box 489, Asella, Ethiopia E-mail: desalegnd@freemail.et Abstract A multilocational trial consisting of 18 bread wheat genotypes including standard and local checks was conducted in 39 locations between 1999 and 2001 across the diverse wheat agro-ecologies of Ethiopia. Yield stability of the genotypes was assessed in high potential, moisture stress and waterlogged growing environments using the AMMI (Additive Main Effects and Multiplicative Interaction) model. The results showed highly significant genotypic and G x E interaction affects in all three enviro nmental categories. Genotypes FH 8-2, HAR 3354, ETBWC 037 showed low positive interaction while HAR 3116, and FH 6-1-7 showed low negative IPCA axis 1 scores indicating stable performance in the high potential wheat growing areas. In the low moisture areas FH 8-1, ETBWC026, HAR 3224 and HAR 2870 showed low positiv e IPCA axis 1 scores. Among these genotypes, HAR 2818 gave the highest mean grain yield and is better suited to moisture stressed areas. The analysis of waterlogged areas showed that HAR 3354, FH 4-2-11 and ETBWC037 had high mean grain yield performance and low G x E interaction. From the results we conclu ded that FH 8-2, HAR 3354, ETBWC 037, HAR 3116 and FH 6-1-7 are widely adaptable; FH 8-1, ETBWC026, HAR 3224 and HAR 2870 perform better in the moisture stressed areas while FH 4-2-11 and ETBWC037 are tolerant to waterlogged vertisols. Introduction In developing crop varieties with high and stable yields, data from multi-environments, representing spatial and temporal target domains play an important role in estimating and predicting yield and yield stability. The pattern of response of genotypes to different environments is used by the plant breeder to sel ect the best varieties for any particular region. Yield response of genotypes tested in different environments is almost always subject to genotype x environment interaction (GxE). To estimate the level of interaction of genotypes to environments and eliminate as much as possible the unexplainable and extraneous variability contained in the data, several sta tistical techniques have been developed to describe G x E and measure the stability of ge notypes. These techniques include conventional analysis of variance (Yates and Cochran, 1938; Cochran and Cox, 1957); the joint regression method proposed by Yates and Cochran (1938),modi fied by Finlay and Wilkinson (1963) and further improved by Eberhart and Russell (1966) ; combined analysis of variance for each pair of genotypes (Plaised and Peterson, 1959) ; ecovalence (Wrickle, 1962;1964); stability variance (Shuka, 1972); cluster analysis (Lin et al. 1986); and cultivar superiority measure (Lin and Binns, 1988). Furthermore, the additive main effect a nd multiplicative interaction method, widely known as AMMI model, has recently been developed and used to analyze multi-

PAGE 85

environmental trials (Gau ch and Zobel, 1988; Zobel et al ., 1988; Crossa et al. 1990). It is a unified approach that fits the additive main effects of genotypes and environments by the usual analysis of variance and then describes the non-additive part by principal components analysis fitted to the AMMI model according to the following equation: Yge = + g +e + n gn en + ge, n=1 where Yge is the yield of genotype g in environment e; is the grand mean; g are the genotype mean deviations (the genotype means minus the grand mean); e are the environment mean deviations; n is the eigenvalue of principal components analysis (PCA) axis n; gn and gn are the genotype and environment PCA scores for PCA axis n; is the number of PCA axes retained in the model; ge is the residual. Wheat is a widely grown crop in the world, produced under a wide range of environments (Hanson et al ., 1982). In Ethiopia, wheat is grown between 6 and 14o N latitudes; and between 35 and 42oE longitude at altitudes from 1500 m to 3200 m. The current total area of production of both durum ( Triticum turgidum var. durum ) and bread wheat ( Triticum aestivum ) is estimated to be between 1.2 to 1.5 million hectares. The environmental variation within the wheat growing range is noticeable ove r short distances. This type of environment obviously affects the adaptation of a particular crop variety and complicates the process of plant breeding. To counterbalance this problem, categor izing environments and conducting multienvironment variety evaluation a nd stability analysis are imperative. In the last two decades, a considerable number of bread wheat varieties we re released in Ethiopia; all were assessed for adaptation and stability over different envir onmental categories before release (Desalegn et al ., 1996; Debebe et al ., 2000; Bedada et al ., 1999; Desalegn et al., 2001). The purpose of this paper is to report on the stability of yield perfo rmance and yield potential of advanced bread wheat genotypes in different environments in Ethiopia. Materials and Methods A national variety trial consisting of 18 bread wheat genotypes was conducted from 1999 to 2001 across the wheat growing agro-ecologies in Ethiopia. The genotypes included advanced materials from a breeding program and selections from introductions. The entries were established on six rows of 2.5m plots with a distance of 20cm between rows. The seeds were drilled into the row at the rate of 150kg ha-1. A randomized complete block design with four replications was used at all locations. Site-s pecific agronomic practices were applied. The central 4 rows were harvested to determ ine the yield potential of each genotype. The testing sites included nine high poten tial locations, two waterlogged sites (Ginchi and Arsi-Robe) and three moisture stress sites (Alemaya, Asasa and Dhera) as described in Table 1. For the three environmental categories, analysis of variance was performed on yield data of individual trials and AMMI analysis was carried out separately on pooled data using Agrobase 99 (Agronomix Software, Inc., 1999).

PAGE 86

Table 1. Environmental characteristics of national bread wheat testing sites in Ethiopia. Location Altitude (masl) Long-term Annual Rainfall (mm) Major production constraint(s) Kulumsa 2200 830 SR Bekoji 2780 1098 YR, STB Adet 2240 1250 SR Holetta 2400 1078 STB Arsi-Negele 1800 763 SR Sinana 2400 1023 YR, SR, STB Hossana 2450 YR, SR Kokate 2150 SR, YR Alemaya 1900 638 SR, drought Ginchi 2100 1093 Waterlogging Arsi-Robe 2400 873 Waterlogging Asasa 2360 644 Terminal drought Dhera 1680 617 Intermittent drought SR= Stem rust; YR=yellow rust; STB=Septoria blotch Results and Discussion In high potential wheat growing areas, the AMMI analysis showed that environments, genotypes and G x E interaction were highly significant. In this set of environment, IPCA1 explained 33 per cent of the G x E interacti on sum of squares and IPCA2 explained 23 per cent of the interaction. The first principal component explained more of the interaction variation and this could be in terpreted in terms of morpholog ical, phenological, agronomic or disease factors that may affect the performance the genotypes (Table 2). In the drought stressed sites, the environments, genotypes and G x E were highly significantly different. About 41 per cent of th e interaction sum of squares was explained by IPCA1. IPCA2 and IPCA3 explained 17% and 14% of the interaction, respectively. The first principal component factor has a high contributi on to the interaction sum of squares while the IPCA2 and IPCA3 are small. This indicates that one primary factor is influencing the G x E interaction in this set of environment (Table 2). Likewise, in the waterlogged areas the envi ronments, genotypes and G x E interaction were significantly different (Table 2). IPCA 1 explained 42% of sum of squares of the interaction, while IPCA2 and IPCA3 were res ponsible for 24% and 17 % of the interaction sum of squares, respectively. This indicates that one fundamental factor affects G x E interaction; this could be either genotypic or environmental in nature. In AMMI analysis, IPCA scores of genotypes are stability indicators. IPCA scores closer to zero (0), either from a positive or negativ e direction, indicate stable performance of genotypes over sampled environments for the particular trait under consideration. Accordingly, in high potential wheat grow ing environments, FH 8-2, which had IPCA score nearest to zero (Table 3), was the most st able variety although it was low yielding. FH 8-2 was developed from a cross between Ethiopian and Germany materials for multiple diseases resistance. Moreover, genotypes HAR 3316, HAR 3354, FH 6-1-7 and ETBWC 037 showed stable and predictable yield performance across the different environments (Figure 1).

PAGE 87

TABLE 2. Analysis of variance fo r multi-environment bread wheat trial over the period of 1999 to 2001. Favorable Environment Low moisture environment Water logged environment Source df MS Prob df MS Prob df MS Prob Total 1799 57 5 43 1 Environments 24 43510329.0 ** 7 98028937.3 ** 5 160558072.6 ** Genotype 17 5002302.5 ** 17 2974331.8 ** 17 6143106.4 ** G x E. 408 1279672.6 ** 11 9 585980.1 ** 85 1158614.0 ** IPCA 1 40 4251121.2 ** 23 1234844.4 ** 21 1969767.6 ** IPCA 2 38 3222205.7 ** 21 565213.7 ** 19 1258203.6 ** IPCA 3 36 1396068.3 ** 19 499430.6 17 1023044.6 IPCA 4 34 1234018.4 ** IPCA 5 32 969418.0 ** IPCA 6 30 979453.2 ** IPCA 7 28 779923.7 ** Residual 1275 414285.4 40 8 284442.936 30 6 433747.8 *, ** =significant at 0.05 and 0.01 levels respectively TABLE 3. Genotype IPCA Axis 1 Scores for High potential environment No Genotype Score Mean No Genotype Score 1 a HAR 1522 20.5423 3 688.48 | 11 11 HAR 2870 29.0941 2 b HAR 3224 16.4640 3908.44 | 10 10 HAR 2818 26.0322 3 c HAR 3116 -2.0432 4091.36 | 8 8 HAR 2812 20.8600 4 d HAR 3354 2.1311 4036.38 | 1 1 HAR 1522 20.5423 5 e ETBWC026 -10.4608 3621.87 | 2 2 HAR 3224 16.4640 6 f ETBWC028 -10.5972 3872.90 | 14 14 FH 9-3-4 10.7835 7 g ETBWC037 2.7556 3560.85 | 9 9 HAR 2814 7.7190 8 h HAR 2812 20.8600 3691.62 | 7 7 ETBWC037 2.7556 9 i HAR 2814 7.7190 3 832.50 | 4 4 HAR 3354 2.1311 10 j HAR 2818 26.0322 3963.50 | 17 17 FH 8-2 1.1397 11 k HAR 2870 29.0941 3746.58 | 3 3 HAR 3116 -2.0432 12 l FH 4-2-11 -17.3445 3787.62 | 16 16 FH 6-1-7 -2.5637 13 m FH 81 -27.9384 4039.08 | 5 5 ETBWC026 -10.4608 14 n FH 9-3-4 10.7835 4117.90 | 6 6 ETBWC028 -10.5972 15 o FH 7-1-5 -24.6687 4059.52 | 12 12 FH 4-2-11 -17.3445 16 p FH 6-1-7 -2.5637 3459.06 | 15 15 FH 7-1-5 -24.6687 17 q FH 8-2 1.1397 3372.47 | 13 13 FH 81 -27.9384 18 r L.CHECK -41.9050 3986.09 | 18 18 L.CHECK -41.9050

PAGE 88

29.094 0 -41.90 2535.8 3824.2 5109.3 key for the letters in the AMMI biplot A= A-NEGELLE 01, B=ADET 01, C=BEKOJI 01, D=DEBRE-ZEIT 01, E=HOLETTA 01, F=KOKATE01, G= KULUMSA 01, H=SINANA01, I=A-NEGELLE 00, J= ADET 00, K= BEKOJI 00, L=D-ZEIT 00, M= HOLETTA 00, N=HOSSAENA 00, O=KOKATE 00, P=KULUMSA 00, Q=SINANA 00, R=A-NEGELLE 99, S=ADET 99, T=BEKOJI 99, U=HOLETTA 99, V=HOSSAENA 99, W= KOKATE 99, X=KULUMSA 99, Y= SINANA 99; a=HAR 1522, b=HAR 3224, c=HAR 3116, d=HAR 3354,e=ETBWC026, f=ETBWC028, g=ETBWC037, h=HAR 2812, I=HAR 2814, j= HAR 2818, k=HAR 2870, l=FH 4-2-11, m=FH 81, n=FH 9-3-4, o=FH 7-1-5, p=FH 6-1-7, q=FH 8-2, r=L.CHECK Figure 1. AMMI biplot for bread wheat vari ety trial consisting 18 genotypes and 25 environments (high potential) over the period of 1999 to 2001. (Note: 1 genotype and 3 environments in place of others with similar means and not shown) On the other hand, FH 8-1 and FH 7-1-5 s howed high negative IPCA1 scores, which is evidence of their specific adaptability to favor able environments. It was noted that the genotypes have considerably high variation around the mean grain yield of 3824 kg ha-1. The test sites also showed year-to-year variation in mean grain yield and were evenly distributed in all four quadrants indicating the importance of seasonal climatic variation. Clearly, two or more seasons of testing are better than a single year. High seasonal variation was observed among the drought stressed environments (Figure 2). This is probably attributable to seasonal ra infall variation both in amount and distribution. The genotypes tended to vary widely in grain yield around the grand mean (3576.8 kg ha-1). Among these genotypes, HAR 2818 gave the highest mean grain yield indicating that it could be an ideal variety for low moisture stress areas; while ETBWC026 was the most stable genotype in this environment (Table 4).

PAGE 89

For water logged environment, genotypes lik e ETBWC037 and FH 4-2-11, which had IPCA score close to zero, were found to perform stably and may be used in such areas as Arsi Robe and Ginchi (Table 5) 39.3 0 -29.43 1998.8 3576.8 5219.8 key for the letters in the AMMI biplot A=AS01, B=DH01, D=AS00, E=DH00, F=AS 99, G=DH99, H=AL99, I=AL01; a=HAR 1522, b=HAR 3224, c=HAR 3116, d= HAR 3354,e=ETBWC026, f=ETBWC028, g=ETBWC037, h=HAR 2812, I=HAR 2814, j= HAR 2818, k=HAR 2870, l=FH 4-2-11, m=FH 81, n=FH 9-3-4, o=FH 7-1-5, p=FH 6-1-7, q=FH 8-2, r=L.CHECK Figure 2. AMMI biplot for bread wheat variety trial consisting 18 genotypes and 9 environments (moisture stress areas) over the period of 1999 to 2001.

PAGE 90

TABLE 4. Genotype IPCA Axis 1 Scores for Moisture stress environment No Genotype Score Mean No Genotype Score 1 a HAR 1522 -8.4258 3571.91 | 8 HAR 2812 30.0680 2 b HAR 3224 -0.9773 3829.22 | 16 FH 6-1-7 18.7863 3 c HAR 3116 -19.9266 4065.53 | 18 L.CHECK 14.0816 4 d HAR 3354 -4.8353 3701.38 | 9 HAR 2814 8.0520 5 e ETBWC026 0.2994 3561.94 | 17 FH 8-2 5.5885 6 f ETBWC028 4.4483 3638.34 | 6 ETBWC028 4.4483 7 g ETBWC037 3.9407 3348.19 | 7 ETBWC037 3.9407 8 h HAR 2812 30.0680 3445.28 | 5 ETBWC026 0.2994 9 i HAR 2814 8.0520 3545.06 | 2 HAR 3224 -0.9773 10 j HAR 2818 -5.8004 4134.16 | 13 FH 81 -2.1974 11 k HAR 2870 -2.9071 3390.59 | 11 HAR 2870 -2.9071 12 l FH 4-2-11 -13.4219 3373.72 | 4 HAR 3354 -4.8353 13 m FH 81 -2.1974 3680.75 | 10 HAR 2818 -5.8004 14 n FH 9-3-4 -13.3237 3891.50 | 1 HAR 1522 -8.4258 15 o FH 7-1-5 -13.4493 3535.34 | 14 FH 9-3-4 -13.3237 16 p FH 6-1-7 18.7863 2938.75 | 12 FH 4-2-11 -13.4219 17 q FH 8-2 5.5885 3054.97 | 15 FH 7-1-5 -13.4493 18 r L.CHECK 14.0816 3676.22 | 3 HAR 3116 -19.9266 TABLE 5. Genotype IPCA Axis 1 Scores for water-logged environment No Genotype Score Mean No Genotype Score 1 a HAR 1522 1.0963 4994.58 | 6 ETBWC028 21.1901 2 b HAR 3224 -3.0194 5094.63 | 17 FH 8-2 18.9806 3 c HAR 3116 -12.8362 5441.63 | 16 FH 6-1-7 12.4562 4 d HAR 3354 -1.9025 5489.25 | 15 FH 7-1-5 11.3533 5 e ETBWC026 -15.6519 4 746.42 | 11 HAR 2870 9.9798 6 f ETBWC028 21.1901 4486.79 | 18 L.CHECK 9.9606 7 g ETBWC037 1.8652 4576.63 | 10 HAR 2818 6.9080 8 h HAR 2812 -9.4631 4733.50 | 7 ETBWC037 1.8652 9 i HAR 2814 -3.0989 4962.17 | 12 FH 4-2-11 1.6042 10 j HAR 2818 6.9080 5120.92 | 1 HAR 1522 1.0963 11 k HAR 2870 9.9798 4760.33 | 4 HAR 3354 -1.9025 12 l FH 4-2-11 1.6042 4129.79 | 2 HAR 3224 -3.0194 13 m FH 81 -16.7825 4533.42 | 9 HAR 2814 -3.0989 14 n FH 9-3-4 -32.6399 5225.58 | 8 HAR 2812 -9.4631 15 o FH 7-1-5 11.3533 4313.71 | 3 HAR 3116 -12.8362 16 p FH 6-1-7 12.4562 4298.67 | 5 ETBWC026 -15.6519 17 q FH 8-2 18.9806 3399.08 | 13 FH 81 -16.7825 18 r L.CHECK 9.9606 4744.17 | 14 FH 9-3-4 -32.6399

PAGE 91

33.7 0 -32.6 Key for the letters in the AMMI biplot A=Arsi-Robe 01, B=Ginchi 01, C=Arsi-Robe 00, D=Ginchi 00, E=Arsi-Robe 99, F=Ginchi 99; a=HAR 1522, b=HAR 3224, c=HAR 3116, d=HAR 3354,e=ETBWC026, f=ETBWC028, g=ETBWC037 h=HAR 2812, I=HAR 2814, j=HAR 2818, k=HAR 2870, l=FH 4-2-11, m=FH 81, n=FH 9-3 -4, o=FH 7-1-5, p=FH 6-1-7, q=FH 8-2, r=L.CHECK Figure 3. AMMI biplot for bread wheat variety trial consisting 18 genotypes and 6 environments (water logging stress) over the peri od of 1999 to 2001. (Note: one environments in place of other with similar means and not shown) The genotypes were ranked for their yield stability in the favorable environments according to each genotype by environment interaction statistical analysis procedure used.(Table 6). Spearmans rank correlation coefficient was computed for each of the possible pairs of the G x E statistics (Table 7). Students t test for the Spearmans rank correlation coefficients showed no signifi cant correlation with AMMI, but was highly significantly associated with the other proce dures. A similar trend was noted for the two extreme moisture stress environments (data not presented). Clearly, the stability analysis procedures do not conform to AMMI analysis. From the results, we can conclude that HAR 3116 and HAR 3354 are high yielding and stable across high potential environments; HAR 3224 and HAR 2818 perform better in the drought stress environment while FH 4-2-11 and ETBWC037 are tolerant to waterlogged vertisols.

PAGE 92

TABLE 6. Estimates of various G x E statistics and stability parameters for 18 bread wheat genotypes grown at 25 environments during 1999 to 2001. Stability Variance No Genotypes Cultivar Superiority Measure, Lin and Binns Ecovalence, Wricke Shukla Finlay and Wilkinson Beta Deviation from Linearity 1 HAR 1522 996717 5817426 1010788 1011008 0.873 121138 2 HAR 3224 735159 4445462 753545 784681 1.035 70843 3 HAR 3116 525681 5887418 1023911 978553 1.180 113926 4 HAR 3354 530726 5952597 1036132 993175 1.178 117176 5 ETBWC026 973059 5052679 867398 835330 0.841 82099 6 ETBWC028 716908 9040414 1615098 1681013 1.047 270028 7 ETBWC037 1024086 2243215 340623 300359 0.858 -36784 8 HAR 2812 1086634 7680015 1360023 1308673 0.801 187286 9 HAR 2814 767383 6031427 1050913 1057356 1.120 131438 10 HAR 2818 828490 7437071 1314471 1372687 0.983 201512 11 HAR 2870 1084103 9258608 1656010 1724961 0.958 279795 12 FH 4-2-11 743175 7003698 1233214 1282976 0.955 181576 13 FH 81 463092 8720217 1555061 1593398 1.105 250558 14 FH 9-3-4 555071 7247264 1278882 1336333 0.995 193433 15 FH 7-1-5 422436 6374056 1115156 1140053 1.095 149815 16 FH 6-1-7 1279747 10782393 1941719 2027627 0.988 347054 17 FH 8-2 1288069 2611873 409747 429482 1.001 -8090 18 L.CHECK 708448 18940766 3471414 3623810 0.987 701761 TABLE 7. Spearmans rank correlation for various stability parameters Cultivar Superiority Measure Ecovalence (Wricke) Stability Variance (Shukla) Stability Variance (Finlay & Wikinson) Deviation from Linearity AMMI Cultivar Superiority Measure 0.88* 0.88* 0.88* 0.66* 0.36 Ecovalence (Wricke) 1.00* 0.99* 0.90* 0.13 Stability Variance Shukla) 0.99* 0.90* 0.13 Stability Variance Finlay & Wikinson) 0.90* 0.14 Deviation from Linearity 0.04 *Students t test is significant at 0.01 level of significance

PAGE 93

References Bedada Girma, Desalegn Debelo and Bekele Geleta. 1999. Evalua tion and characterization of some bread wheat genotypes for waterlogging tolerance in Ethiopia. African Crop Science Proceedings, vol.4.pp .113-116, Uganda. Crossa, J. 1990. Statistical analysis of multilocation trials. Advances in Agronomy 44 : 55-86. Cochran, W. G. and Cox, G. M. 1957. Experimental designs. John Wiley &Sons Inc., New york, 2nd Ed. Debebe Masresha, Desalegn Debelo, Bedada Girma, Solomon Gelalcha and Balcha Yaie.2000. Bread wheat yield stability and environmental clustering of major wheat growing zones in Ethiopia. pp. 78-86. In : CIMMYT. 2000. The Eleventh Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. Desalegn Debelo, Bekele Geleta, Balcha Yaie and Zewdie Alemayehu. Grain Yield Response of Some Bread Wheat Cultivars in Diverse Environments of Ethiopia. pp.382-86. In : Tanner, D. G., Payne, T. S., and Abdalla, O. S. Eds. 1996. The Ninth Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. Desalegn Debelo, Bedada Girma, Zewdie Alemaye hu and Solomon Gelalcha. 2001. Drought tolerance of some bread wheat genotypes in Ethiopia. African Crop Sci. J 9(2): 385-392. Eberhart, S. A. and W. A. Russell. 1966. Stability parameters for comparing varieties. Crop sci. 6 : 3640. Finlay, K. W. and G. N. Wilkinson. 1963. The analysis of adaptation in plant breeding programs. Austral. J. Agri. Res. 14: 742-754. Guach, H. G. 1988. Model selection and valid ation for yield trials with interaction. Biometrics 88: 705715. Guach, H. G. and R. W. Zobel.1988. Predictive and po stdictive successes of statistical analysis of yield trials. Theor. Appl. Genet 76 : 1-10. Hanson, H., Borlaug, N. E. and Anderson, R. G. 1982. Wheat in the third world. Boulder, CO, USA, Westview Press. Lin, C. S. and Binns, M. R. 1988. A superiority m easure of cultivar performance for cultivar x location data. Can. J. Plant Sci. 68: 193-198. Plaised, R. L. and Peterson, L. C. 1959. Atechnique for evaluating the ability of selection to yield consistently in different locations and seasons. Amer. Potato J. 36: 381-385. Shukla, G. K. 1972. Some statistical aspects of partitioning genotype-environmental components of variability. Heredity 28: 237-245. Wright, A. J. 1971. The analysis and prediction of some two factors interactions in grass breeding. J. Agric. Sci 76: 301-306. Yates, E. and Cochran. 1938. The analysis of grouped experiments. J. Agric. Sci 28: 556-580.

PAGE 94

Seedling and Adult Plant Resistance in Ethiopian Wheat Varieties to Local Puccinia graminis Isolates Emebet Fekadu1, Belayneh Admassu2 and Zerihun Kassaye Ethiopian Agricultural Research Organization, Plant Protection Research Center, P.O. Box 37, Ambo, Ethiopa: E-mail: 1 f_emebet@yahoo.com 2 belay120@yaho.com Abstract Wheat entries showing resistance to st em rust at the seedling stage do not necessarily possess adult plant resistance, and vice-versa. This investigation was therefore undertaken to ev aluate some wheat varie ties from Ethiopia for their resistance to four virulent stem rust is olates collected from Ambo and Debre Zeit (two isolates from each local ity) at seedling as well as ad ult plant stages. Out of 15 durum ( Triticum turgidum ) and 14 bread wheat ( T. aestivum ) varieties tested, 10 durum and 7 bread wheat varieties were found to be resistant both at the seedling and adult stages to the first Ambo isolate. Whereas 11 durum and 13 bread wheat varieties were resistant to the second isolate at both growth stages. Eight durum and 8 bread wheat varieties exhibited a resistant infection type to the two Debre Zeit isolates at both growth stages. Some mate rials, for instance durum wheat varieties Kilinto and Gerardo were resistant at the seedling stage but were susceptible at the adult stage, whereas variety DZ 169505 exhibited a suscep tible and resistant reaction at the seedling and adult stages respectively to the first Debre Zeit isolate. Our results indicate that selection for resistance must involve rigor ous field evaluation with a range of isolates to achieve effective resistance. Introduction Stem rust, caused by Puccinia graminis f. sp. tritici occurs every year in the wheat growing regions of Ethiopia. For over 20 years, stem rust has been controlled through the use of resistant cultivars. Appearance of new races a nd conditions favorable for diffusion of the pathogen enhance the likelihood of epidemics that may significantly reduce wheat yield. Although host resistance is the best method to avoid epidemics and minimize yield losses, changes in pathogen virulence and the attendant loss of resistance complicates this control strategy. Goludan et al. (1928) reported that adult plant resistance was inherited independently of seedling resistance. Mutkekar et al. (1985) also reported similar results stating that entries showing resistance at the seedling stage do not necessarily posses adult plant resistance and vice-versa. The present inv estigation was undertaken to evaluate selected wheat varieties for their resistance to stem ru st at seedling as well as adult plant stages. The resistant varieties are recommended as the best sources for breeding programs. Materials and Methods Stem rust field populations from Ambo and Debr e Zeit were multiplied on the adult plants of the susceptible variety Morocco in a gree nhouse under conditions favourable for the development of the pathogen. Two most virulent isolates from each location were selected for the present study.

PAGE 95

Four sets of the seedlings and adult plants of 29 bread and durum wheat cultivars were grown in 4*4 cm size earthen pots in a glass house where they were maintained disease free. After inoculating seedlings at 2-leaf stage and adult plan ts at 5-leaf stage with individual races, each set were placed in a moist chamber for 24 hour s for incubation and then transferred to glass house benches for the development of uredinia ( pustules). Rust reactions were recorded about 14 days after inoculation by following the 0-4 scale (Stakman and Levine, 1922) where 0 ; = immune, 1-2 = resistant, and 3 and 4 = susceptible. Results Seedling and adult plant reactions of the cultivars are given in Tables 1 and 2 for isolates from Ambo and Debre Zeit, respectively. Variety 2 exhibited highly resistant reactions to both isolates at both growth stages with Ambo isolates whereas varieties 5, 6, 7, 9, 10, 11, and 14 exhibited moderately resistant reactions and he nce are suggested to be used as alternatives (Table 1). Bread wheat varieties exhibited highly r esistant reaction compared to durum wheats. Some varieties were immune to the pathogen is olates at either of the two growth stages. Varieties 8 and 9 were highly resistant with immune or resistant reactions to the Ambo isolates at both growth stages. Varieties 3, 7, 10, 13, 14 had a reaction type ranging from immune to moderately resistant (Table 1). He nce, these varieties could provide additional resistance sources. With Debre Zeit isolates varieties 8, 9, 13 had immune to highly resistant reaction type. As alternatives 2, 3, 10, 12 and 14 could be used as their reactions ranged from immune to moderately resistant to the two pathogen isolates (Table 2). Some materials, for instance durum wheat vari eties Kilinto and Gerardo were resistant at the seedling stage but were susceptible at th e adult stage, whereas variety DZ 169505 exhibited a susceptible and resistant reaction at the seedling and adult stages, respectively to the first Debre Zeit isolate (Table 2). Resistance that could be to all is olates can be generated by carefully choosing the right parents b ased on the data we have presented here. Table 1. Reaction of seedling and adult wheat plants to p.graminis isolate from Ambo Isolate 1 Isolate 2 Plot no. Variety Seedling infection type(04scale) Adult infection type (0-4scale) Seedling infection type(0-4scale) Adult infection Type (04scale) Durum 1 Assassa 2 3 2 2 2 Tob-66 1 1 1 1 3 Kilinto 1 3 2 2 4 Foka 3 2 1 1 5 Gerardo 1 1 2 2 6 DZ1640 2 1 2 2 7 DY1050 2 2 2 2 8 Cocorit-71 1 3 1 3 9 Yielma 2 1 1 1 10 DZ1928-2 1 1 1 2 11 DZ 1691 2 2 2 2 12 DZ 1695-5 2 2 3 2 13 DZ1543 1 2 1 3 14 Boohai 2 2 2 2 15 Bichena 3 2 1 3

PAGE 96

Local check 4 4 4 4 Bread Wheat 1 Kubsa 2 4 2 3 2 Wabe 3 3 0 1 3 Galama 2 2 2 2 4 Tuse 3 3 0 1 5 Katar 3 2 2 ; 6 Shina 3 3 0 2 7 Hawi 0 ; 0 0 8 Simba 0 ; 0 0 9 Wetera 0 0 0 ; 10 HAR 2192 1 2 0 2 11 HAR 2508 3 1 ; 1 12 F-H-6-1-7 1 3 ; 1 13 Abola ; 2 0 0 14 Tura 2 1 0 ; Local check 4 4 4 4

PAGE 97

Table 2. Seedling and adult plant reactions (0-4 scale) of wheat varietie s when tested with to P. graminis isolates from DebreZeit Isolate 1 Isolate 2 No. Variety Seedling reaction Adult plant reaction Seedling reaction Adult plant reaction Durum wheat 1 Assassa 2 2 2 3 2 Tob-66 1 2 1 ; 3 Kilinto 2 3 ; 1 4 Foka 1 2 1 1 5 Gerardo 2 3 1 1 6 DZ1640 1 2 1 2 7 DY1050 2 2 2 2 8 Cocorit-71 ; 1 1 2 9 Yielma 3 4 1 2 10 DZ1928-2 2 2 1 2 11 DZ 1691 1 2 2 2 12 DZ 1695-5 3 2 1 2 13 DZ1543 2 2 1 2 14 Boohai 3 3 3 1 15 Bichena 3 4 2 2 Susceptible check 4 4 4 4 Bread Wheat 1 Kubsa 1 3 1 0 2 Wabe 2 2 1 0 3 Galama 2 2 ; 2 4 Tuse ; 2 1 3 5 Katar 3 3 1 0 6 Shina 2 4 1 2 7 Hawi 3 0 ; 0 8 Simba ; 0 0 0 9 Wetera ; 0 ; 0 10 HAR 2192 1 2 1 0 11 HAR 2508 2 3 1 3 12 F-H-6-1-7 1 2 1 ; 13 Abola 1 ; ; 0 14 Tura ; 2 ; 0 Susceptible check 3 4 3 4 REFERENCES Goluden, C.H., Neatby, K.W. and Welsh, J.N. (1928). The inheritance of resistance of Puccinia graminis tritici in a cross between two varieties of Triticum vulgare Phytopathology 18:631658. Mutkekar, M.L., Bhangale, G.T., Patil, J.V. and Kalekar, A.R. (1985). Source of resistance to stem rust of wheat. Cereal Rust Bulletin 13:19-22. Stakeman, E.C. and Levine, M.N. (1922). Th e determination of physiologic forms of Puccinia graminis on Triticum spp. Tech. Bull. 10 Univ. Minn. Agric. Exp. Stn. 1-10pp.

PAGE 98

Evaluation of Kenyan Wheat ( Triticum aestivum L.) Lines for Bread Making Quality Kimani E.N., J. Ndungu, M.G. Kinyua and J. Owuoche KARINjoro P.O. Private Bag20107 Njoro Abstract The grain composition and baking quality of wheat ( Triticum aestivum L.) are important in determining end-us e product acceptability, although some of these factors are influenced by environmenta l conditions. This study investigates the bread making quality and protein content of Kenyan wheat advanced lines. Eleven lines and one wheat variety (control) were planted in four locations in 2001 and 2002, and evaluated for flour extraction rate; protein content; farinographic dough water absorption and dough development time (DDT); and bread loaf volume, using official methods of the American Association of Cereal Chemists. There were significant (p< 0.01) effects associated with year of planting for protein, DDT; water absorption and loaf volu me. Effects due to genotype were significant (p< 0.01) for flour yield and DDT. No significant (P>0.05) genotype year interaction was observed. Genotype K7972-1 had 5.94% higher mean FY than the control K. Heroe. Lines 92B20 and R932 showed 4.5 and 2.67mins, respectively, longer DDT than K. Heroe. R899 had the highest mean flour Water Absorption (WA), 2.18% more than the control. R946 had the highest Loaf Volume (LV) of 619.38cm3 (20.1% higher than K. Heroe). K. Heroe recorded the lowest loaf volume of 515.63 cm3. 92B19, 92B24 and R946 had higher protein content, 19.2%, and 17.1% and 15.9%, respectively, than K. Heroe. Protein content and loaf vo lume correlated significantly (r = 0.657** and 0.707**, respectively) with DDT. Protein content and loaf volume correlated significantly (r= 0.642**) while flour yield correlated negatively (r= 0.462**) with flour water absorption. From this study, we suggest that 92B9, 92B19 and R946 can be used in the breeding program to improve the quality of wheat. Introduction Grain and baking quality of wheat ( Triticum aestivum L ) refers to its suitability for the type of processing or utilization for which the raw ma terial is destined (Morris and Rose, 1996). Grain composition and baking quality are variab le factors that depend on both genotype and the growing conditions. Environmental conditi ons such as weather-related factors, soil fertility, temperature, and soil moisture regim es have a major influence on grain and end use quality of wheat (Peterson et al, 1998). Therefore, there is genetic variation in wheat baking quality although some of the quality factors ar e influenced by the environment (Busch et al. 1969; Baezinger et al. 1985; Bassett et al. 1989; Peterson et al. 1992). Wheat strength is usually associated with flour protein. Both prot ein quantity and quality are considered to be primary factors in measuring the quality of flour in relation to bread making. The quantitative expression of crude protein is related to tota l organic nitrogen in the flour, whereas quality evaluations relate specially to physicoche mical characteristics of the gluten-forming component. Protein quality criteria are related pr imarily to the gluten portion of the flour protein. Glutenins and gliadins are the major components of the storage protein in wheat and make a significant contribution to dough rheology and baking quality (Payne, 1987; Weegels et al. 1996). Quality is appraised largely by subj ecting the flour to several physical testing devices, which measure various rheological charac teristics. The tests are performed usually on flour-water dough. They characterize the dough as related to the properties of the gluten

PAGE 99

portion of the protein. Resistance to extension of dough at rest, hydration time, maximum development time and tolerance or resistance to breakdown at a predetermined consistency during mechanical mixing, are some of the most common dough parameters evaluated. Mixing characteristics are usually related to gluten quality, and can be reasonably well defined by the use of recording dough mixers such as the farinograph. Mixing requirements of bread dough correlate with some of the measur ements obtained with recording dough mixers. Performance of flour is evaluated experime ntally under conditions that are similar to those applied by the millers and bakers. The ev aluation of flour is carried out by test baking depending on intended use. Tests are perfor med under rigidly controlled and standardized condition (AACC methods). Flour water absorption is an important factor in production of all types of baked goods. Usually high water absorption values are desirable, since they tend to increase unit yields of baked products. It is measured as the am ount of water required to yield dough of predetermined consistency (Pratt, 1971). In Kenya wheat is mainly used for chapatti and bread making but rarely used for pasta production. The increased demand for these products has prompted the breeding of more new varieties with good qualities. Although yield and resistance to diseases and pest have been the main objective of breeding work, recently, there has been an increasing emphasis on improvement of quality of the grain for milling a nd baking. The objective of this study was to investigate the bread making quality of advanced wheat lines, evaluated at four locations in Kenya. Materials and Methods Two wheat genotypes developed in Kenya (K 7972-1 and K7872), five CIMMYT lines (R946, R891, R892, R932, R899) and four from Kenya Seed Company (92B9, 92B19, 92B20 and 92B24) were used in this experiment. K. Heroe a variety was used as a control. The genotypes are described in Table 4. The experiment was conducted at Lanet (0'S; 36'E), Naivasha (0'S; 36'E), Mau Narok (0'S; 35'E) in Rift Valley Province a nd at Katumani (1'S; 37'E) in the Eastern Province of Kenya. The cultural practices were followed according to the recommendations in all sites in order to sustain plant growth and production of grain. The seeding rate was maintained at 138 Kg ha-1. Diammonium phosphate fertilizer was applied at the rate of 231 Kg ha-1 in order to supply 42 Kg N and 100g of DAP (18% N + 20% P) per plot during planting time. Pre-emergence herbicide STOMPR (Pendimethalin 500g/litre) was applied at sowing time to restrict early weed growth. The e xperiment was laid out in a partially balanced lattice design, replicated three times and in each re plicate, there were 4 blocks with 4 entries. At physiological maturity, the crop was harvested by Hege combine Model 140, seed cleaned and composited. Due to inadequate seed for entri es, the sites were considered as replicates. About 800-1000g of wheat samples were sieved and cleaned, tempered to 15% moisture content (MC) and left overnight at room temperature about 20-21 C. These samples were then milled using a Buhler mill (Buhler Brot hers Ltd., Switzerland) to determine flour extraction. The milled products were withdrawn and weights of all six flour streams (blended) and of shorts and bran were determined. Flour yield was calculated using equation: 100 ccov products ered milled total re Flour FlourYield Chemical, rheological, and baking tests were carried out to evaluate the baking quality of the wheat. The flour water absorption (WA) and dough development time (DDT) were analysed

PAGE 100

using a small bowl (50g) Brabender Farinogr aph (brabender OHG. Duisburg Germany) as described in the AACC method 54-21 (AACC, 1983). The amount of flour used was 50g (14% moisture basis). Flour water absorption is the amount of water required to centre the farinograph curve on the 500 brabender unit line. Dough development time (DDT) is the durati on required to develop dough visco elastic properties. It is the interval from first addition of water to that point of maximum consistency range immediately before first indication of weak ening. This value has also been referred to as peak or peak time. Bread making was performed using a modified basic straight dough method. The baking formula used included 100g flour at 14% moistu re basis (mb), 1.2g yeast, 2.5g sugar, 1g salt, 3g dry milk powder, 3g fat and 1ml of bromat e phosphate solution. Dough ingredients were mixed until optimum mixing time and then the dough was fermented (30C and 80% relative humidity) for about four hours, with intervals of moulding the dough and finally proofing for 55minutes. This was then baked for 25 minutes in the oven at 220C. Loaf volume (LV) was measured using the Rapeseed displacement method 10 minutes after the loaf was removed from the oven, using a Loaf Volumeter (National Mfg. Co., Lincohl, Nebraska.) Crude flour protein (PC) analysis was done using the Kjeldahl method (N X 5.7) using a Tecator Kjeltec system. SAS system was used to do statistical analys is using the General Linear Model. Analysis of variance for all the parameters, involving sites and years was done for each quality trait. Pearson correlation was done between so me of the baking quality parameters. Results The mean year effects were significantly different (p< 0.01) for all the traits (Table 1). The effects due to genotype were also significant (p< 0.05) for FY and DDT. No significant (P>0.05) effects due to year genotype interaction were observed. A comparison of the genotypes for the two years showed much difference across the traits (Table 2). The means of the second year were lower for flour yield and higher for DDT, LV and PC than those of the first year. Among the tested genotypes, K7972-1 produced 5.94% more FY than the control K. Heroe (Table 4). 92B20 and R932 showed longer differences of DDT of 4.5 and 2.67min, respectively, as compared to K. Heroe. R899, 92B19 and 92B24 showed higher water absorption than the control (Table 4). The loaf volumes were generally low for a ll the genotypes (lower than the desirable 700cm3 and above). The highest loaf volume was shown by R946 with 619.38 cm3. This was higher than K. Heroe by 20.1%. Genotype 92B 19, 92B24 and R946 exhibited protein ranging from 118.30 .01 g/kg. This was 19.2%, 17.1% and 15.9% than K. Heroe (Table 4). Table 1: Mean Squares for flour yield, Dough De velopment, Flour Water absorption, loaf volume and protein content of the 12 wheat genotypes ( Triticum aestivum L.) Source DF FY (%) DDT (min) WA LV (cm3) PC g.Kg-1 Year 1 343.01** 0.90** 32.67** 112951.40** 1889.08** Genotype 11 29.68* 0.37** 3.92 9422.79 237.93 Rep 3 107.78 7.22 12.23 314602.39 8021.20 Gen*Year 11 29.19 0.13 9.14 7482.89 138.81 Error 64k 12.86 0.10 4.14 5492.40 167.14 C.V 5.54 13.72 3.36 13.26 11.75 *, ** P< 0.05, P< 0.01

PAGE 101

portion of the protein. Resistance to extension of dough at rest, hydration time, maximum development time and tolerance or resistance to breakdown at a predetermined consistency during mechanical mixing, are some of the most common dough parameters evaluated. Mixing characteristics are usually related to gluten quality, and can be reasonably well defined by the use of recording dough mixers such as the farinograph. Mixing requirements of bread dough correlate with some of the measur ements obtained with recording dough mixers. Performance of flour is evaluated experime ntally under conditions that are similar to those applied by the millers and bakers. The ev aluation of flour is carried out by test baking depending on intended use. Tests are perfor med under rigidly controlled and standardized condition (AACC methods). Flour water absorption is an important factor in production of all types of baked goods. Usually high water absorption values are desirable, since they tend to increase unit yields of baked products. It is measured as the am ount of water required to yield dough of predetermined consistency (Pratt, 1971). In Kenya wheat is mainly used for chapatti and bread making but rarely used for pasta production. The increased demand for these products has prompted the breeding of more new varieties with good qualities. Although yield and resistance to diseases and pest have been the main objective of breeding work, recently, there has been an increasing emphasis on improvement of quality of the grain for milling a nd baking. The objective of this study was to investigate the bread making quality of advanced wheat lines, evaluated at four locations in Kenya. Materials and Methods Two wheat genotypes developed in Kenya (K 7972-1 and K7872), five CIMMYT lines (R946, R891, R892, R932, R899) and four from Kenya Seed Company (92B9, 92B19, 92B20 and 92B24) were used in this experiment. K. Heroe a variety was used as a control. The genotypes are described in Table 4. The experiment was conducted at Lanet (0'S; 36'E), Naivasha (0'S; 36'E), Mau Narok (0'S; 35'E) in Rift Valley Province a nd at Katumani (1'S; 37'E) in the Eastern Province of Kenya. The cultural practices were followed according to the recommendations in all sites in order to sustain plant growth and production of grain. The seeding rate was maintained at 138 Kg ha-1. Diammonium phosphate fertilizer was applied at the rate of 231 Kg ha-1 in order to supply 42 Kg N and 100g of DAP (18% N + 20% P) per plot during planting time. Pre-emergence herbicide STOMPR (Pendimethalin 500g/litre) was applied at sowing time to restrict early weed growth. The e xperiment was laid out in a partially balanced lattice design, replicated three times and in each re plicate, there were 4 blocks with 4 entries. At physiological maturity, the crop was harvested by Hege combine Model 140, seed cleaned and composited. Due to inadequate seed for entri es, the sites were considered as replicates. About 800-1000g of wheat samples were sieved and cleaned, tempered to 15% moisture content (MC) and left overnight at room temperature about 20-21 C. These samples were then milled using a Buhler mill (Buhler Brot hers Ltd., Switzerland) to determine flour extraction. The milled products were withdrawn and weights of all six flour streams (blended) and of shorts and bran were determined. Flour yield was calculated using equation: 100 cov products eredmilled total re Flour FlourYield Chemical, rheological, and baking tests were carried out to evaluate the baking quality of the wheat. The flour water absorption (WA) and dough development time (DDT) were analysed

PAGE 102

Table 4. Pedigrees and grain characteri stics (hardness and colour) of wheat ( Triticum aestivum L.) genotypes evaluated for milling, grain and ba king qualities and Ls means of the 12 wheat genotypes over the two years and four locations Discussion Lill D Van (1995) in his study on multivariate as sessment of environmental effects indicated that climatic factors contributed most to variation in yield and baking and milling quality characteristics. This explains the variation due to year effects, as the locations did not change. The change due to environmental effects also caused the differences in the parameters between the two years for all the genotypes. The correlation between loaf volume and DDT indicates that a longer dough development time is required to obtain large loaf volumes. It suggests that one can predict the volume of loaf from the dough development time obtained from a farinogram. The highly significant correlation between DDT and protein observed in this study concurs with findings of Bushuk et al. (1969). This suggests that there is a linear relationship between protein content and dough development time. As the protein content increases, there is increase in duration for de velopment of dough. DDT relates to changes occurring in gluten proteins during mixing (Preston and Kilborn, 1984.) The dough of the two genotypes, R932 and 92B20, with long DDT started weakening after a long duration of time. They showed exceptionally strong dough thus the high dough development time. This was especially so for the Katumani site. Variation in loaf volume is mainly influenced by protein content (Pomeranz, 1971). In this experiment there is a linear relationship between protein content and loaf volume (Table 3). Lill D. Van (1993) found that loaf volume actually correlates significantly with gliadin and gluten in, the storage proteins in wheat. It can therefore be concluded that within a genoty pe, DDT and loaf volume both depend on the protein content. Flour yield is an essential pa rameter for milling profitability. Table 3 shows that there is negative correlation between flour yield and flour water absorption. Starch when damaged during milling takes up greater quantities of water than the 35% by intact starch. The quantity of additional water absorbed by dough as a result of starch damage depends on the degree and extent of starch damage (Sands tedt, 1955). Cultivars with a high proportion of Genotype Pedigree Grain Colou r Grain Hardnes s FY (%) DDT (min) WA (%) LV (cm3) PC (g/kg) K7972-1 PASA/WED 186//K7837/WED365 Ambe r Hard 69.13 4.07 60.10 603.75 110.58 K7872 K.CHIRIKU/WED354//BOUNTY White Soft 67.50 3.88 59.78 546.25 112.58 R946 OASIS/SKAUZ//4*BCN CMSS93Y04054M2M-OY Red Hard 63.00 4.97 60.88 619.38 115.01 R891 ATTILA CM83836-5OY-OM-OY-3M-05Y Red Hard 62.88 3.56 60.10 520.63 103.96 R892 IL-752264/4/CAR//KAL/BB/3/NAC/5/GAA/CM11 2735-OTOPY-18M-020Y-010M-3Y-0101YOM Ambe r Hard 65.00 4.50 60.14 523.75 111.98 R932 PRL/SARA//TSI/VEE\50 CM103443-41M030Y-020Y-010M-2Y-OM-OSY Ambe r Hard 62.88 8.09 59.51 570.85 103.87 R899 PGO/SARA CM95414-OM-IY-030M-020Y010M-3Y-010Y-OM Red Hard 64.71 4.44 61.85 540.63 106.46 92B9 Red Hard 64.93 5.62 59.89 591.47 113.37 92B19 Red Hard 63.75 5.78 61.26 530.00 118.30 92B20 Red Hard 64.97 6.26 60.29 572.21 110.91 92B24 White Hard 62.80 4.71 61.39 576.27 116.24 K.Heroe MBUNI/SRPC 64//YRPC1 Red Hard 65.25 3.59 60.53 515.63 99.24 Mean 64.73 4.96 60.48 559.23 110.21 SEM 0.38 0.19 0.21 7.77 1.37

PAGE 103

starch damage (common in hard wheat) produce flours with high water absorption (Blackman and Payne, 1987). This would explain the negativ e correlation of flour yield (FY) to water absorption (WA). However, water absorption is also influenced by other factors as protein content and particle size of the flour. Flours with smaller average particle size would have higher absorption values because of greater specific surface. From the Dunnets means of squares, the cultivars were not significantly differe nt for this parameter but were of desirable flour yield Conclusion Our aim was to evaluate the advanced lines in the wheat-breeding program for quality. We classified the 12 bread wheat cultivars for crude protein content and only 4 cultivars including the control had values below 12%. Genotypes 92B19, 92B24 and R946 had the highest values. These varieties showed high loaf volum es as compared to the others. 92B9 had high protein values as well as loaf volume. Protein content was used for classification as it correlates with most of the other parameters. References American Association of Cereal Chemists. 1983. Approved methods of the AACC. 8th Edition St Paul, MN. Baezinger.P.S., R.L. Clements, M. S. McIntosh, W.T. Yamazaki, T. M. Starling, D.J. Sammons, and J.W. Johnson. 1985. Effect of cultivar, environm ent and their interaction and stability analyses on milling and baking quality of soft red winter wheat. Crop Sci 22: 5-8. Bassett, L.M., R.E. Allan and G.L. Rubenthaler. 1989. Genotype x environment interactions on soft white winter wheat quality. Agronomy Journal 81:955-960. Blackman, J.A. and P.I. Payne. 1987. Grain Quality. In Lupton F.G.H. (ed), Wheat Breeding. Chapman and Hall, London and New York, pp. 455-485 Busch, R.H., W.C. Shuey, and R.C. Frohberg. 1969. Response of hard red spring wheat ( Triticum aestivum L .) to environments in relation to six quality characteristics. Crop Sci 9:813-817. Bushuk, W. K.G. Briggs, and L.H. Shebeski. 1969 Protein quantity and quality as factors in the evaluation of bread wheats. Cereal Chem. 49:113-122 Finney, W.T., V.L. Yamazaki, and G. L. Rubenthaler. 1987. Quality of hard, soft and durum wheats. In Heyne, E.G. (Ed) Wheat and wheat improvement (2nd Edition), ASA, Madison, WI pp. 677748. Katunzi A.L., T.E. Maganga, and A. Mrema. 1991. The Effects of Genotype, Environment and their Interaction on Soft Bread Wheat Qu ality. In Proceedings of the. 7th Regional Wheat Workshop, Tanzania, pp. 64-72 Lill D. van, J.L. Purchase, M.F.Smith, G.A Agen bag, and O.T. De Villie rs. 1995. Multivariate assessment of environmental effects on hard re d winter wheat. II. Canonical correlation and canonical variate analysis of yield, biochemical and bread-making characteristics. S. Afr. J. Plant Soil 12(4):164-169. Morris,C.F. and S.P. Rose. 1996. Wheat. In: Henr y, R.J.and P.S. Kettlewe ll, (eds) Cereal Grain Quality. Chapman and Hall, London. Pp. 3-54. Peterson C.J., R.A. Graybosc h, P.S. Baezingger, and A.W. Grombacher 1992. Genotype and environment effects on quality characteristics of hard red winter wheat. Crop Sci 32:98-103. Peterson C.J., R.A.Graybosch D.R. Shelton, and P.S. Baenziger. 1998. Baking quality of hard winter wheat: Response of cultivars to environment in the Great Plains. Euphytica 100:157-162. Pomeranz Y. 1971. Wheat Chemistry and Technology: Compositional and functionality of wheat flour components Chapter 12: 602-603 Pratt Jr D.B. 1971. Wheat Chemistry and Technology : Criteria of Flour Quality. Chapter V:201-225 Preston, K.R., and R.H. Kilborn. 1984 The Fa rinograph handbook, B.L. DApplonia, and W.H. Kunerth (ed.) 3rd ed. AACC. St.Paul, MN.) Sandstedt, R.M. 1955 Photomicrographic studies of wheat starch. III. Enzyma tic digestion and granule structure. Cereal Chem. 32 (incomplete reference)

PAGE 104

Physiological Traits Associated with Drought Tolerance in Bread Wheat ( Triticum Aestivum L.) under Tropical Conditions Kimurto, P.K.1, M.G. Kinyua2, J.B.O. Ogola3, J.M. Macharia4, and P.N. Njau2 1 Department of Agronomy, Egerton Uni versity, P.O. Box 536, Njoro, Kenya 2National Plant Breeding Research Center, Njoro, P.O. Njoro, Kenya 3Department of Plant Production, The University of Venda for Science and Technology, Private Bag X5050, Thohoyandou 950, South Africa. 4 Department of Botany, Egerton University, P.O. Box 536, Njoro, Kenya Abstract Although it is generally accepted that drought tolerance is a critical agronomic trait, efficient and predictable improvement in drought tolerance in bread wheat ( Triticum aestivum L.) has not yet been achieved. Evaluating the responses of physiological traits associated with drought tolerance in bread wheat varieties will enhance selection for tolerance in wheat varieties grown in marginal rainfall areas. This study assessed the drought responses of bread wheat using physiological traits associated with drought tolerance that may be used for selection. Two experiments (under the rain shelter and in the fiel d) were carried out, each for two seasons (2001/2002). The rain shelter experiment simulated rains at three watering regimes: low (210mm), medium (240mm), and high (270mm). Significant genotypic variation (P<0.01) was observed for water use efficiency (WUE), harvest index (HI), stomatal conductance, transpiration rates, and CO2 assimilation. Under dry conditions, these were identified as key control points in determining the drought resistance of tolerant genotypes. These traits could also be resp onsible for sustained survival and facilitated recovery after rewatering. Thus, their use as selection criterion in breeding for drought tolerance is promising. In these respects, the genetypes R960, KM14, R963, and R965 were the most promising candidate s with superior physiological traits and high grain yield. All of the genotypes we re compared with the commercial checks Duma and Chozi. A need exists to determine the heritability of these traits to realize their potential usefulness in a breeding program. Introduction Water scarcity is increasingly becoming a majo r limitation for agricultural production and food security in sub-Saharan Africa (Turner, 2001). In Kenya, for example, the major constraint limiting wheat production in marginal rainfall areas is inadequate and erratic rainfall (Jaetzold and Smith, 1983; Mugo et al ., 1998). Development of wheat cultivars with improved adaptation to drought has thus been a major goal in most of national wheat breeding programmes. However, breede rs have traditionally applied methods where grain yield comparison is used as the main selection criteri a for drought tolerance. The approach has often succeeded in the absence of in-depth know ledge about physiological basis for superior performance of existing germplasm. The effectiveness of selection for grain yield per se is, however, low because of the large number of genes involved, hence low heritabilities (Acevedo, 1993) and large genotype year, genotype location and genotype year location interactions (Calhoun et al ., 1994; Van et al ., 1998). These make genetic progress for drought and heat tolerance extremely difficult and new varieties are released, for commercial production, after long periods (up to eight y ears). To improve genetic gains and realize

PAGE 105

production increases, more efficient screening and selection methodologies and tools need to be developed (Pfeiffer et al ., 2000; Reynolds et al ., 2001). Ludlow and Muchow (1990) noted that grain yield (under drought) is dependen t upon many phenological, morphological, and physiological characters. Therefore, the use of physiological traits as an indirect selection would be important in augmenting yield-based selection procedures (Acevedo, 1993) and hence result in more precise targeting of factors limiting yield and consequently faster rates of yield improvement and broadening of genetic base (Richard et al 2002). Physiological traits take less time to measure than grain yield and can be observed gradually at seedling stage before flowering. The breeder can thus elim inate susceptible lines fro m the crossing nursery and shorten the selection cycle time. By using physiological traits, it may be possible to provide an easy estimate of yield potential before the final harvest. (Edmeades et al ., 1996). The objective of this study was to evaluate droug ht responses of physiological traits in bread wheat that control grain yield and hence may be used for selection in dry tropical environments. Materials and Methods Experimental sites Two experiments (experiment I and experime nt II) were undertaken in 2001 and 2002. Experiment I was carried out under field cond itions at the National Dryland Research Centre, Katumani (1S, 3714E, altitude 1560m). Katumani lies in the semi-arid zone low potential dryland area within Agro-Ecozone UM 4 in the Eastern Province of Kenya. The average annual rainfall is about 716mm, with mean minimum and maximum temperatures of 13.9C and 24.7C, respectively. Water loss th rough evaporation is about 1800mm per year, creating an annual water deficit of about 1048 mm (ICRAF, 1988). The soils are Ferral Chromic Luvisols, which are well-drained, deep sandy loam to clay loam (Jaetzold and Schimdt, 1983). In contrast, experiment II was carried out under rain shelter, by simulating early season drought at seedling stage, at the National Plant Breeding Research Centre, Njoro (0 20S, 35 56E, altitude 2160 m). Njoro is in the Greater Rift Valley, Kenya, and it receives an average annual rainfall of 931mm. Mean maximum and minimum temperatures are 22.7C and 7.90C, respectively. The soils are well draine d Mollic Andosols with sandy loam (Jaetzold and Schmidt, 1983). The shelter is similar to th at described by Jefferies (1993) and Upchurch et al. (1983) and is 15.5m long 7.5m wide. Translucent sheets (which allow up to 90% of photosynthetic photon flux density to pass through) covered the roof. Wheat genotypes Seventeen bread wheat genotypes were evaluated in the field in experiment I (Table 1). Of these, 12 genotypes (R963, R965, R962, R960, R966, R970, 94b01, R917, KM20, Chozi, Duma and Heroe) were evaluated in experiment II under the rain shelter (Table 1). Four of the genotypes originated from Kenya while thirteen were introductions from CIMMYT and other international nurseries showing resistance to dr ought and having different phenotypic traits, maturity periods, and yield potential (Table 1) Mutants were developed by irradiating the seeds of parents with different levels of radi ation at International Atomic Energy Agency (IAEA), Vienna Austria (Table 1).

PAGE 106

Table 1. The origin, pedigree, and drought tolera nce level for the seventeen wheat genotypes used in the study. Genotype Origin Pedigree Drought tolerance response 1. Chozi Kenya F12.71/COC//GEN Tolerant 2. Duma Kenya SW 53 = BUCK BUCK S Tolerant check 3. Kenya Heroe Kenya MBUNI/SRPC 64//YRPC1 Susceptible check 4. R913 CIMMYT Mexico KMN/BOW/OPATA Unknown 5. R917 CIMMYT Mexico URES/BOW/OPATA Moderately tolerant 6. R920 CIMMYT Mexico PJN/BOW//OPATA Moderately tolerant 7. R960 CIMMYT Mexico PASTOR Moderately tolerant 8. R962 CIMMYT Mexico KLEIN CHAMACO Moderately tolerant 9. R963 CIMMYT Mexico BOW//URE S//KEA Moderately tolerant 10. R965 CIMMYT Mexico BOW//BUC/BUL/3/KAUZ Moderately tolerant 11. R966 CIMMYT Mexico FILIN Moderately tolerant 12. R970 CIMMYT Mexico PIPED/5PATIO/ALD//PAT 72300/3PUN/4/BOW/ 6BAW 898 Moderately tolerant 13. R840 CIMMYT Mexico PIPED/5PATIO/ALD// PAT 72300/3PUN Moderately tolerant 14. KM14 Kenya PASA MUTANT (BUC S/CHAT S Tolerant 15. KM15 Kenya PASA MUTANT Moderately tolerant 16. KM20 Kenya PASA MUTANT Moderately tolerant 17. 94b01 Kenya PUN// BOW/BAW Moderately susceptible Experimental design Experiment I was sown on 25 October 2001 (season I) and 20th April 2002 (season II). The test germplasm was drilled in 4 rows (each 6m long) spaced 0.2m apart (at a seed rate of 125 kg ha-1) in a randomized complete block design, rep licated three times. Fertilizer was applied at the recommended rate (70 Kg P205 ha-1 and 37 Kg N ha-1) as Diammonium phosphate (4618-0). Aphids were controlled by applicati on of Metasystox at the rate of 1L ha-1. The experiments were kept weed-free throughout the growing season by manual weeding. Physiological data was taken 26-31 Dec 200 1 (61 DAE) and 27-30 June 2002 (65 DAE). Experiment II was a split-plot design sown on 15 September 2001 (season I) and 5 January 2002 (season II) with water regime as main plots (size) replicated three times and the 12 wheat genotypes as sub-plots (size). Wheat seed s were drilled in rows 0.2m apart at the recommended seed rate of 125 kg ha-1. Fertilizer rates were similar to those in experiment I. Plots were shielded from rainfall by covering w ith the rain shelter at all rainy times and at night. To ensure good germination and cr op establishment before imposing irrigation treatments, all plots were watered to field capac ity (30-32% moisture content) at planting. In addition, all the plots received 30mm of water at emergence and at 7 days after emergence (DAE). The watering regime treatments were imposed by withholding water supply for a period of 2, 4, and 6 weeks, respectively (up to 21, 35, and 49 DAE, respectively). In total, the low watering regime received 210 mm of moisture, medium received 240 mm, and high water regimes received 270 mm during the growing season. The amount and frequency of water application simulates the am ount and nature of ra infall pattern usually received in most marginal areas during cropping season (Mugo et al ., 1998; Jaetzold and Smith, 1983). Drip irrigation was the method used to water the plots (Chapin Watermatics, 1999).

PAGE 107

Measurement of weather variables An automatic weather station located 100m from the site of experiment I recorded rainfall (mm), maximum and minimum air temperatures (C), solar radiation (MJ m-2 per day), and relative humidity (RH, %) each day during the experiment (Table 2) Also, instantaneous weather variables like atmos pheric temperature, RH (%), and photosynthetically active radiation (PAR) were recorded in experiment I (Table 3). Measurement of water use Total crop water use was determined in experime nt II by monitoring changes in volumetric moisture content throughout the season. Soil moisture content was measured at 7-day intervals using a neutron probe (Troxler Mode l 4300, New York). Measurements were taken between 7 and 91 DAE (season I) and 7 and 98 DAE (season II). Volumetric water content at each depth was calculated using calibra tion equations for this site (Ooro et al ., 2001). Table 2. Monthly total (rainfall and Epot) and daily mean of weather variables (temperature, solar radiation and rel ative humidity) during the 2001 and 2002 growing seasons at Katumani. Mean Total Maximum Minimum Mean Solar Relative Rainfall Epot. Daily Daily Temperature Radiation Humidity (mm) (mm) Temperature (0 C) Temperature (0 C) (o C) (MJ m-2 d-1) (%) 2001 January 244.5 115.4 24.5 14. 0 19. 3 609.1 71.0 February trace 161 26.4 14. 3 20. 4 694.4 60.0 March 113 158 26.7 14.3 21.5 658.1 60.0 April 88.9 115.4 24.9 15.1 20.0 549.2 68.5 May 15.3 123.8 25.0 14.0 19.5 533.9 63.5 June 4.3 79.4 23.6 11.9 17.8 498.2 65.0 July 4.3 100.3 21.2 10.9 16.2 490.7 66.0 August 2.5 134.5 24.5 11.0 17.8 527.5 59.0 September trace 169.5 26.7 12 .5 19.6 611.2 54.5 October 73 180.3 27.1 13.6 20.4 630.0 51.5 November 169 126.1 24.0 14.6 19.3 573.9 69.0 December 43.6 127.6 24.2 14 .4 19.3 552.7 72.5 Mean/Total 758.4 1591.3 24.9 13.1 19.2 577.4 63.4 2002 January February 79.5 148.2 25.9 14. 1 20. 0 624.4 65.5 March 98.9 7.5 179.0 27.1 13. 9 20. 0 676.4 53.0 April 120.4 141.6 26.3 15.3 21.3 634.1 65.5 May 125.6 151.9 25.8 15.8 20.3 572.4 67.0 June 94.9 111.3 24.4 14.2 19.3 475.5 70.0 July trace 94.9 23.4 12.1 17.8 447.8 66.5 August 0.2 101.3 23.9 13.2 18.6 445.6 65.0 Mean/Total 527.0 120.3 24.1 12.0 18.1 435.6 64.5 1048.5 25.1 13.8 19.4 539.0 64.6 The mean of daily maximum and minimum temperature

PAGE 108

Table 3. Diurnal variation in mean temperature, relative humidity (RH) and photosynthetically active radiation (PAR) at 61 (season I) and 65 (season II) DAE. Time Atmospheric RH PAR Of day (hours) Temperature (0C) (%) ( E m-2s-1) 0800 23.0 52.1 780.8 1000 28.1 39.8 1369.7 1200 31.3 37.1 1866.8 1300 34.0 36.7 1941.3 Mean 29.1 41.4 1489.7 Total crop evapotranspiration (ET) was estimat ed using the soil water balance equation: ET = S + I D R. 1 where S is the change in storage (the differen ce in volumetric water content of the entire profile between the start and the end of the expe riment), I is irrigation, D the drainage, and R is the runoff. Drainage and runoff were assumed to be negligible. The soil moisture content of the deepest layer (1.0m) showed little change during the crop-growing season. There was no runoff because drip irrigation was used. W/ET Water use efficiency (WUE); the ratio of the total above ground dry matter produced (DM) to the total amount of water used by the cr op was determined using the equation: WUE =DM/ET 2 Measurement of gas exchange parameters Stomatal conductance (g) and instantaneous tran spiration (T) rates were measured on both the upper and lower surfaces of the uppermost fully expanded leaves at two-hour intervals (between 1000 and 1300 hrs) on clear sunny days, at booting stage, using a steady-state porometer (LI-1600Lico Inc. Lincoln, NE, USA). Stomatal resistance (rs) was calculated as inverse of stomatal conductance (Burrows a nd Milthorpe, 1976; Cowan, 1977). Net leaf CO2 exchange rates (CER) was measured on selected leaves using a portable Infrared Gas Analyzer (IRGA), fitted with Parkinson Leaf chamber. Statistical analysis Analysis of variance (ANOVA) was used to evaluate the treatments using a general linear model (GLM) (SAS, 1996). Statistical differences for the watering regimes and means of all variables measured for wheat genotypes were separated using Fishers least significance test (LSD) at P<0.05. Results Crop biomass, grain yield, and harvest index The interaction between watering regimes a nd genotypes affected biomass such that the increase in biomass due to irrigation varied w ith genotypes (Table 4). The medium watering regime increased biomass of all genotypes, fro m a 15% increase in R963 to 122% in R965. On average, the increase in biomass due to medium watering regime was greater for drought susceptible (68%) and medium tolerant (59%) genotypes than drought tolerant ones (41%). Similarly, the increase in biomass due to high watering regime varied with genotypes (from 50% to 215%). The increase in biomass due to high watering regime was greater (on average) for drought susceptible (131%) and medium tolera nt (121%) varieties than drought tolerant ones (95%). On average, irrigation increased crop biomass by 56% (medium watering regime) and 116% (high watering regime) (Table 4).

PAGE 109

The interaction between moisture regimes a nd wheat genotypes affected grain yield such that medium watering regime decreased (by 14%) the grain yield of genotype R917 but increased the grain yield of the other varie ties (by 11% 113%, respectively for 94b01 and R970) (Table 4). The increase in grain yield due to the medium watering regime was greater for susceptible varieties (60%) than medium to lerant (55%) and drought tolerant (56%) ones (Table 4). In contrast, the high watering regime increased grain yield of all genotypes, but the magnitude of the increase varied by genotype For example, the increase was greatest for R963 (157%) and lowest for R917 (13%). On average, the increase in grain yield due to the high watering regime was greater for susceptib le varieties (121%) than drought tolerant (117%) and medium tolerant (104%) genotypes (Table 4). Harvest index was not affected by the in teraction between water regimes and genotypes (Table 4). Genotype affected harvest index unde r all moisture regimes (Table 4). Averaged over the moisture regimes, HI varied from 0.21 (R917) to 0.46 (R960). Drought tolerant varieties had greater (by 25%) HI than droug ht susceptible and medium tolerant varieties (Table 4). Irrigation increased HI by 26% (medium moisture) and 22% (high moisture regime) (Table 4).

PAGE 110

Table 4. Response of grain yield, biomass, ev apotranpiration (ET), wa ter use efficiency of biomass production (qd), and grain yield (qg) and harvest index (HI) to three moisture regimes for two seasons (2001/2002)/ Genotype/ ET Biomass Grain yield WUEg WUEb HI Irrigation (mm) (kg ha-1) (kg ha-1) (kg ha-1mm-1) (kg ha-1mm-1) Low Moisture Chozi 97.4 2993.3 703.2 5.7 30.2 0.24 94b01 83.2 1199.2 605.6 7.2 14.1 0.40 Duma 93.9 1480.8 652.8 7.9 16.7 0.44 Heroe 86.0 2303.3 394.9 4.8 26.2 0.18 KM20 75.8 1420.0 494.4 4.5 24.4 0.34 R917 81.7 1454.2 622.7 4.1 16.3 0.40 R960 88.1 1230.8 661.9 7.5 13.8 0.51 R962 93.6 2311.3 570.6 5.7 23.8 0.22 R965 108.9 1470.8 650.1 4.8 14.7 0.45 R963 79.1 1630.0 682.9 8.1 20.2 0.41 R966 100.9 2171.3 598.1 5.6 21.3 0.36 R970 85.3 1407.5 447.7 5.8 15.0 0.34 Medium moisture Chozi 131.0 3831.3 1061.1 8.1 29.2 0.35 94B01 110.7 2125.8 783.3 7.7 23.0 0.37 Duma 107.7 2277.9 1049.1 9.6 22.2 0.46 Heroe 111.9 3142.9 820.4 6.4 30.0 0.23 KM20 117.2 2612.9 709.7 5.7 26.8 0.28 R917 120.9 1933.3 685.6 5.6 17.6 0.31 R960 104.9 2003.7 1089.4 10.5 22.2 0.54 R962 127.3 3342.5 1023.8 7.2 26.8 0.31 R963 115.4 2372.1 1043.0 9.1 21.7 0.43 R965 116.2 3261.3 946.3 7.6 28.8 0.31 R966 112.5 2582.9 861.6 7.0 24.9 0.32 R970 107.3 2650.0 952.3 8.0 22.1 0.33 High moisture Chozi 165.8 5337.9 1616.8 8.9 32.5 0.31 94B01 149.3 3057.5 1405.1 11.5 26.1 0.44 Duma 127.8 3143.3 1323.1 12.1 24.6 0.42 Heroe 157.1 4194.6 960.2 8.4 31.4 0.26 KM20 180.4 3502.9 1056.9 6.0 32.4 0.29 R917 131.0 3244.2 904.6 6.9 24.1 0.27 R960 134.4 2868.3 1505.7 11.4 21.2 0.53 R962 158.0 5060.4 1298.2 8.5 32.0 0.25 R963 179.5 3082.9 1346.1 8.4 17.2 0.44 R965 155.7 4638.7 1184.7 7.6 30.2 0.26 R966 158.8 3758.6 1229.2 7.7 30.7 0.35 R970 132.4 2855.0 995.7 7.5 21.0 0.34 S.E.D 146.1 1351.0 1072 0.93 11.5 0.003 Lsd 438.19 458.9 107.21 1.78 2.67 0.03 Variety P<0.001 P<0.001 P<0.001 P<0.005 P<0.001 P<0.001 Water P<0.001 P<0.005 P<0.001 P<0.05 P<0.001 P<0.001 W x G P<0.05 P<0.05 P<0.05 ns P<0.05 ns W x G = water regime x genotype interaction, ns, not significant

PAGE 111

Crop water use (ET) The interaction between water regimes and geno types affected total crop ET such that the increase in crop ET due to irrigation varied with varieties (Table 4). For example, the medium watering regime increased crop ET by 7% (R965) to 55% (KM20). Increase in crop ET due to the high watering regime ranged from 36% (Choz i) to 138% (KM20). Also, the increase in ET due to irrigation was greater for the drought susceptible (32% and 81%, respectively for the medium and high watering regimes) and medi um tolerant (33% and 77%, respectively for the medium and high watering regimes) varieties than for drought tolerant varieties (25% and 53%, respectively for the medium and high watering regimes) (Table 4). Water use efficiency (WUE) The interaction between moisture regimes and wh eat genotypes affects water use efficiency of biomass production (qd, Howell et al. 1998). The increase in qd due to the medium moisture regime varied respectively from 14-17% (K M20 and R966) to 47-50% (R970 and R960). Similarly, the increase in qd due to the high moisture regime varied with genotype (ranging from 4% to 94%, respectively for Heroe and R917) (Table 4). On average, the increase in qd due to the medium moisture regime was greater (32%) for medium tolerant genotypes than for tolerant (28%) and susceptible (29%) genot ypes (Table 4). Also, the increase in qd due to high moisture regime was greater for medium tole rant (65%) and tolerant (36%) genotypes than drought susceptible ones (22%) (Table 4). The interaction between watering regime and genotype did not affect water use efficiency of grain production (qg, Howell et al ., 1998; Ogola et al ., 2002) (Table 4), but qd varied with genotypes (from 5.4 kg ha-1mm-1 to 9.9 kg ha-1mm-1 respectively, for R917 and Duma) across the three moisture levels (Table 4). On average, the drought tolerant genotypes had a greater qg (8.9 kg ha-1mm-1) than susceptible (7.9 kg ha-1mm-1) and medium tolerant (6.9 kg ha-1mm-1) genotypes (Table 4). Averaged over all genotypes, the medium watering regime increased qg by 28% (from 5.8 to 7.4 kg ha-1mm-1), while the high watering regime increased qg by 53% (from 5.8 to 8.9 kg ha-1mm-1). Gas exchange parameters Results showed similar trends in both seasons. Genotypes affected leaf temperature in field experiment I; Duma, R960, and Chozi had the lo west leaf temperature, and R920, R963, and KM20 the greatest (Table 5). This resulted in mean lowest and highest leaf temperature depressions, respectively for these genotypes. On average, leaf temperature was lower (by 3%) for drought tolerant genotyp es (28.6C) than medium tolerant (29.3C) and susceptible (29.5C) genotypes. The difference between the lowest and the greatest leaf temperature was 7%. The greatest leaf temperature depression was recorded in R920 and the lowest in Duma (Table 5). Leaf temperature depression was greater for drought tolerant genotypes (-0.6C) than medium tolerant (0.4C) and susceptible (0.2C) genotypes. Instantaneous transpiration (IT) differed by genotype; KM14, R963, Chozi, Duma, and R960 had the greatest IT (ranging 10.3-11.2 mmol m-2 s-1) and R913, R840, R970, and KM20 the lowest (ranging from 7.0-7.7 mmol m-2 s-1) (Table 5). On average, IT was greater (by 21% and 44%, respectively) for drought tolerant ge notypes than medium tolerant and susceptible genotypes. Also, average IT was 20% greater for susceptible than medium tolerant genotypes. There was variation among the genotypes in CER (from 7.9 to 15.1 mol m-2 s-1) with KM14, R960, R966, R965, and Duma having the lowest values (Table 5). Drought tolerant genotypes recorded 16% and 20% greater CE R than medium tolerant and susceptible genotypes, respectively. The results of CER ranked susceptible genotypes (like R966) higher than tolerant genotypes (Duma).

PAGE 112

Stomatal conductance (g) and resistance rs were affected by genotypes; Chozi and Heroe had the greatest g and lowest rs, respectively (1.3 mol cm-2 s-1 and 2.1 s cm-1). KM20 the lowest g (0.6 mol cm-2 s-1) while R840 had highest rs (4.9 s cm-1) (Table 5). Stomatal conductance was greater in drought tolerant than susceptible genotypes by an average of 27% and nonsignificant with medium tolerant genotypes. The genotypes ranking based on g and rs was inconsistent (Table 5). Table 5. Gas exchange parameters of 17 wheat cultivars measured at Katumani for two seasons (2001/2002). Cultivar LT LTD g TR rs CER Yield (0C) (0C) ( mol cm-2 s-1) (mmol m-2 s-1) (s cm-1) ( mol m-2 s-1) Kg/ha R963 30.0 0.4 0.9 10.0 3.5 10.5 1269 KM14 29.29 -0.2 1.1 11.2 3.2 15.1 1185 Heroe 29.2 -0.2 1.3 9.5 1.5 9.5 676 R920 30.0 0.8 0.7 8.6 3.9 10.1 1034 R840 29.8 0.6 0.7 7.1 4.9 12.3 1028 R966 29.2 0.4 1.0 9.8 2.9 13.0 1033 KM15 29.8 0.7 0.7 7.8 2.2 9.4 615 R960 28.7 -0.1 0.8 10.3 2.9 13.4 1315 R970 29.5 0.4 0.8 7.4 2.8 9.7 1059 94B01 29.6 0.5 0.8 9.0 3.4 12.9 676 R965 29.2 0.1 0.7 8.1 4.0 13.1 1364 Duma 28.1 -0.9 1.1 10.3 1.2 12.9 1258 KM20 29.7 0.7 0.6 7.7 3.9 10.1 776 Chozi 28.3 -0.6 1.3 10.4 2.1 12.3 1267 R917 29.6 0.8 0.9 8.2 1.5 8.4 719 R962 28.7 -0.4 1.0 8.6 1.9 11.4 848 R913 29.4 0.7 0.7 7.0 2.3 7.9 898 Mean 29.3 0.2 0.9 8.9 3.1 11.1 1003 P(F-ratios) P<0.05 P<0.01 P<0.001 P<0.001 ns P<0.001 P<0.001 Lsd 1.13 0.36 0.15 0.57 3.95 3.80 661.4 Discussion The bread wheat genotypes used to examine va riation in physiological responses to early season drought could be classified into three groups: relatively drought tolerant (Chozi, Duma, R960, R963, and KM14); moderately droug ht tolerant (R840, R970, R920, R965, and R966) and relatively drought susceptible (H eroe, KM20, KM15, R917, 94B01, R913, and R962) based on physiological changes and grain yields measured. The production of biomass varied considerably among the genotypes across the moisture regimes. Chozi, R962, Heroe, and R965 produced the highest biomass, while Duma, 94B01, R960, and R970 had the lowest biomass. This variation was closely associated with the amount of water transpired (ET) by each genotype as expected (Sinclair et al ., 1984 and Ogola et al ., 2002). The response to increasing moisture among the former group of cultivars enabled them maintain high ET resulting in increased water use for dry matter production (qd) at the expense of water use for grain yield production (qd). These represent inherent ability of high yield potential by way of a good biomass source. These findings are in agreement with earlier findings (van Ginkel et al ., 1998 and Kirigwi et al ., 2004) that yield potential is a useful criterion in breeding for superior pe rformance in drought environments. Ceccarelli et al ., (1987) disagreed with these findings and concluded that yield potential is not a useful criterion in breeding for superior performance in drought environments. In semi-arid areas of Kenya where seedling stage drought stress h as been found to cause serious yield losses

PAGE 113

(Mugo et al ., 1988 and Kimurto et al ., 2003), the latter group of cultivars could be recommended as they sustained growth with initia l moisture. Studies in winter wheat indicate the importance of early-season dry matter producti on in early-drought situations (Entz and Fowler (1990). They also tended to invest mo re water for grain yield production (high WUE) than biomass, which may be used as an indicat or of drought tolerance, as earlier reported (Ehdaie, 1995). These cultivars could be grouped into those susceptible to early season drought stress which utilize more water extracted in biomass production, and those tolerant to drought with more water used efficiently for grain yield production The exception to this is Chozi, which possessed a relatively high yield and biomass. The former group of cultivars could do well in areas with high amounts of rain fall to support high ET and longer growth duration. The latter group of genotypes was shor t and early maturing with high HI, as a result of early growth rate and vigour, which allowe d more efficient initial water use during the cooler part of growing season when vapor pr essure deficits are low. This was clearly demonstrated by Duma and R963 in this study, suggesting that high seedling vigor and faster biomass accumulation enhance early-season drought survival. ET was affected by an interaction between genotype and moisture regime such that the WUE decreased with decreasing irrigation. This is in agreement with earlier findings (Ehdaie and Waines, 1994 and Ehdaie, 1995) that drought significantly reduced mean WUE for nearisogenic lines of wheat. Harvest index provides an estimate of the conversion efficiency of dry matter to grain yield. The interaction between water regimes and genotypes did not affect HI, but drought tolerant cultivars (Chozi, Duma, R960, and R963) had a higher mean HI (0.38) than susceptible genotypes (0.23) (Heroe, KM20, R9 62 and R917). These observations concur with earlier statement (Richards et al ., 2002) that drought tolerant genotypes convert more of biomass into grain raising the HI and final yield. Siddique et al., (1990) also reported that improved mechanisms of partition between st raw and grain (improved HI) are the main causes of increased yield in a Mediterranean-type climate. Similarly, Donmez et al ., (2001) noted that wheat grain yields have been asso ciated with increases in harvest index as opposed to increases in biomass. These results further s howed that while HI of the same cultivars (except Chozi which was tall and 94B01, a shor t cultivar) grown under the high moisture regime approached and sometimes exceeded 0. 40, it was generally much less under drought conditions with the same cultivars. Therefore, it appears that the HI of the study cultivars is well below the potential, and perhaps the maxi mum grain yield is not being achieved. These findings are in disagreement with Edhaie (1995) who reported that mean HI for two cultivars were similar in well-wate red and dry conditions. Field data was taken during anthesis on flag leaf, which was expected to be the most active physiological growth stage. The resu lts showed that despite the higher ambient temperatures, tolerant genotypes were able to create cooler conditions around the leaf, which was essential for photosynthesis. Similarly, Reynolds et al. (1994) reported that high stomatal conductance (g), which permits leaf cooling th rough evapotranspiration, was associated with heat tolerance in CIMMYT wheat. For wheat plants to absorb carbon dioxide (CO2) for photosynthesis, they must expose wet surfaces (stomata) to dry atmosphere and in consequence suffer evaporative water loss. The drought tolerant genotypes (R963, Duma, KM 14, R960, and Chozi) had higher values of transpiration than susceptible genotypes. Drought tolerant varieties also exhibited a smaller stomatal resistance though not more significant than the second group, which was susceptible to drought stress. These findings are in agreement with those reported by El Hafid et al ., (1998), that susceptible genotypes exhibit a higher stomatal resistance (rs) than tolerant genotypes on exposure to stress. The resultant being higher transpiration rates, suggesting that the ability of a cultivar to keep its stomata open despite internal water stress and could

PAGE 114

therefore be considered a form of drought resi stance and for possible use as selection trait, is promising. Similar observations have been reported earlier (Johson et al ., 1987). This is because greater transpiration rates may lead to lower leaf temperatures, creating ambient temperatures for C02 uptake, resulting in higher yield. Gummuluru et al. (1989) disagreed and reported that higher stomatal resistance has been related to greater drought resistance. Although significant, cultivar differences in stomatal conductance (g) were quite small ranging from 0.6-1.1 mol cm-2 s-1. Similarly, Gutierrez et al. (1998) recorded stomatal conductance values ranging between 1.22-1.73 mol cm-2 s-1 under water stress conditions in Tuxpeno maize population. CER differed among genotypes with drought tolerant group of genotypes exhibiting higher rates than susceptible genotypes. Gummuluru et al.( 1989) reported similar findings and noted that leaf photosynthesis and stomatal conductance differed significantly between the drought tolerant and susceptible genotypes, where higher stomatal conductance was related with drought resistan ce. These results suggest that stomatal conductance and transpiration rates are traits used by tolerant genotypes to enhance high CER and final grain yield in stress conditions. These genotypes were among the highest yielding in the field and rain shelter conditions (Table 4 a nd 6). This is in agreement with results by Omanyin et al. (1996) in sorghum and Cornish et al. (1991) in Pima cotton, where drought resistant genotypes had higher yield as they suffered less yield reduction because they exhibited higher leaf stomatal conductance a nd relative water content with associated heat tolerance. Omanyin et al. (1996) reported that drought tole rant sorghum varieties exhibited higher leaf RWC and suffered less yield reduction because of maintenance of turgidity, hence high stomatal conductance led to continued phot osynthesis. Further, enhanced productivity was expected because increased st omatal conductance increases CO2 partial pressures within the leaf and therefore also increases assim ilation rates and final yield (Jones, 1987). The data was taken during a very dry peri od (at anthesis, in Dec 2001 and June 2002) when the crop had received low rainfall agains t high evapotranspiration (Table 2). Despite severe drought stress, high radiation, and low RH at the time of data taking (Table 3) the tolerant genotypes continued ph otosynthesising. These results are consistent with earlier findings (Chaves, 1991) that plant survival durin g and after drought stress is, in part, possibly due to the maintenance of photosynthesis during drought stress, which allows rapid recovery of the plant after dehydration. The derived values of instantaneous CER to water loss (TR) ratio (A/E) (Table 5) showed that tolerant genotypes had higher ratios (1.1-1. 59) as compared to susceptible genotypes. However, under drought stress conditions when survival is more important than optimal functioning, the A/E ratios may not remain cons istent among drought tolerant genotypes. For example, R840 recorded the highest ratio (1.7), indicating that it was very efficient in fixing large quantities of CO2 while loosing small quantities of water. Similar observations were reported earlier in durum and winter wheat (Van Resburg and Kruger, 1993 and El Hafid et al ., 1998). El Hafid et al. (1998) further noted that drought-resistant group of cultivars had a lower decrease in net photosynthetic rate to tr anspiration ratio (A/E), showing that they can continue with C-assimilation despite partial stomata closure. Richards et al. (2001) noted that power of CO2 sucking by the leaf is determined by the amount of photosynthetic machinery per unit leaf area and leaf stomatal conductance. In same study, El Hafid et al. (1998) noted that under drought stress, drought tolerant wheat cultivars exhibited lower internal CO2 concentration to ambient CO2 concentration (Ci/Ca ratio), suggesting that they can create CO2 diffusion gradient into the cells. The increase in stomatal conductance could have accounted for most of the increased carbon assimilation rate. Due to a strong correlation between A/E ratio and grain yield, aboveground drymatter, WUEg, and WUEgm under drought (Van Resburg and Kruger, 1993 and El Hafid et al ., 1998), the value of determining A/E ratio of wheat cultivars during stress for possible use as a selection trait is promising.

PAGE 115

Higher stomatal conductance was not correlated with increased yield in cassava (Githunguri et al., 1998), hence it may only be one of the drought mechanisms for drought tolerance and not necessarily for increased yi eld production. In contrast, Omanyin et al. (1996) reported that in sorghum drought resistant genotypes suffered less yield reduction because they exhibited higher leaf stomatal conductan ce and relative water content (RWC). Reynolds et al. (1994) reported that high stomatal c onductance permits leaf cooling through evapotranspiration. These traits can ther efore be used for selection in wheat. Conclusion The major hypothesis tested in this study was th at high grain yield (under drought stress) of bread wheat is controlled by physiological factor s and that genotypic differences in such traits exist in wheat. This hypothesis was based on the assumption that a drought ideotype would combine all traits controlling yield and thus once identified could be used to select for yield. The hypothesis has not been disapproved. Genotypes that had high WUEg, HI, early biomass accumulation, stomatal conductance, concomitant high transpiration, a nd photosynthetic rate were the highest yielding under drought stress conditions during seedling stage. These factors were therefore identified as key control points in determining the drought resistance of bread wheat genotypes. These traits may have facilitate d maintenance of survival and facilitated the recovery upon rewatering. In these respects, Duma, Chozi, R960, KM14, and R963 appeared to be the most promising genotypes. These tra its can be used to separate susceptible and tolerant wheat germplasm, and their use as a possible selection tools is promising. There is need to determine the heritability of these traits to know the potential usefulness of these traits in a breeding program. References Calhoun, D.S., G. Gebeyehu., A. Miranda., S. Rajaram and M. van Ginkel. 1994. Choosing evaluation environments to increase wheat grain yield under drought conditions. Crop Science 34: 673678. Ceccarelli, S. 1987. Wide adaptation: How wide? Euphtica 40 : 197-205. Chaves, M.M. 1991. Effect s of water deficits on carbon assimilation. Journal of Experimental Botany 42 : 1-16. Donmez, E., R.G. Sears, J.P. Shroyer and G.M. Paulsen. 2001. Genetic gain in yield attributes of winter wheat in the Great Plains. Crop Science 41 : 1412-1419. Edaie, B. 1995. Variation in water-use efficiency and its components in wheat II: Pot and field experiments. Crop Science 35 : 1617-1626. Edhaie, B., J.G. Waines and A.E. Hall. 1988. Differ ential responses of landr ace and improved spring wheat genotypes to stress environments. Crop Science 28 : 838-842. Edmeades, G.O., J. Bolanos and S. C. Chapman. 1996. Value of secondary traits in selecting for drought tolerance in tropical maize In: G.O. Edmeades, M. Banzinger, H.R. Mickelson and C.B. Pena-Valdivia (Eds). 1997. Developing drought and low N tolerant maize. Proceedings of a symposium, March 25-29, 1996, CIMMYT, El Batan, Mexico, D.F: CIMMYT. El Hafid, R., D.H. Smith., M. Karrow and K.Samir. 199 8. Physiological attributes associated with early season drought resistance in spring durum wheat. Canadian Journal of Plant Science 78 : 227237. Gummuluru, S., S.L.A. Hobbs and S. Jana. 1989. Genotypic variability in physiological characters and its relationship to drought tolerance in durum wheat Canadian Journal of Plant Science 69: 703-711. Howell, T.A., J.A. Tolk., A.D. Schneider, S.R. Ev ett. 1998. Evapotranspiration, yield, and water use efficiency of corn hybrid s differing in maturity. Agronomy Journal 90 : 3-9. ICRAF. 1988. International Centre for Agro-fores rty (ICRAF) Field Station, Machakos, Part 1I. General Account, ICARF, Nairobi, Kenya Jaetzold, R. and H. Schimdt. 1983. Farm management Handbook of Kenya. Natural conditions and farm management information Vol. II/B. Central and Western Kenya. Government Printers,

PAGE 116

Nairobi.KARI, 1984. Kenya Agricultural Resear ch Institute Annual Report. KARI, Nairobi, Kenya. Johnson, R.C., D.W. Morhinwerg, D.M. Ferris an d J.J. Heitholt. 1987. Leaf photosthesis and conductance of selected Triticum spp at different water potentials. Plant Physiology 83 : 10141017. Kimurto, P.K. M.G. Kinyua and J.M. Njoroge. 2003. Response of bread wheat genotypes to drought simulation under a mobile rain shelter in Kenya. African Crop Science Journal 11 :16-25. Kirigwi, F.M., M. van Ginkel., R. Trethowan., R. G. Sears., S. Rajaram and G.M. Paulsen. 2004. Evaluation of selection strategies for wheat adaptation across water regimes. Euphytica 135 : 361-371. Ludlow, M.M. and R.C. Muchow. 1990. A critical evaluation of trait for improving crop yields in water-limited environments. Advances of Agronomy 43: 106-153. Mugo, S., M. Smith, M. Banzinger and T. Setter. 1998. Performance of early maturing Katumani and Kito composites under drought at the seedling stage and flowering stages. African Crop Science Journal 6(4): 329-344. Ogola, J.B.O., T.R. Wheeler., P.M. Harris. 2001. Water use of maize in response to planting density and irrigation. European Journal of Agronomy. In Press. Pfeiffer, W.H., R.M. Tr ethowan and T.S. Payne. 2000. CIMMYT's approach to address production constraints in marginal areas-Global Project 5. In: The 11th Regional wheat workshop for Eastern, Central and Southern Africa. Ad dis Ababa, Ethiopia, CIMMYT. Reynolds, M.P., Balota, M., Delgado, M.I.B., Aman i, I., Fischer, R.A. 1994. Physiological and morphological traits associated with spring wheat yield under hot, irrigated conditions. Austrl. J. Plant Physiol. 21:717-30. Reynolds, M.P., S. Nagarajan, M.A. Razzque and O.A.A. Ageeb. 2001.Heat tolerance In. M.P. Reynolds, J.I. Ortiz-Monasterio and A. McNabs (eds.). Application of Physiology in wheat Breeding. Mexico, D.F.: CIMMYT. Richards, R.A., A.G. Condon and G.J. Rebetzke. 2001. Traits to improve yield in dry environments. In. M.P. Reynolds, J.I. Ortiz-Monasterio and A. Mc Nabs (Es.). Application of Physiology in wheat Breeding. Mexico, D.F.: CIMMYT. Skovmand, B. and M.P. Reynolds. 2000. Increasing yield potential for marginal areas by exploring genetic resources collections. The 11th Regional wheat workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia, CIMMYT. Richards, R.A., G.J. Rebetzke., A. G. Condon and A.F. van Herwaarden. 2002. Breeding opportunities for increasing the efficiency of water us e and crop yield in temperate cereals. Crop Science 42 : 111-121. Siddique, K.H.M., D. Tennant., M.W. Perry and R.K. Belford. 1990. Water use and water use efficiency of old and modern wheat cultivars in a Mediterranean environment Australian Journal of Agricultural Research 41 : 432-447. Sinclair, T.R., C.B. Tanner and J.M. Bennett. 19 84. Water use efficiency in crop production. Bioscience 34 : 40-60. Turner, N.C. 2001. Optimizing water use. In: J. Nosb erger., H.H. Geiger and P.C. Struik (eds.). Crop Science Congress Proceedings, Australia. CAB International. Van Ginkel, M., D.S. Calhoun., G. Gebeyehu., A. Miranda., C. Tian-you., R. Pargas Lara., R.M. Trethowan., K Sayre., J. Crossa and S. Rajaram. 1998. Plant traits related to yield of wheat in early, late or continuous drought conditions. Euphytica 100 : 109-121. Van Rensberg, L and G.H.L. Kruger. 1993. Compara tive analysis of differential drought stress-induced suppression of and recovery in carbon dioxide fixation, stomatal and non-stomatal limitation in Nicotiana tabacum L Journal of Plant Physiology 142 : 296-306.

PAGE 117

Allelism of Resistance Genes to Phaeosphaeria nodorum in Wheat C. A. Kuwite1 and G. R. Hughes2 1Selian Agricultural Research Institute,P.O BOX 6024, Arusha, Tanzania 2 Crop Science Department, University of Saskatchewan, Saskatoon, S7N OWO, Canada. Abstract Disease reaction of the F1, F2 and F3 populations of crosses made between six winter wheat cultivars with resistance to Phaeosphaeria nodorum showed that they possessed a major gene for resistance in common. Cultivars EE 8, Hadden, 821WWMN 2019 and Atlas 66 may carry two genes for resistance to P. nodorum The second resistance gene in EE 8 is non-allelic to the second resistance gene in 821WWMN 2019 or Hadden. Observations made on F2 plants derived from single crosses or crosses in various combinations revealed the presence of minor resistance genes or modifier genes in the resistant cultivars. None of the F3 families derived from susceptible F2 plants were susceptible. These findin gs suggest that host resistance to P. nodorum in wheat is controlled by both major and minor. Introduction Septoria nodorum blotch caused by Phaeosphaeria nodorum (Mller) Hedjaroude anamorph Stagonospora nodorum (Berk.) Castellani and Germano has become an increasingly important disease of wheat. Recommendations for disease control include crop rotation, fungicide sprays and the use of resistant cultivars. The most economical way of controlling the disease under Saskatchewan conditions is the use of resistant cultivars. However, very little success has been achieved in breeding for resistance to P. nodorum due to lack of information concerning its inheritance. Monogenic and polygenic control, and recessive, dominant, epistatic, pleiotropic and additive gene action have been reported in studies of seedling resistance to P. nodorum in wheat (Eyal et al ., 1987; Frecha, 1973; Kleijer et al ., 1977; Nelson and Gates, 1982; Wong and Hughes, 1989). Most studies have indicated that resistance to P. nodorum is under polygenic control (Nelson, 1980; Nelson and Gates, 1982), but monogenic control has been reported (Frecha, 1973; Kleijer et al ., 1977; Wong and Hughes, 1989). Wong and Hughes (1989) found that genetic control of resistance to L. nodorum in three winter wheat cultivars 811WWMN 2095, Coker 76-35 and Red Chief was due to a single recessive gene. Similar gene action but with dom inant effect was reported by Frecha (1973) in Atlas 66. Kleijer et al., (1997) located the resistance gene in Atlas 66 on chromosome 1B. Ma (1993) showed that the resistance genes to P. nodorum are located on chromosomes 2A and 3A in the common wheat cv. EE 8 and on chromosome 3A in cv. Red Chief. Genes for resistance that are at different loci can be combined through breeding for increased levels of resistance. The main obj ective of this study was to determine the independence of genes controlling resistance to P. nodorum among selected resistant winter wheat cultivars. Materials and Methods Plant Material Wheat plants used in this study were grown in a 3:1:1(v/v) sterilized soil, vermiculite and peat mixture in 15cm diameter pots in a walk-i n growth room with a 16-h photoperiod and

PAGE 118

220C/160C day/night temperatures. They were fe rtilized with 20:20:20 (N:P:K) water-soluble fertilizer at weekly intervals. Crosses were made among six P. nodorum resistant parents (Table 1). Except for cv. Atlas 66, previous studies had indicated monogenic or oligogenic control of seedling resistance in these cultivars (Wong and Hughes, 1989; Ma 1993). Some F1 plants were selfed to produce F2 plants, others were inoculated and rated for disease reaction. After being rated for disease reaction, 50 randomly selected F2 plants of the Red Chief x 831WWMN 2051, EE 8 X 831WWMN 2051, Atlas 66 x 821WWMN 2019 and Hadden x Red Chief crosses were selfed to produce F3 families. The parental, F1 and F2 generations of a cross and the susceptible cv. Kenyon were included in each test. Reciprocal F1 plants of certain crosses were also tested. Testing of the F3 families of the selected crosses was done subsequently. Inoculum Preparation A single pycnidiospore of P. nodorum isolate KVTN, obtained from field-infected leaf material at Kelvington, Saskatchewan was used in all experiments. The isolate was increased on solid V8 juice agar containing 15g agar, 1.5g calcium carbonate, and 150 ml of V8 juice and 850 ml of water. The cultures, incubate d for 14 d under continuous fluorescent light grew as the anamorph ( S.nodorum ). Pycnidiospores were collected in distilled wa ter by gently rubbing with a toothbrush. The mixture was then blended in an Oster blender for 1 min. This process was repeated once. The pycnidiospore suspension was th en filtered through four layers of cheese cloth. Inoculum concentration was determined by a hem acytometer and adjusted to 1.5-3.0 x 106 mL-1. Two drops of the surfactant Tween 20 we re added per 100 mL inoculum. Inoculation Test plants were inoculated two weeks after planting, when the second leaf had completely unfolded. The inoculum was applied to the leaves of test plants by a hand sprayer until near runoff. In each experiment, un-inoculated plan ts were included to monitor natural senescence and physiological leaf spotting. Where these occurred, the data was discarded. The inoculated plants were left in a humid ity chamber at 90-100 RH for 48 h, after which they were removed and placed on a growth room bench with a 16 h photoperiod and 22oC/16oC day/night temperature. Disease Assessment Inoculated plants were rated for disease reaction seven days after inoculation on a 0-9 scale in which 0= highly resistant and 9= highly susceptib le (complete necrosis of the leaves). This type of rating took into consideration the size of lesions and the amount of chlorotic tissue (Wong and Hughes, 1989). In situations wh ere disease categories overlapped,, it was necessary to group the reactions into three broa d classes, resistant (ratings 0-3), moderately resistant (ratings 4-5) and susceptible (ratings 6-9) for easier interpretation of the data. Results and Discussion The F1 plants of all the crosses had disease ratings within the parental range. Some crosses produced F2 populations that had all plants rated within that parental range while others had slight or obvious devia tions, but in every case, most of th e plants fell within the parental range. Based on the disease ratings of the F2 populations of these crosses, three types of segregation were became apparent. The first t ype involved the EE 8 x Red Chief and Hadden

PAGE 119

x 821WWMN 2019 crosses (Table 2). In both crosses, the ratings of plants in the F2 populations fell within the parental range. Th is suggested that parents involved in these crosses possess genes that are allelic. These results support Mas (1993) finding of a resistance gene common to both EE 8 and red Chief. Three F2 plants of the EE 8 and Red Chief cross-rated outside the parental range. Howe ver, these plants had a rating of 3, which is within the resistant range. Occasionally, a few Red Chief plants were rated 3. Table 1. Cultivars used in studies of allelism of resistance genes to Leptosphaeria nodorum and their disease reaction. Cultivar Disease reaction Red Chief Resistant EE 8 Resistant 821WWMN 2019 Resistant Hadden Resistant Atlas 66 Resistant 831WWMN 2051 Resistant Kenyon Susceptible Table 2. Distribution of disease ratings of parental, F1 and F2 populations of crosses involving EE 8, Red Chief, Hadden and 821WWMN 2019 and the susceptible control cv. Kenyon using lesion type ratinga for resistance to Leptosphaeria nodorum tested at second leaf stage under controlled environmental conditions. Lesion Type Frequency a 0 1 2 3 4 5 6 7 8 9 Total plants Mean SD EE8 X Red Chief EE8 17 15 16 48 0.97 0.83 F1 2 3 5 0.60 0.54 F2 20 23 29 3 75 1.20 0.88 Red Chief 38 5 7 50 0.38 0.72 RC(F1) 2 2 1 1 3 3 4 0.50 0.52 Kenyon 8 0.80 1.06 Hadden X 821WWMN2019 Hadden 3 6 23 32 1.62 0.65 F1 15 15 2.00 0.00 F2 8 42 66 116 1.50 0.62 821WWMN2019 9 19 2 2 19 12 30 1.76 0.56 Kenyon 33 8.30 0.58 a Rated on a 0-9 scale

PAGE 120

The second type of segregation was observed in the F2 populations of the Atlas 66 x Red Chief and Red Chief x 821WWMN 2019 crosses (Table 3). Most of the F2 plants of these crosses had disease ratings within the parental range suggesting allelism between the Atlas 66 and Red Chief, and between Red Chief a nd 821WWMN 2019 resistance genes. A small proportion of moderately resistant or susceptible plants occurred in the F2 populations of these crosses. However, the susceptible plants obser ved were never as susceptible as the susceptible control Kenyon and thus were not consider ed to represent susceptible genotypes. The observed deviations coul d have been caused by misclassification or environmental variability within the experiment resulting from e.g. non-uniform inoculation of the plants. The third type of segregation involved F2 plants of the Atlas 66 x 821WWMN 2019, Hadden x Red Chief and Red Chief x 831WWMN 2051 and EE 8 x 831WWMN 2051 crosses (Table 4 and 5). While most of the F2 plants of these crosses had disease ratings within the parental range, the proportion of plants which had disease ratings outside the parental range was large enough for both allelism and modifier gene effects to be suggested. The distribution of plants outside the parental range was continuous and the susceptible plants in these populations were not as susceptible as Kenyon. The presence of modifier genes influencing the expression of resistance to S nodorum has been reported by other workers (Laubscher et al ., 1966; Kleijer et al ., 1977; Ma, 1993). Table 3. Distribution of disease ratings of parental, F1 and F2 populations of crosses involving Atlas 66, Red Chief and 821WWMN 2019 and susceptible control cv. Kenyon using lesion type rating for resistance to L. nodorum tested at second leaf stage under controlled environmental conditions Lesion Type Frequency a 0 1 2 3 4 5 6 7 8 9 Total plants Mean SD Atlas66 X Red Chief Atlas 66 7 14 2 23 1.78 0.59 F1 3 6 9 0.66 0.50 F2 5 19 32 43 4 4 107 2.32 1.07 Red Chief 12 6 6 3 27 1.00 1.07 Kenyon 2 7 15 24 8.54 0.65 Red Chief X 821WWMN2019 Red Chief 10 4 5 4 23 1.13 1.17 F1 2 7 2 11 1.00 0.63 F2 22 11 22 42 2 2 1 102 1.96 1.28 821WWMN2019 1 8 9 2 1 21 1.71 0.90 Kenyon 2 7 15 24 8.54 0.65 a Rated on a 0-9 scale

PAGE 121

Table 4. Distribution of disease ratings of parental, F1 and F2 populations of crosses involving Atlas 66, 821WWMN 2019, Hadden and Red Chief and the susceptible control cv. Kenyon using lesion type ratinga for resistance to L. nodorum tested at second leaf stage under controlled environmental conditions. Lesion Type Frequency a 0 1 2 3 4 5 6 7 8 9 Total plants Mean SD Atlas66 X 821WWMN2019 Atlas 66 7 14 2 23 0.78 F1 2 13 15 0.86 0.35 F2 2 16 21 39 12 3 7 100 2.80 1.37 F3 260 276 181 57 3 7 3 4 791 1.14 1.13 821WWMN2091 1 8 9 2 1 2 7 15 24 8.54 0.66 Kenyon Hadden X Red Chief Hadden 3 5 15 1 24 1.58 0.77 F1 5 6 4 15 0.93 0.78 F2 11 36 23 13 12 3 1 1 100 2.07 1.41 F3 110 219 29 67 17 2 5 3 452 1.34 1.30 Red Chief 10 7 1 4 22 0.95 1.13 Kenyon 2 7 15 24 8.54 0.65 a Rated on a 0-9 scale

PAGE 122

Table 5. Distribution of disease ratings of parental, F1 and F2 populations of crosses involving Red Chief, 831WWMN 2051 and EE 8, and the sus ceptible control cv. Kenyon using lesion type ratinga for resistance to L. nodorum tested at second leaf stage under controlled environmental conditions. Lesion Type Frequency a 0 1 2 3 4 5 6 7 8 9 Total plants Mean SD Red Chief X 821WWMN2051 Red Chief 22 4 4 1 22 1.66 0.70 F1b 3 3 2.00 0.00 F2b 12 2 36 12 6 4 4 2 78 2.46 1.68 F3 93 236 279 73 15 29 19 9 2 0 755 1.86 1.45 821WWMN2091 6 10 14 7 37 1.59 0.98 RC (F1) 7 4 11 1.36 0.58 Kenyon 3 15 14 32 8.34 0.65 EE8 X 821WWMN2051 EE8 15 15 4 34 0.67 0.68 F1b 4 4 2.00 0.00 F2 9 3 60 9 10 4 2 97 2.30 1.40 F3b 82 59 155 43 14 8 10 1 372 1.77 1.39 821WWMN2091 6 10 14 7 37 1.59 0.98 RC (F1) 5 7 2 14 0.78 0.69 Kenyon 2 7 15 24 8.54 0.65 a Rated on a 0-9 scale. b=Reciprocal cross progeny not included, RC= Reciprocal cross F2 plants with disease ratings outside the parental range did not occur in the EE 8 x Red Chief cross (Table 2), but were observed in the Red Chief x 831WWMN 2051 and EE 8 x 8311WWMN 2051 crosses (Table 5). However, ev en where such plants were observed, the ratings of the most F2 plants were within the parental ra nge. This suggested that EE 8, Red Chief and 831WWMN 2051 have a common resistan ce gene. Allelism of one resistance gene in these parents is further supported by the fact that the F1 disease ratings were within the parental range. In the Atlas 66 x 821WWMN 2019, Atlas 66 x Red Chief and Red Chief x 821WWMN 2019 crosses (Table 3 and 4) most of the F2 plants were rated within parental range. However, there were more plants in the F2 population of the Atlas 66 x 821WWMN 2019 that deviated from parental range than in the other crosses. Th is suggested the presence of modifying genes or other undetected resistance genes. There is a possibility that Atlas 66 also possesses other resistance gene(s) apart fro m the one reported by other workers (Frecha, 1973; Kleijer et al ., 1977). Data from the Atlas 66 x Red Chief cross (Table 3) suggested that Atlas 66 possessed the Red Chief resistance gene which is located on chromosome 3A. The allelic relationship that was proved in this st udy, plus the fact that cultivars EE 8 and Red Chief each has a resistant gene located on ch romosome 3A (Ma, 1993) suggests that the allelic gene in all the resistant cultivars is al so located on chromosome 3A. Allelism has been implicated in other host-pathogen systems when resistant parents produced F2 populations that did not segregate ( Zink and Gubler, 1986; Hibberd et al ., 1987; Sigh and McIntosh, 1988; Potts, 1990; Sykes and Bernier, 1991). The F3 generation of those crosses in which susceptible F2 plants had occurred were tested. Although susceptible F3 plants were observed (Table 6), none of the F3 families were susceptible (Table 7). These observations support the conclusions made from the F1 and F2

PAGE 123

data that the resistance genes in the parents are allelic. If there was no allelism, susceptible F3 families would have occurred. Table 6. Distribution of disease ratings of parental, F3 populations of crosses involving EE 8, 831WWMN 2051, Red Chief, Atlas 66, 821WWMN 2019, Hadden and Red Chief, and susceptible cv. Kenyon using lesion type ratinga for a resistance to L. nodorum tested at the second leaf stage under controlled environmental conditions. Lesion Type Frequency a 0 1 2 3 4 5 6 7 8 9 Total plants Mean SD EE8 X 831WWMN2051 F3 82 59 155 43 14 8 10 1 372 1.77 1.39 831WWMN2051 1 0 6 2 9 2.00 0.86 EE8 6 2 1 9 0.44 0.72 Kenyon 2 7 9 7.77 0.44 Red Chief X 831WWMN2051 F3 93 236 279 73 15 29 19 9 2 755 1.86 1.45 Red Chief 4 6 1 11 0.72 0.64 831WWMN2051 7 7 1 15 1.60 0.63 Kenyon 3 5 9 17 8.30 0.78 Atlas 66 X 821WWMN2019 F3 260 276 181 57 3 7 3 4 791 1.14 1.13 Atlas 66 8 6 2 16 0.62 0.71 821WWMN2019 2 10 4 2 18 1.33 0.84 Kenyon 5 3 9 17 8.23 0.90 Hadden X Red Chief F3 110 219 29 67 17 2 5 3 452 1.34 1.30 Hadden 5 6 1 3 15 2.13 1.12 Red Chief 3 12 4 19 1.05 0.62 Kenyon 2 4 8 14 8.42 0.75 a Rated on a 0-9 scale Table 7. F3 family segregation for resistancea to L. nodorum in seedling tests of four crosses. Cross homozygous Total resistant Segregation Susceptible Familie s Red Chief x 831WWMN 2051 25 23 0 48 EE 8 x 831WWMN 2051 20 19 0 39 Atlas 66 x 831WWMN 2051 43 8 51 Hadden x Red Chief 28 17 0 45 a 0-3=Resistant; 4-5=Moderately resistant; s= Susceptible esion Type Frequence a One explanation for these observations is to hypothesize that all the parents have a common resistance gene. This hypothesis is supported by the fact that the various cross combinations of these parents produced F1 plants with disease ratings within the parental range. Most or all of the F2 generations of these crosses had pl ants rated within the parental range. Furthermore, no segregation was observed in the F3 families (Table 7).

PAGE 124

The occurrence of plants that rated outside the parental range in the F2 population of the Hadden x Red Chief cross but not of the Hadden x 821WWMN 2019 cross suggested the presence of a second resistance gene in Hadden and 821WWMN 2019. A second resistance gene is suggested because the disease ratings of all F2 plants of the Hadden x 821WWMN 2019 cross fell within the parental range (Table 2) and, except for three plants, the same was observed for the Red Chief x 821WWMN 2019 cross (Table 3). This gene(s) may be different from that in EE 8 because deviations from parental range were observed in the F2 population of the Hadden x Red Chief cross (Table 4), but not in the EE 8 x Red Chief or Hadden x 821WWMN 2019 crosses (Table 2). The implication of these results is that although it appears that there is a common resistance gene in all the resist ant cultivars used in this study, other resistance genes may be present in certain cultivars. The existence of different genes for resistance provides the opportunity to combine resistance genes from different sources to enhance the level of resistance to septoria nodorum blotch in a breeding program. References Eyal, Z., Scharen, A.L., Prescott, J.M., and van Ginkel, M. 1987. The septoria diseases of wheat: Concepts and methods of disease management. CIMMYT, Mexico. 46pp. Frecha, J.H., 1973. The inheritance of resistance to Septoria nodorum in wheat. Bol. Genet. Fitotec. Castelar 8: 29-30 Hibberd, A.M., Bassett, M.J., and Stall, R.E 1 987. Allelism tests of three dominant genes for hypersensitive resistance to bacterial spot of pepper. Phytopathology 77:1304-1307 Kleijer, G., Bronnimann, A., and Fossati, A 1977. Chromosomal location of a dominant gene for resistance at the seedling stage to Septoria nodorum Berk. In the wheat variety Atlas 66. Z. Pflanzenzuichtg 78:170-173. Laubscher, F.X., von Wechmar, B., and van Schalkwyk, D. 1966. Heritable resistance of wheat varieties for glume blotch (Septoria nodorum Berk.). Phytopathol. Z. 56:260-264. Ma, H. 1993. Genetic and cytogenetic studies of resistance to Septoria nodorum in tetraploid and hexaploid wheat. Ph.D thesis. University of Saskatchewan, Saskatoon. Nelson, L.R. 1980. Inheritance of resistance to Septoria nodorum in wheat. Crop Sci. 20:447-449. Nelson, L.R., and Gates, C.E. 1982. Gene tics of host plant resi stance of wheat to Septoria nodorum Crop. Sci. 27:771-773. Potts, D.A. 1990. Expression and genetics of resistance to Septoria tritici in wheat. PhD Thesis. University of Saskatchewan, Saskatoon. Primard, S.J., Morris, R., and Papa, C.M. 1991. Cytogenetic studies on a heterozygous reciprocal translocation in the wheat ( Triticum aestivum ) cultivar Atlas 66. Genome 34:313-316. Singh, S.J., and McIntosh, R.A 1988. Allelism of two genes for stem rust resistance in triticale. Euphytica 38:185-189. Sykes, E.E., and Bernier, C.C. 1991. Qualitative in heritance of tan spot resistance in hexaploid, tetraploid and diploid wheat. Can. J. plant Pathol. 13:38-44. Wong, L.S.L. and Hughes, G.R. 1989. Genetic control of seedling resistance to Leptosphaeria nodorum in wheat. Pp 136-138, in:Fried, M.P., ed. Septoria of cereals. Proc. Workshop, July 4-7, 1989, Zurich, Switz. 189pp. Zink, F.W., and Gubler W.D. 1986. Inheritance of resistance to races 0 and 2 of Fusarium oxysporum f. sp melonis in gynoecious muskmelon. Plant Disease 70:676-678.

PAGE 125

Evaluation of Kenyan Breadwheat ( Triticum aetivum L.) Varieties for Resistance to Russian Wheat Aphid in Multi-location Trials J. Malinga1, M. G. Kinyua1, A. Kamau2, J. K. Wanjama 3, P. Njau1, and J. Kamundia1 1National Plant Breeding Research Center. P.O. Njoro 20107,. Kenya; 2Egerton University, Department of Agronomy, P.O. Box 536, Njoro Kenya; 3Ministry of Agriculture, P.O. Box 30028, Nairobi, Kenya. Abstract Russian wheat aphid (RWA, Diuraphis noxia: Kurdjumov) is a serious pest of wheat (Triticum aestivum L.) worldw ide. The purpose of this investigation was to evaluate commercial bread wheat cultivars in different environments in Kenya to identify RWA resistance. Seven wheat cultivars (Pasa, Mbuni, Kenya Heroe, Kenya Fahari, Chozi, Duma and Kwale) were planted at five locations over two years in Kenya in a split plot RCBD experimental design. The main plot consisted of insecticide treatment at two levels (untre ated and insecticide treated). The subplot was the varieties. Results showed significant (P<0.05) genotypic differences across location. Effects due to insecticide year location, and genotype year location interaction were significant (P< 0.05%) fo r thousand-kernel weight and grain weight. Positive correlation was observed between plant height and yield (r = 0.935 **), and thousand-kernel weight and yield (r = 0.876**). Significant negative correlations were recorded between RWA damage and plant height (r = -0.662**); and yield (0.785**); and kernel weight -0.667**). No significant association was observed for percentage plant height reduction and RWA damage. The first two principal components accounted for 83% of the vari ability. The study shows that K. Fahari, which has been previously reported to be resistant to green bug may have some resistance to RWA. Introduction Russian wheat aphid (RWA), Diuraphis noxia (Kurdjumov), is a major pest of small grained cereals in the world (Liu et al ., 2001). RWA is endemic to Eastern Europe, and has been invading other regions of the world after being identified in South Africa in 1978, (Walters, 1984; Marassas et al ., 1995). It rapidly spread to Central, North and South America, and most parts of Europe. In Africa, it has maintained minor status in Egypt, Sudan and Ethiopia (Miller and Pike, 2002) but flared up in Kenya in 1995 where it remains the most important cereal pest of wheat and barley (Anon., 2002). The pest is difficult to detect until symp toms occur on the plant. RWA may cause yield losses of up to 90% (Du Toit and Walters, 1984: Hewit, 1988), and in Kenya most wheat varieties appear to be susceptible. Farmers, even with the use of insecticide application, have reported high yield losses of upto 40% (Kinyua et al. 2002). Despite this, information on performance of wheat varieties infested by RW A is lacking. Preliminary greenhouse studies indicate that all Kenyan varieties are susceptible to RWA (Malinga et al ., 2001). Despite the susceptibility of wheat varieties to RWA, it may be possible to identify low levels of RWA resistance that can be useful to farmers as they await the development of RWA resistant varieties. Varieties previously thought to be susceptib le have been found elsewhere to have low resistance (Hein, 1992; Smith et al. 1991). Low resistance also called tolerance has been

PAGE 126

defined as qualities that show less damage th an the average crop (Thomas and Waage 1996; Butts and Pakendorf, 1984; Painter, 1951). Tolera nce is identified on the ability of the crop to recover despite pest attack (Thomas and Waage, 1996). Yield reduction is one of the measures that may indicate tolerance (Painter, 1951). However, although yield is highly correlated with tolerance it should be carefully treated, as it is the output of many other factors. Quantifying degree of yield reduction attributed to RWA in crop loss assessment studies is useful measures to aid identify varieties with resistance or tolerance to RWA. The quantification RWA damage on wheat varieties, across locations, would therefore be useful in the latter manageme nt of the pest. The objective of this study was to quantify reactions due to RWA infestati on on selected varieties and identify those with low levels of RWA resistance. Materials and Methods Site descriptions The experiment was conducted at five sits described below. Njoro, Kenya (0 20'S; 35 56'E; 2166m above sea level (asl)) is located in the lower highlands with a mean annual rainfall of 936 mm (Kinyanjui, 1979) and temperatures of 7.9 oC to 21.9o C and a mean of 14.9oC. The soils are fertile vitric mollic Andosols that are well drained, deep to very deep, dark reddish brown in colour, consisting of heavy textured friable silty clay to clay humic top soils (Jaetzold a nd Schmidt, 1983a). These are well drained, deep to very deep dark reddish brown friable and s ilt clay to clay soils with humic topsoil. Timau, Kenya (0 05'S; 37 20'E; 2640m asl) is located in the upper highlands in the pyrethrum-wheat zone. The site receives a nnual rainfall of 1170 mm with temperatures ranging from 7.5 to 18.36o C. Absolute minimum temperatures of 0-2o C are recorded. The soils are chromi c and ferric Luvisols with Lithosols These soils are well-drained deep, dark reddish brown and red friable sandy clay, which are shallow to deep. Eldoret, Kenya site was located at Chepkoilel Campus, Moi University ( 0 N; 35 W; 2133 m asl) with mean annual rainfall of 885 mm and maximum and minimum temperatures of 23.5 and 9.5 respectively. Naivasha, Kenya is about 70 km from Nairobi in Nakuru province, at an altitude of around 1800 m asl, with a warm and dry climate. and daytime temperatures are up to 30C, with rainfall of 500-700mm per annum. Katumani, Kenya (1 35S; 37 14E; 1575m asl) is loca ted in the upper midland zones. The site receives 700mm of rainfall annually with mean temperatures of 19.3o C. The soils are well drained, deep, dark reddish brown, friable sand clay, classified as ferral chromic Luvisols (Jaetzold and Schmidt, 1983b; FURP, 1988). Rearing RWA aphid cultures A total of three different colonies of RWA a phids were established one each from three key wheat-producing areas of Kenya (Njoro, Timau and Eldoret). Aphids collected from the specified locations were multiplied into col onies. Viviparous female RWA aphids were placed on clean and non-infested plant raised under aphid free conditions in the greenhouse. The females were caged inside low cost plex i cage (50cm x 55 cm x 48cm). The aphids were multiplied on wheat seedlings of the variety Mb ega under greenhouse conditions. The aphids were harvested after one month of caging usi ng a paintbrush into Petri dishes dusted with talcum powder. Aphids were used to infest test varieties in regions were there were collected.

PAGE 127

Experimental design and field layout The trial was planted in a randomized complete block in split plot arrangement replicated three times. Two factors were investigated: insecticide treatment (insecticide treated vs. untreated) as main plots and seven wheat genotyp es as sub plots. The genotypes tested were Kenya Fahari, Duma, Chozi, Kenya Heroe, Kwa le, Pasa, and Mbuni (Table 1). The main plots (24m long x 3m wide) separated by 0.5 m alley were randomly allocated to insecticide treatment. Variety was allocated to the sub-plots measuring 3m x 1m and separated by 0.5m paths. Each subplot had 4 rows spaced 20 cm apart. Genotypes were manually drilled in hill plots spaced at 20cm seed rate of 125kg/ha and di-ammonium phosphate (DAP) 18: 46:0 (NPK) fertilizer applied at the rate of 60 kg/ha in them prior to planting of the seed. Hand weeding was carried out when plants reached growth stage 16 (Zadok et al ., 1974). Foliar diseases were controlled at Njoro and Eldoret wh ere the crop succumbed to rust infection by spraying with one fungicide spray. Table 1: Wheat varieties, released in Kenya, used in this study. Variety Released by Pedigree Chozi NPBRC F12.7/COC//GEN Duma NPBRC AU/VP/301//GLL/SX/3/PEW'/4/MAYA/PEW Fahari NPBRC Tob66/SRPC527/CI8155/2*Fr K. Heroe NPBRC Mbuni/SRPC64/YRPC1 Kwale NPBRC KINGLET Mbuni NPBRC ZA75/LD357E-Tc3xGU, CM30520-1B-1B-3Y-0Y Pasa NPBRC FINK Infestation procedure At the two to three leaf stage (Zadoks 12-14) the test plots were infested with greenhouse reared aphids and aphids picked from surroundi ng crop using a paint brush. The purpose was to ensure infestation. Approximately three to five aphids were placed on plants at 30 cm intervals. The aphids were allowed to multiply freely. Un-infested plots were protected at planting with seed application of Gauc ho 350 FS (200ml/100kg seed ) and subsequent monthly application of metasystox (0.5 l/ha) for 2 months. Although insecticide treatments would be used to control other insects as well, RWA was by far the predominant insect pest species present. The crop was not caged. Parameters measured The following parameters were measured during the growth period of the crop: Plant height was taken at tillering, jointing and maturity to determine the number of plants stunted due to RWA damage. Damage rating scores were taken at tille ring and jointing from 5 randomly selected plants in each plot to determine the degree of RWA damage. A visual 1-6 scale (1=resistant to 6=very susceptible) using a nondestructive method was used. Number of infested plants per m2 per plot was recorded on naturally infested plots at harvest. Type of natural enemies and observed predators on 5 random plants was observed at tillering and jointing to determine the numbe r of predators reducing aphid numbers. Number of parasitized (mummies) and diseased (cadavers) per five random plants sampled at Zadok's 12 (8 DAE), tillering 26 (29 DAE), to harvest was measured. Number of spikes per plant at harvest, with average taken from 3 plants in 3 replicates. Total grain weight was determined in 3 replicates by harvesting the entire plot.

PAGE 128

Thousand Kernel weight (TKW) at harvest was determined by taking the weight of 1000 seeds. Reductions in grain yield, spike number, tiller number and 1000-kernel weight were calculated using the formula XR=(1-Xi /Xu ) 100 where X is the component of yield in question, Xi is the value of X in infested plots and Xu is the value in un-infested plots (Calhoun et al ., 1991). Field temperatures, humidity, and rainfall were collected over the test period. Data collection and analysis Data was collected on using a quadrant (0.5 m X 0.5m) taken from the two middle rows at three stages (tillering, booting and harvesting) fo r RWA damage, plant height, tiller number, and spike length. However, all rows constitute d the harvest plot from which both thousand kernel weight and grain yield were derived. Data was analyzed using SAS package and means separated using Duncan mean range test. Corre lations were also carried out to determine significant associations. Results and Discussion Across location and year analysis The analysis of insecticide treatment and genot ypic variation within and across locations and years is shown in Table 2. Highly significan t differences were found on all traits between years except on tiller number. Significant differen ces were reported on locations and location x year interaction on all traits. Significant In secticide treatment x year, Insecticide treatment x location interaction and Insecticide x year x location was recorded on RWA damage, thousand kernel weight and grain yield. Environment means for RWA damage, plan t height, tiller number thousand kernel weight and grain yield over locations over two years are presented in Tables 3 and 4. There were highly significant (P<0.05) differences am ong insecticide treatment and genotypes at all locations in both years for all traits except tiller number. Thousand kernel weights and yields were reported in insecticide treated plots with significantly (P<0.05) lower damage. When the means over two years were considered, RWA damage reduced yields by 41% in untreated plots than in control plots. Thousand kernel weights recorded 11% reduction. Table 2: Combined mean squares for Russian wheat aphid damage, agronomic traits and yield for seven bread wheat commercial varietie s at five locations, 2002 and 2003. Source of variation df RWA damage (score 1-6) Plant height Tiller number Thousand kernel weight Yield Years (Y) 1 2.6** 20847.8*** 0.003ns 1260.1*** 12.3*** Locations (L) 4 18.9*** 14031.6*** 35.7*** 2186.2*** 59.1*** LxY 4 6.5*** 3655.7*** 107.3*** 820.6*** 22.6*** Rep(LxY) 20 0.8ns 115.2ns 3.3ns 15.3ns 0.2ns Factor A 1 226.2*** 2033.2** 7.0ns 1529.6*** 39.3*** YxA 1 9.7*** 205.9ns 0.4ns 269.6*** 4.6*** LxA 4 5.4*** 216.6ns 1.9ns 34.1* 2.5*** YxLxA 4 3.1*** 132.5ns 4.5ns 84.4*** 1.5*** Error 8 6.1 195.3 2.6 11.0 0.2

PAGE 129

Factor B 6 4.3*** 6037.8*** 3.2ns 467.2*** 3.4*** YxB 6 0.9* 65.4ns 5.3ns 13.1* 0.2ns LxB 24 0.9*** 130.4*** 2.9ns 86.0*** 0.9*** YxLxB 24 0.4ns 121.6*** 1.5ns 30.0*** 0.6*** AxB 6 1.0** 12.8ns 1.6ns 4.8ns 0.1ns YxAxB 6 0.7ns 27.2ns 0.9ns 1.8ns 0.2ns LxAxB 24 1.0*** 49.6ns 2.0ns 5.2ns 0.1ns YxLxAxB 24 0.5ns 73.9ns 1.3ns 8.1ns 0..2ns Error 240 0.3 48.5 1.8 5.2 0.2 Total 419 586.7 158960.5 1393.5 23403.7 508.7 Genotypic means over locations and years showed that one variety, K. Fahari significantly (P<0.05) out-yielded all others, and significantly (P<0.05) recorded the highest thousand kernel weight, tallest plant height. However, significantly lower RWA damage (2.3) was found on K. Fahari. RWA damage on Duma and Chozi (2.5) did not differ from that on K. Fahari. Both Duma and Chozi are varieties bred for dry areas. Significantly low RWA damage indicates low resistance to RWA. K. Kwale was significantly (P<0.05) the most susceptible to RWA and had the most the shortest plant heights, thousand kernel weights and yields.

PAGE 130

Table 3. Mean response and percentage loss fo r agronomic traits on seven wheat cultivars infested with Russian wheat aphid, at five locations, 2002 and 2003. Damage Plant Height % Red. Tiller no % Red. TKW % Red. Yield % Red. Chozi 3.2d 92.8b 5 6.5 10 36.0b 12 1.7b 35 Duma 2.9e 84.9cd 4 5.8 2 34.5c 9 1.8b 39 Fahari 2.9e 103.2a 4 6.2 5 37.6a 10 2.1a 29 Heroe 3.5bc 87.2c 5 6.1 0 31.2e 13 1.6bc 31 Kwale 3.9a 77.0e 7 6.5 9 32.8d 14 1.4c 43 Mbuni 3.3cd 82.7d 4 5.9 5 32.9d 11 1.6bc 37 Pasa 3.6b 72.8f 4 5.5 -2 30.0f 13 1.4c 43 Se (0.05) 0.1 1.2 ns 0.4 0.1 CV 21.9 8.3 22.9 7.2 28.2 Combined over sites *** *** *** *** *** Combined over years ** *** ns *** *** Table 4. Means of Russian wheat aphid damage, yield and yield parameters at each of the 5 locations in Kenya, 2002 and 2003. RWA damage Plant height Tiller no Thousand Kernel weight Yield Eldoret 2.7c 86.1b 5.3ab 31.0c 1.0c Njoro 1.9e 86.5b 6.9c 27.8d 0.7d Timau 3.0b 92.5a 6.2ab 37.6a 2.3a Naivasha 2.5d 91.9a 5.5b 36.3b 2.1b Katumani 3.1a 61.1c 5.7a 26.0e 0.5e Mean 2.6 83.6 5.9 31.7 1.3 Se (0.05) 0.1 1.5 0.2 0.4 0.1 CV% 21.9 8.3 22.9 7.2 28.3 Significant at P< 0.05

PAGE 131

Table 5. Pearsons correlation coefficient matrix Damage % Dam. Red. Plant ht. % Plant ht. Red. Tiller no. % Tiller Red. TKW % TKW Red. Yield % Yield Red. Damage 1.00 0.50 -0.73 0.71 0.13 0.02 -0.72 0.90* 0.89* 0.57 % dam Red. 1.00 -0.57 0.37 -0.10 -0.26 -0.66 0.27 -0.61 0.28 Plant ht. 1.00 -0.26 0.43 0.32 0.85 0.479 0.940.89* % Plant Ht Red. 1.00 0.71 0.55 -0.13 0.70 -0.46. 0.31 Tiller no. 1.00 0.87* 0.52 0.33 0.21 -0.29 % Tiller no. Red. 1.00 0.60 0.15 0.20 -0.04 TKW 1.00 -0.54 0.89* -0.54 % TKW Red. 1.00 -0.73 0.33 Yield 1.00 -0.76 % Yield Red. 1.00 Significant at 5% Correlation between parameters Pearsons correlation coefficients across tw o seasons for RWA damage, plant height, tiller number, thousand kernel weight and yield ar e presented in Table 5. There were highly significantly strong and positive associations were recorded over the two years between damage and percent TKW reduction (r=0.90) ; an d, damage and yield (r=-0.89). In addition significantly strong and positive associations were also recorded between plant height and kernel weight (pooled r=0.85, P<0.001); plant height and yield (pooled r= (0.94, P<0.001) and plant height and percentage yield reduction. Thus, taller plants had heavier kernels and higher yields. Subsequently a significant st rong and positive association was recorded between percentage TKW reduction and yield (p ooled r=0.89, P<0.001). However there was no significant association recorded between per centage plant height reduction and damage. Conclusion In this study, K. Fahari and Duma suffered the lowest RWA damages. In addition, K. Fahari a tall wheat variety sustained better yields when infested with RWA. Perhaps the ability to recover through rapid and extended stem growth period enables the plants of this variety to compensate for the attack borne on its growing shoot. Environment has a bearing on degree of RWA damage and final yields received in the field. This is seen from high yield losses due to RWA found in drier areas Katumani and Naivasha where particularly high RWA damage was recorded. This is important considering the fact that the future of wheat expansion in Kenya lies within the marginal lands. The higher yield losses in marginal areas were somewh at expected. Higher temperatures lead to shorter lifecycles for RWA and under extremel y high temperatures paedogenesis (nymphs

PAGE 132

giving birth to young) occurs (Wanjama, 1986). The implications are that taking wheat to dry land regions must be accompanied by a package and development of resistant and adapted varieties must be a cornerstone. Acknowledgments -This work is part of PhD study at Egerton University. The funds used were from The World Bank through Kenya Agricultural Research Institute (KARI). References Du Toit, F. 1989. Inheritance of resistance in two Triticum aestivum lines to Russian wheat aphid (Homoptera; Aphididae). J. Econ. Entomol. 82: 1251-1253 Du Toit. 1988. Another source of Russian wheat aphid (Diuraphis noxia) resistance in Triticum aestivum Cereal Res. Commun. 16:105-106. Ehdaie B and C. A. Baker. 1999. Inheritance and a llelism for resistance to Russian wheat aphid in an Iranian spring wheat. Euphytica 107: 71-78. Elsidaig A. and P.K. Zwer. 1993. Genes for resist ance to Russian wheat a phid in PI 294994 wheat. Crop Sc. 33: 998-1001. Gomez K.A. and A.A. Gomez. 1984. Statistical procedure for Agricultural research. John Wiley and Sons. Inc. Jaetzold R and Schmidt 1983. Farm Management handbook of Kenya Ministry of Agriculture. Kenya. Central Kenya (Rift Valley and Central Province). Vol II /B Liu X.M., Smith C. M., Gill B. S. and Tolmay V. 2001. Micro satellite markers linked to six Russian wheat aphid resistance genes in wheat. Theor. Appl. Genet. 102: 504-510 Malinga J. N., M. Kinyua, L. Karanja L and E. Alomba. 2001. Multi-locational evaluation of wheat lines for resistance to cereal aphids with pa rticular reference to Russian wheat aphid. In Proceedings of the first Annual Conference of National Plant Breeding Research Centre, Njoro. 26th-30th November 2001: 44-47 Marais G.F. and F. Du Toit. 1993. A monosomic analysis of Russian wheat aphid resistance in the common wheat PI 294994. Plant Breed 111:246-248. Marasas C.; Anandajayasekeram, P.; Tolmay V.: Ma rtella, D; Purchase J.; Prinsloo, G: 1997. Socioeconomic impact of the Russian wheat aphid control Research Program SACCAR. Gaberone, Botswana.147 p. Miller, C. A., A. ALtinkut and N.L.V.Lapitan. 2001. A micro satellite marker for Tagging Dn, a wheat gene conferring resistance to the Russian wheat aphid Crop Sci 41:1584-1589. Meyer W. L., K.K Nkongolo., F. B Peairs. and J. S. Quick 1989. Mechanism of resistance in the wheat line PI 37129 to the Russian wheat aphid. In: D. Baker (ed) Proceedings of the 3rd Russian wheat aphid conference. Albuquerq ue, New Mexico. 25-27 Oct. 1989 : 23-24 Nkongolo K.K., J.S. Quick and F.B. Peairs and W. L. Meyer. 1991. Inheritance of P1 372129 wheat to Russian wheat aphid. Crop Sc i. 31: 689-692. Robinson J. 1994. Identification an d characterization of Resistance to the Russian wheat aphid in small grain cereals: Investigations at CIMMYT, 1990-1992. CIMMYT Research Report No 3 Mexico D. F. CIMMYT. 44p Tolmay V., van Deventer C. S. and van der Westhuizen M. C. 1999. Inheritance to resistance to Russian wheat aphid (Diuraphis noxia (Homoptera: aphididae) in two wheat lines, In: Proceedings of the tenth regional wheat workshop for Eastern, Central and Southern Africa. Pretoria University, South Africa 13-17, January 1997: 408-417. Walters, M.C.1984. Progress in Russian wheat aphid ( Diuraphis noxia Mord.) research in the Republic of South Africa. Proceedings of a meeting of the Russian Aphid Task Team held at the University of the Orange free State. Bl oemfontein 5-6 May 1982. (ed) M.C. Walters. Technical Communication No 191. department of agriculture, republic of South Africa. Zadok J.C. Chang T.T. and Konzak C. F. 19 74. A decimal code for growth stages of cereals Weed Sci 14: 415-421

PAGE 133

On-Farm Evaluation and Comparison of New and Old Wheat Varieties R. V. Ndondi1, C .A. Kuwite1 and R.Shekibula1 Selian Agricultural Research Institute, P. O .Box 6024, Arusha, Tanzania Abstract Progress in new wheat variety development is measured by how much the new wheat varieties excel old varieties in terms of disease resistance, yield, consumer preference and other desirable agronomic traits. In 2002 eight old wheat varieties were evaluated and compared with 22 new advan ced wheat lines on farmers fields in different agro-ecological zones in Morogoro, Iringa, Arusha and Kilimanjaro wheat growing regions of Tanzania. Data on reaction to diseases, grain yield and farmers variety preference was obtained from various sites. Most of the old wheat varieties were susceptible to stem, leaf and yellow rust, while many new wheat varieties were resistant to most pr evailing wheat diseases. Combined analysis across locations indicated that mean grain yields ranged from 1.0 t/ha to 1.87 t/ha with five new wheat varieties having the highest mean grain yields. During field days at various sites, farmers selected seven new wheat varieties and only one of the old wheat varieties. The most preferred new wheat varieties will be further evaluated on farm to confirm on their superiority over old wheat varieties. Introduction Wheat varieties currently used by farmers in Tan zania were released more than fifteen years ago and are now susceptible to nearly all three rusts (stem, leaf and yellow rust) and their yield potential has been greatly reduced (Ndondi et al. 2001). Therefore old wheat varieties need to be replaced with new improved disease resistant and high yielding wheat varieties. Such cultivars are available, however, these vari eties have not been tested extensively in different agro-ecological zones of Morogoro, Iringa, Arusha and Kilimanjaro regions, mainly due to lack of funds. These varieties have consequqntly not been exposed to small scale farmers and farmer preferences have not been determined. Variety evaluation in different environments is vital to determine the perfo rmance of varieties across locations (Sariah et al. 1990, Chirot et al. 1990, Eberhart et al. 1966). In order to identify new, well adapted wheat varieties for small scale farmers in these regions, on-farm trials were conducted in the different agro-ecological zones with full participation of small scale farmers and village extension officers. The purpose of these trials is to ensure availability of new improved well adapted wheat varieties for small scale farmers in Tanzania. Materials and Methods During February and March 2002, thirteen v illages representing different agro-ecological charactaristics of the wheat growing regions were identified and selected in collaboration with District Agriculture and Livestock Devel opment Officers (DALDOs), Village Extension Officers (VEOs), Village leaders and farmers in respective districts and trial sites in Iringa, Morogoro, Arusha and Kilimanjaro regions. Th ese sites represented medium to high altitude with marginal, medium to high rainfall agro-eco logical zones. Four villages namely Itungi, Dabaga, Kinenulo and Ulembwe were selected in Iringa, two villages (Lumbiji and Vidunda)

PAGE 134

in Morogoro, four villages (Musa, Kilimatembo, Kainam and Tloma) in Arusha and three villages (Ngare,Olmolog and Moshas). Sites in Arusha and Kilimanjaro regions were selected in February while site selection in Iringa and Morogoro was done in March. Eight old and twenty two new improved wheat cultivars were seeded on-farm at 13 villages in a randomized complete block design with two replications. Plots were 2.5m long with 10 rows each and 0.25m between rows Seeding was done by hand. Weeds were controlled by applying post emergence herbicide 2 4 D Amine or Buctril Mc at five leaf stage at a rate of 1.25 l ha-1, also hand weeding was done to rem ove late emerging broad leaf weeds and grass weeds respectively. Data was collected on reaction to diseases, grain yield and farmers variety preference. The wheat cultivars and their pedigree are presente d in Table1. Entries1-8 are old cultivars; entries 9-30 are new cultivars. Table1. Name and cross of whea t cultivars evaluated on-farm No. Entry Pedigree or Cross 1 Selian 87 BN-YR 70/T.aestivum x KAL-BB 2 Mbayuwayu KVZ-K4500 LA4 3 Kware BB-GALLO xCj71/T.aestivum x KAL-BB 4 Juhudi HAHN S Cm33682 5 Njombe 7 CMH 78.409-37Y-7B-1Y-1 PTZ-OY 6 Azimio PAVON 76 (CM x CNO 67-7C / KAL-BB) 7 Tausi VEERY S-8 8 Mbuni TROPHY x K6106-1 9 Chiriku TSI / VEE S 10 W9811 TIA.2/4/CS/TH.CU//GLEN/3/ALD/PVN 11 W9905 BJY/COC//PRL/BOW 12 W9920 GIM/LIRA//TURACO/CHIL/3/IRENA 13 W9921 TJB368.251/BUC//TURACO 14 RV10 SW99-5124*2 /FASAN 15 RV212 NING MAI 96019 16 RV227 BATAN F96 17 RV292 BR23/PF869107 18 RV427 R37/GHL121//KAL/BB/3/JUP/MUS/4/2*YM1#6/5/CBRD 19 RV428 R37/GHL121//KAL/BB/3/JUP/MUS/4/2*YM1#6/5/CBRD 20 RV635 HXL 8088 / DUCULA 21 RV768 SITELLA 22 RV992 PSN /BOW //KAUZ 23 RV1163 LD*6KVZ//LD*6/AGE/3/LD*6/KVZ//LD*6/WTP/4/IAS63/ALDAN/5/RIVADENEIRA7 24 RV1165 LD*6KVZ//LD*6/AGE/3/LD*6/KVZ//LD*6/WTP/4/IAS63/ALDAN/5/RIVADENEIRA7 25 RV1166 LD*6KVZ//LD*6/AGE/3/LD*6/KVZ//LD*6/WTP/4/IAS63/ALDAN/5/RIVADENEIRA7 26 RV1183 TNMU / PF85487 // DUCULA 27 RV1267 CBRD/POS//ESDA/LIRA 28 RV1270 GALVEZ S 87 29 RV1421 ALUBUC / BUC // PRL / VEE # 6 30 RV1434 LAJ3302 /2*BORL95 Results Nearly all trial sites received medium to high rainfall at the beginning of the season. In Arusha region rains started in February, continue d into March and tailed off in mid April. In Kilimanjaro region, rains started in March ending in mid April. Rains in Iringa and Morogoro regions continued in March and stopped in the third week of April. Data obtained from six sites in Arusha and Kilimanjaro sites wi ll be presented and discussed in this paper

PAGE 135

Arusha and Kilimanjaro region sites 1.0 Grain Yield Grain yields from trial sites in Arusha and K ilimanjaro regions are presented in Table 2. Grain yields at Kilimatembo site varied from 0.53t/h a to 2.60t/ha. W9912 gave the highest yields (2.60t/ha) followed by RV227 (2.58t/ha), RV212 (2.47tha-1), W9905 (2.40t/ha), RV1166 (2.33t/ha), Chiriku (2.32t/ha), RV1165 (2.28t/ha) and RV1434 with 2.27t/ha. At this site 13 of the new improved wheat varieties had mean grain yields of over 2 t/ha, while the eight old wheat varieties had mean grain yields of less than 2 t/ha. The low mean grain yields of the old wheat varieties may be due to diseases mainly leaf and stem rust (Table 3). During the field day held on 14.6.02 at Kili matembo site, farmers selected the following entries: W9905, RV212, RV227, RV1165, RV1166, RV10 and RV1434 (Appendix2), which are all new improved wheat cultivars. Mean grai n yields at Kainam site ranged from 0.75t/ha to 2.17t/ha. W9811 yielded 2.17 t/ha, followed by W9912 (2.07t/ha), RV227 (2.03t/ha), RV635 (2.0t/ha), W9905 (1.92t/ha) and RV1434 with a mean grain yield of 1.91t/ha. During the field day at Kainam, on 20.6.02 farmers selected following entries: W9905, W9912, W9811, RV10, RV212, RV427, RV1166 a nd Selian 87. Four of the entries selected by Kilimatembo farmers (W9905, RV10, RV212, RV1166) were also selected by Kainam farmers. At Tloma site yields ranged from 1.55t/ha to 2.70t/ha. Cultivar RV428 had the highest mean grain yield of 2.70 t/ha followed by RV1183 (2.60 t/ha), RV593 (2.58t/ha), Chiriku (2.57 t/ha), Juhudi (2.50 t/ha), RV427 (2.45 t/ha) and RV1234 with 2.43 t/ha. At Musa site, grain yields ranged from 0.53t//ha to 2.97t/ha. Grain yields at Musa village were lower than those at Kilimatembo and Kain am sites. This may be due to drought and severe stem and leaf rust infection at this site. Musa site was the best screening site for resistance to wheat diseases in the 2002 crop season. Grain yields at Ngare site were generally lo w, mainly due to drought. Yields ranged from 0.33t/ha to 1.93t/ha. Cultivar W9905 gave the highest yields (1.93t/ha). The trial at Olmolog received only one ra in, which allowed seeds to germinate. No further rain was received, providing an exce llent opportunity to identify drought tolerant wheat cultivars. Grain yields ranged from 0.38t/ha to 1.10t/ha. Juhudi gave the highest grain yield of 1.10t/ha followed by W9921 (0.87t/ha), RV1166 (0.87t/ha), Chiriku (0.85t/ha), RV10 (0.83t/ha), RV1434( 0.82t/ha) and RV292 with 0.73t/ha. These seven wheat cultivars are either early maturing or drought tolerant. In the combined analysis across six locations mean grain yields ranged from 1.01t/ha to 1.87 t/ha. Cultivars Cultivar Chiriku had the hi ghest mean grain yield (1.87) followed by RV1166 (1.72 t/ha), W9921 (1.70 t/ha), W9905 (1.70 t/ha), and W9912 with 1.67 t/ha (Table 2).

PAGE 136

Table2. Mean grain yield of wheat varieties in Arusha and Kilimanjaro regions in 2002 crop season. Location and Mean grain yield (t/ha) No. Variety KILMBO KAINAM NGARE TLOMA MUSA OLMOL Combined 1 Selian 87 1.83 b-f* 1.16 g-i 0.83 de 1.67 de 1.80 ab 0.40 f 1.11 fg 2 Mbayuwayu 1.58 e-g 1.48 b-g 1.23 a-e 2.10 a-e 2.97 a 0.43 ef 1.60 a-d 3 Kware 1.90 a-g 1.73 a-g 1.23 a-d 2.12 a-e 2.15 ab 0.58 b-f 1.60 a-d 4 Juhudi 1.90 a-f 1.38 d-h 0.72 de 2.50 a-c 2.03 ab 1.10 a 1.66 a-c 5 Njombe 7 1.98 a-f 1.33 d-i 0.68 de 1.73 c-e 1.57bc 0.47 ef 1.39 b-f 6 Azimio 1.93 a-f 1.63 a-g 0.86 c-e 2.00 a-e 1.80ab 0.65 b-f 1.51 a-e 7 Tausi 1.07 gh 0.87 hi 0.63 de 1.87 b-e 1.57 bc 0.55 b-f 1.01 g 8 Mbuni 1.50 fg 1.40 c-h 0.88 c-f 1.82 b-e 1.52 bc 0.50 d-f 1.28 d-g 9 Chiriku 2.32 a-d 1.88 a-e 1.33 a-d 2.57 ab 1.83ab 0.85 ab 1.87 a 10 W9811 2.00 a-f 2.17 a 1.08 b-f 2.07 a-e 1.65 bc 0.45 ef 1.62 a-d 11 W9905 2.40 a-c 1.92 a-d 1.93 a 2.00 a-e 1.55bc 0.68 b-f 1.70 ab 12 W9912 2.60 a 2.07 ab 1.37 a-d 2.13 a-e 1.42bc 0.58 b-f 1.67 a-c 13 W9921 1.92 a-f 1.33 d-i 1.68 ab 2.27 a-e 2.03 ab 0.87 ab 1.70 a-c 14 RV10 2.05 a-f 1.69 a-g 1.50 a-c 2.23 a-e 1.77-c 0.83 a-c 1.62 a-d 15 RV212 2.47 ab 1.23 f-i 1.23 a-d 1.55 e 1.45 bc 0.52 c-f 1.41 b-f 16 RV227 2.58 a 2.03 ab 0.77 c-f 2.03a-e 1.52 c 0.43 ef 1.57 a-e 17 RV292 2.07 a-f 1.37 d-h 0.82 c-e 2.58 ab 1.70bc 0.73 be 1.47 b-f 18 RV427 1.88 a-f 1.68 a-g 0.75 c-e 2.45 a-d 1.70bc 0.40 f 1.40 b-f 19 RV428 1.90 a-f 1.28 e-i 0.90 c-e 2.70 a 1.62bc 0.37 f 1.36 b-g 20 RV635 1.67 d-g 2.00 a-c 0.33 e 1.83 b-e 1.77 a-c 0.37 b-f 1.36 b-g 21 RV768 2.13 a-f 1.77 a-f 0.38 e 2.32 a-e 1.70 bc 0.57 b-f 1.51 a-e 22 RV992 2.03 a-f 1.70 a-g 0.77 c-e 2.13 a-e 1.45 bc 057 c-f 1.53 a-e 23 RV1163 1.65 d-g 1.30 e-i 0.93 b-e 2.32 a-e 1.35 bc 0.52 b-f 1.34 b-g 24 RV1165 2.28 a-e 1.63 a-g 1.07 b-e 2.08 a-e 1.97 ab 0.62 b-f 1.62 a-d 25 RV1166 2.33 a-d 1.62 a-g 1.22 a-d 2.20 a-e 123 bc 0.87 ab 1.72 ab 26 RV1183 1.87 a-f 0.75 i 0.83 c-e 2.60 ab 1.47bc 0.58 b-f 1.32 c-g 27 RV1267 2.18 a-f 1.35 d-h 0.83 c-e 2.28 a-e 1.07 bc 0.60 b-f 1.43 b-f 28 RV1270 0.53 h 1.41 c-h 0.43 e 2.38 a-d 1.10*bc 0.38 f 1.09 fg 29 RV1421 1.68 c-g 1.15 g-i 0.43 e 2.43 a-d 1.03 bc 0.50 d-f 1.20 e-g 30 RV1434 2.27 a-e 1.91 a-d 0.72 de 2.25 a-e 0.53 c 0.82 a-d 1.52 a-e MEAN 1.95 1.54 0.95 2.17 1.61 0.59 1.47 LSD.05 0.73 0.59 0.76 0.80 0.12 0.32 0.38 CV% 18.2 18.9 39.3 17.8 37.7 26.9 22.64 Means within a column followed by the sa me letter(s) are not significantly different according to DMRT. Disease Reaction Wheat diseases were recorded at Musa, Tloma and Kilimatembo sites. .These sites allowed good screening for stem, leaf and yellow rust Powdery mildew occurred at Ngare and Olmolog sites. The disease reaction of the old a nd new wheat varieties is presented in Table 3. Almost all of the old wheat cultivars are susceptible to stem and leaf rust, while several of the new wheat cultivars have either trace to low le vels of infection to stem and leaf rust.

PAGE 137

Table 3. Disease reaction of old and new wheat varieties in 2002 crop season. Diseases No. Variety Plant height ( cm) Stem rust (SR) Leaf rust (LR) Yellow rut (YR) 1 Selian 87 84 80S TR D 2 Mbayuwayu 90 R 50S E 3 Kware 91 10MR 10MR L 4 Juhudi 73 70S 80S E 5 Njombe 7 96 30MR 40MS T 6 Azimio 101 40MR 20MR E 7 Tausi 86 70S 30S 8 Mbuni 87 50MS 40MS 9 Chiriku 72 40MR 40MR T 10 W9811 95 80S 80MS H 11 W9905 86 80S 50MR I 12 W9912 86 60MR 30MR S 13 W9921 70 10R 5R 14 RV10 69 10MR 30MR C 15 RV212 84 70S 60MS O 16 RV227 97 40MR 30MR L 17 RV292 119 40MR 20MR U 18 RV427 84 80MS 30MS M 19 RV428 92 TR 10MR N 20 RV635 97 R 30MR 21 RV768 92 30MR 10MR 22 RV992 100 R TR 23 RV1163 90 R 5MR 24 RV1165 86 R 5MR 25 RV1166 89 R 5MR 26 RV1183 79 R R 27 RV1267 110 20MR TR 28 RV1270 89 5R 5MR 29 RV1421 102 TR R 30 RV1434 85 5R R Key: R = Resistant MR = Moderately Resistant Tr = Trace MS = Moderately Susceptible S = Susceptible 3.0 Field Days and Farmers Assessment of Wheat Varieties 3.1 Field days As part of new wheat variety promotion and farmers wheat variety assessment, field days were planned at all sites so that participa ting farmers and farmers from surrounding villages could contribute in identifying well adapted wh eat varieties in their zone. Due to severe drought and low temperatures at some trial sites the wheat varieties in these sites could not depict the intended messages easily. Field days were held at Kainam, Kilimate mbo, and Musa., and were attended by 60 to 150 farmers. Farmers using their own criteria selected wheat varieties presented in Table 4. Several cultivars were selected at more than one location. The differences in variety choice at some sites may be due to differen ces in agro ecological zones.

PAGE 138

Table 4. List of varieties selected by farmers and their pedigree/parentage No. Variety Pedigree or Cross 1 RV10 SW99-5124*2 /FASAN 2 W9905 BJY/COC//PRL/BOW 3 RV212 NING MAI 96019 4 RV227 BATAN F96 5 RV427 R37/GHL121//KAL/BB/3/JUP/MUS/4/2*YM1#6/5/CBRD 6 RV1165 LD*6KVZ//LD*6/AGE/3/LD*6/KVZ//L D*6/WTP/4/IAS63/ALDAN/5/RIVADENEIRA7 7 RV1166 LD*6KVZ//LD*6/AGE/3/LD*6/KVZ//L D*6/WTP/4/IAS63/ALDAN/5/RIVADENEIRA7 8 W9920 GIM/LIRA//TURACO/CHIL/3/IRENA 9 W9811 TIA.2/4/CS/TH.CU//GLEN/3/ALD/PVN 10 RV1434 LAJ3302 /2*BORL95 11 RV992 PSN /BOW //KAUZ 12 RV1263 MILAN/DUCULA//ATTILA 13 W9921 TJB368.251/BUC//TURACO 14 RV1395 CHIL/ESDA/3/HEI/3*CNO79//2*SERI 15 Chiriku TSI / VEE S 16 Selian BN-YR 70/T.aestivum x KAL-BB 3.2 Farmers assessment On June 14th, 2002, 107 farmers at Kilimatembo site assessed the wheat cultivars. Staff from Socio-economics department and extension st aff from Karatu facilitated this evaluation. Fifteen farmers (5 females and 10 males) were chosen to assess the wheat varieties. The 15 farmers had either participated in on-farm wh eat trials or had been observing the performance of the wheat varieties in the trials. Farmers were first asked to examine all the wheat varieties in the trial. After thorough examination of the varieties they were asked to rank them in order of preference (Table 5) Table 5. Absolute ranking of selected wheat varieti es at Kilimatembo village in 2002 crop season. Wheat varieties Rank RV 212 2 RV 227 4 W 9905 6 RV 1165 3 RV 1166 1 RV 1434 7 RV 10 5 Selian 87 8 RV 1166 was the most preferred cultivar, follo wed by RV212 RV 1165 and RV227. Criteria applied by farmers to select cultivars were good to excellent resistance to diseases, many spikes, early maturity, strong stem and plumb big grains. Selian 87, an old wheat variety was the least preferred because of its susceptib ility to diseases as indicated in Appendix 1. Farmers were asked to mention and rank criteria used to select good wheat cultivars as indicated in Table 6.

PAGE 139

Table 6. Farmers criteria for selecting good wheat varieties at Kilimatembo in 2002 crop season. Criteria Rank Resistance to diseases 1 Many spikes 3 Strong stem 4 Carried over seed 2 Drought tolerance 5 Grain Size 7 Baking quality 6 Early maturity 8 A matrix ranking (Table7) was done using all the eight (8) important criteria and the eight selected wheat varieties as shown in Table 8. The eight selected wheat varieties were ranked excellent to average in most criteria. Va rieties RV1166, RV 1165, RV212 and RV227 ranked good to excellent in disease resistance. Vari eties RV 10, RV1166 and RV1165 were ranked excellent in earliness. Table 7. Matrix ranking of wheat varietie s at Kilimatembo site in 2002 crop season. Criteria Varieties RV 10 RV1434 RV1166 RV1165 RV212 RV227 W9905 Selian Total Rank Disease resistance 2 3 4 4 5 5 4 2 24 8 Spike number 2 3 3 4 5 5 5 5 32 4 Baking Quality 3 3 4 4 3 4 4 3 28 7 Grain Size 5 4 4 4 5 5 4 4 35 1 Drought tolerance 5 4 5 5 5 4 3 2 33 3 Carried over seeds 5 3 5 3 5 4 4 3 32 4 Strong stems 5 5 5 5 4 3 4 3 34 2 Early maturity 5 3 5 5 3 3 3 2 29 6 Total 32 28 35 34 35 33 31 23 Rank 5 7 1 3 2 4 6 8 Key: 5 = Excellent, 4 = Good, 3 = Average, 2 = Satisfactory, and 1 = Poor Pair wise ranking of the selected eight wheat cultivars was done (Table 8). During pair wise comparison the variety RV212, RV227 and Selian were the most preferred. Table 8. Pair wise ranking of farmer selected wheat varieties RV10 RV1434 RV1166 RV1165 RV212 RV227 W9905 Selian Total Rank RV10 RV1434 RV1166 RV1165 RV212 RV227 RV1434 Selian 0 8 RV1434 RV1434 RV1434 RV212 RV227 W9905 Selian 4 4 RV1166 RV1166 RV212 RV227 W9905 Selian 2 6 RV1165 RV212 RV227 RV212 Selian 1 7 RV212 RV212 RV212 RV212 7 1 RV227 RV227 RV227 6 2 W9905 Selian 3 5 Selian 5 3

PAGE 140

4.0 Discussion The performance of old and new wheat cultivars in 6 on-farm trials in different agroecological zones in Arusha and Kilimanjaro regi on (Table 2) 2 show significant differences in grain yield.. Severe drought at some sites cau sed low grain yields, but enabled selection for drought tolerance. Cultivars with highest yiel d under severe drought conditions were Chiriku, Juhudi, W9921, RV10, RV292, RV1166 and RV14 34. These cultivars were also the earliest. At sites without drought, mean grain yields were high for most cultivars and differences in grain yield among cultivars were significant. At locations with high disease pressure (Tloma and Musa), grain yield of susceptible vari eties were reduced. Most old wheat varieties are susceptible to stem and leaf rust. Sowing the trials at different but well selected agroecological zones allowed identifying cultivars with disease resistance, tolerance to drought and high yield potential. Farmers and extensi on staff in these regions participated in the evaluation and selection of well adapted disease resi stant wheat cultivars. At sites where field days were held, farmers selected the new im proved wheat cultivars, while the old cultivars were mostly discarded due to their susceptibility to diseases. The new improved wheat cultivars have high levels of disease resistance to stem and leaf rust and have high yield potential. These cultivar s will further evaluated in on-farm trials in Iringa, Morogoro, Arusha and Kilim anjaro regions to verify th eir agronomic superiority and disease resistance over the old cultivars, before fina lly being released for small scale farmers. Acknowledgement -The financ ial and material support from ASPS-Seed Unit is highly acknowledged, without which these activities would have been impossible. References Allard, R.W., and A.D. Bradshaw. 1964. Implications of genotype-environment interaction. Crop Sci. 4: 503-507 Chirot, Y. and Hailu, B. 1990. On farm evaluation of three bread wheat vari eties in the Wolmera red soil zone. In Tanner, D.G., M. van Ginkel, and W.Mwangi, eds. 1990. Sixth Regional Wheat Workshop for Eastern, Central and Southern Africa. Nakuru ,Kenya: CIMMYT Sariah, M. A., R.V.Ndondi, and M. J. Mollel. 1990. Grain yield potential and adaptation of ten bread wheat varieties in Tanzania. In Tanner, D.G., M. van Ginkel, and W.Mwangi, eds. 1990. Sixth Regional Wheat Workshop for Eastern, Central and Southern Africa. Nakuru Kenya: CIMMYT Ndondi, R.V., H. Mansoor, C.A. Kuwite, M. Mugendi and R. Shekibula. 2001. On farm evaluation of five pre-released wheat varieties at Karatu. In Wheat Research Prog ramme Annual Progress Report 2000/ 01.

PAGE 141

Appendix 1. Farmers observations in the field on the characteristics of the selected wheat varieties were: 1. RV 1166 Good disease resistance Many spikes Early maturity Drought resistance Heavy seeds Strong stems Good baking quality 2. RV 212 Good disease resistance Many spikes Medium maturity Drought resistance Heavy seeds Strong stems Good baking quality 3. RV 1165 Good disease resistance Many spikes Medium maturity Drought resistance Heavy seeds Strong stems Good baking quality 4. RV 227 Moderate disease resistance Many spikes Medium maturity Drought resistance Heavy seeds Strong stems Good baking quality 5. RV 10 Moderate disease resistance Many spikes Early maturity Drought resistant Heavy seeds Strong stems Good baking quality 6. W 9905 Moderate disease resistance Many spikes Early maturity Drought tolerant Heavy seeds Strong stems Good baking quality 7. RV 1434 Moderate disease resistance Many spikes Early maturity Drought resistance Heavy seeds Strong stems 8. Selian 87 Moderate disease resistance Many long spikes Late maturity Poor drought tolerance Heavy seeds Strong stems.

PAGE 142

Evaluation of Improved Wheat Varieties Under Different Management Practices in Eastern Wallagga Highlands Tolera Abera1, Daba Feyisa1, Girma W. Tsadik1, Hasan Yusuf1 and Gemechu Keneni2 1Oromiya Agricultural Research Institute Bako Agricultural Research Center P.O. Box 03, Bako, West Oromiya, Ethiopia, Email: akthirpha@yahoo.com 2Holetta Agricultural Research Center, P. O. Box 6282, Holetta, Ethiopia Abstract A trial was conducted during the 1997-1999 cropping seasons at two locations. Four improved wheat ( Triticum aestivum ) varieties and one local variety were planted in a factorial design with two (research and farmers') management practices. Yield of the wheat varieties wa s significantly different across the two locations. Improved varieties performed better under research management practices. Higher grain yield was observed in the im proved management practices compared to farmer practice. Improved management prac tices gave a higher marginal rate of return (57.51%) whereas the local variety ga ve higher mean grain yield under farmer management practices. Improved management practices resulted in a yield advantage of 7.25%. The use of improved varieties with improved management practices is essential if wheat grain yield is to be maximized. Maximum grain yield and higher net return were realized from improved agronomic packages fo r wheat production. The variety HAR-1685 gave better yield with research management practices and is recommended for production. More importantly, the use of improved management techniques targeted at increasing yield a llowed wheat producers to greatly increase economic returns. Thus, extension agents should consider this recommendation for wheat production in Eastern Wallagga highlands. Introduction The introduction of new crop varieties on soils with improved management packages has increased food production in some smallholder farms. The trend, however, has been the opposite in most smallholder fa rms due to poor soil fertility and crop management (Woomer and Ingram, 1990). The rising cost of commercia l fertilizers has limited smallholder use. According to Buresh and Giller, (1998), in the smallholder farming systems of Africa nutrient outputs exceed nutrient inputs. Management options to increase yield and ameliorate productivity of crops are urgently needed for wheat production. Wheat ( Triticum aestivum ) grain yield potential has significantly increased due to the release of improved high yielding varieties (Amsal et al ., 1995). Generally high yielding crop varieties require more intensive agronomic management practices to express their potential yield. Amsal et al. (1999) and Tanner et al., (1993) indicated that recently released cultivars of wheat are highly responsive to improved ma nagement practices and require high rates of nutrient application. They further stated that the productivity of improved varieties of wheat is more sensitive to management practices compared to local varieties The adoption of high yielding improved varieties without improved management practices may not boost the productivity of wheat. To date, however, wh eat production with improved management practices is limited due to the cost and in accessibility of fertilizers and the overlapping farming activities to smallholder farmers. In a ddition, the low current market price of wheat makes wheat production unprofitable. Furtherm ore, improved varieties have failed to adequately meet the needs and requirement of marginal environments(Hardon ,1996). Plant

PAGE 143

breeding programs mainly direct their effort s at increasing yield in more favorable environments. Looking for broad adaptability to different environments with alternative management practices for sustainable production of wheat is urgently needed. Giving value to wheat varieties because of their adaptati on to alternative management practices will encourage sustainable production. Therefore, the objectives of this study were to estimate the extent of cultivars x management level inte raction, identify cultivars that suit alternate management practices and identify stable cultiv ars that can better perform under both farmer and improved management practices Materials and Methods The experiment was conducted during the 1997, 1998 and 1999 croppi ng seasons on farmers fields in the eastern Wallagga highlands (Shambo and Arjo). Shambo lies between 9o34'N latitude and 37o06'E longitude at an altitude of 2 400 meter above sea level with a mean annual rainfall of 1,695 mm (NMSA, 2003). It has a cool humid climate with mean minimum, mean maximum, and average air temperatures of 8.15oC, 15.72oC, and 11.94oC, respectively. Arjo lies between 8o45'N latitude and 36o40'E longitude at an a ltitude of 2400 meter above sea level. Mean annual rainfall is 1,330 mm (N MSA, 2003). It has a cool humid climate and the mean minimum, mean maximum, and average air temperatures are 9.33oC, 17.85oC, and 13.59oC, respectively. The experiment was laid out as split plot in a randomized complete block design with variety as main plots and management practices as sub-plots. The improved varieties tested were HAR-710, HAR-1685, HAR -1709, ET-13 and a local variety was also chosen. Management practices were farmers' management and research management. The improved cultural practice was the research recommendation for production of wheat. The local cultural practices were local farmers' cultural practices for wheat production in that area. The seed rate used was 160 kg ha-1 for local and 150 kg ha-1 for improved cultural practices. Sowing dates followed farmers' prac tices, which are between mid June to early July. The plot size used was 4 m x 4 m. The recommended fertilizer rates of 100 kg ha-1 DAP and 100 kg ha-1 Urea was applied at planting. For th e improved cultural practices, two times hand weeding was practiced as per recommendati on for wheat production. For the farmer's cultural practices, hand weeding was done once 25 days after planting. The data were analyzed using Mstatc Computer Softwa re. Mean separation was done using least significance difference (LSD) at 5 % probability level. For economic evaluation, partial budget, values to cost ratio (VCR) and marginal analyses were used. To estimate economic parame ters, wheat grain yield was valued at an average open market price of EB 143.00 100 kg-1 for the last five years. The yield was adjusted down by 10 % to reflect actual production environments (CIMMYT, 1988). The seed cost of wheat was EB 3.75 kg-1 for the improved varieties and EB 1.43 kg-1 for the local variety. Urea and DAP were valued at the official prices of EB 269.65 and 303.35 100 kg-1, respectively. The cost of labour for weeding was EB 3.50 day-1.

PAGE 144

Results and Discussion Cropping season significantly (P < 0.05) affected me an plant height and grain yield (Table 1). This might be likely due to the distribution pa ttern of rainfall. Environmental factors had a positive impact on the productivity of wheat In addition to management practices, environmental factors also greatly influence the sustainability of wheat production. Locations and year x location interaction had significant (P <0.05) effects on mean plant height and grain yield (Table 1). This indicates that the varieti es performed differently at the two locations and across cropping seasons. Varieties were significantly (P<0.05) differe nt for mean plant height and grain yield (Tables 1,2 and 3). Year x variety, location x va riety and year x location x variety interaction were significant for plant height and grain yi eld (Table 1), indicating that different wheat varieties performed differently within and among locations. Clearly, different varieties have different yield potential at each location. Yield of wheat can therefore be improved by providing different varieties to wheat producers in a given area. Management practices significantly (P<0.05) a ffected mean plant height and grain yield (Tables 1, 2 and 3). This indicated that manage ment practice influenced the yield potential of different wheat varieties. Higher mean plant height and grain yield of wheat was obtained from improved management practices. Clearly, higher grain yield of wheat with improved agronomic practices is possible for the area. All varieties showed a significant response to management practices at both locations. Improved wheat varieties showed greater response to improved management levels across the two fields. This result agrees with Amsal et al. (1999) who reported that improved cultivar s are highly responsive to improved crop management systems. Improved agronomic practices gave yields 7.25% higher than the farmers' local practices (Table 3). Farmers using improved agronomic packages can maximize the grain yield of wheat and their profit. Interaction effects of varieties and management practices were significant (P<0.05) for mean grain yield but non-significant for plant height (Tables 1, 2 and 3). Different varieties had different responses to management practices. Greater grain yield of wheat was achieved from improved management practices for all varieties. These findings are similar to those of Katyal (1999) who found that improved practices, improved cultivars, and NPK application gave the highest yield, returns, and profitabilit y. Similarly, Bhagat and Singh (1998) reported that improved agronomic practices gave a 47 % increase in rice-equivalent yield compared with local farmers practices (local cultivars and 30 kg N ha-1). The improved wheat variety HAR-1685 gave higher mean grain yield compared to all other varieties. This variety (HAR1685) had wi de adaptability compared to other varieties. Our results show that improved agronomic practices will significantly improve wheat production and the sustainability of pro duction in this region. The technologies for sustainable crop management practices relevant to small wheat farmers outlined by Sayre (1999) can become a reality. Providing improve d wheat varieties with improved management packages to farmers is es sential practices for maximum production and profit. Economic analysis for management practices indicated that the highest net benefit of EB 2391 ha-1 with a marginal rate of return of 66.22 % and a value to cost ratio of EB 3.02 profit per unit of investment for wheat was obtained from improved management practices (Table 4). The net benefit for farmers' cultural practices was EB 2305 ha-1 with a value to cost ratio of EB 3.49 profit per unit investment for wheat production. The value to cost ratio with both management practices includes the price of pro duction and provides an estimate of profit for wheat producers in the area. Using the average current price of inputs, both management technologies were economically viable.

PAGE 145

Table 1. Mean square of straw and grain yield of wheat due to variety and management levels across years and location at Shambo and Arjo. Mean square Source of variation DF Plant height (cm) Grain yield (kg/ha) Year 2 2590.557** 23111592.129** Location 1 1988.929** 48860448.004** Year x Location 2 1724.407** 10284810.779** Variety 4 14691.787** 7477252.869** Year x variety 8 150.72** 2442707.478** Location x variety 4 565.227** 560040.515** Year x Location x variety 8 201.859** 787328.493** Error 72 39.931 150684.932 Management levels 1 651.092** 1664167.604** Year x Management levels 2 219.934** 298755.204 Location x Management levels 1 102.573 124556.337** Year x Location x Management levels 2 583.403** 348527.587 Variety x management levels 4 53.162 473592.135** Year x Variety x Management levels 8 115.474** 67066.532 Location x variety x Management levels 4 14.534 181936.223 Year x Location x Variety x Management levels 8 55.2 194826.426 Error 90 41.923 117243.876 ** Significant at 1% and 5% level of probability respectively Table 2. Combined mean effects of varieties an d management practices on plant height (cm) of wheat at Shambo and Arjo Treatments Varieties Management practices HAR-710 HAR-1685 HAR-1709 ET-13 Local Mean Farmers management 89 81 103 102 127 100 Improved management 91 84 106 109 129 104 Mean 90 82 104 105 128 Varieties Management prac tices Varieties vs. Management practices LSD (5%) 2.57 1.66 Ns CV % 6.20 6.35 Table 3. Combined mean effects of variety and management practices on grain yield (kg/ha) of wheat at Shambo and Arjo Treatments Varieties Management practices HAR-710 HAR-1685 HAR-1709 ET-13 Local Mean Farmers management 2134 3043 2117 1976 2256 2305 Improved management 2293 3145 2262 2462 2197 2472 Mean 2213 3094 2190 2219 2226 Varieties Manage ment practices Varieties vs. Management practices LSD (5%) 58 87.79 196.4 CV % 16.25 14.34

PAGE 146

Table 4. Partial budget and marginal rate of return (MRR) analysis for the effects of management practices on the mean grain yield of wheat comb ined over locations Management practices Items Farmers management Research management Average yield (kg/ha) Wheat 2305 2472 Adjusted yield (kg/ha) Wheat 2074.5 2224.8 Gross field benefit of Wheat 2966.54 3181.46 Seed cost (EB/ha) Weeding cost (EB/ha) 414.40 246.95 388.5 402.15 Total costs that vary (EB/ha) 661.35 790.65 Net benefit 2305.19 2390.81 Values to cost ratio 3.49 3.02 Marginal rate of return (MRR) 66.23% Note: D= dominated treatment, Grain price= EB 1.43 /kg, Seed price = EB 3.75/kg for improved variety, Seed price = EB 1.43/kg for local variety, Yield was do wn adjusted with 10% coefficient Conculsion Improved management practices significantly boost the yield of wheat. Therefore, use of improved management practices is agronomically and economically feasible. These recommendations should be extended to wheat pr oducers in the Shambo and Arjo highlands to ensure sustainability of production and to increase profits. AcknowledgmentThe authors thank Oromiya Ag ricultural Research Institute for funding the project. We are also grateful to Mr. Tesfaye G/Gorges and Yosef kenea for their assistance in carrying out the experiment and e fficiently collecting the data. Bako Research Center Management is also thanked for f acilitating the execution of the field trial. References Amsal Tarekegne, Tanner, D.G. and Getinet Gebeye hu. 1995. Improvement in yield of bread wheat cultivars released in Ethiopia from 1949 to 1987. African crop Science Journal 3: 41 49. Amsal Tarekegne, Tanner, D.G., Taye Tessema and Chanyalew Mandefro. 1999. A study of variety by management interaction in bread wheat varieties released in Ethiopia. In The Tenth Regional Wheat Workshop for Eastern, Central and Southern Africa. pp 196 212. Addis Ababa Ethiopia: CIMMYT Bhagat, R. K. and Singh, R. S. 1998. Crop mana gement in rice-wheat cro pping system. Research Journal of Birsa Agricultural University.10 (1): 30-33. Buresh, R.J. and Giller, K.E. 1998. Strategies to replenish soil fertility in African smallholder agriculture. In Soil fertility Research for Ma ize-based Farming Syst ems in Malawi and Zimbabwe. Proceedings of the Soil Fert Net Result and Planning Wo rkshop. Waddington, S.R., Murwira, H.K., Kumwenda J.D.T., Hikwa, D. and Tagwira, F. (eds.). pp 13-19. Africa University, Mutare, Zimbabwe. Soil Fert Net and CIMMYT-Zimbabwe, Harare, Zimbabwe. CIMMYT. 1998. From Agronomic Data to Farmer Recommendations. An Economics Training Manual. Completely Revised Edition. CIMMYT, Mexico, D.F., Mexico. 79 pp Hardon, J. 1996. The Global Context: Breeding and Crop genetic Diversity. pp 1 3. In : Eyzaguirre, P and Iwanaga, M. (eds.). Participatory Plan t Breeding. Proceedings of a Workshop on Participatory Plant Breeding, 26 29 July 1995, Wageningen, The Netherlands. IPGRI, Rome, Italy. Katyal, V.; Gangwar, K. S. and Gangwar, B. 1999. Influence of input use and management practices on sustainability and economics of rice cultivation. Journal of the Andaman Science Association. 15: 56-59. NMSA (National Meteorological Service Agency). 2003. Meteorological data of Shambo area for 1969-2003. NMSA, Addis Ababa, Ethiopia.

PAGE 147

Sayre, K. D. 1999. Ensuring the use of sustainabl e crop management strategies by small-scale wheat farmers in the 21st century. In The Tenth Region al Wheat Workshop for Eastern, Central and Southern Africa. pp. 119 141. Addis Ababa Ethiopia: CIMMYT Tanner, D. G., Amanuel Gorfu and As efa Taa. 1993. Fertilizer effects on sustainability in the wheat-base smallholder farming systems of southeastern Ethiopia. Field Crops Research 33: 235 248. Woomer, P and Ingram, J.S.I. 1990. The biology an d fertility of tropical soils. TSBF Report. 1992. MARVEL EPZ, Nairobi, Kenya. 44 pp.

PAGE 148

Cell Membrane Stability (CMS) as Screening Technique for Drought Tolerance in Bread and Durum Wheat Genotypes Alemayehu Zemede1, H. Martens2, and M.T. Labuschagne2 1Ethiopian Agricultural Research Organization, Debrezeit Center, P.O.Box 32, Ethiopia 2Department of Plant Sciences, University of Free State, P.O.Box 339, Bloemfontein, South Africa Abstract Genetic improvement of drought resistance in crop plants requires identification of relevant drought resistance mechanisms and the development of a suitable methodology for their measurement in large breeding populations. Bread and durum wheat genotypes/cultivars were used in study to determine their level of injury using cell membrane stability as a measure of drought tolerance. The drought tolerance test is based on the measuremen t of the electro-conductivity of aqueous media containing leaf discs that were previously water stressed in vitro by exposure to a solution of polyethylene glycol 6,000 (PEG). Genotypes differences were highly significant for percent injury level. Wheat leaves of bread and durum genotypes exposed under conditions of moisture stress varied significantly in their membrane injury level. Injury levels among bread wh eat genotypes were lower (6.8 to 21.5%) than among the durums (11.5 to 24.55%) when the genotypes were screened artificially using a 20% PEG solution as a dehydration medium. Although minimum injury levels among genotypes of both types of wheat were recorded, the test would be used for initial screening for drought tolerance along with the other artificial screening methods. Introduction Drought is a multidimensional problem and cove rs large areas throughout the world (William, 1989). Gupta (1997) estimated that about 26% (17, 255,700 square miles) of the worlds total cultivable land is arid or semi-arid, where wate r is the limiting factor to crop production. An estimated 32% of the 99 million hectares of wheat grown in developing countries experiences varying levels drought stress (Rajaram et al. 1996) Consequently, the development of drought to lerant varieties became a major objective in plant breeding programs for crops grown in tropical regions without irrigation. Evaluating drought tolerance based on yield compared to standard cultivars over several years and location where drought is likely to occur is a common procedure in countries, including Ethiopia. This process is dependent on year-toyear changes in weather and extremely time consuming. Plant breeders have long desired simple and rapid measurement techniques for use on early generation material to identify potential drought resistance (Matin, M.A. et al. 1989). The role of cell membrane stability under conditions of moisture stress could be considered a major measurement of electrolyte leakage from the cells. The rate of injury to cell membranes is commonly used as a measure of tolerance to additional plant stresses, such as freezing and heat (Chen et al. 1982). It would therefore be important to determine the drought tolerance of genotypes using simple laboratory techniques like cell membrane stability (CMS) test to supplement the data obt ained through field screening so that breeders would have additional information for selec ting genotypes for to drought tolerance.

PAGE 149

This study attempted to differentiate va rietal differences of durum and bread wheat genotypes of Ethiopian and South African origin in response to moisture stress using the cell membrane stability (CMS) test. Materials and Methods Ten bread wheat and 10 durum wheat genotypes we re used for this study. The bread wheat genotypes/cultivars used in this study are of Ethiopia and South Africa origin, where as the durum wheat cultivars were of Ethiopian origin (Table1). Table 1.Designation and origin of two species of wheat genotypes used for the study. Bread wheat Origin Year of release Durum wheat Origin Year of release BD1-10 SA Cocorit-71 Et 1976 Bdl-41 SA Gerardo Et 1976 BD1-48 SA LD-357 Et 1979 Bdl-24 SA Boohai Et 1982 BDl-20 SA Foka Et 1993 HAR-1685 Et Kilinto Et 1994 HAR-604 Et Tob-66 Et 1996 Dereselign Et Quamy Et 1996 Et-13 Et Tob-2 Et 2002 Cadu-57 Growing conditions The materials used in this study were planted in pots containing 3kg of soil in a glasshouse with three replications. Sampling Leaf samples of about 10mm in diameter were taken from fully expanded young leaves. Five samples were taken from two or three leaves per genotype. Samples were kept in an airtight test tube, wetted a drop of water, and tran sferred to the laboratory within an hour. Drought tolerance test The method of Sullivan (1972) was followed to test the drought tolerance. Samples were washed with three changes of distilled water to remove surface-adhered electrolytes. Five leaves samples for the stress treatment were placed in test tubes with a 10cc solution of a 20% concentration of Polyethylene Glycol (600 PGC) Five samples for the control treatment were placed in 10cc of distilled water. All the samp les were incubated at 10C for 24 hours and then equilibrated in a water bath at 25C. Conductivity of the incubation medium was read using a conductivity meter. After reading, the samples were autoclaved for 15min to kill the leaf tissues and a second conductivity reading w as made after the samples reached room temperature of 25oc. Calculation of percentage injury due to desiccation was made as follows: % injury=1-[1-(T1/T2)/1-C1/C2)]*100 Where T and C refer to mean of the treatment and control reading, respectively, and the subscripts1 and 2 refer to initial a nd final conductivities, respectively.

PAGE 150

Statistical analysis The data were subjected to a variance analysis (ANOVA) using SAS computer software (SAS Institute 1996). Results and Discussion Analyses of variance for percent injury of the wheat genotypes are presented in Table 2. Genotypes differences were highly significant (P<0.01) for percent injury. A significant difference (P<0.05) between groups of bread and durum wheat genotypes was also found. Analysis of variance data between groups of wheat genotypes of different origins indicated significant difference (P<0.05) in percent injury level. Mean values of percent injury for the 20 wheat genotypes are given in Table 3. The highest injury level was obtained from the durum wheat genotype Cocorit-71, while the lowest was from Israel (6.8%), significantly different from all the other genotypes. Although the contrast comparisons between bread and durum wheat were significant, there were inconsistencies between genotypes of the two species. The durum wheat cultivars had relatively high levels of injury compared to the bread wheat genotypes/ cultivars. For instance, the three highest percent injury levels, as well as and the second, fourth, and fifth low percent injury level were obtained from the durum wheat cultivars (Table 3). Similarly, based on the LSD values, signifi cant differences were observed between South African and Ethiopian bread wheat genotypes. The percent injury levels of bread wheat genotypes of Ethiopian origin ranged from 19.8 to 6.8, with a mean of 14.5. The South African bread wheat genotypes varied from 21.48 to 13.28, with a mean value of 16.8 (Table 4). It is highly suspected that this difference could be due to environmental variation between the countries of origin. Table 2. Analysis of variance and contrast comparison between two species groups and three origin groups of wheat genotypes/cultivars Source of Variation df Mean squares F-value Genotypes 19 86.591868 4.67** Error 58 Species 1 131.07200 6.86** Origin 2 91.642125 4.94* ** Significant at the 0.01 level Significant at the 0.05 level

PAGE 151

Table 3.Percent injury by dehydration for all bread and durum wheat genotypes/cultivarsGenotyp es that exhibited a low percent of injury could be considered as drought tolerant (Blum and Ebercon, 1981: Mark et al, 1991) According to this study, genotypes with a minimum percent injury by dehydration are considered more tolerant. Hence, among bread wheat genotypes, Israel, Et-13, Bdl-24, and Bdl-41 were tolerant, whereas Bdl-10, Dereselign, and Bdl-48 all had highest level of injury and could be classified as non-tolerant genotypes. On the other hand, among the durum wheat cultivars Tob-66, Boohai, and Fetan showed low level of injury after dehydration and coul d be considered tolerant. Foka and Cocorit-71 had high levels of injury; therefore they are nontolerant genotypes. In contrary to this theory, the performance of Cocorit-71 (early maturing t ype) was promising, thus making it a drought tolerant cultivar, which has also been seen un der field experiments carried out in Ethiopia when yield and its components were used as a selection criteria for drought tolerance. This could due to major drought tolerance m echanisms, probably drought avoidance. In a similar study, Blum and Ebercon (1981) explained this adaptation as osmotic adjustment when the cell membrane adjusts to drought stress. The degree of injury to cell membrane stability by controlled dehydration was found to decrease in plants that were subjected to a period of drought stress. The percentage cell membrane injury level obt ained in this study was very low less than 30% compared to 70% injury level reported by Blum and Ebercon (1981). The reason is probably that the PGE concentration used in this study was low, 20 % to induce osmotic stress, as opposed to the 40% PGE concentration used in their studies. In general, based on this study, the perfo rmance of durum wheat genotypes in terms of drought tolerance seemed to be progressive and showed some level of improvement over time. For instance, the old cultivars, Cocorit 71 and Gerardo showed rela tively higher level of Table 4.Ranges and mean va lues of percent injury for groups of genotypes Groups Species Ranges Mean Bread Wheat 21.486.83 15.69 Durum Wheat 24.55-11.5 18.25 Origin SABW 21.48-13.28 16.8 ETBW 19.78-6.83 14.5 ETDW 24.55-11.5 Bread Wheat Genotypes/cultivars Percent injury Durum Wheat Genotypes/cultivars Percent injury BD1-10 21.48 Foka 23.98 Bdl-20 16.83 Kilinto 20.55 BD1-24 13.28 Tob-66 11.50 Bdl-48 18.48 LD-357 17.65 BDl-41 14.10 Cadu-17 18.05 Israel 6.83 Cocorit-71 24.55 Dereselign 19.78 Gerardo 22.18 HAR-1685 16.20 Quamy 19.58 HAR-604 17.93 Fetan 12.38 Et-13 12.0 Boohai 12.08 Mean 15.69 Mean 18.25 LSD(0.05) 6.096

PAGE 152

injury than the recently released cultivars Tob-66 and Fetan, despite a few inconsistancies. These could be due to the environmental cond itions under which they were evaluated and developed, which changed over time. Bread wheat genotype Bd-l0 demonstrated a high level of injury has also been categorized as a non-tolerant line in a screeni ng trial carried out to differentiate genotypes based on yield and yield component for drought tolerance (Alemayehu, 2001) Bdl-24 showed a low level of injury and was considereda tolera nt line. This indicates that osmotic adjustment could be correlated with drought tolerance. In conclusion, although the injury levels among genotypes was minimum, it was possible to differentiate the genotypes reaction to drou ght tolerance using this test. Along with other artifical screening methods, this test coul d be useful for initial screening for drought tolerance. Using many simple screening techniques to generate various types of data is very important when classifying drought tolerant a nd susceptible genotypes. In addition, these tests should be correlated with the field results. References Alemayehu Zemede, 2001. The Characterization of Ethiopian and South African bread and durum wheat genotypes for Drought Tolerance. M.SC. Thesis, Department of Plant breeding, Faculty of Natural Sciences and Agriculture Univers ity of Free State Bloemfontein, South Africa. Beweley, J.D., 1979 Physiological aspects of desi ccation tolerant, Annu. Rev.Plant Physiology 30:195238 Blum, A and A.Ebercon, 1981. Cell Membrane Stability as a measure of drought and hear tolerance. Crop Science 21:43-47 Chen.H.H., Zheng-Yan Shen, and P.H.Li., 1982 Adapta bility of crop plants to high temperature stress. Crop science 22:719-725. Gupta, U.S. 1997. Crop improvement : Stress tole rance volume 2 Science Publishers, Inc., New York, London. Mark,R.,A.A. Kenneth and H.D. Stanley, 1991. Leakag e of Intracellular substances as an indicator of freezing injury in Alfalfa. Crop Science 31:430-435 Matin, M.A., H.B., Jarvis and F.Hayden, 1989. L eaf water potential, Relative Water content, and Diffusive Resistant as a screening Techniques for drought Tolerance in Barley. Agronomy Journal 81:100-105. SAS Institute 1996: SAS for analysis of variance. SAS Institute, Cary. Sullivan, C.Y. 1972. Mechanisms of heat and drought resistance in sorghum and methods measurement. In N.G.P Rao and L.R House (ed.), Sorghum in the seventies. Oxford and IBH Publishing CO., New Delhi, India, pp. 65-69. Rajaram, S.H.J, Braun, and M. Van Ginkel, 1996. CIMMYT's approach to breeding for drought tolerance. Euphytica 92:142-153. William J.R. 1989. The dimensions of drought : Drought Resistance in cereals. In Baker, F.W.C. (ed.) CAB International pp.1-13.

PAGE 153

Physiological Races and Virulence Diversity of Puccinia graminis f. sp. tritici on Wheat in Ethiopia Belayneh Admassu1, Emebet Fekadu and Zerihun Kassaye Ethiopian Agricultural Research Organization, Plant Protection Research Center, P. O. Box 37, Ambo, Ethiopia; E-mail: belay120@yahoo.com Abstract Physiologic races of Puccinia graminis f. sp. tritici (Pgt) isolates collected in Ethiopia were determined on seedlings of wheat stem rust differential cultivars following the international system of nomenclature for Puccinia graminis 39 different races were iden tified from 49 isolates studi ed. The 16 rust isolates collected in 2001 belonged to 16 different race groups, whereas the 33 isolates collected in 2002 be longed to 23 different race gro ups. Most of the single pustule isolates selected from a single population showed variability, whereas some of them belonged to the same race group. The co mposition of physiologic races differed greatly between 2001 and 2002. Only races TTR and JGH were identified in both cropping seasons. Races such as TTT, TTR, TPT and RTT have wider spectra of virulence. Races such as TTR, TRR, TPR and DBL have wider spatial distribution whereas the rest are confined to certain localities. Generally, Pgt populations in Ethiopia appear to be highly variable, and this would be an important consideration for the breeding programme in the country. Introduction Wheat ( Triticum aestivum L.) is the fourth largest food crop in Ethiopia, covering more than one million hectares of land and making up abou t 13% of the total crop production with an average yield of 1.4 ton per hectare (CSA, 2002). Demand of wheat has steadily increased in the last decades in Ethiopia. Though over 30 funga l diseases of wheat have been identified in Ethiopia, stem rust (caused by Puccinia graminis f. sp. tritici ) is a major production constraint in most wheat growing areas of the country, causing yield losses of up to 100% in epidemic outbreaks. One such outbreak was reported in 1993/94 in the Southwestern part of Ethiopia in the Arsi and Bale regions, which are majo r wheat producing areas of the country. The previously resistant bread wheat cv. Enkoy became highly susceptible during the epidemic (Ayele, 2002). Outbreaks of stem rust were al so reported in 2003 in the same part of the country, the cause of which is as yet unknown. Cultivation of resistant high yielding varieties is the most economical method of controlling rusts; however, due to sudden change in the rust race-pattern, commercial varieties often become vulnerable to rust attack. Breedi ng for vertical resistance, i.e. resistance to certain races of a pathogen has been main part of cereal breeding since earlier times (Hoerner, 1919; Stakman and Levine, 1922). Breeding varie ties with a specific r esistance has demanded systematic studies of race composition of path ogens. Thus, to study the prevailing and virulent races and the dynamics of race co mposition of the stem rust pathogen has a significant role in attaining sustainable disease c ontrol. This study was conducted to study the variations in the pathogen and to determine the prevalence and distribution of physiologic races of P. graminis f. sp. tritici in the major wheat growing regions of Ethiopia.

PAGE 154

Materials and Methods Field surveys were conducted in North Shewa, Ar si and Bale regions of Ethiopia in 2001and 2002 cropping seasons to collect samples of wheat stem rust. In both years the surveys were done during mid October when wheat plants were at flag leaf stage. The surveys followed main and feeder roads on preselected routes where wheat is important and stem rust is known to be present. Samples were randomly collected from commercial fields every 10 km or at the first field thereafter. In addition, biased sampling was done from strategically located Ethiopian Wheat Rust Trap Nursery sites thro ughout the regions. Infected leaf or stem samples were collected from breeding materials or commercial cultivars grown in the vicinity of those locations exhibiting infections. Th ree samples were collected per field including geographic information, and kept in paper bags. A total of 49 stem rust isolates (16 in 2001 and 33 in 2002) were collected from Shewa, Arsi and Bale regions of Ethiopia. Seven da y old seedlings of Morocco (highly susceptible genotype to stem rust) were inoculated with bulked spore populations collected from each field. Two to three single pustules were is olated from each sample, and were increased on morocco in separate pots in a greenhouse adjusted at 25 2oC to produce enough inoculum for the race study. Spores from each single pustule isolate was collected in separate test tubes and stored at 4oC until they were inoculated on the standard differential sets of P. graminis f. sp. tritici A suspension of the spores (prepared by mixing urediospores with lightweight mineral oil) was inoculated onto seven day old seedli ngs of the standard differential cultivars. Immediately after inoculation, the seedlings were placed in a humid chamber in the dark for 24 hrs at 19-21oC. After 24 hours, they were transferred to a greenhouse where the temperatures varied between 20 to 26oC. In addition to the 12 sta ndard differential hosts (that carried genes Sr5, Sr6, Sr7b, Sr8a, Sr9b, Sr 9e, Sr9g, Sr11, Sr17, Sr21, Sr30 and Sr36 ) used in determining Puccinia graminis f. sp. tritici races, several other known stem rust resistance genes were also included in the experiment. Infection types displayed by the differentia l lines were scored 14 days after inoculation using the 0 4 scale of Mains and Jackson (1926). Races of Puccinia graminis f. sp tritici were assigned using the International code of Roelfs & Martens (1988). Results and Discussion Using the International system of nomenclature for Puccinia graminis f. sp. tritici (Roelfs and Martens 1988), 39 different races were identified from the 49 isolates. The 16 rust isolates collected in 2001 were identified as 16 different races and the 33 isolates collected in 2002 belonged to 23 different race groups. Most of the single pustules selected from a single population showed variability, whereas some of them belonged to the same race. The avirulence/virulence formulae for the 37 races ar e given in Table 2. Plants with genes Sr9b and Sr17 displayed consistently high infection ty pes to all races identified in 2001 except to isolates GDB, and RRG and GDB respectively. On the other hand, plants with genes Sr30 displayed low infection types with the exception of races RTK and TTT. In 2002 plants with Sr9b and Sr9e genes were susceptible to most of the is olates identified in that year. Similar to the 2001 reaction of plants with the Sr30 gene was found to be resistant to most of the collections made in 2002. Most of the races identified in Ethiopia durin g the two seasons are virulent to most wheat differentials. For instance, a race like TTT identif ied in 2001 was virulent to all 12 standard differentials, which is a threat to Ethiopian wheat production. Similarly, race TTR which has got a wide spectrum of distribution (in the three regions) was virulent to all differentials except gene Sr30 It is also important to note that gene Sr31 which is known to confer

PAGE 155

resistance to races recorded up to now, has ma intained its resistance to the races identified during 2001 and 2002 in Ethiopia. The composition of physiologic races diffe red greatly between 2001 and 2002. Only races TTR and JGH were identified in both cropping seasons. Races such as TTT, TTR, TPT and RTT have wider spectra of virulence. Races such as TTR, TRR, TPR and DBL have wider spatial distribution whereas the rest are confined to certain localities. In all, 19 races from Bale, 15 from Arsi, and 7 from Shewa regions were recorded during the two years. This finding is indicative of high variability in o ccurrence of races across regions and over time. Some of the races, e.g. RTR, TTR and KTR, differed in virulence or avirulence for only one gene. In case of such minor variations the more virulent race could have arose by simple mutation as has been stated by Singh (1991). In general, absence of virulen ce in all rust pathotypes towards Sr30 gene is a good indication that this genotype could serve as a s ource of resistance to the prevailing races in Ethiopia. Contrary to this, the breeding programmes in Ethiopia should avoid using materials with genes Sr9b, Sr9e and Sr17 to develop wheat varieties for production purpose as they are susceptible to most of the races prevailing in the country. In addition, hi gh virulence diversity observed in this study across regions and over time is assumed to be the cause for resistant materials to lose their resistance in a very s hort time. Therefore, the breeding programmes in the country should focus towards maintaining high genetic diversity in the cultivars and deploy cultivars with different resi stance genes int different regions. Table 1. P. graminis tritici races in Arsi, Bale and Shewa regio ns of Ethiopia during 2001 and 2002 Year Region Races 2001 Arsi DRR, DTR, HG R, HRR, LPR, RTK, TTT Bale JGH, KKR, KTR Shewa CPR, GDB, KJR, RRG, RTR, TTR 2002 Arsi DGG, DPR, JGH, JGQ, KGH, SGH, TRK, TTR Bale DBG, FGG, FGR, FGQ, JGR, KGQ, KGR, KKQ, KQQ, RRT, RTT, TPR, TPT, TRR, TRT, TTR Shewa KTR

PAGE 156

Table 2. Avirulence/vi rulence formulae on Sr genes based on seedling reactions, for 37 races of Puccinia graminis f. sp. tritici identified in Ethiopia during 2001 and 2002 Number Race Avirulence/virulence formulae 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 CPR DBG DGG DPR DRR DTR FGG FGR FGQ GDB HGR HRR JGH JGQ JGR KGH KGQ KGR KJR KKQ KKR KQQ KTR LPR RRG RRT RTK RTR RTT SGH TPR TPT TRK TRR TRT TTR TTT 5, 6, 9e, 21, 30 / 7b, 8a, 9b, 9g, 11, 17, 36 5, 6, 7b, 8a, 9g, 11, 17, 21, 30, 36 / 9b, 9e 5, 6, 7b, 9b, 11, 21, 36 / 8a, 9e, 9g, 17, 30 5, 6, 7b, 21, 30 / 8a, 9b, 9e, 9g, 11, 17, 36 5, 7b, 8a, 21, 30 / 6, 9b, 9e, 9g, 11, 17, 36 5, 7b, 21, 30 / 6, 8a, 9b, 9e, 9g, 11, 17, 36 5, 8a, 9g, 11, 17, 21, 30, 36 / 6, 7b, 9b, 9e 5, 8a, 9g, 11, 21, 30 / 6, 7b, 9b, 9e, 17, 36 5, 8a, 9g, 11, 17, 21, 30 / 6, 7b, 9b, 9e, 36 5, 6, 7b, 9b, 9e, 9g, 11, 17, 30, 36 / 8a, 21 5, 8a, 9e, 9g, 11, 30 / 6, 7b, 9b, 17, 21, 36 5, 8a, 9e, 30 / 6, 7b, 9b, 9g, 11, 17, 21, 36 5, 7b, 8a, 9g, 11, 30, 36 / 6, 9b, 9e, 17, 21 5, 7b, 8a, 9g, 11, 17, 30 / 6, 9b, 9e, 21, 36 5, 7b, 8a, 9g, 11, 30 / 6, 9b, 9e, 17, 21, 36 5, 8a, 9g, 11, 30, 36 / 6, 7b, 9b, 9e, 17, 21 5, 8a, 9g, 11, 17 30 / 6, 7b, 9b, 9e, 21, 36 5, 8a, 9g, 11, 30 / 6, 7b, 9b, 9e, 17, 21, 36 5, 9g, 11, 30 / 6, 7b, 8a, 9b, 9e, 17, 21, 36 5, 11, 17, 30 / 6, 7b, 8a, 9b, 9e, 9g, 21, 36 5, 11, 30 / 6, 7b, 8a, 9b, 9e, 9g, 17, 21, 36 5, 8a, 9g, 17, 30 / 6, 7b, 9b, 9e, 11, 21, 36 5, 30 / 6, 7b, 8a, 9b, 9e, 9g, 11, 17, 21, 36 6, 7b, 9e, 21, 30 / 5, 8a, 9b, 9g, 11, 17, 36 8a, 9e, 17, 30, 36 / 5, 6, 7b, 9b, 9g, 11, 21 8a, 9e / 5, 6, 7b, 9b, 9g, 11, 17, 21, 30, 36 9e, 36 / 5, 6, 7b, 8a, 9b, 9g, 11, 17, 21, 30 9e, 30 / 5, 6, 7b, 8a, 9b, 9g, 11, 17, 21, 36 9e / 5, 6, 7b, 8a, 9b, 9g, 11, 17, 21, 30, 36 7b, 8a, 9g, 11, 30, 36 / 5, 6, 9b, 9e, 17, 21 6, 30 / 5, 7b, 8a, 9b, 9e, 9g, 11, 17, 21, 36 6 / 5, 7b, 8a, 9b, 9e, 9g, 11, 17, 21, 30, 36 8a, 36 / 5, 6, 7b, 9b, 9e, 9g, 11, 17, 21, 30 8a, 30 / 5, 6, 7b, 9b, 9e, 9g, 11, 17, 21, 36 8a / 5, 6, 7b, 9b, 9e, 9g, 11, 17, 21, 30, 36 30 / 5, 6, 7b, 8a, 9b, 9e, 9g, 11, 17, 21, 36 5, 6, 7b, 8a, 9b, 9e, 9g, 11, 17, 21, 30, 36

PAGE 157

Table 3. Distribution of Puccinia graminis f. sp. tritici races in Arsi, Bale and Shewa regions of Ethiopia in 2001 2002. Bale Arsi Shew a Pathotype 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 Total Percentag e CPR 1 1 2.7 DBG 2 2 5.4 DGG 1 1 2.7 DPR 1 1 2.7 DRR 1 1 2.7 DTR 1 1 2.7 FGG 1 1 2.7 FGR 1 1 2.7 FGQ 1 1 2.7 GDB 1 1 2.7 HGR 1 1 2.7 HRR 1 1 2.7 JGH 1 1 2 5.4 JGQ 1 1 2.7 JGR 1 1 2.7 KGH 1 1 2.7 KGQ 1 1 2.7 KGR 1 1 2.7 KJR 1 1 2.7 KKQ 1 1 2.7 KKR 1 1 2.7 KQQ 1 1 2.7 KTR 1 1 2.7 LPR 1 1 2.7 RRG 1 1 2.7 RRT 1 1 2.7 RTK 1 1 2.7 RTR 1 1 2.7 RTT 1 1 2.7 SGH 1 1 2.7 TPR 1 2 3 8.1 TPT 1 1 2.7 TRK 1 1 2.7 TRR 1 1 2.7 TRT 1 1 2.7 TTR 1 1 3 1 1 1 8 21.6 TTT 1 1 2.7 1 = Adaba, 2 = Asasa, 3 = Robe, 4 = Sagure, 5 = Adali, 6 = Lemo, 7 = Sinana, 8 = Agarfa 9 = Kofele, 10 = Wabe, 11 = Bekoji, 12 = Kulums a, 13 = Iteya, 14 = Era, 15 = Debre Zeit 16 = Negelle, 17 = Ankober References Badebo, A. 2002. Breeding bread wheat with multip le disease resistance and high yield for the Ethiopian highlands: Broadening the genetic basis of yellow rust and tan spot resistance. Cuvillier Verlag. Goettingen. Germany. Ph.D. Thesis. CSA. 2002. Report on the preliminary results of area, production and yield of temporary crops. Addis Ababa. Ethiopia. Hoerner, G. R. 1919. Biologic forms of Puccinia recondita on oats. Phytopathology 9: 309-314. Mains, E. B. and Jackson, H. S. 1926. Physiolo gic specialization in the leaf rust of wheat, Puccinia triticina Erikss. Phytopathology 16: 89-92.

PAGE 158

Roelfs, A. P. and Martens, J. W. 1988. An international system of nomenclature for Puccinia graminis f. sp. tritici Phytopathology 78: 526-533. Singh, R. P. 1991. Pathogenicity variation of Puccinia recondita f. sp. tritici and P graminis f. sp. tritici in wheat-growing areas of Mexico during 1988 and 1989. Plant Disease 75: 790-794. Stakman, E. C. and Levine M. N. 1922. The determination of physiologic forms of P. graminis on Triticum spp Tech. Bull. 10. Univ. Minn. Agric. Exp. Stn. 1-10.

PAGE 159

Participatory Evaluation of Bread Wheat Varieties in the Central Highlands of Ethiopia Kassa Getu, Kassahun Zewdie, Yeshimebet Gebrehiwot and Addisu Alemayehu Ethiopian Agricultural Research Organization, Holetta Agricultural Research Center P.O.Box 2003, Addis Ababa, Ethiopia Abstract Six bread wheat varieties were evaluated on a total of 26 non-water logging environments in four districts and on 20 water logging farms at the Ginchi watershed site in the central highlands of Ethiopia in 1999 and 2000. Combined analysis of variance over environments sh owed that Galema and Ketar significantly ( P = 0.05) outyielded the rest of the varieties in non-water logging environments. At Ginchi, four varieties, namely Wetera, Shina, Galema and Ketar gave significantly ( P = 0.05) higher yield than Tura and ET-13. Among the superior varieties, Wetera had the highest mean grain yield. In the partic ipatory evaluations, farmers used their own criteria to select varieties. Weighted accep tance rank analysis showed that Galema scored first in their preference followed by Ketar, Wetera, Shina, ET-13 and Tura in descending order. The significance of such participatory evaluation of varieties on the seed systems in the region is discussed in the paper. Introduction In Ethiopia, bread (60%) and durum (40%) wheat occupy about 1.02 million ha of the total cultivated land, with a total annual produc tion of 1.14 million tones (NSIA, 1998). The average national yield is low, and ranges from 1.1 t/ha for peasant farmers to about 2 t/ha on state farms (Hailu et al. 1991). Other sources and personal communication with farmers in the central highlands, particularly Northern Sh ewa farmers, have revealed that they produce less than the national average yield. The use of unimproved varieties by subsistence farmers, for reasons that include lack of awareness, or the improved varieties themselves, (Hailu, 1988), is one of the major reasons for low productivity. Therefore, increasing the productivity of bread wheat is an important approach towa rds improving the living standards of low imput farmers, and could resolve, in part, the incr eased food demand of Ethiopias rapidly growing population. The national wheat improvement program, now split into the national bread and durum wheat improvement projects, has been conductin g research in the central highlands of Ethiopia. This research indicates that there is a possibility of increasing wheat production by using improved technologies such as high yieldi ng varieties, optimum fertilizer rate, integrated pest management, improved dr ainage on waterlogged areas and other crop management practices. Although selection, h ybridisation and evaluation of bread wheat varieties under diversified soil conditions ha ve taken place at various stages (Bekele et al. 1994), evaluation of these elite bread wheat varieties for wider adaptation and sustainability under on-farm conditions, though th e participation of farmers received little attention. In addition to these, the seed multiplication scale of the Ethiopian Seed Enterprise (ESE) does not meet current demand. The short life span of improved varieties in production is also another constraint and it was estimated that the mean number of years a new variety stays in production is low, and ranges from 2.5 to 4 years (Reggassa et al. 1998), beyond which the yield potential or disease resistance starts to decline. If a variety tak es long to multiply and disseminate, it may not be as useful at the time it reaches the farmers. In this regard, more researchers are suggesting other avenues for seed production and multiplication that can supplement the ESEs

PAGE 160

production. These include the use of irrigation, informal secondary seed production, and the participation of the private sector. Informal seed production on farmers fields could contribute to meeting the farmers seed demand, as well as facilitate the dissemina tion of technologies to the farmers (Gemechu et al. 2004). As a result, research on adaptability and acceptance of new varieties in different localities with participation of farmers is desi rable. With this in mind, the bread wheat breeding project at Holetta Agricultural Rese arch Centre conducted a participatory bread wheat evaluation trial in the central highlands of Ethiopia. The objectives of the experiments were to evaluate the adaptation and acceptability to farmers, of different bread wheat varieti es under non-waterlogged and waterlogged sites, and to inform participating farmers about the bene fits of secondary seed multiplication, both for seed, and income generation. Materials and Methods On-farm evaluation studies on six bread wheat varieties were conduc ted on non-waterlogged and waterlogged environments in the central hi ghlands of Ethiopia for two cropping seasons (1999 and 2000). Thirteen (13) sites were selec ted from four districts: Guder, Jeldu, MetaRobi and Selale in Northwest and West Sh ewa to represent non-waterlogged environments, and 10 farms from Ginchi to represent waterlogge d environments. In the case of Ginchi each farm was considered as a block, whereas in other locations, each site had a complete experiment in RCB design with 4 replications. The varieties were sown on 10 x 10 m, and 4 x 5 m plots in the waterlogged and non-waterlogged sites respectively. DAP and urea fertilizers were applied at rate of 60:69 kg N: P2O5/ ha for non-waterlogged site, and 69:60 kg N: P2O5/ ha at Ginchi waterlogged farms. A seed rate of 175 kg/ha was used for all sites. Except for data collection, the farmers themselves handled all management practices. In a field day organized by Wheat Research and Vertisol Project in 2000, farmers and researchers came together and evaluated the si x varieties based on qualitative parameters, i.e. by seed color, seed size, baking quality, bread taste, nifro (boiled grain) and kolo (roasted grain) quality, with the full participation of farmers. The panelist farmers were asked to prioritize/ rank their selection criteria according to their importance. Average values of ranks were then used to give weight s during ranking of the varieties against each selection criterion. Each panelist farmers (35 in number) were then given a chance to evaluate the varieties independently on a 1-10 scale (where 1 refers to the best and 10 refers to the least preferred) across each qualitative criterion. Finally, the accep tance of the varieties by the farmers was determined by analyses of weighted acceptance rank matrix i.e. by the sum of the product of the weight given to the criterion and ranks assigned to the varieties against each selection parameter. For the combined ANOVA for grain yield in non-waterlogged environments, environments were assumed to be the combinati ons of locations, years and sites and only five of the varieties common for both years were analyzed using the SAS statistical software (SAS, 1998). For waterlogged sites combined ANOVAs by years were analyzed using MSTAT-C statistical software. Results and Discussion Grain yield The results of combined ANOVA over environments in the no n-waterlogged environments for grain yield revealed that there was significant ( P = 0.05) yield difference among the tested varieties, environments and the interaction of variety x environment, suggesting a relatively inconsistent yield ranking of the varieties across a range of the test environments (Table 8).

PAGE 161

With mean grain yields of 26.82 and 26.54 q/ ha respectively, varieties Galema and Ketar yielded significantly better ( P = 0.05) than the other varieties. Shina and the local variety (ET-13 A2) had statistically equivalent yield and gave significantly higher grain yield than Tura (Table 1). The analyses of variance for th e varieties grain yield performance across the years have also showed that there were highly significant ( P = 0.05) variations among genotypes, environments and their interaction, except for year 1999, where the interaction was insignificant (Table 2). In most of the environments, except five cases, Galema and Ketar out yielded the local check(Table 3 and 4). Th e third best varieties for some environments was Shina, whereas Tura had lower yield than ET13-A2 in most of the tested environments. The analyses of variance for grain yield across the environments showed that environment15 is the most suitable environment for wheat production with higher yield of 39.46 q/ha than the rest of testing environments and followed by environment 21,7 and 1 as the second, third and forth consecutively (Table 3 and 4). Table 1. Mean grain and biomass + yield (q/ha) of bread wheat varieties grown in nonwaterlogged and waterlogged environments after combined analysis over years and environments Non-waterlogged sites Waterlogged farms Varieties Grain yield Grain yield Biomass yield Galema (HAR 604) 26.82 26.50 a 106.9 a Tura (HAR 1775) 20.89 20.98 b 87.03 c Shina (HAR 1868) 23.09 26.82 a 90.09 c Ketar (HAR 1899) 26.54 26.27 a 98.16 b Wetera (HAR 1920) 27.43 a 92.00 bc Local check (EA-13-A2) 22.93 20.99 b 98.36 b Mean 24.1 24.8 95.4 CV (%) 21.3 13.4 13.5 LSD (5%) 2.0 8.0 Root MSE 5.11 Combined analysis over locations refers only to non-waterlogged sites + Biomass yield was taken only for waterlogged environments Note: For non-waterlogged environments combined analysis was made only for five of the varieties, which were common for both years. Table 2. Analysis of Variance of mean grain yield of bread wheat varieties grown in 13nonwaterlogged environments for year 1999 and 2000 Year1999 Year 2000 Source DF Mean Square Source DF Mean Square Env 12 1109.44** Env 12 1613.83** Gen 4 325.42** Gen 5 441.61** Env*gen 48 29.07NS Env*gen 60 34.85* Error 195 27.42 Error 234 24.94 Mean 23.90 Mean 23.59 Root MSE 5.24 Root MSE 4.99 C.V (%) 21.92 C.V (%) 21.17 NS, *, **= non significant, significant (p<0.05), and highly significant (p<0.01) respectively

PAGE 162

Abbreviations: gen= genotype, Env= Environment Table 3. Mean grain yield performance of bread wheat varieties grown in 13 non-water logged environments for year 1999 Varieties Env1 Env2 Env3 Env4 Env5Env6 Env7 Env8 Env9 Env10 Env11 Env12 Env13 ET_13 34.00 30.14 28.06 26.19 16.0427.93 32.03 16.41 11.28 20.63 17.10 15.55 18.10 HAR1775 36.99 22.95 29.70 28.54 15.1922.76 30.85 9.45 11.30 25.00 14.32 13.39 16.11 HAR1868 30.28 26.89 29.24 23.71 13.7324.94 35.79 18.80 16.24 20.00 14.47 15.42 24.57 HAR1899 38.61 28.13 32.49 32.35 16.3631.90 43.38 17.70 12.87 26.25 25.94 21.63 27.01 HAR604 38.88 29.35 31.18 30.43 15.8523.80 36.70 19.11 17.06 24.75 19.17 21.41 26.70 Grand Mean 35.43 27.39 29.96 28.12 15.6626.00 35.43 16.50 14.03 23.34 18.35 17.65 22.54 C.V (%) 4.57 5.92 5.41 5.76 10.35 6.18 4.57 9.82 11.55 6.94 8.83 9.19 7.19 Table 4. Mean grain yield perf ormance of bread wheat varieties grown in 13 non-water logged environments for year 2000 Varieties Env14 Env15 Env16 Env17 Env18Env19 Env20Env21 Env22 Env23 Env24 Env25 Env26 ET_13 25.65 36.98 21.94 16.22 18.94 24.36 12.05 41.77 25.06 18.41 22.24 19.06 20.03 HAR1775 20.03 38.09 22.66 14.70 16.37 21.01 9.81 28.58 19.73 18.36 19.38 17.63 20.27 HAR1868 22.16 40.89 27.99 15.77 15.64 23.06 10.88 43.27 19.30 18.87 26.23 16.64 24.06 HAR1899 22.66 42.38 25.98 16.82 18.48 23.01 14.74 47.64 21.59 22.85 28.05 21.91 29.52 HAR1920 14.83 37.12 19.96 14.03 15.88 20.70 10.84 33.17 18.42 16.17 17.44 21.73 26.42 HAR604 24.70 45.80 33.32 17.76 19.25 25.06 17.25 46.12 22.67 20.83 23.87 33.09 33.24 Grand Mean 21.76 39.46 25.32 16.23 17.71 22.90 13.09 39.35 21.24 19.44 22.90 21.76 25.50 C.V (%) 9.42 5.20 8.14 12.63 11.58 8.95 15.65 5.21 9.65 10.55 8.95 9.42 8.00 At Ginchi, in waterlogged farms, results of the combined analysis of variance over year revealed that there was significant difference in grain yield among the varieties (Table 9). However, year and year-by-variety interac tions were non-significant. Among the tested varieties, Wetera gave the highest mean grai n yield, 35.76 q/ha. Three varieties, namely Galema, Ketar, and Shina had statistically equi valent grain yield with Wetera, and showned significant yield advantage over Tura (27.62 q/ ha) and ET-13A2 (27.32 q/ha) (Table 1). It seems that Shina could be a variety of choi ce under both conditions than the local check. Results of the combined analysis of variance over years for biomass yield in waterlogged farms showed that there was insignificant diffe rences among the varieties at the 5% level of significance. Unlike grain yield, variety-by-year interaction was also significant (Table 9). The highest mean biomass yield was obtained from Galema, which gave significantly higher biomass yields of 106.9 q/ha than the rest of the varieties followed by ET-13-A2, Ketar and Wetera in descending orders. Like the grain yield, the least biomass yield was registered in Tura (Table 1).

PAGE 163

Table 5. Importance of whea t quality variables as ranked by farmers at Ginchi Farmers Variables 1 2 3 4 5 6 Total Rank Seed colour (SC) 4 5 3 3 4 4 23 4 Seed size (SS) 2 2 5 2 3 2 16 2 Backing quality (BQ) 7 3 4 5 7 5 31 5 Bread taste (BT) 3 7 2 4 2 3 21 3 Nifro quality (NQ) 8 8 7 7 8 7 45 8 Kolo quality (KQ) 6 4 6 6 6 6 34 6 Grain yield (GY) 1 1 1 1 1 1 6 1 Biomass yield (BY) 5 6 8 8 5 8 40 7 Table 6. Sum of scores given to the varieties a gainst each selection crit erion by 35 farmers and ranks (in parenthesis) determined accordingly Varieties Variables Galema Tura Shina Ketar Wetera Et-13 Seed colour 90(3) 188(6 ) 66(1) 68(2) 143(4) 179(5) Seed size 109(4) 176(5 ) 73(1) 73(1) 88(3) 209(6) Backing quality 61(1) 207(6 ) 114(4) 114(4) 91(2) 103(3) Bread taste 92(1) 191(6 ) 106(3) 106(3) 98(2) 116(4) Nifro quality 116(3) 161(6 ) 84(1) 84(1) 126(4) 156(5) Kolo quality 82(1) 162(5 ) 107(4) 107(4) 96(3) 200(6) Field performance* 92(3) 113(4 ) 63(1) 122(5) 75(2) 157(6) Grain yield** (3) (5) (2) (4) (1) (6) Biomass yield** (1) (5) (6) (2) (4) (3) It was taken from 30 panellists; ** Ranks are determined from mean yield values Table 7. Weighted acceptance valu es obtained by multiplying of weights given to each selection criterion and the ranks of the varieties again st each parameters used to determine final acceptability rank Variables with their weights Variety GY (1) SS (2) BT (3) SC (4) BQ (5) KQ (6) BY (7) NQ (8) Total Final rank Galema 3 8 3 12 5 6 7 24 68 1 Tura 5 10 18 24 30 30 35 48 200 6 Shina 2 4 15 4 25 12 42 16 120 4 Ketar 4 2 9 8 20 24 14 8 89 2 Wetera 1 6 6 16 10 18 28 32 117 3 Et-13 6 12 12 20 15 36 21 40 162 5

PAGE 164

Table 8. Combined analysis of variance over envi ronments for grain yield of five bread wheat varieties grown in 26 non-waterlogged environments in year 1999 and 2000 Grain yield Source of variations DF Mean square Env Gen Env x Gen Error 25 4 100 390 1210.108** 675.991** 33.555* 26.12 Mean 24.05 Root MSE 5.11 C.V (%) 21.25 NS, *, **= non significant, significant (p<0.05), and highly significant (p<0.01) respectively Abbreviations: gen= genotype, Env= Environment Table 9. Combined analysis of variance over years for grain and biomass yield (q/ha) for six bread wheat varieties tested at Ginchi watershed farmlands Grain yield Biomass yield Source of variations DF Mean square Mean Square Year Farmlands within year Variety Year x variety Error r 1 18 5 5 90 105.1 NS 83.73 NS 180.6** 10.8NS 9.2 113.1* 989.1NS 1031.1* 525.6* 164.8 Mean 24.8 95.4 C.V (5%) 13.4 13.5 LSD 2.0 8.0 NS, *, **= non significant, significant (p<0.05), and highly significant (p<0.01) respectively Abbreviations: gen= genotype, Env= Environment Farmer participatory evaluation of the six wheat varieties considered both qualitative and quantitative parameters. Qualitative parameters de termining farmers varietal preferences, as identified by farmers themselves, were seed colo r, seed size, baking quality, bread test, nifro (boiled grain) and kolo (roasted grain) quality. Grain and biomass yield were also considered. Farmers also evaluated field performances of th e varieties before harvesting. The participant farmers made the prioritization of these preand post-harvest parameters in field days organized by the vertisols project and the bread wheat research project at Holetta Agricultural Research Center. Grain yield was unanimously considered to be of paramount importance by all farmers reflecting that it has been a prime selection crite rion for the subsistence farmers in particular. Quality of boilded grain ranked lowest as a criterion for evaluating the varieties (Table 4). Results of variety-acceptance evaluation from more than 30 participant farmers/ panelists for

PAGE 165

all qualitative and quantitative criteria indicated that Galema was preferred over the local check. Ketar and Wetera were also ranked higher than the local check for all qualitative and quantitative parameters, except baking quality and biomass yield respectively (Table 5). The final weighted direct matrix for accepta nce rank showed that, all the varieties except Tura ranked higher than the local check, and th at Galema was the overall favorite variety, followed by Ketar, Wetera and Shina (Table 6). During field days, farmers recognized the importance of secondary seed multiplication as a path to obtaining improved seeds. These farmer s had sown their preferred varieties (Wetera, Galema and Ketar) from their first year harvest on larger plots to use as seeds for subsequent years, and to sell or barter with fellow farmers. Access to improved technology has been cons trained by many factors contingent with poorly developed seed industry. Among theses factors is smallholder farmers inability to pay for the improved seed and fertilizer. Through par ticipatory variety evaluation, however, the farmers were spent less money on seed purchase, and obtained the newly released variety, which otherwise would take many more years to reach them, ahead of time. Kassa et al. (2000) have indicated the positive impact of such an interven tion based on the success observed in Gudar, one of the study areas of th is work. In that study, it was possible to disseminate variety Galema around Gudar and ot her areas, in less than 4 years, with the distribution to remoter areas being mediated by NGOs. Although we did not take actual figures as to the extent of dissemination, we did witness the fascination of the workshop participants with their fellow farmers who had ta ken the lead to work with the researchers in seed production. The participating farmers were very eager to obtain the improved wheat varieties. Although the ESE has been showing comp rehensive and rapid progress in seed production through contractual seed production agreements with private enterprises, Ethiopian Agricultural Research Organization (EARO), Regional Agricultural Research Centers (RARC), farmers and state farmers, it has not been meeting the escalating demand for improved seed in the country (Demesie, 2004 and Girma et al. 2004). According to CTA (1999), only about 7% of the total seed requirement for wheat has been supplied by ESE. The majority of the national seed demand still de pends on the informal community based seed system (Abdissa et al. 2001). Hailu (1992) has shown that the informal farmer-to-farmer exchange system contributes 85% of the total seed requirement. In addition to the aforementioned problems, lack of information about improved seeds has been crucial in addition to the complex socioeconomic problems of the stakeholders. Farmer participatory evaluation and demonstrati on of improved crop varieties, therefore, has a vital role to play in effective diffusion of technology among farmers rather than any others technology transformation methods (Tesfaye et al. 2004). Some earlier reports have also emphasized that informal seed systems, wh ere by farmers themselves produce seeds with some technical assistance from seed specialist, extension workers or breeders should be strengthened to overcome seed availability constraints (Frew and Davids, 1999). Conclusions and recommendations In non-waterlogged areas, Galema and Ketar yiel ded well, and could be promoted on a large scale in the central highlands of Ethiopia a nd in other areas with similar agro-ecologies. In waterlogged areas, Wetera, Shina, Galema and Ketar performed almost equally, and could be promoted there. Shina yielded better, and was less sensitive to the environment than the local check in waterlogged areas. However, in non-waterlogged sites Shina yielded lower that the local check.

PAGE 166

The performance of Tura, in both waterlogged and non-waterlogged environments was poorer than the local check, and hence it would not be a good choice for production in the study areas. Based on our experience in the study areas, if strengthened, informal farmer-to-farmer seed exchange system could be of great importance in disseminating improved seed, of wheat and other crops. Acknowledgement The authors acknowledge the staff of Ethiopian Rural Self Help Association (ERSHA) at Gudar and Food for Hung er International (FHI) at Metarobe for their help in organizing farmers for participatory work We are also grateful to Vertisols project team for the support in financing the field days. The contribution of technical assistants of bread wheat improvement project, namely Assefa Yilma, Chanyalew Mandefro, Adane Meresa, Emebet Admasu, Kassech Birhanu, Assefa Gabisa and Hirut Yirga is sincerely appreciated. References Abdisa Gemeda, Girma Aborha, H. Verkuijl and W. Mwongi. 2001. Farmers Maize Seed Systems in Western Oromia, Ethiopia. Mexico, D.F, International Maize and Wheat Improvement Center (CIMMYT) and Ethiopian Agricultural Research Organization (EARO). p. 18 20. Bekele Geleta, Ammanuael Gorfu and Getnet Gebeyehu 1994. Wheat Production Research in Ethiopia: Constraints and Sustainability. In: Tanner, D.G (eds.). Developing Sustainable Wheat Production Systems: The Eighth Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. p. 12 18 CTA. 1999. The Role of Smallholder Farmer s in Seed Production Systems. Report and Recommendation of Study Visit to Zimbabwe. 15 16 February 1999. Syee Publishing, London, United Kingdom. Demesie Chanyalew. 2004. Improved Crop Varieties, Food Deficit, Seed and Land Use in Ethiopia: Trend and Gap Analysis. In: Asefaw Zelleke Getachew Belay, Belay Simane, Bulcha Weyessa and Nigussie Alemayehu (eds.). Cr op Science Society of Ethiopia (CSSE). Sebil.Vol.10. Proceedings of the tenth conference, 19 21 June 2001, Addis Ababa, Ethiopia. p. 279 292 Frew Mekibib and S. David. 1999. Informal Bean ( P. Vulgaris .L) Seed System in Eastern Ethiopia: Implication for Establishment of Sustainable S eed System in Ethiopia Alemaya University, Research Report. Series No 1, Alemaya, Ethiopia. Gemechu Keneni and Adugna Wakjira. 2004. Genetic Uniformity of Crop Cultivars: Challenges and Opportunities. In: Asefaw Zelleke, Getachew Belay, Belay Simane, Bulcha Weyessa and Nigussie Alemayehu (eds.). Crop Science So ciety of Ethiopia (CSSE). Sebil.Vol.10. Proceeding of the tenth conference, 19 21 J une 2001, Addis Ababa, Ethiopia. P. 1 9. Girma Abera, Mathewos Belissa, Shimellis Dejene, Hailu Gudeta and Gebremedhin W/Giorgis. 2004. Enhancing Food Security Through Farmers Based Seed System, The Case of Improved Potato Production Technology Transfer in Western Ethi opia. Research Report. Oromiya Agricultural Research Institution, Bako ARC, Oromiya, Ethiopia. Hailu Gebre.1988. Crop Agronomy Research on Vertis ol in the Central Highlands of Ethiopia: In: S.C. Jutzi, L. Haque, J. Mc Intire and J.E. Stares (eds.). Management of Vertisols in Sub-Saharan Africa. Proceeding of a Conference, 31 Aug4 Sept. 1987. ILCA, Addis Ababa. p. 263-283. Hailu Beyene, Wilfred Mwangi and Workneh Negatu. 1991. Research Conducted on Wheat Production Constraints in Ethiopia. In: Hailu Gebremariam, Tanner, D.G. and Mengistu Hulluka (eds.). 1991. Wheat Research in Ethiop ia: A Historical Perspective, Addis Ababa, IAR/CIMMYT. p. 17 32 Hailu Gebremariam. 1992. Availability and Use of Seed in Ethiopia. Addis Ababa, Ethiopia. Program Support Unit, Canadian International Development Agency. Kassa Getu, Kassahun Zewdi, Amsal Tarekegne and Girma Taye. 2000. Farmer Participatory Evaluation of Bread Wheat Varieties and Its Im pacts on Adoption of Technology in West Shewa Zones of Ethiopia. In: The Eleventh Regi onal Wheat Workshop fo r Eastern, Centeral and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. p. 427-431.

PAGE 167

NSIA (National Seed Industry Agency). 1998. Crop Variety Registration. Issue Number 1. Addis Ababa, Ethiopia. p. 12. Reggassa Ensermu, Wilfred Mwangi, Hugo Verkujil, Mohammed Hassena and Zewdie Alemayehu. 1998. Farmers Wheat Seed Source and Seed Management in Chilalo Awraja, Ethiopia. Mexico, D.F: IAR and CIMMYT. SAS. 1998. The SAS System for WindowsTM. SAS Institute Inc. Tesfaye Zegeye, Girma Taye, Douglas Tanner, Hugo Verkuiji, Aklilu Agidie and Wilfred Mulangi. 2004. Adoption of Improved Bread Wheat Varieties and Inorganic Fertilizer by Small Scale Farmers in Yelmane Bensa and Forta Districts of North West Ethiopia. In: Asefaw Zelleke, Getachew Belay, Belay Simane, Bulcha Weyessa and Nigussie Alemayehu (eds.). Crop Science Society of Ethiopia (CSSE). Sebil.Vol.10. Proceeding of the Tenth Conference, 19 21 June 2001, Addis Ababa, Ethiopia. p. 246 258

PAGE 168

Part 3. PROTECTION

PAGE 169

Survey of Natural Enemies of the Russian Wheat Aphid, Diuraphis Noxia (Kurdijimov) in Kenya M.Macharia, M. Njuguna, and I. Koros KARINational Research Centre, P.O. Private Bag, Njoro, Post Office Code-20107, Njoro, Kenya Abstract The Russian wheat aphid, Diuraphis noxia (Kurdijimov), was first noticed in Kenyan farmers' wheat ( Triticum aestivu m L.) fields in 1995. Currently, Russian wheat aphid control relies mainly on use of insecticides to kill aphids already established on the crop. In an effort to enhance the integration of natural enemies in an integrated pest management control strategy, surveys were initiated to document and determine the effectiveness of natura l enemies that attack the Russian wheat aphid. A number of predators and parasitoids were observed to attack cereal aphids, but none of these bio-control agents exerted adequate controls. The observed predators were Coleoptera: Adonia variegata and Cheilomenes spp.; Diptera: Syrphidae (hover flies); Arachnoidea (s piders) and Neuroptera (lacewings). Parasitoids observed were Hymenoptera: Aphidius spp. and Aphelinus spp. Field observations and results of research carried out indicated that peak population densities occurred after peak percentage infestation by the wheat aphid. It was observed that the increased numbers of na tural enemies did not have any noticeable impact on RWA as they arrived late in th e season when the pest had already caused damage to the wheat crop. Russian wheat aphid was observed to be attacked by several predators and parasitoids commonly associated with other cereal aphids. Efforts should therefore be made to conserve these bio-control agents. They are of great importance in controlling other cereal aphids, and may as well reduce Russian wheat aphid populations. Introduction In Kenya, the six important cereal aphids species that attack wheat are Diuraphis noxia Schizaphis graminum Sitobion avenae, Rhopalosiphum padi, R. maidis and Metopolophium dirhodum (Wanjama, 1990; Macharia et al, 1993; Macharia et al, 1997). Of these six species of cereal aphids, D. noxia, the Russian wheat a phid (RWA), is the most destructive followed by S. graminum The other species are less important. Besides their direct damage to wheat crops through sucking of plants sap, the aphids transmit barley yellow dwarf virus (Wangai and Torres, 1990). Currently in Kenya, RWA control relies mainly on use of insecticides to kill cereal aphids already established on the crop and so pr event or reduce the risk of further spread. However, chemical control of RWA has proven di fficult because of its habit of colonising the inner surfaces of tightly curled leaves of damaged plants. Worldwide, the use of biological control ag ents is seen as a desirable alternative to insecticides because of its low cost, sustainabilit y and environmental friendiness. In Kenya, most of the growers are subsistence farmers, and neither the crop, nor the resources of the farmers can warrant the use of repeated insec ticidal applications against the cereal aphid pests. Expense and possible environmental po llution from insecticide use are of major concern to the farmers. Farmers may prefer to reduce losses through the use of resistant cultivars and/or in combination with effective natural enemies.

PAGE 170

Majority of wheat growers in Kenya are not aware of cereal aphid biological control measures. However, majority of farmers will be willing to stop spraying pesticides should the biological control strategy be effective. Due to the wide range of aphid species that attack wheat and barley (Hordeum vulgare L.) in Kenya, causing substantial yield losses of 10100% (Macharia et al, 1997), biological contro l strategies must be developed that will enhance the integration of these control agents in an Integrated Pest Management (IPM) control strategy. Therefore surveys were initiated to document and determine the effectiveness of natural enemies that attack RWA. In addition, studies were also initiated to evaluate the effectiveness of these natural enemies. Materials and Methods Surveys in the farmers' fields around Njoro, Kenya were carried out in 2001-2002 cropping season. The surveys were carried out at late tillering stage (G.S. 25; Lancashire et al ., 1991) of the wheat crop. The aphids were sampled by examining 20 plants randomly along diagonals within the fields at approximately 50m intervals. The aphid species were collected, identified for confirmation and recorded (Blackman et al, 1985). Predators, parasitized aphids and parasitoids were also collect ed for species identification. Field studies to determine the effectiveness if the natural enemies were initiated in 2002. In June 2002, a 20 x 20m strip was seeded with the bread wheat cultivar Kwale at 75 kg/ha. This late planting was timed to coincide with p eak aphid infestations at Njoro. Plots were fertilized at the rate 18kg nitrogen and 46kg P2O5 /ha. Plots were kept weed free by application of Buctril MC (Bromoxynil +MCPA) at the rate of 1.25 l/ha. Leaf diseases were controlled by applying Folicur 250 EW (Te buconazole) at the rate of 0.75 l/ha. Assessment of the RWA and endemic natura l enemies' [predators and parasitoids] occurrence was done by taking a random sample of 20 plants from the wheat field, once per week. Sampling commenced 2 weeks after germina tion and continued for 8 weeks. The data taken included: number of RWA/plant; number of Adonia variegata and coccinellid larvae; incidence of parasitism; and percentage of plants damaged by RWA. The infested wheat plants were examined in the field for the pr esence of natural enemies. The aphid species were identified (Blackman and Eastop, 1985) and the particular predator/parasitoid identified and recorded. Results and Discussion The survey revealed that wh eat is attacked by several species of cereal aphids. The aphid species collected during the survey in decreasing order of importance were, D. noxia, M. dirhodum, R. maidis, R. padi, S. avenae and S. graminnum Similar observations have been reported by Macharia et al (1997). Several types of predators and parasitoids (Tab le 1) attacked most of these cereal aphids. Predators (a) Coccinellid beetles, Adonia variegata (Goeze) and Cheilomenes species were the most important. Their populations were rather low a nd tended to occur late in the season, at heading stage of the wheat crop. Population dens ity of the coccinellid beetles was observed to be about 1.8 beetles per plant. Haile and Mege nasa (1987) have reported similar observations. (b) Hoverflies (Syrphidae) were also important predators with their population peaks coinciding with peak aphid infestation at heading stage of wheat. They were second in importance to coccinellid beetles. Population de nsity of the hoverfly was observed to be about 0.6 larvae per plant

PAGE 171

(c) Lacewings (Neuroptera) and spiders (Ar achnoid) were also observed in very low population levels and tended to occur from late tillering to heading stage of the crop. Population density of the Lacewi ngs was observed to be about 0. 3 and 0.1 spiders per plant. Parasitoids The most important parasitoids were Aphidius spp and Aphelinus spp both of which attack D. noxia. Their population densities tended to be low an d appeared late in the season after RWA population peak and showed only about 5% par asitism. All these natural enemies somehow occur at different times during the cropping seas on and they may be collectively contributing to season long control of RWA and other cereal a phids in the wheat crop. The survey revealed that many of the natural enemies (predators and parasitoids) of other cereal aphids also attack RWA although none of them exerted adequate control (Table 1). Peairs (1998) and Gary et al. (1998) have reported similar results. The absence of successful aphid predators and parasitoids may explain the rapid spread of RWA. Field observations also revealed that the natural enemies of RWA were only present late in the crop season when damage to wheat and barley crops had already taken place. Data presented in (Table 2) indicates that peak percentage infestation by RWA occurred at the end of mid-July to mid-August which co incided with start of stem elongation to full flag leaf emergence (GS 37; Lancashire et al ., 1991). This was the period of exponential increase in percentage infestation. Population density of RWA peaked in midA ugust [135.3 aphids per plant] (Table 2). Field observations revealed that initial field infestations of wheat by RWA started from the edges of the field followed by high natural enemy activity. The number of coccinelids A. variegata and Aphidius spp rose from end of July onwards reaching population peak at the end of August. This is probably in response to the increase in aphid numbers (Table 2). This indicates that the natural enemies had their gr eatest impact at the end of the season when RWA population was on the decline. It was obser ved that the increased numbers of natural enemies did not have any noticeable impact on RWA as they arrived late in the season when the pest had already caused damage to the crop. Table 1. Predators and parasitoids of aphids recorded in wheat fields around Njoro, Kenya, 2001-2002. Cereal aphid species Predators and parasitoids D. noxia M. dirhodum R. maidis R. padi S. avenae S. graminum Predators Coleoptera [Beetles] Adonia variegata Cheilomenes spp. Diptera Syrphidae [Hover flies] Arachnoidea [ Spiders] Neuroptera [Lacewings] Parasitoids Hymenoptera Aphidius spp. 2. Aphelinus spp. x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

PAGE 172

Table 2. Mean counts of percentage Russian wheat aphid (RWA) infestation, No. of RWA per plant [apterae and alates], Adonia variegata and coccinellid larvae, and No. of mummies [Aphidius spp] at Njoro, 2002. Sampling dates 30 June 5 July 30 July 15 Aug. 30 Aug. 15 Sept. Growth stage 21 25 29 37 51 59 % RWA infestation 4.2 23.9 82.5 100.0 100.0 100.0 No. of RWA/plant 5.2 32.5 100.2 135.3 10.3 1.1 No. of Adonia variegata & larvae/plant 0.0 0.0 5.6 22.5 30.3 20.1 No. of mummies [Aphidius spp.]/plant 0.0 0.0 2.1 10.5 12.2 9.2 Conclusion Several natural agents, including predators such as coccinelids beetles, hoverflies, lacewings, spiders and parasitoids including Aphidius spp and Aphelinus spp were recorded during the survey as important natural enemies of cereal aphid populations. Due to their low numbers, they were not effective controls and were unable to keep RWA populations below damaging levels. Efforts should be made to conserve th ese natural enemies as they of great importance in controlling the cereal aphids and may as well reduce RWA populations. It was also observed that at earlier seedling stages, the population of natural enemies was too low to exert effective cereal aphid cont rol. The populations of these natural enemies increased only with the rise of cereal aphids. The increased numbers of natural enemies were observed not to have any effective impact on the population control of RWA. Biological control strategies must be devel oped that enhance the integrated operation of these natural enemies of cereal aphids. RWA ma nagement will become in creasingly reliant on adoption of wheat crop production systems that enhance and improve in the effectiveness of biological control agents. References Blackman, R.L., and V. F. Eastop. 1985. Aphids on the world's crops, An Identification Guide. Wiley, Chichester. U.K. Gary, L.H., F. P. Baxendale, J. B. Campbell, A. F. Hagen, and J. A. Kalisch. 1989. Russian Wheat Aphid. Neb Guide G89-936-A. Cooperative Exte nsion, Institute of Agriculture and Natural Resources, University of Nebraska-Lincoln. Haile A. and T. Megenasa. 1987. Survey of aphids of barley in parts of Shewa, Welo and Tigrai, Ethiopia. Ethiop. J. Agric. Sci. Vol 9 No. 1: 39-53. Lancashire, P.PD., Bleiholde, H., v.d. Boom, T.,Langeludedeke, P.,Staus, R., Weber, E. and Witzenberger,A. 19991. A uniform decimal code fo r growth stages of crops and weeds. Ann appl. Biol.119,561-601 Macharia, M., P.M. Muthangya and J.K. Wanjama. 1997. Management of Russian wheat aphid, Diuraphis noxia, on wheat in Kenya by use of seed dressing insecticides. African Crop Science Conference Proceedings Vol. 3. Pp. 1191-1198. Macharia, M., P.M. Muthangya, A.W. Wangai and J.K. Wanjama. 1993. Barley Yellow Dwarf Virus Report 1990-1993. Kenya Agricultural Research Institute, National Plant Breeding Research Centre, Njoro, Kenya. Peairs, F.B. 1998. Russian Wheat Aphid Management. In: The Inaugural National Wheat Industry Research Forum Proceedings .San Diego, California. January 14 and15, 1998

PAGE 173

Wanjama, J.K. 1990. Ecology of cereal aphids transmitting Barley yellow dwarf virus in Kenya. Pages 240-243. In: P.A. Burnnet (ed.), World perspectives in Barley Yellow Dwarf. CIMMYT, MEXICO, D.F. Wangai, A.W. and E. Torres. 1990. Barley yellow dw arf virus situation report for Eastern Africa with special emphasis in Kenya. Page 71 In: P.A. Burnnet (ed.), World perspectives in Barley Yellow Dwarf. CIMMYT, MEXICO, D.F.

PAGE 174

Evaluation of the Herbicide Monitor, Alone and in Tank-Mixes for Weed Control in Wheat ( Triticum aestivum L.) D.O.K. Amadi, J.N. Kamwaga and G.O. Mbanda Abstract Experiments were conducted at Wangu Embori Farm (Timau) and Purko Farm (Mau Narok), Kenya to evaluate th e effect of the herbicide Monitor [1-(2ethylsulfonylimidazo [1,2-a] pyridin-3-ylsulfonyl)-3-(4,6-dimethoxypyrimidin-2-yl) urea] alone and in tank-mixes with Derby (a.i. Flurasulam + Flumetsulam) and Buctril MC (a.i. Bromoxynil + MPCA), for management of brome grass ( Bromus spp.) and broadleaf weeds in wheat ( Triticum aestivum L.). Monitor, was screened at 20, 25, 30, 35, 40, and 40g ha-1 (two applications of 20g ha-1 each with an interval of three weeks) and 80g ha-1, all with Agral 90 as a surfact ant at 0.25% v/v. At 40g ha-1 Monitor was tank mixed with Derby and Buctril MC at the rates of 50 and 375ml ha-1 respectively, but in the second and third se asons the rate of Buctril MC was increased to 700ml ha-1. Monitor was also tank mixed with Derby at 80g ha-1 and 100ml ha-1 respectively. Cossack (a.i. Mesosulfuron + iodosulfuron) was used as the standard herbicide at the rate of 300g ha-1 + 150ml ha-1 of Agral 90. Monitor at 30, 40 and 80g ha-1 gave excellent control of Bromus sterilis G alium spurum Emex australis and Chenopodium album. The activity of Cossack on the target weeds at the site was comparable to that of Monitor at 80g ha-1 and its tank mix partners. Some significant differences were observed among treatm ents with regard to efficacy and phytotoxicity (crop height). Crop emergence (m-2), TKW (g), spike density m-2 and wheat grain yield (t ha-1) in certain seasons did not have any significant differences between the herbicide treatments. Herbicide phytotoxicity was influenced by the rate of application of Agral 90. Introduction Wheat is the second-most-important cereal crop in Kenya, but local production does not satisfy demand, and the deficit must be imported. It is therefore vital to place emphasis on production strategies that will inrease wheat grai n yield. Improved weed management is one such strategy. Under farm conditions in Kenya, the weed flora is in a state of continuous change. Human activities influence and alter the balan ce of the weed species in specific localities. Cultural practices such as liming, increased fer tilizer usage and changing cropping patterns all favour some weeds, and reduces others. The more widespread use of herbicides that selectively kill broadleaf weeds has favoured profuse multiplication of some annual grasses, previously not important agronomic weeds. Bromus spp. and other annual grasses are now very important weeds, and require proper management. As is always the case weeds are competitive, persistent and pernicious. Because of these attributes, efficient and successful weed control should integrate different methods. Farmers efforts to reduce the growth and proliferation of weeds in their wheat fields by means of rotations, and other cultural practices do much to reduce the number of weeds able to set seed and so improve the quality of grain harvested. But these activities are not enough to ensure a wheat crop that is not adversely affected by weeds. Economic weed control in cereals requires st rategies based on intelligent or judicious use of herbicides, coupled with appropriate cultu ral practices. Weed populations at which an

PAGE 175

economic yield response can be expected from the use of herbicides are difficult to define and vary according to the time of weed germination and the growth of weeds in comparison to the crop. This evaluation trial focused mainly on the effect of Monitor at different application rates, alone and in tank-mixes with Derby and Buctril MC, on brome grass ( Bromus spp.) and broad-leaf weeds. The pernicious effects of the herbicide treatments on the wheat crop were also assessed. Monitor (sulfosulfuron) has the chemical name: [1-(2-ethylsulfonylimidazo [1,2-a] pyridin-3ylsulfonyl)-3-(4,6-dimethoxypyrimidin-2-yl) ur ea]. It is a post-emergence herbicide that enters the plant through both foliar uptake (60% ) and root uptake (40%). Monitor kills the target plants by inhibiting the formati on of amino acids essential for cell division. Materials and Methods The experiment was conducted in three seasons The first season trial was done at Purko Farm, Mau Narok. It was planted on 28-8-02, using the wheat cultivar Mbuni at a seed rate of 90kg ha-1. Basal fertilizer (DAP = 18:46:0 N:P2O5:K2O) was applied at the time of planting at the rate of 100kg ha-1. The second season experiment was superimposed on a commercial wheat crop (cv. K. Kwale) planted on 3-4-2003 at Wangu Embori Fa rm in Timau. The seed rate was 110kg ha-1. Monoammonium phosphate (MAP 11:52:0 N:P2O5:K20) was applied at seeding at the rate of 147kg ha-1. The site was previously under commercial wheat. The third season trial was also conducted at Wangu Embori Farm. Wheat cultivar, Kenya Mbuni was seeded on 16-10-2003 using a crop seeder at a seed rate of 100kg ha-1. MAP was drilled with the seed at 150kg ha-1. The trial site was previously under commercial wheat where weeds were controlled using Attr ibut 75 WG (a.i. Propoxycarbozone). The herbicide treatments applied in the three seasons are shown in Tables 1 and 2. Herbicide treatments were applie d at 48 days, 55 days and 28 days after planting in the first, second and third seasons respectively. Table 1. Herbicide treatments included in th e first (Purko Farm, Mau Narok) and second (Wangu Embori Farm, Timau) seasons. Herbicide treatment Rate ha-1 Monitor + Agral 2% 20g + 3L Monitor + Agral 2% 25g + 3L Monitor + Agral 2% 30g + 3L Monitor + Agral 2% 35g + 3L Monitor + Agral 2% 40g + 3L Monitor + Agral 2% Two applications 3 weeks apart 20g + 3L and 20g + 3L Monitor + Buctril MC + Agral 2% 40g + 375ml + 3L Monitor + Derby + Agral 2% 40g + 50ml + 3L Cossack + Agral 1% 250g + 1.5L Cossack + Agral 1% 300g + 1.5L Monitor + Hussar OF + Agral 2% 30g + 150g + 3L Untreated control Vol/vol.

PAGE 176

Table 2. Herbicide treatments in the th ird season. Wangu Embori Farm, Timau. Herbicide treatment Product Rate ha-1 Monitor + Agral 0.25% v/v 30g + 375ml Monitor + Agral 0.25% v/v 40g + 375ml Monitor + Agral 0.25% v/v 20g + 375ml (Repeated 3 wks later) Monitor + Derby + Agral 0.25% 40g + 50ml + 375ml Monitor + Agral 0.25% v/v 80g + 375ml Monitor + Derby + Agral 0.25% v/v 80g + 100ml + 375ml Monitor + Buctril MC + Agral 0.25% v/v 40g + 700ml + 370ml Cossack + Agral 0.1% v/v 300g + 150ml Untreated check Hand weed check In all trials, the herbicide treatments we re arranged in a random ized complete block replicated four times. The trial plots measured 3m X 6m. The herbicide treatments were applied using a hand operated knapsack spra yer with a 2m spray swath and 150L ha-1 nozzle delivery volume. Measurements of relative humidity, ambient temperatures and wind speed were not made due to lack of instruments. The soil was moist at the time of herbicide application in all the trials and the crop and w eeds were growing vigorously: a prerequisite for effective performance of soil acting herbicides. Before treatment application the most preval ent weed species at the sites were identified and counted in 1m2 quadrats in each plot. These sampling areas were marked and changes in weed density monitored three times at interval s of fourteen days. Crop reaction to herbicide treatments was first assessed visually three da ys after application and subsequently on each day weed counts were taken. Weed species that were biologically alive 28 days after herbicide application were deemed to have escaped herbicidal effect. Treatment e ffects on weeds were expressed in terms of percentage mortality, a quantitative response that reflects the degree of weed control (efficacy) achieved. The formula below was used to derive mortality. Percentage mortality (efficacy) = x 100 where, X is the mean number of weeds per plot in untreated plots and Y is the mean number of weeds per plot receiving specific treatment. Crop emergence (plants m-2) and crop height (5 plants/plot) in all the experimental units were taken 14 days after planting and at physiologic maturity respectively. A 9m2 central area of the treated plot was harvested, threshed and grain yield and moisture content determined. Yields are expressed as grain yield at 12.5% moisture. The percentage efficacy, wheat grain yiel d, thousand kernel weight (TKW), spike density m-2 and crop height were subjected to anal ysis of variance. Treatment means were separated using Duncan Multiple Range Test (DMRT). Results and Discussion The trial sites had a narrow spectrum of weeds as shown in Table 3 below. The weed growth stages at the time of application are defined in terms of the number of leaves. It should be noted that the weed growth stages in certa in instances were beyond those recommended for effective performance of Monitor. (X Y) X

PAGE 177

The efficacy of the individual herbicide treatments in controlling the principal weeds present is shown in Tables 4, 5 and 6 for the th ree seasons. At Purko Farm in the first season, brome grass, the main target weed, did not germinate at the right time. In Timau (2nd Season) Monitor alone (20g ha-1, 25g ha-1, 30g ha-1, 35g ha-1 and 40g ha-1) did not satisfactorily control Bromus sterilis (Table 6). The 40g ha-1 rate gave significantly better control of the grass than the lower rates, but it was still not adequate This treatment however provided somewhat satisfactory control of Galinsoga parviflora (75.5%), and gave some control of Emex australis, Chenopodium album and Cortula abysinica A tank mix of 40g ha-1 Monitor with Derby or Buctril MC did not control brome grass any better than Monitor alone at 40g ha-1. However, the Table 3. Weed species and growth stages at time of herbicide application in the three seasons. No. of leaves Weed species 1st Season (Purko Farm) 2nd Season (Timau) 3rd Season (Timau) Bromus sterilis 2-mid tillering 3-4 Cyperus diformis 6-12 Emex australis 4-9 2-4 Chenopodium album 7-8 4-8 2-4 Cortula abysinica 4-10 Galinsoga parviflora 4-8 4-8 Raphanus raphanistrum 6-8 Polygonum convolvulus 5-9 Lolium temulentum 3-early tillering Galium spurium 2-4 Table 4. Control of individual weed species by the various herb icide treatments on. Purko Farm, 2002, Weed control (Percentage Efficacy) Herbicide treatment Galinsoga parviflora Chenopodium album Raphanus raphanistrum Polygonum convolvulus Lolium temulentum Cortula abysinica Mean %-efficacy Monitor 20g ha-1 + Agral 2% 60.5 47.8 78.2 20.0 23.3 64.6 49.1 E Monitor 25g ha-1 + Agral 2% 64.8 75.1 100.0 65.0 37.9 67.7 68.4 D Monitor 30g ha-1 + Agral 2% 86.9 95.2 100.0 71.4 52.2 89.5 82.5 C Monitor 35g ha-1 + Agral 2% 88.6 94.4 100.0 44.3 48.9 93.3 78.3 C Monitor 40g ha-1 + Agral 2% 86.7 87.5 100.0 40.0 57.7 80.0 75.3 C Monitor 20g ha-1 + Agral 2% (2 applications 3 wks apart) 55.6 68.3 91.7 44.3 36.9 90.4 64.5 D Monitor 40g ha-1 + Buctril MC 375ml ha-1 + Agral 2% 88.5 100.0 100.0 79.2 42.0 92.7 83.7 BC Monitor 40g ha-1 + Derby 50ml ha-1 + Agral 2% 100.0 100.0 97.0 100.0 47.7 100.0 90.7 AB Cossack 250g ha-1 + Agral 1% 98.2 94.4 100.0 93.3 72.4 100.0 93.1 A Cossack 300g ha-1 + Agral 1% 99.5 100.0 100.0 89.7 78.1 100.0 94.6 A Monitor 30g ha-1 + Hussar 150g ha-1 + Agral 2% 0.0 0.0 0.0 0.0 0.0 6.7 1012 F Untreated control 0.1 0.0 0.0 0.0 0.0 0.0 0.0 F C.V 8.71% LSD at alpha=0.05 9.6 S.E 3.26

PAGE 178

tank mixes did give fair control of the othe r broadleaf weeds. None of the treatment combinations gave any control of Cyperus diformis Percentage efficacy results in the second season (T able 5) show that Cossack treatments were superior to all other treatments in overall w eed control and suppression. At the rates of 300g ha-1 + Agral 1% and 250g ha-1 + Agral 1% the levels of weed kill were 79.3% and 77.9% respectively. These percentages represent average performance on all the weed species at the site. Table 5. Control of individual weed species by the various herbicide treatments on Wango Emburi Farm, Timau. 2002 (2nd season) Weed Control (Percentage Efficacy) Herbicide treatment Bromus sterilis Cyperus diformis Emex australis Chenopodium album Cortula abysinica Galium spurium Mean Monitor 20g ha-1 + Agral 2% 30.9 F 0.1 38.1 54.1 56.4 66.8 41.08 F Monitor 25g ha-1 + Agral 2% 31.8 F 0.03 38.0 54.2 58.2 66.9 41.53 F Monitor 30g ha-1 + Agral 2% 35.4 E 1.1 39.0 55.4 61.8 67.6 43.39 EF Monitor 35g ha-1 + Agral 2% 37.8 D 1.2 39.5 60.9 62.8 68.8 45.16 E Monitor 40g ha-1 + Agral 2% 43.6 C 1.7 41.1 66.5 73.4 75.5 50.29 D Monitor 20g ha-1 + Agral 2% (2 applications 3 wks apart) 39.1 D 0.35 42.0 56.9 58.6 72.0 44.82 E Monitor 40g ha-1 + Buctril MC 375ml ha-1 + Agral 2% 43.2 C 1.5 71.0 72.9 76.8 78.5 57.31 C Monitor 40g ha-1 + Derby 50ml ha-1 + Agral 2% 43.4 C 0.8 81.8 93.1 94.9 98.4 68.73 B Cossack 250g ha-1 + Agral 1% 87.5 B 1.7 85.9 96.3 96.7 99.3 77.90 A Cossack 300g ha-1 + Agral 1% 91.8 A 1.9 87.6 96.4 97.9 99.9 79.30 A Untreated check 0.15 G 0.1 0.0 0.1 0.13 0.25 0.12 G C.V S.E. LSD at alpha =0.05 3.53 0.77 2.25 2.90% 0.7207 2.081 Treatments followed by the same letter(s) are not significantly different. At the time of treatment application in the second season in Timau most of the brome grass seedlings were at mid-tillering stage. Monitor is reported to be most effective on brome grass when the weed has between 2 and 4 leaves, a nd therefore the application was too late and poor control was obtained with all rates of Monitor and with the tank mixes with Derby and Buctril MC. The split application (20g ha-1 twice at an interval of three weeks) of Monitor gave significantly lower brome grass control than a single application of 40g ha-1, and

PAGE 179

therefore was not economic. Broadleaf weed contro l was also poor at all rates of Monitor, but the tank mixes with Derby and Buctril MC e nhanced its performance on the broadleaf weeds at the site (Table6). Table 6. Mean grain yield, crop height and mean levels of weed control (% efficacy) in the trial in the second season Wango Emburi Farm, Timau Herbicide treatment Crop height (cm) Grain yield (ton ha-1) Efficacy % Monitor 20g ha-1 + Agral 2% 95.75 B 1.305 G 41.08 F Monitor 25g ha-1 + Agral 2% 92.50 BC 1.330 G 41.53 F Monitor 30g ha-1 + Agral 2% 92.75 BC 1.523 F 43.39 EF Monitor 35g ha-1 + Agral 2% 92.25 C 1.560 F 45.16 E Monitor 40g ha-1 + Agral 2% 93.25 BC 1.832 E 50.29 D Monitor 20g ha-1 + Agral 2% (2 applications 3 wks apart) 92.25 C 1.563 F 44.82 E Monitor 40g ha-1 + Buctril 37ml ha-1 + Agral 2% 95.50 BC 2.575 D 57.31 C Monitor 40g ha-1 + Derby 50ml ha-1 + Agral 2% 93.00 BC 2.820 C 68.73 B Cossack 250g ha-1 + Agral 1% 94.75 BC 3.463 B 77.90 A Cossack 300g ha-1 + Agral 1% 93.00 BC 3.560 A 79.30 A Untreated check 107.00 A 1.100 B 0.12 G C.V S.E. LSD at alpha =0.05 2.45 1.1624 3.357 2.57 0.0266 0.07910 2.90% 0.7207 2.081 Treatments followed by same letter(s) are not significantly different. The improvement in the control of broadl eaf weeds when Monitor was tank mixed with Buctril MC and Derby was statistically signifi cant. The Monitor/Derby tank mix was the most effective, and therefore it would be worthwh ile to tank mix Buctril MC and Derby with Monitor in situations where the spectrum of br oadleaf weeds is wide. Increasing the rate of Buctril MC from 375ml ha-1 to about 700ml ha-1 could significantly improve the degree of control of broadleaf weeds. Cossack did not control Cyperus diformis but this is a weed whose growth habit and morphological characteristics do not favour aggressive competition for growth resources. All the broadleaf weeds at the site were vulnerable to Cossack. It is also instructive to note that despite the delayed application Cossack still gave impressive control of Bromis sterilis Statistical analysis (p=0.05) showed that there was no significant difference between the Cossack rates of 250g ha-1 and 300g ha-1 in terms of their activity on weeds. This suggests that the lower rate should be used for early applications. In this second season trial visual assessment of crop tolerance to the herbicide treatments showed remarkable crop reaction symptoms in vi rtually all the treatments. General chlorosis was observed three days after treatment applica tion. The wheat crop (cv. K. Kwale) became even more chlorotic seven days after applica tion and at this time general crop stunting was discernable. Crop recovery was evident fourteen days after application, and by 28 days after application the yellowing of wheat foliage ha d drastically decreased. Forty two days after treatment the crop had fully recovered but height differences between the treated areas and the adjacent untreated buffer strips could be seen. Th e intensity of chlorosis appeared uniform in all the treatments. This observation suggests that Agral 90 (2% and 1% v/v) was perhaps responsible for the crop reaction. To further investigate herbicide phytotoxicit y, wheat crop height was measured in all the plots: treatment means are shown in Table 6. The untreated plots on average had the tallest plants (107cm). This was probably due to compe tition for light with the weeds, mainly brome grass. Plots treated with Monitor at the rates of 35g ha-1 and 20g ha-1 (same treatment repeated after three weeks) had the shortest plants (92.25cm). There was no evidence that tank mixing Monitor with Derby or Buctril MC gave a greater degree of phytotoxicity.

PAGE 180

Results in Table 6 indicate that grain yield w as to a large extent dependent on the level of weed suppression attained by individual herb icide treatments, whereas a negative effect of phytotoxicity (crop height) is not evident. Late application of treatments probably reduced the activity of Monitor on brome grass, stressing th e importance of timely application of this particular product. Table 7: Control of individual weed species by the various herbicide treatments on Wango Emburi Farm, Timau. 2002 (3rd season) Weed Control (Percentage Efficacy) Herbicide treatments Emex australis Galium spurium Bromus sterilis Chepnoodium album Mean efficacy % Monitor 30g ha-1 + Agral 0.25% 72.2 82.6 84.2 46.7 71.4 E Monitor 40g ha-1 + Agral 0.25% 74.4 93.7 95.2 66.7 82.5 C Monitor 20g ha-1 + Agral 0.25% 80.4 73.1 79.5 68.8 75.5 D Monitor 40g ha-1 + Derby 50ml ha-1 + Agral 0.25% 95.5 96.0 96.4 100 97.0 AB Monitor 80g ha-1 + Agral 0.25% 87.3 94.3 100 100 95.4 B Monitor 80g ha-1 + Derby 100ml ha-1+ Agral 0.25% 100 100 100 100 100 A Monitor 40g ha-1 + Buctril MC 700ml ha-1 + Agral 0.25% 100 100 95.6 100 98.9 A Cossack 300g ha-1 + Agral 0.1% 97.3 100 100 100 99.3 A Untreated check 3.4 0 3.1 4.6 2.8 G Hand weeded check 57.0 56.5 52.2 12.5 44.6 F C.V S.E LSD at alpha =0.05 2.78% 1.07 3.1 All values followed by the same letter within columns are not significantly different at 5%. In the third seasons experiment (Timau) the tim ing of treatment application was correct. The rates of the wetting agent, Agral 90, were also altered (Table 2). Results indicate that Monitor at 30g ha-1, 40g ha-1 and 80g ha-1 with Agral 90 (0.25% v/v) gave significant control of Bromus sterilis (Table 7). At 80g ha-1 Monitor gave complete kill of brome grass (%-efficacy =100). Lower rates of 30g ha-1 and 40g ha-1 had %-efficacy of 84.2 and 95.2 respectively. These levels of c ontrol fall within the acceptability limit. When the 40g ha-1 rate was applied in two split applications of 20g ha-1 at an interval of three weeks, the level of performance was not as good as a single application. This suggests that the second application was not effective due to the advanced stage of weed growth. Analysis of variance (p=0.05) showed significant differences among the Monitor rates in terms of their activities on brome. Emex australis, Galium spurium and Chenopodium album were satisfactorily controlled by Monitor at 80g ha-1. Galium spurium was the most vulnerable broadleaf weed to all the test rates of Monitor (Table 7) The inclusion of Derby and Buctril MC as tank mixes with Monitor enhanced the level of control of the broadleaf weeds at the site. Monitor + Derby + Agral 90 at 40g ha-1 + 50ml ha-1 + 375ml ha-1 respectively attained a mean efficac y of 97.2% on broadleaf weeds. At double rates of the products (80g ha-1 and 100ml ha-1) the level of performance rose to 100%. The Monitor and Buctril MC combination at 40g ha-1-1 and 700ml ha-1-1 respectively gave complete control of all the broadleaf weeds present.

PAGE 181

Some weeds in the untreated control plots probably died from natural causes, hence the 2.8% mean %-efficacy. Hand weeded plots had late germinating weeds which accounted for the low average %-efficacy. Crop tolerance to herbicide treatments was firs t assessed three days after application. No discernable crop reaction to the treatments was noted at this time or subsequently. The wheat cultivar Mbuni, used in the tria l, appears to have good tolera nce to Monitor treatments. In a separate trial designed to determine the sens itivity of popular wheat cultivars to Monitor, Mbuni was one of the cultivars which tolerated Monitor at 80g ha-1. Mean crop height, plant populations, grain yield and yield components, together with the efficacy of weed control of the treatments are shown in Table 8. There were some significant differences in crop height in plots receiving specific herbicide treatments. However, it is difficult to attribute these differences to treatment effects. For instance, the wheat crop in plot s treated with Monitor at 30g ha-1 was shorter than those in plots that received 40g ha-1 of Monitor. If the crop height depression was dependant on product rate then the converse should have been true. Plots treated with Monitor at 80g ha-1 and its tank mix with Derby at 100mls ha-1 had significantly shorter plants than those from plots treated with lower Monitor rates. The untr eated control plots had the tallest wheat plants. Table 8. Crop height, plant populations, grain yi eld, spike numbers, kernel weights and efficacy of weed control in the third season Wangu Embori Farm, Timau. Herbicide treatments Crop height (cm) Plants/m2 Grain yield (t ha-1) Spikes/m2 TKW (g) %efficacy Monitor 30g ha-1 + Agral 0.25% 91.00 C 180.8 A 3.720 A 379.5 AB 40.95 A 71.40 E Monitor 40g ha-1 + Agral 0.25% 91.25 C 148.3 A 3.478 A 379.3 AB 41.10 A 82.50 C Monitor 20g ha-1 + Agral 0.25% 91.50 BC 186.0 A 3.592 A 429.0 A 40.70 A 75.50 D Monitor 40g ha-1 + Derby 50ml ha-1 + Agral 0.25% 92.50 B 152.0 A 3.695 A 389.3 AB 40.35 A 97.00 AB Monitor 80g ha-1 + Agral 0.25% 88.25 D 181.0 A 3.553 A 403.3 A 38.70 A 95.40 B Monitor 80g ha-1 + Derby 100ml ha-1+ Agral 0.25% 88.50 D 159.0 A 3.767 A 425.5 A 38.80 A 100.00 A Monitor 40g ha-1 + Buctril MC 700ml ha-1 + Agral 0.25% 91.25 C 174.5 A 3.717 A 362.8 AB 39.15 A 98.90 A Cossack 300g ha-1 + Agral 0.1% 92.50 B 171.0 A 3.680 A 360.3 AB 39.70 A 99.30 A Untreated check 94.00 A 166.5 A 2.456 B 321.0 B 40.42 A 2.800 G Hand weed check 92.00 BC 169.3 A 3.459 A 355.5 AB 40.50 A 44.42 F C.V. S.E LSD at p=0.05 0.89% 0.04084 1.184 18.22% 15.3909 44.66 18.37% 0.3225 0.9358 13.63% 25.9232 75.22 4.42% 0.8858 2.570 2.78% 1.0659 3.098 Crop emergence counts taken 14 days after planti ng were not significantly different (p=0.05). There were only slight differences in spike de nsity, with untreated plots having the lowest number of spikes (321 m-2), and no significant treatment effects on TKW (p=0.05). Similarly, there were few treatment effects on grain yiel d was only significantly lower in the unweeded check. Monitor (80g ha-1) + Derby (100ml ha-1) gave the highest grain yield (3.78 t ha-1). It was noted that TKW as well as spike density m-2 did not directly influence wheat grain yield (t ha-1).

PAGE 182

Results of the experiment in the three seasons indicate that the activity of Monitor alone on brome grass was dependent on the stage of growth of the weeds and the soil hydrological status. In conclusion, Monitor at 30g ha-1, 40g ha-1 and 80g ha-1 will significantly suppress brome grass if applied before the grass exceeds the four-l eaf stage. But this is only achievable if the target brome is vigorously growing and the so il has sufficient moisture. Derby and Buctril MC have been shown to be compatible w ith Monitor and tank mixes with these products broaden the spectrum of weeds controlled. Care should be taken in the use of Agral 90 as a wetting agent: applications at rates above th e recommended rate are antagonistic as was evidenced in the second season (Timau), when 2% (v/v) of Agral 90 suppressed the crop and caused general chlorosis. At 0.25% (v/v) th e crop reaction attributable to Agral 90 in combination with Monitor was not detectable.

PAGE 183

Control of the Russian Wheat Aphid, Diuraphis Noxia (Kurdijumov) in Wheat Using Systemic Insecticides in Kenya M. Macharia, M. Njuguna and I. Koros KARINational Plant Breeding Research Centre, P.O. Private Bag, Njoro, Post Office Code-20107, Njoro, Kenya. Abstract Wheat crops in Kenya are subject to widespread and heavy infestations of the Russian wheat aphid (RWA). Control by applications of contact insecticide sprays is difficult because the aphid feeds inside tightly rolled leaves. Initial RWA infestations are usually caused by migrant winged aphids settling on wheat crop. Subsequent spread results from the movement of the original colonizing aphids and their offspring. Effective Russian wheat aphid management strategies require control at the initial stage of infestation. In the cu rrent work, four systemic insecticides were screened to determine their effectiveness in controlling RWA. Evaluation tests carried out in Mau Narok (2835 masl) demons trated that using systemic insecticides could substantially avoid RWA damage. In Mau Narok average yield gain was 26.2, 21.9, 19.9 and 13.0 % for Gaucho 350FS, Cruiser 350FS, Metasystox 250EC and Furadan 350ST treatments, respectively, compared with the untreated control. Systemic insecticides are effective in th e control of RWA and such treatment can prevent heavy production losses. Seed dressing treatments with Furadan 350ST was not effective against RWA because of inco mplete protection of the crop. The cost benefit analysis indicated that foliar application of Metasystox 250EC was more beneficial than those of Gaucho, 350FS and Cruiser 350FS. Introduction About 135,000 ha are sown to wheat annually in Kenya (FAO, 2003) with average yields being about 2.3 t /ha. Among the main constraints that limit wheat production in Kenya are the cereal aphids (Wanjama, 1990). Important cereal aphids that attack wheat include greenbug, S chizaphis graminum (Rondani); rose grain aphid, Metopolophium dirhodum (Walker); bird cherry aphid, Rhopalosiphum padi (Linneus); cereal leaf aphid, Rhopalosiphum maidis (Fetch); and English grain aphid, Sitobion avenae (Walker) (Wanjama, 1990; Macharia et al. 1993; Muthangya et al., 1994). Apart from the damage they cause by directly feeding on crops, cereal aphids transmit barley yellow dwarf virus (BYDV) disease as they feed (Wangai and Torres, 1990b). Yield losses of 47% and 27% have been recorded in wheat and barley, respectively (Wangai, 1990a). The Russian wheat aphid (RWA) Diuraphis noxia is a recent introduction in Kenya. Having invaded the country in mid-1995, it has become one of the most important pests of cereal crops, potentially threatening all of the countrys wheat production (Macharia et al., 1997a). Although an important pest, however, D. noxia is poor in transmitting BYDV (Dickey, 1998). All commercial wheat varieties currently being grown by farmers are susceptible to RWA (Macharia et al., 1997a). Current control strategies rely mainly on the use of contact aphicides to kill RWA already established on crops. However RWA feeds inside rolled

PAGE 184

leaves, making control by contact insecticides difficult (Valiulus, 1986; Dickey, 1998 and Peairs, 1998). Russian wheat aphid is causing considerab le losses to wheat in Kenya and crop losses range from 10-100% depending on the stage of infestation (Macharia et al. 1997a). Toxins injected by RWA into the host plant as it feeds cause damage (Summers et a l., 2004). Damage is also often associated with reduc tion in photosynthetic area caused by rolling of leaves when attacked (Valiulus, 1986, and Peai rs, 1998). In addition, crops attacked by RWA tend to retain other cereal aphids longer, since they are able to hide inside the rolled leaves of the attacked crops. In an effort to minimize damage, studies we re conducted to evaluate the effectiveness of systemic insecticides on RWA. Early sown wheat crops, in Kenya, can also be severely damaged by BYDV introduced and spread by migrant winged aphid vectors early in the season. The systemic insecticides will also be evaluated for control of the migrant winged cereal aphids as a preventive measure against BYDV infection. Materials and Methods In 2001/2002, a series of field trials were set out to investigate the effectiveness of systemic aphicides on cereal aphids and BYDV control. The trials were conduc ted in Mau Narok (2835 m asl), a site that is particularly favoured by severe BYDV infections and a composition of different cereal aphid species. Four systemic aphicides were evaluated, namely Gaucho 350FS [imidacloprid], Cruiser 350FS [thiom ethoxam], Furadan 350ST [carbofuran] and Metasystox 250EC [oxydemeton-s-methyl] and untreated control. All the aphicides were assessed for signs of phytotoxicity from crop emergence to 4 weeks and crop vigour was assessed at heading stage. Details of seed-d ressing insecticides admixed with the seed at planting and foliar aphicide evaluated and the rat es of application are indicated in Table 1. Mbuni variety planted at 100 kg/ha was used as the test crop, and was fertilized using DAP at 130 kg/ha. The trial was designed, as a randomized block design with four replicates and the plot size was 1.5 x 6.0m [9m2]. The trials were seeded using Oyjord seed drill. The systemic insecticides Gaucho 350FS, Cruiser 350FS and Furadan 350ST were admixed with the seed before planting. The systemic foliar system ic Metasystox 250EC was applied immediately after early symptoms of RWA infestation were noticed on the wheat crop after scouting. The foliar spray treatment was applied using a kna psack CP3 sprayer system set to emit 200 l/ha-1 spray volume. Aphid severity was determined by counting the number of cereal aphids on five randomly selected plants at 2 weekly interval s for 8 weeks. The incidence of BYDV was assessed at crop heading based on visual symptoms using a 1-9 scale, where 1= no infection and 9= very severe infection (Macharia et al ., 1997b) In addition, crop vigour was also assessed using 1-9 scale [where 1 = not vigorous and 9 = very vigorous]. Harvesting was done using a small plot Hege combine and yields fro m each plot were recorded and adjusted to a standard moisture content of 12.5%. Plot yields were later converted to kg/ha. Analysis of variance [ANOVA] was used to analyse all the data and means separated using Duncans multiple range test (P<0.05) Statistical analysis was done on aphid population means transformed using square root ( x+1) and the means were transformed back to original values. Comparative efficacy of insecticides on cereal aphid populations was calculated as percentage aphid population reducti on compared to the control. The effect of insecticides on grain yields was also assessed as the percentage grain yield increase over the control. In addition, the cost: benefit analysis was calculated for the systemic aphicides using 2003/2004 pricing structure.

PAGE 185

Results and Discussion In Mau Narok, the most abundant cereal aphid species was the Russian wheat aphid, D. noxia followed by M. dirhodum, R. maidis and R. padi. However infestation by last three aphid species was comparatively lower. All the treatments resulted into improved RWA control compared to the untreated control. The best RWA control was achieved by Gaucho 350FS followed by Cruiser 350FS and Metasystox 250EC foliar spray. (Table 1). As reported previously (Macharia et al. 1997a; Macharia et al. 1997b; Macharia et al. 2001) systemic insecticides Gaucho 350FS and Cruiser 350FS were more effective than Furadan 350ST against RWA. Field observations also revealed that applications of Gaucho 350FS, Cruiser 350FS and Metasystox 250EC are not phytotoxic and gave better control of RWA than Furadan 350ST, the standard treatment. No effect on crop growth or marked phytotoxicity by any of the treatment was observed. Previous conclusions (Macharia, 2002) were confirmed that with the exception of Furadan 350ST, applications of Gaucho 350FS and Cruiser 350FS were effective as control measures against RWA. The Metasystox 250E C treatments performed well when applied before the establishment of RWA on the new crop on detection of early symptoms of infestation. These results suggest that applica tions of Gaucho 350FS, Cruiser 350FS, Furadan 350ST and Metasystox 250EC controlled the initial RWA population, thereby preventing further spread of the aphid and improved on th e BYDV control. Similar results have been reported on the control of cereal aphids in Kenya [Muthangya et al., 1994; Macharia et al., 1997b]. The crop vigor was improved by the use of systemic insecticides [Table 2]. The incidence of BYDV ranged from low to moderate during the trial period. All the treatments resulted into adequate BYDV control except fo r the untreated control (Table 2). Field observations revealed that Furadan 350ST was not effective against RWA but was effective against M. dirhodum, R. maidis and R. padi which are good vectors of BYDV Gaucho 350FS and Cruiser 350FS were effective on all species of cereal aphids, therefore resulting in good control of BYDV. The various control options resulted into improved wheat yields. Significantly higher yields were recorded for Gaucho 350FS, Cr uiser 350FS and Metasystox 250EC foliar spray (Table 3). The cost benefit analysis (Table 4) i ndicated that Metasystox 250EC foliar spray application was more beneficial, with cost be nefit ratio of 1:17.7 than the other systemic insecticides. Gaucho 350FS and Cruiser 350FS were not significantly different, with cost benefit ratios of 1:11.2 and 1:11.3 respectively. Furadan 350ST had the lowest cost benefit ratio of 1:4.6. Table 1. Effect of systemic insecticides on th e control of Russian wheat aphid in Mau Narok Product Rate/ 100 kg seed Rate/ ha Mean no. of RWA/ 5 plants Percentage reduction over untreated control Gaucho 350FS Cruiser 350FS Furadan 350ST Metasystox Untreated control LSD [5%] CV(%) 200 ml 150 ml 714 ml 0.5 l 0.2 a* 0.4 a 3.6 b 0.9 a 8.5 c 1.5 20.61 97.6 95.3 57.6 89.4 *Means in the same column followed by the same letter do not differ significantly at the 5% level.

PAGE 186

Table 2. Effect of systemic insecticides on cro p vigour and control of Barley yellow dwarf [ BYDV] in Mau Narok. Product Rate/ 100 kg seed Rate/ ha Crop vigour [1 9 scale} Barley yellow dwarf score [0-9 scale] Gaucho 350FS Cruiser 350FS Furadan 350ST Metasystox Untreated control LSD CV (%) 200 ml 150 ml 714 ml 0.5 l 7.1c* 7.4c 6.9c 6.0b 4.0a 0.8 10.0 1.2a* 1.6a 2.0a 2.6a 4.2b *Means in the same column followed by the same letter do not differ significantly at the 5% level. Table 3. Effect of systemic insecticides on wheat grain yields in Mau Narok .Product Rate/ 100 kg seed Rate/ ha Mean grain yields [kg/ha] Percentage yield increase over untreated control Gaucho 350FS Cruiser 350FS Furadan 350ST Metasystox Untreated control LSD [5%] CV (%) 200 ml 150 ml 714 ml 0.5 l 4430.0 a* 4379.0a 3967.0c 4288.0a 3508.0d 138.0 25.60 26.2 21.9 13.0 19.9 *Means in the same column followed by the same letter do not differ significantly at the 5% level. Table 4. Cost benefit analysis for systemic ins ecticides on the control of Russian wheat aphid Product Rate/ 100 kg seed Rate/ ha Cost/ ha [K.shs] Yield increase [bags/ha] Value* [K.shs] Cost: Benefit ratio Gaucho 350FS Cruiser 350FS Furadan 350ST Metasystox 200 ml 150 ml 714 ml 0.5 l 1,400.00 1,320.00 1,713.00 750.00 9.2 8.8 4.6 7.8 15,640.00 14,960.00 7,820.00 13,260.00 1: 11.2 1: 11.3 1: 4.6 1: 17.7 *Price of wheat at K.shs 1,700.00 per 100kg bag. The price of systemic insecticides for 2003/2004 cropping season was: Gaucho 350FS @ K.shs 7,000.00 per litre. Cruiser 350FS @ K.shs 8,800.00 per litre. Furadan 350ST @ K.shs 2,400.00 per litre. Metasystox @ K.shs 1,500.00 per litre. Conclusion Adequate RWA control was achieved by the u se of systemic insecticides, Gaucho 350FS, Cruiser 350FS and Metasystox 250EC. Seed treatment with Gaucho 350FS and Cruiser was effective in the management of RWA and improved grain yields. Foliar spray with Metasystox 250EC increased grain yields, but its satisfactory control depends on early detection of infestation through periodic scouti ng. Furadan was effective against other cereal

PAGE 187

aphids (S. graminum M dirhodum, R. padi R. maidis and S. avenae ) but not RWA. This approach will therefore offer a cheaper strategy for cereal aphid and BYDV control. Acknowledgement -We acknowledge Kenya Agricultural Re search Institute for financial support in carrying out this work. References Dickey, E.C. (1998). Russi an Wheat Aphid: Neb Guide In: Ins ects and Pests. C-31, Field Crops. Institute of Agriculture and Natural Reso urces. University of Nebraska. FAO (2003). Wheat production in Kenya. FAO Crop Protection Compendium. CAB International 2003, Wallingford, Oxon 0X10 8DE, UK. Macharia, M., P.M. Muthangya, A.W. Wangai and J.K. Wanjama. 1993. Barley Yellow Dwarf Virus Report 1990-1993. Kenya Agricultural Research Institute, National Plant Breeding Research Centre, Njoro, Kenya. Macharia, M., Muthangya P.M. and Wanjama J.K. (1997a). Management of Russian wheat aphid, Diuraphis noxia, on wheat in Kenya by use of seed dressing insecticides. African Crop Science Conference Proceedings Vol. 3. pp. 1191-1198. Macharia, M., Muthangya P.M. and Wanjama J.K. (1 997b). Barley Yellow Dwarf Virus Report 19941997. Kenya Agricultural Research Institute, Na tional Plant Breeding Research Centre, Njoro, Kenya. Macharia, M. P. Muthangya and J. K. Wanjama. (2001) Effectiveness of Foliar applied aphicides in the control of cereal aphids and barley yellow dwarf. Egerton University, Faculty of Agriculture/KARI-NPBRC Symposium, 14-15 Nove mber 2001, ARC, Eg erton University. Macharia, M., J.N. Malinga and M.G. Kinyua [2002 ] Occurrence of Russian Wheat Aphid in Kenya: The problem and control strategies. Egerton University/ KARI -NPBRC Symposium, 2526 November 2002, NPBRC-Njoro. Muthangya, P.M., Migui, S.M., Macharia, M. & Wa njama, J.K. (1992). Survey of cereal aphid predominance and barley dwarf virus disease inci dence in wheat and barley growing areas of Kenya. Seventh Regional Wheat workshop for Eastern, Central and Southern Africa. Nakuru, Kenya. CIMMYT. Muthangya, P.M., Migui, S.M., Macharia, M. & Wa njama, J.K. (1994). Ef fectiveness of Gaucho [imidacloprid] seed dressing and foliar aphicide to control cereal aphids and barley yellow dwarf virus disease on barley in Kenya. Ninth Regional Wheat Workshop for Eastern, Central and Southern Africa. October 1995. Addis Ababa, Ethiopia. Peairs, F.B. (1998). Russian Wheat Aphid Management. In: The Inaugural National Wheat Industry Research Forum Proceedings .San Diego, California. January 14 and 15, 1998. Summers, C.G., Godfrey, L.D. and Gonzales, F. (2004). Russian Wheat Aphid. UC IPM. Pest Management Guidelines: Small Grains, UC ANR Publication 3466. Statewide IPM Program, Agriculture and Natural Resources University of California. Valiulus, D. (1986). Russian Wheat Aphid: A new pe st that may be here to stay. Agrochemical age 30,10-11. Wangai, A.W. (1990 a). Effects of barley yellow dw arf virus on cereals in Kenya. Pages 391-393. In P.A. Burnet (Ed.), World perspectives on Barley yellow dwarf virus. Mexico CIMMYT, D.F. Wangai, A.W. and E. Torres. (1990 b). Barley yellow dwarf virus situation report for Eastern Africa with special emphasis in Kenya. Page 71 In: P.A. Burnnet (ed.), World perspectives in Barley Yellow Dwarf. CIMMYT, MEXICO, D.F. Wanjama, J.K. (1990). Ecology of cereal aphids tr ansmitting Barley yellow dwarf virus in Kenya. Pages 240-243. In: P.A. Burnnet (ed.), World perspectives in Barley Yellow Dwarf. CIMMYT, MEXICO, D.F.

PAGE 188

Part 4. ECONOMICS

PAGE 189

Analysis of Marketing and Pricing Policies on Technology, Input Use and Production of Wheat in the Sudan Abbas Elsir M. Elamin Agricultural Research Corporation, Socioeconomic Research Program, Wad Medani, Sudan Abstract This paper discusses the situation of wheat production with respect to its financial and economic profitability and international competitiveness in The Sudan. It analyses government reform policies and their role in wheat production, consumption and input use in the 1990s. Policy Analysis Matrix (PAM) was used to measure the domestic resource cost ratio to international value added, nominal and effective protection coefficients. Results indicated that wheat production utilized the countrys resources inefficiently, thus enjoyed no competitive position during the last decade. The implementation of economic liberalization polic y and removal of subsidy in 1990s has negatively affected the adoption of improved technology and led to low crop productivity. The government policy has also resulted in escalated prices of tradable inputs, and consequently, the cost of production. During the last decade wheat prices did not increase proportionately with production costs, leading to lower returns and an accumulation of debt for the majority of wheat producers in Sudan. Introduction Since the mid-seventies, most sub-Saharan Afri can countries have been experiencing decline in economic performance. The economic decline in Africa has largely been attributed to poor domestic policies which resulted in economic dist ortions rather than to external factors (Ogbu, 1991). As in most African countries, Sudan embarked on Structural Adjustment Program aimed at eliminating the economic distortions, adapting the economy to market forces, and improving the efficiency of resour ce allocation and utilization. The adjustment process involved the realignment of the exchange rates, changes in relative prices, improving incentive structures and institutional reforms. The agricultural sector is perhaps one sector that has been extensiv ely affected by the adjustment process. Agricultural marketing has al so been deregulated and liberalized with the abolition of marketing boards. Agricultural i nputs including fertilizers are witnessing the desubsidization since 1992 country-wide. Fertilizer procurement, distribution and pricing policy reform is intended to increase the consumpti on of the commodity and thereby raise the productivity of the farmers and aggregate agri cultural outputs. However, even with the removal of subsidy, the total consumption of fe rtilizers in Sudan is relatively low compared to its large endowment of 85 million hectares of arable land (Table 1). It is worth noting that the agricultural sector supports over 75 percent of the population, contributes to over 90 percent of the foreign exchange earnings before oil extraction and 30 45% of the GDP (Appendix1). Wheat is the main staple cereal in Sudan especially in urban areas and second to sorghum in many irrigated rural areas. It is traditionally produced and consumed in northern Sudan. However, since the early 1960's, its production has expanded into the large public projects of central Sudan, which contribute 90% to dom estic production since mid nineties. Most of production decisions in the public projects are de termined by the government, but wheat has received greater attention in the food sector. The government intervenes at all stages of

PAGE 190

production, processing and distribution This is largely because wheat constitutes the staple food for the urban population which has strong political power to influence bread prices. Table 1. Quantities and values of imported fertilizers in Sudan, 1994 998. Year Fertilizers quantity (tone) Value (000 US$) 1994 1995 1996 1997 1998 58,586 47,397 261,447 183,838 70,882 12,726 15,745 74,711 47,349 10,624 Source: Bank of Sudan annual reports. Sudans domestic wheat production has always r un short of satisfying the rising domestic demand. The average per capita consumption increased from 10.5 kg in 1960 to 20.4 kg in 1971 and from 31.7 kg in 1986, to a bout 40 kg in 1999 (Fig. 1). Wheat consumption has been increasing sharply and the gap between cons umption and domestic production levels has widened (Fig. 2 and Appendix 2). The cont ribution of the domestic production to wheat demand has been fluctuating, ranging from 97% in 1992 to about 12.5% in 1999 and 17% in 2001 (Elamin, 2002). This has been mainly du e to faster growth in demand and drastic reduction in domestic production, as a result of reduction in wheat acreage in central Sudan especially in the Gezira project, which used to be the main producer, coupled with a reduction in the productivity of the crop (Table 2). A national wheat program was developed in the late 1990s to make the country selfsufficient in wheat. This program involved hor izontal expansion of wheat production in the middle and high terrace soil of northern Sudan a nd vertical expansion through promotion of improved wheat technologies, mainly improved seeds and nitrogenous-phosphorus. The liberalization policies implemented by the government affected the adoption of these technologies in different ways. Table 2. Wheat cultivated area, pr oductivity and production in the Gezira project, 1992 2002. Season Area grown (acres) % to total ar ea Production (tons) Yield (ton/acre) 1991/92 532,813 58.7 499.779 0.94 1992/93 514,033 64.4 629,868 0.53 1993/94 522,783 85.6 273,938 0.52 1994/95 392,690 55.9 230,116 0.59 1995/96 390,490 56.5 256,552 0.66 1996/97 389,801 50.1 250,642 0.64 1997/98 301,925 50.2 211,348 0.70 1998/99 123,016 37.9 37,766 0.31 1999/2000 58,627 28.3 29,314 0.50 2000/2001 70,409 26.6 56,327 0.80 2001/2002 80,818 26.5 68,695 0.80 Source: SGB, Department of Statistics, MANRAW and Socio-Economic Research. Objectives and Methodology The objectives of this study were to (i) examin e the situation of wheat production with respect to its financial and economic profitability and international competitiveness when using domestic resources; and (ii) analyze the govern ment economic policies and their role in the adoption of wheat technology, production and consumption in the Sudan. Firstly, the study hypothesized that the refo rms have not significantly made fertilizers and seeds available and accessible to farmers at a ffordable prices. Secondly, that there is no significant impact on the productivity of the farms at the microlevel. That is, farmers do not

PAGE 191

readily have access to fertilizers and seeds, and where available, they cannot afford the prices offered and do not actually use fertilizers and or seeds, and where they do, the quantities used are not adequate as to make noticeable impact on production. Financial and economic eval uations were conducted using the theory of comparative advantage and policy analysis matrix. Domestic Resource Cost (DRC) analysis was employed as a procedure for tracing the competitiveness of wheat production over a lternative productive uses of Sudans resources. DRC ratio indicates whether it is cheaper for a region or a country to produce wheat locally or import it. The method can also be used to rank production alternatives showing comparative advantage ove r trade options. The crop activity with the biggest comparative advantage is the one which has the most efficient use of the region's or country's resources. The results from these analyses are used to propose appropriate intervention and strategy for the expansion of wheat production in northern Sudan. Both primary and secondary data were collect ed to achieve the stated objectives. Results and Discussion Production policy Wheat is entirely grown under irrigation in the public projects, mainly in central Sudan, or in private pump projects in northern Sudan. The irrigated public projects, where the cultivators are tenants of the government, are characterized by fairly extensive use of modern inputs such as machinery, fertilizers, pesticides and seeds. These inputs in addition to irrigation water are provided on credit basis. The tenants are res ponsible for all manual work and obliged to follow the prescribed cropping pattern determined by the project administration. On the other hand, private irrigated farms are owned and managed by small semi-commercial families. According to Elamin (2001), Sudans local wheat production has always run short of domestic demand. Wheat consumption has risen sharply, increasing the gap between consumption and local produc tion (Appendix 2). Excessive consumption pattern requires rationalization and regulation that positively contributes to narrowing the wheat consumption gap. Most of the wheat consumed in Sudan is imported but increased wheat imports are likely to erode the country's balance of payments. For instance, in 1998 wheat imports reached 984,000 tons with total value of US $ 132 million or 22 percent of the countrys total foreign exchange earnings and 77% of the total value of agricultural exports (Tables 3 and 4). Table 3. Wheat and wheat flour import in the Sudan, 19902001. Year Amount (000 ton) Value (million US $) 1990 132.573 21.762 1991 361.134 74.809 1992 213.334 31.195 1993 212.449 45.795 1994 488.127 112.928 1995 307.810 89.847 1996 354.345 97.859 1997 576.623 138.401 1998 984.108 131.945 1999 549.483 123.333 2000 1013.400 207.942 2001 650.232 138.096 Source: Ministry of External Trade and Bank of Sudan.

PAGE 192

Table 4. Value of wheat import and agricultural export, 1997-2000. Year Wheat and wheat flour imports (US $) Agricultural exports value (US $) % 1997 138.402 133.372 103.7 1998 131.945 171.370 76.9 1999 123.333 142.566 86.5 2000 207.942 91.187 228.0 Source: Bank of Sudan, Annual Reports. The increased wheat consumption may have enc ouraged the policy makers to increase public resources devoted to wheat production. However, successful implementation of this policy shift could only occur if domestic resources are being used efficiently and farmers find it profitable to grow more wheat. This is possi ble if and only if wheat is using domestic resources efficiently both financially and econom ically at the farm and national levels, respectively. The pricing policies have played a negative role in promoting domestic wheat production. Wheat farmers used to receive an indirect subsidy on the price of imported inputs (fertilizer and fuel) but at the same time they had to de liver their produce at substantially lower prices than local market clearing prices. This policy taxes wheat producers. The economic liberalization policy and removal of subsidy in the 1990s has resulted in increasing the price of tradable inputs and consequently the costs of production. The increase in wheat price did not offset the effect of rising costs of pr oduction (Table 5). Consequently, some wheat producers have fetched low returns and the majo rity of them accumulated debts during the last decade (Table 6). When the national economy adopted the free market system, the government withdrew from direct financing of the agricultural sector. A consortium of commercial banks was formed to serve as a financial resource fund le nding production and marketing of agricultural commodities. They mostly cater for crops grown in the irrigated sub-sector but the costs of lending have been found to be high. The farmers union with the assistan ce from the government has acquired the Commercial Bank of Sudan and renamed it the Farmers Bank. The farmers union also acquired the Cotton Company, formerly government cotton marketing owned company. These two institutions have considerably in creased farmers' access to credit and other financial resources. However, the huge financ ial services required by farmers are beyond the capacity of the Farmers Bank and the Cotton Co mpany. Moreover, farmers financial base is poor and insufficient to finance their agricultural operations. Most of the agricultural operations in the pub lic projects in central Sudan (e.g. the Gezira and Rahad) are provided by the project administration with costs paid after harvest or deducted from the farmers individual account (from his crop receipts). In this production relationship, farmers often get credit to be paid in form of produce at an agreed upon price significantly below the expected market price. Labor and incidental costs are usually met from the farmer own resources. When the harvest is poor, the farmer may go bankrupt. For the private projects, the credit situation is even worse where failure of repayment is prominent. Farmers in these private projects express their dissatisfaction with the amount, timing and conditions of credit which is not co nducive for adopting technology and sustaining production. Again, after liberalization of the mark et for grains, imported wheat out-competed domestic wheat, with a relative imported price of 0.77 of the local wheat price in 1999. Therefore, high production costs, low productivity and less competition with imported wheat

PAGE 193

have made wheat an unattractive crop. Eventu ally, wheat yields in the Gezira project decreased from 0.94 ton/acre (2.2 ton/ha) in 1992 to 0.31 ton/acre (0.7 ton/ha) in 1999. In Rahad project, they have declined by 575 percen t over the same period. This situation has led to low or even negative wheat profits and he nce increased indebtedness of the farmers during the last decade. As a result, wheat production has been discontinued in the Rahad project and the area under wheat drastically reduced in the Gezi ra project to only 11 percent of the area in 1992. Wheat became an optional crop enterp rise starting from 2000 season (Elamin, 2001). Countrywide, the total wheat cultivation has declined to about 123,000 hectares (294,000 acre) and production to 266,000 tones, representing only 17 percent of total wheat consumption (Appendix2). This widening ga p between supply and demand can not be significantly reduced in the very short run. To bridge the consumption gap, the government has two options: to rationalize wheat consumption by reducing wheat import or to rely again on a crash program for promoting wheat production in its traditional areas. Since the latter option should reduce pressure on the limited foreign exchange available to Suda n, the government launched a national wheat production program in late 1990s with the objec tives of restoring areas of wheat cultivation needed for self-sufficiency thro ugh horizontal expansion in th e middle and high terrace soil of northern Sudan. This comes through the rehab ilitation of the existing projects, establishing new projects and amalgamating small ownership into large scale farms that economize irrigation water. The target areas are 184,000 and 113,000 hectares in the Nile River and Northern States, respectively. This program exte nds for five years with a total budget of US$ 440 million to be secured by the Agricultural Bank of Sudan (ABS). Some serious questions against the successful completion of this wheat relocation program may arise. Private profitability, soil reclamation, and provision of water to distant areas from the Nile are critical at the farm level. At the national level, comparative advantage in terms of domestic resource costs is a per tinent issue. To utilize the countrys resources efficiently, local wheat must generate forei gn exchange earnings (savings) that exceed the value of traded inputs used in its production. Efficiency indicators In the central Sudan (Gezira project) where wheat used to contribute more than 60% of the Sudan wheat production during early 1990s and 10% in early 2000, wheat is grown in rotations with cotton, sorghum, groundnuts and to some extent vegetables and fodder with more or less similar inputs levels. Despite these si milarities, wide crop yields variations exist due to variability in irrigation-water, climate, soil types, weeds intensity, and the levels of farm management. As a result of recent economic changes following the adoption of liberal economic policies and changes in different agricultura l policies, production returns, comparative advantages and competitiveness of different crops were affected at varying degrees (Elamin, 2002). This resulted in an increase in prices of tradable inputs and consequently an increase in the cost of production. Product prices have also increased but by lowe r percentage than the increase in cost resulting in a decrea se in returns to the producers. Fertilizer price index in the Gezira Scheme has increased from 100 in 1990 to 48,249 in 1999 that is 482 times during the period 1990-1999. By contrast, increases in wheat price indices are only 16,000 during the same period. In other words, between 1989 and 1999, the price index of fertilizer has moved by as high as 3 times that of wheat. This means that nominal fertilizer price index has increased much more compared to that of wheat (Table 5).

PAGE 194

Table. 5. Index of wheat price, cost, return and fertilizer price in the Gezira project, 1990-1999. Season 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Producer Price (Ls/t) Nominal Net Return Wheat Cost (Ls/fed) Fertilizer Price (Ls/t) Wheat Price Index Net Return Index Wheat Cost Index Fertilizer Price Index 3000 903 960 1340 100 100 100 100 6000 862 2110 3340 200 95 220 249 8750 3164 5015 8040 292 350 522 600 19000 -1872 8710 26160 633 -207 907 1952 70000 5580 18598 50720 2333 618 1937 3785 75125 10433 34567 105707 2504 1155 3601 7889 220000 82628 86552 249700 7333 9150 9016 18634 380000 65721 191479 581720 12667 7278 19946 43412 380849 27772 238228 646540 12695 3076 24815 48249 480000 -97591 227191 646540 16000 -10807 23666 48249 Source: Calculated by the author. Table 6. Wheat productivity, average cost, net retu rn per feddan in the Gezira project, 1993 1999. 1993 1994 1995 1996 1997 1998 1999 Productivity (ton/acre) 0.525 0.524 0.586 0.657 0.643 0.7 0.31 Break-even point 0.667 0.403 0.450 0.336 0.479 0.627 Average cost/acre. 8710 18598 34567 86552 191479 241225 244951 Total revenue/acre 6838 24178 45000 169180 257200 266594 147360 Net revenue/acre -1872 5580 10433 82628 65721 27772 -97591 Rate of net revenue (%) -21% 30% 30% 95% 34% 12% -39% Source: Sudan Gezira Board

PAGE 195

The product/input price ratios for wheat in the Gezira project during the last decade is calculated based on productivity per unit area k eeping inputs used other than fertilizers constant. The wheat/nitrogen price ratio used to be high at the export market indicating its high comparative advantage, in terms of domestic resource costs, and in utilizing the imported fertilizer using foreign exchange capital. Neve rtheless, the crop product/input price ratio is deteriorating over time from early 1990s on wards. This serious deterioration of this crucial measurable economic and pricing indicator is behind the observed fluctuating cultivation of this crop in the Gezira project in the late 1990s (Table 7). The effect of government policies on comp arative advantage for wheat were assessed using PAM which identifies the divergences be tween existing market (private) values and social (real) optimal values. Results in Table 8 showed that nominal procurement prices offered by the government reflected taxation (neg ative transfers) to producers, who were at the same time enjoying subsidies on tradable a nd domestic factors. This is reflected by the Effective Rate of Protection (ERP), which indicates the net level of transfers. It shows high net positive protection (implicit subsidy) with a coefficient of more than one (1.3) during 1996 season. This means that, at this price level, the effect of agricultural policy is to increase producer benefit. While for the period 1997 2000, this coefficient has a value of less than one which means that the effect of the govern ment policy is to reduce producer benefit. The nominal protection coefficient (NPC), which shows the divergences between social and private values, indicates that private prices were lower than social ones, which implies taxed output (its value is less than one du ring 1996 2000). The domestic resource cost (DRC) ratio assesses social returns to domestic factors and serves as a measure of relative efficiency of domestic resource use. The yards tick of comparative advantage is a DRC ratio of unity, which is a break-even point. DRC values of less than unity imply comparative advantage, and greater than unity values indicat e no comparative advantage at the given levels of yield and prices. This, therefore, indicates th at it is more efficient to import wheat and use domestic resources in more competitive crops. It is obvious that even with the escalated cost of production; wheat had a comparative a dvantage during 2000 2 002 growing seasopns, due to the improvement in wheat productivity. Improvement in wheat productivity is due to the fact that wheat cultivation is confined to areas where the soil is more productive and farmers are financially capable as wheat became an optiona l crop in central Sudan. It is worthy noting that low productivity before 2000 season coupled wi th escalated cost of production resulted in high value of the domestic resource cost and thus the crop enjoyed no comparative advantage. These results indicate that the agricultural reform program designed by the government to promote greater self-sufficiency in wheat c ould be achieved if more protection is provided for farmers. This comes through reducing c onstraints that hinder the production process, maintaining low cost of production and ensuring marketing of produce at competitive prices. In addition, provision of improved seeds, fertili zers, irrigation water after the establishment of the proposed dams in the Northern States and full mechanization of the crop may act to lower the costs of production. This when coupled with the program of soil reclamation should improve wheat competitiveness, at both the farm a nd national levels, and pa ve the road for the dream of self-sufficiency in wheat.

PAGE 196

Table 7. Wheat product and fertilizer input price ratio, Gezira project, 1990-1999. Season Yield (ton/acre) Wheat price (Us $ /tone) Wheat price (SD/tone) Wheat value (SD/acre) Fertilizer cost (SD/acre) Product-input price ratio 1991/92 0.94 291 875 823 95 8.66 1992/93 0.53 139 1,900 1,007 230 4.38 1993/94 0.52 210 7,000 3,640 609 5.98 1994/95 0.59 179 7,500 4,425 1,080 4.10 1995/96 0.66 271 22,000 14,520 2,578 5.63 1996/97 0.64 150 2,000 14,080 6,672 2.11 1997/98 0.70 234 40,000 28,000 6,579 4.26 1998/99 0.31 188 45,000 13,950 7,221 1.93 1999/2000 0.50 233 60,000 30,000 5,847 5.13 2000/2001 0.80 213 55,000 44,000 6,936 6.34 2001/2002 0.80 171 45,000 36,000 7,066 5.09 Table 8. Results of wheat comparative advantage, 1996 2002 Season NPC ERP DRC 1995/96 0.56 1.3 1.12 1997/98 0.81 0.68 0.91 1998/99 0.6 0.5 2.4 1999/00 0.81 0.7 0.97 2000/01 0.7 0.55 0.4 2001/02 0.81 0.79 0.8 Procurement of seeds and fertilizers For improved seeds, the breeder seeds are produced by the Agricultural Research Corporation whereas the foundation and registered seed s are produced by the Seeds Propagation Administration (SPA) of the Ministry of Agricultural but their production was discontinued because of poor budgets and unavailability of breeder seeds (Elamin, 2002). Production of improved seeds has faced by many difficulties th at resulted in unavaila bility of these seeds, and when available their prices are high relative to local seeds. Sudan consumes about 200 thou sand metric tons of fertilizers, mainly in the forms of nitrogen and phosphorus; it is used in both pr ivate and public agricultural sectors, annually. This amount of fertilizers costs the economys balance of trade a sum of US $60 million per year. Generally, the use of chemical fertilizers is mostly confined to national irrigated projects such as Gezira, Halfa, Rahad and Suki and a number of privately owned irrigated commercial farms in northern parts of Sudan. Urea fertilizer is usually used by public irrigated projects in more than 90 percent of the area at varying levels but the recommended rate is 190 kg per hectare for wheat. However, in the Northern States of Sudan, where wheat is traditionally irrigated, the rate of urea and supper phosphate application is still low. However, the use of these fertilizers is dependent on the availabilit y and accessibility to credit financing. Before 1992 and during the seventies, urea fertilizer was acquired through bilateral cooperation between Sudan and Kuwait and gove rnment of Japan and the Netherlands. In 1980 the European Foundation for Development and the international Development Agency (IDA) were also engaged in financing agricultu ral inputs. After international financial aid declined, Sudan resorted to financing importation of agricultural inputs on its own. The system of Consortium Banks was introduced to finance irrigated public projects. These banks were responsible for providing local currency to agricultural projects whereas the Bank of Sudan provided the foreign currency. Neverthe less, and because of the inexperience of Banks in the newly introduced forms of finance and the high costs of finance through the

PAGE 197

Consortium of Banks, the agricultural projects resorted to a system of deferred payment through forward sale of cotton to finance agri cultural inputs. The new system resulted in several problems which hampered the efficient distribution of inputs. In the private wheat sector, financing of wheat is based on farmer's individual collateral, which includes personal guarantees, the tenancy itself, and farmers un ions guarantees. This method of finance works as a disincentive for technology transfer reduced costs, increased productivity and production. Impact of technology adoption Technologies, if adopted are expected to have impacts on producers, consumers and the society as a whole. Technology impact could be measured by several indicators at the farm and national levels. The need for improve d and modern farm technologies to increase agricultural production in Sudan has been recognized and highlighted in many studies since the early 1980s. Implementation of the libera lization policies has negatively affected the adoption of recommended technologies. Removal of credit facilities had an adverse effect on farmer financial capabilities. This is coupled w ith the removal of subsidies has resulted in high cost of essential inputs and consequently lo w adoption rates of technology. These factors have played a major role in lowering the productivity of wheat crop. Low productivity of the cultivated crops coupled with high cost of production inevitably entailed low and unstable farm incomes and deteriorating standa rd of living for the rural households. At the farm level, the real impact attri buted to the adoption of technology could be observed in wheat yield improvement after the adoption of technology. The recommended technology proved to be superior to farmers' traditional practices. Among the most influential components of the package were sowing date, improved seeds and fertilizer application. These components are totally dependent upo n their availability and their prices. According to research recommendation, November is the recommended date for sowing. However, due to engagement of farmers with the cultivation of other crops during November and their tendency to delay sowing with the objective of reducing risk of birds attack if they grow wheat early, the sowing date extended fro m November to January and thus resulted in a wide range of variability in wheat productivit y in the Northern States. About 12% of the farmers cultivated wheat during November while 35% and 53% cultivated wheat during December and January, respectively. It is obvious from Table 9 that farmers who cultivate wheat during November obtain higher yield (4,760 kg/ha) than those who cultivate wheat during December (4,046 kg/ha) and January (2, 858 kg/ha), respectively. This means that delaying the sowing date from November to D ecember and January has resulted in a 15% and 40% reduction in yields, respectively. On the other hand, delaying the sowing date from December to January has resulted in a 29% reduction in yields. Regarding the improved seeds, 60% of the farmers used improved seeds while 40% use local seeds. In the northern Sudan, the adoption rate of the improved varieties in the Northern States is higher than that of the Nile State due to the availability of impr oved seeds coupled with the vital role of the extension service in technology dissemination. In central Sudan, only about 12 percent of the farmers used the recommended fertilizer rate while about 88 percent did not. Fertilizer application varied from 60 to 400 kg/ha. Financial inability was the main problem f acing farmers in using recommended rate of fertilizer. About 11% of the sampled farmers in the Northern States applied the recommended doses of fertilizers while 22% and 55% applied less or more than the recommended dose, respectively (Table 10). Non-adopters of the recommended dose (11%) claimed that fertilizer is not available and is expensive.

PAGE 198

Economic evaluation was conducted to assess the profitability of the package against that of farmers' practices. This is necessary becau se the economic ranking is often different from the agronomic ranking due to differences in factor costs. Whole farm budget was employed using yield and cost data from both groups to appraise the effect of technology adoption on the net benefits between adopters and non-adopt ers of the package. Only costs that vary between the two groups were considered in the analysis. The field price (SD 70 /kg of wheat) was used to derive total benefit which equa ls price times yield. Then net benefit was calculated as the difference between total benefit and the cost of production. The adopters incurred 45% additional costs than non-adopters but obtained 50% yield advantage and 57% increase in net benefits over non-adopters. (Table 11). This means that the additional cost of the package was more than compensated for by its high yield and hence higher returns. To relate the benefits accrued to costs incurred, a marginal rate of return (MRR) was calculated and found to be 181%. This means that with adoption of the package, a farmer is expected to retain every Sudan ese Dinar (SD) invested in the package plus additional SD 1.8. Total yield advantage in the Northern Sudan was estimated at about 228,245 tones of wheat as a result of technology adoption. In terms of monetary value, the total incremental benefit as a result of tec hnology adoption in wheat crop will be about 3 billion Sudanese Dinar (Elamin, 2004). Table 9. Adoption of wheat sowing date and their respective yield and area in the Northern State, 2004. Sowing date Cultivated area (ha) Percentage Average yield (kg/ha) % reduction in yield During November 6,237 12% 4,760 During December 19,049 35% 4,046 10 % During January 28,948 53% 2,858 40% Total area 54,234 100% 3,887 Table 10. Adoption of fertilizer applic ation in the Northern State, 2004. Sowing date Percentage Apply the recommended dose 11% More that the recommended dose 55% Less than the recommended dose 22% Do not apply fertilizers 11% Table 11. Cost of package and no ne package winter crops, 2004 Crops Package Farmers practices Differences % difference Cost of production 147,860 101,788 46,072 45 % Yield (kg/fed) 3,572 2,381 1,191 50 % Total return 250,005 166,670 83,335 50% Net return 102,145 64,882 37,263 Marginal rate of return 181 Total yield advantage in Northern Sudan (tone) 228,245 Total net return for wheat area in Northern Sudan ( Billion SD) 3 Source: Elamin, 2004

PAGE 199

Conclusion Agricultural crop pricing policies followed by th e Government of Sudan, during the last decade, have had detrimental effects in prom oting crop production and meeting the stated food security goal. The unfair pricing policy coupl ed with an inefficient fertilizer procurement and distribution policy have led to deterior ating output/input ratios for wheat crop and declining production at the aggregate levels. This situation has resulted in deterioration of the comparative advantage of wheat in central Suda n. Hence wheat acreage has been reduced and production shifted to northern Sudan. After liberalization of the market for grai ns in 1999, imported wheat price out-competed domestic wheat price by a large margin of 23 % mainly due to taxation of output and the removal of subsidies. Consequently, high costs of production, low productivity and less competition with imported wheat have made wh eat an unattractive crop in Sudan. A serious program for relocation of wheat from its original areas of production in the River Nile and Northern States is currently underway. However, for this agricultural reform program to be successful it needs more protection in terms of reduced constraints to accessing fertilizer, seed and water inputs, reduced costs of production and marketing at competitive prices. The results of the economic analysis showed th at substantial benefits were realized through the adoption of the recommended package. At the farm level, adoption of improved technology resulted in 50% yield advantage and 57% increase in net benefit over the traditional practices. At the state level, it r esulted in about 228 thousand tones yield advantage and about 3 billion SD as an incremental benefit. The government withdrawal from direct fina ncing of the agricultural sector after the orientation of the national economy towards fr ee market economy coupled with poor financial capability of farmers to finance the agricultural operations has affected farmers behavior towards input use (especially improved seeds and fertilizer) and his performance in conducting different agricultural ope rations. This calls for the government to revisit its input and credit subsidy policy and support to extension services if the dream of self-sufficiency in wheat is to be realized.

PAGE 200

Appendix 1. Contribution of agricultural sector in the annual gross do mestic production current producer prices in million LS. Year GDP total Agricultural Sector contribution % 1980 4792.7 1650.5 34.4 1981 6217.9 2314.7 37.2 1982 6330.8 2915.8 35.0 1983 10403.1 3641.1 35.0 1984 11777.0 3650.9 31.0 1985 16425.3 5748.9 35.0 1986 20763.8 7059.7 34.0 1987 36471.0 12764.9 35.0 1988 46791.0 15909.0 34.0 1989 82562.0 27245.4 33.0 1990 100863.0 30545.6 30.2 1991 190827.1 54701.2 28.7 1992 401840.0 136087.0 33.9 1993 857477.0 326834.0 38.1 1994 2386330.0 973571.0 41.1 1995 4233912.0 1776417.0 43.0 1996 10215174.0 4594675.0 45.0 1997 15929308.0 759828.0 47.6 1998 19916123.0 9699153.0 48.7 1999 24488851.0 12195447.0 49.8 2000 2969452.4 624.4 46.4 2001 3380555.0 653.7 45.6 Source: MOF and Bank of Sudan, Annual Reports. Appendix 2. Cultivated area, production, consum ption, food gap and self-sufficiency ratio. Source: Ministry of Agriculture and Forests. Ministry of External Trade. Year Area (000 acres) Production (000 tons) Consumption (000 tons) Gap (000 tons) Rates of production to consumption (%) 1980/81 437 218 550 332 39.6 1981/82 329 142 580 438 24.4 1982/83 225 176 610 434 28.8 1983/84 335 157 640 283 24.5 1984/85 111 79 670 591 11.7 1985/86 36 199 700 501 28.4 1986/87 282 157 730 573 21.5 1987/88 34.3 181 740 559 24.4 1988/89 393 247 780 533 31.6 1989/90 614 409 850 441 48.1 1990/91 1104 686 870 184 78.8 1991/92 930 895 920 25 97.2 1992/93 805 445 976 525 45.8 1993/94 881 492 1020 528 48.2 1994/95 741 498 1070 572 46.5 1995/96 774 575 1130 555 50.8 1996/97 782 640 1200 560 53.3 1997/98 646 535 1200 675 44.5 1998/99 405 172 1370 1198 12.5 1999/2000 244 250 1460 1210 17.1 2000/2001 294 266 1500 1234 17.7

PAGE 201

Fig. 2. Wheat production, consumption and self-sufficient ratio (SSR) 1981-2001 0 200 400 600 800 1000 1200 1400 16001980/81 83 85 87 89 91 93 95 97 99 2001Year(000) tone0 20 40 60 80 100 120SSR production Consumption SSR 0 5 10 15 20 25 30 35 40Per capita consumption 1960196819711983199319961999 YearFig. 1 Per capita annual wheat consumption in The Sudan selected years

PAGE 202

References Elamin, Abbas Elsir Mohamed (2001). Constraints of Wheat Production in Central and Northern Sudan. NVRSRP Newsletter, the International Ce ntre for Agricultural Resources in the Dry Areas (ICARDA), Cairo, Egypt (i n press ). Elamin, Abbas Elsir Mohamed (2001). Evolution of Wheat Production Policy in the Sudan. NVRSRP Newsletter, the International Centre for Agricultural Resources in the Dry Areas (ICARDA), Cairo, Egypt (i n press ). Elamin, Abbas Elsir Mohamed (2001). Economic Situation of Wheat Production in Central and Northern Sudan, the National Coordination Meeting of the NVRP, Wad Medani, Sudan, August 2001, Annual Report of the ARC, 2001. Elamin, Abbas Elsir Mohamed (2002). Constraints of Wheat in the Gezira and Northern States, the National Coordination Meeting of the NVRP, Sudan, Khartoum, Aug. 2002, Annual Report of the ARC, 2002. Elamin, Abbas Elsir Mohamed (2002). Evaluation of Farmers Attitude and Input Use on Wheat Productivity in the Northern State, the National Coordination Meeting of the NVRP, Sudan, Khartoum, Aug. 2002, Annual Report of the ARC, 2002. Elamin, Abbas Elsir Mohamed (2004). Economics of Wheat Production in Sudan. A paper to be presented in the National Coordination Meeti ng of the Nile Valley and Red Sea Regional Program ICARDA, Wad Medani, Sudan, Aug. 2004. Elamin, Abbas Elsir Mohamed (2004). Monitoring and Evaluating Improved Wheat Production Technologies in Northern Sudan. A paper to be presented in the National Coordination Meeting of the Nile Valley and Red Sea Regional Program ICARDA, Wad Medani, Sudan, Aug. 2004. Ogbu, O (1991). Structural Adjustment and Agricultural Supply Response in Sub-Saharan Africa: Synthesis and Limits of Current Analysis.Research for Development Vol. 6 No. 2 pp 1-2.

PAGE 203

Opening Speech 12th Regional Wheat Workshop for Ea st, Central and Southern Africa Merica Hotel, Nakuru 22-26 November 2004 by Hon. John Koech, Minister for East Afri can and Regional Cooperation, Kenya Participants, distinguished guests, ladies and gentlemen: It gives me much pleasure to preside over the opening of the 12th Regional Wheat Workshop for East, Central and Southern Africa here in Nakuru. I acknowledge the presence of our international visitors whom we wish a pleasant stay and a fulfilling experience in the short period they will be in the country. On behalf of the Government and the people of Kenya and on my own behalf, I wish to welcome you to this country and in particular to this workshop. We would however, wish that you take some time after the workshop to savour and experience some of the attractions this country offers that makes it one of the premier destinations for tourists from all over the world. The task that you have before you in the comi ng days of the workshop is to share your research findings with one another for the benefit of the people in the region and beyond. As you deliberate on the many scientif ic issues impacting on agriculture in the region, I would wish to remind you that the adoption of the technologies that you develop is what will impact positively on farm productivity and not the generation of the technologies per se It is therefore important that concerted efforts be made to disseminate this information and generated technologies to as many farmers as possible. In this context, I would wish that the future workshops should involve other stakeholders and service providers especially those involved in agricultural extension and if possible farmers. In line with the workshop theme, Development and dissemination of sustainable wheat production technologies for improved livelihoods in sub-Saharan Africa, it is important that technologies developed in the region be shared throughout the region to enhance food security and alleviate hunger. I note with satisfaction that scientists in the region apar t from generating technologies for increased productivity also ensure that these technologies do not impact negatively on the environment and therefore lead to sustainable exploitation of our natural resources. Although maize is still the most important cereal in the region, wheat has become increasingly important due to urbanization and changed dietary habits. Regrettably, none of the sub-Saharan countries is self sufficient in this commodity. In Kenya demand has been growing at 5% per annum and this has not been matched by production. I believe that the situation in other countries represented here is not any different. We as a region face similar constraints as we try to improve food production to sustain our people. It is imperative then that we look at opportunities for improving agricultural productivity together. There is need for us to pool together and strategize collectively on ways of increasing wheat productivity. This forum gives us an excellent opportunity to exchange ideas, experiences and knowledge and ensures that duplication of effort is avoided. This will conserve the meagre resources available. May I briefly talk about the challenges that face wheat production in Kenya and by extension the region. Apart from environmental factors such as drought, the major

PAGE 204

constraint impacting negatively on food production in the region is use of inappropriate production technologies by farmers. Despite development over the years of high yielding widely adapted and disease resistant wheat varieties, adoption of these varieties by farmers has been done. There is therefore need to catalyze the adoption of these varieties by farmers to give impact to wheat production and reduce importation bill for wheat. The potential of high yielding varieties devel oped by the national programs in the region cannot be exploited unless improved managem ent technologies are also adopted by farmers. We must not only develop agronomic packages for wheat production but must vigorously promote their adoption by as many farmers as possible. Critical areas for agronomic research which need your attention are: Land preparation Soil fertility management Seeding technologies Weed control Post harvesting Quality assessment and value adding. Another major concern but which Government s in the region are addressing is provision of suitable marketing structures. To ensure farmers derive maximum benefit from their endeavors, the Government of Kenya has ta ken measures to protect Kenyan farmers from unfair trade practices and ensuring that wheat farming is profitable. In this context there is renewed interest in wheat farming in the last two years. The Government will encourage this trend by making marketing structures in place and improving them. Ladies and gentlemen, I note with satisfaction the role that CIMMYT has played in improving livelihood of so many of our people. Maize and wheat research has immensely benefited from CIMMYT with many different high yielding varieties being developed through CIMMYT support. CIMMYT has also promoted and still does on production aspects such as land preparation, disease and pest control and socioeconomics. It is only recently that CIMMYT inaugurated the Food security and sustainable livelihood for African households project which will enhance food supplies, food security for the rural and urban poor in sub-Saharan Africa. We in Kenya and indeed all in the region welcome this bold initiative and will work with CIMMYT to ensure the success of this project in improving the welfare of our people. We salute CIMMYT for sponsoring this workshop which goes a long way in ensuring that research in the region remains focused on common constraints and synergies and relative strength of national programs are fully exploited. Lastly, I wish to thank local sponsors Bayer East Africa, Brazafric, Osho Chemical industries and Monsanto who have contributed to the success of this workshop. I also thank the local organizers and CIMMYT staff who have made arrangements to make the workshop a success. It is now my pleasure and privilege to declare the 12th Regional Wheat Workshop officially open. Thank you.

PAGE 205

Closing Speech 12th Regional Wheat Workshop for Ea st, Central and Southern Africa Merica Hotel, Nakuru 22-26 November 2004 by Hon. Moses Akaranga, Assistan t Minister for Agriculture Distinguished Participants, Guests, Ladies and Gentlemen: It gives me great pleasure to be with you today at the closing of the 12th Regional Wheat workshop here in Nakuru. It is my belief that you have had fruitful deliberations on the way forward for wheat production in the region. I am informed that 6 countries within Eastern Africa (Ethiopia, Kenya, Rwanda, Sudan, Tanzania and Uganda) are represented in this workshop. Other participants from Mexico and India are also in attendance. This workshop therefore has provided an excellent opportunity to share expertise and experience on issues affecting development and dissemination of sustainable wheat production technologies for improved livelihoods in Sub-Saharan Africa. Ladies and Gentlemen, In order for the regional countries to collectively ensure food security there is urgent need to foster closer collaboration and partnerships among research institutes in technology development and dissemination. This has been demonstrated during this workshop. I am informed that the scientif ic data presented showed that factors which afflict effective production are not restricted within political boundaries. In our country Kenya, food production has not matched population increment partly due to unpredictable weather patterns, high cost of inputs, marketing constraints and other factors. I believe the situation is not any different in countries represented here. In order to feed growing populations, our countries will increasingly rely on research to generate and transfer technologies in collaboration with stakeholders. Ladies and Gentlemen, In the last six days, you have been deliberating on strategies to increase wheat production in the region. I am informed that you have discussed, identified, ranked priority constraints and suggested possible inte rvention measures. Some of the identified constraints include drought, diseases (in particular stem rust), pests (especially the Russian wheat aphid), unsatisfactory wheat quality and dissemination channels.

PAGE 206

You have also proposed intervention strategi es that include development of collaborative regional research projects to address these constraints. This is indeed a commendable achievement. The challenge ahead is to implement these strategies and to ensure research technologies and products reach the intended end user namely the farmer to improve productivity. It is my wish that this initiative will be fully supported by the respective governments and development partners. On our part, my Mini stry will do everything in its mandate to support this noble undertaking. It is only in this way that we can slowly but surely move towards self sufficiency in food production and create wealth. I would urge you all to pass over the deliberations and recommendations to your respective relevant government system for implementation, so that we can all pull together in addressing the constraints in wheat production within the region. Finally, I would like to thank the sponsors of this workshop namely CIMMYT, Bayer Crop Science, Brazafric, Osho Chemicals and Monsanto for their invaluable contribution. I would also like to appreciate the role played by the Ministry of Agriculture, KARI and the local organizing committee for successfully organizing this workshop. I congratulate the participants for the job well done. Let me take this opportunity to wish our visitors a good and safe trip to their respective destinations. With these remarks, it is now my great pleasure and duty to declare the 12th Regional Wheat Workshop officially closed. Thank you.

PAGE 207

Question and Answer Sessions 12th Regional Wheat Workshop for Ea st, Central and Southern Africa Merica Hotel, Nakuru 22-26 November 2004 SESSION NO. 1 Experiences in wheat research in Eastern Central and Southern AfricaMiriam Kinyua Comment: C.A. Kuwite (Tanzania) Production has dropped due to privatization polic y whereby large scale parastatal farms which used to produce 60% of wheat that was produced in the country have stopped producing. Once private farmers set in, wheat production is expected to increase. Research is now more focused on small scale production. Staff: Co-ordinator (Breeder) R.V. Ndondi Breeder Elizabeth Maeda (In studies) Weed scientist Tuael Mmboga Agronomist Mbayi Mugendi (In studies) Comment: William Wamala Wagoire (Uganda) The wheat programme in Uganda has since time ba ck been based in varietal development through selections which has had very limited staff. Priority wise wheat is not highly considered therefore, resources from the government have been put to wheat. There are regions which still demand wheat seeds i.e. South West and Eastern Uganda. Therefore, forum is now on maintenance of varieties and seed germination. Comment: Nyirigira Ruzindana Antoine Wheat in Rwanda in not a priority crop. The r esearch consists of screening wheat materials on soil acidity tolerance, bread maker qualities, high yielding characteristics. The only researcher that was appointed in National wheat programme left the programme two weeks ago. Another researcher named Innocent h as been appointed in the programme, but since he is a young scientist, he needs training to let him lead wheat research activities. Programmes (crop research programmes) in Rwanda have only one researcher and that is due to the 1994 genocide. The country (Rwanda) still su ffering from the deep gap of the skilled people as one among may 1994 genocide conse quences. This is why Capacity Building is an important element for my country.

PAGE 208

I think wheat can also be called cash crop sin ce when produced in an acceptable quantity, it should let the government save money that was put in wheat flour importation, and to contribute on people welfare. Comment: Abbas Elsir (Sudan) Wheat research in Sudan is carried out by the agricultural corporation which is centered in Central Sudan. We have a national co-ordinat or for wheat programme in which different disciplines are contributed to the generation of wheat production technology. Comment: P. Njau (Kenya) The wheat research programme is divided into sub programmes. Breeding protection, Agronomy Quality and Socio-economics. The main challenge today is stem rust, Russian wheat aphids, drought tolerance and acid soil problems. Question: Pfeiffer Wolfgang Participation of South African countries (Zim babwe, Zambia, and South Africa) Why not present? Answer: They did not send their papers for review. Comment: E.M. Njoka Njoro (Kenya) We need to come out and give what we have to farmers as we proceed with research. Let the farmers get involved especially the small scale farmers. Question: Solomon Gelalcha Why did you not mention the importance of colla boration of ECAMAW with ICARDA, just like that of CIMMYT? Answer: The importance of collaboration with ICARDA is not overlooked. We had invited Dr. Osman Abdalla to attend the work shop but due to unknown reason he could not appear, but still the collaboration is expected to be relevant. Question: A.M. Kibe Why has the National Agricultural Institution of Lear ning (universities, colleges, etc) been left out of research programmes? Can funding be channeled to and through universities? Answer: It is a challenge for us to consider to how we ought to integrate the educational institutions in our research programmes. For some reasons we have somehow forgotten them. Let us select a representative to seat in our forums of discussions, when we chart out the way forward. Comment: Owuoche Being one of the beneficiary of CIMMYT training, I would urge CIMMYT not to forget or scale down the training component since it has benefited the third world countries.

PAGE 209

Question: Ravi Singh Changes in programme structure and people in each countries of the region. Answer: Research management structure and changes in people at CIMMYT was discussed. Two flyers were distributed, which describe the curre nt programme structure and activities related to what research. Comment: Desalegn Debelo The status of ECAMAW, the thematic teams formed under ECAMAW according to the current priority not mentioned. Thematic area leaders and countries need to be included. Application of GIS in Wheat Improvement and ManagementDave Hodson Question: T.C. Riungu Where can scientists get the training in GIS? Answer: CIMMYTs GIS group has conducted extensive traini ng in the region, especially in user-friendly simple software tools e.g. Maize Atlas and Coun try Alumna training in advanced tools has not been undertaken as CIMMYT may not have the comp etitive advantage to do so. In most regions, there should be active GIS workers e.g. at unive rsities, NGOs etc and priv ate companies. These local suppliers may be best options for training. If not then CIMMYT would be willing and happy to work with local partners to increase capacity if feasible. Question: M. Kinyua Could it be necessary risking the introduction of wheat in the densely populated areas? Answer: A needs assessment would need to be undertaken to identify if suitable technologies for these areas existed, combined with assessment of comp arative advantage of wheat vs. other options (wheat may not be the most suitable crop for all areas) and different technologies/materials would be required for different farming types. Question: Owuoche How do you determine poverty levels in different countries? Answer: Poverty maps are created using advance econ-me tric analysis and GIS typically a technique called small-area estimation. These usually pred ict income or expenditure which is then compound to national poverty lines. These vary in monetary terms between countries, but are often determined by the cost of buying a basic food budget that meets minimal nutrition requirements.

PAGE 210

Question: If wheat productions increased in marginal areas, mainly rural poor benefit. If increased in intensive, high productive areas mainly urban a nd land less rural poor benefit from increased surplus production. Answer: Yes both are important. Quest i on: Is it possible to estimate risk of production when wheat area is expanded by using GIS combined with crop modeling. Answer: In principle yes (although lost capacity in CIMMYT). Access to data inputs for models may be limiting factor. Question: How can GIS be used to identify actual wh eat production areas where wheat should not be produced or is not competitive with other crops. Answer: GIS can identify areas of potential interests that could serve as a framework for detailed econ surveys on the ground to determine competitive a dvantage of wheat. GIS alone can not do this, bar introducing limiting biophysical factors e.g. acid soils. Question: During mapping the poverty map, will you integrate th e levels of potential (yield level) of wheat at different places with poverty re duction? e.g. using crop tests? Answer: There is a need to do it in the future. Yes, inter esting idea to investigate in the future. Investigated at present. Comment: Solomon Gelalcha The distribution of the potential wheat producing areas does not exactly match distribution of the needy people in the country (Ethiopia). Question: T.C. Riungu GIS is a very useful tool for scientists and other users. How can one get the training. Answer There is no organized training but there is an a ttempt to involve all those dealing with GIS, possibly contact software producers and through universities and other in stitutions to develop some training.

PAGE 211

Question: Desalegn Debelo Is that possible to see other profitable options in wheat growing areas? What is your comment on the correlation of wheat growing areas and the di stribution of research centres? What is your recommendation to the region concerning the price of satellite usage data? SESSION NO. 2 Social financial evaluation of gender in the adoption and impact of wheat production in KenyaZ. Nyakwara Comment: Prof. E.M. Njoka It is important to have information on the contri bution of straw to the soil physical and chemical composition factored into the cost/benefit-ratio. Answer: We will look into that. Question: Is value of straw factored in cost/benefit ratio? Answer: No: Because this is a new technology that has b een operated in the area due to persisted drought but there is no factored cost/benefit ratio. Question: P. Njau The wheat straw is mainly used to feed livestock dir ectly in the fields. What is the effect of land hiring to both the farmer and the landlord? Answer: It is a mutual agreement during hiring that when you harvest your crop the straw is for the cattle. Those who hire land find no negative effect (impact) by not harvesting straw. Comment: M.G. Kinyua During drought straw of the wheat has become hardy for the communities living in Narok. Analysis of marketing and pricing policie s on technology input use and production of wheat in the SudanAbbas Elsir Question: A. Kibe Why is the yield of wheat at the Gezira irrigation scheme so low and with irrigation you should get more than double the yield.

PAGE 212

Answer: The environment at Gezira is very harsh and the yields, however low, give better economic value than the imported commodity. Comment: Pfeiffer Wolfgang Smallholder wheat production needs to be subsidi zed until it is competitive. In countries which have large and small producers, not all farmers n eed subsidies at the same level e.g. not all through subsidized prices. Comment: Subsidies in countries such as India are given to the industries (e.g. chemical, fertilizer, mechanical, etc) by giving tax rebates on the ra w materials used for producing the commodity. Comment: In this way, the government is able to protect its farmers because the product is processed at a lower cost. In Africa we should probably try to de velop our Agro-industries from this perspective. Question: C.A. Kuwite As a result of liberalization policies there are wh eat imports that are cheaper than local wheat. What plans are there to ensure that local wheat out competes imported wheat so that production increases in Sudan and in the region at large? Answer: The solution is to develop a profitable and cheap er technology that cut or reduce the cost of production starting from high yielding cultivar that are adaptable to low or zero tillage. Another thing is to develop a local industry to reduce the cost of imported fertilizer. Question: Solomon Gelalcha In developing countries the government defend, jus tifies the removal of subsidy by investment on basic infrastructures (road, schools, etc), so what policy options is there to support the poor farmer? Answer: The government has to think for long-term effe ct before completely removing subsidies. The scientists have to think of technological options for reducing production cost (Ex. Minimum tillage etc). Question: Prof. E.M. Njoka What do you think about subsidies in the a lleviation of poverty in African country? Answer: Subsidies are very essential especially for small scale farmers who are financially incapable. It is the only way of keeping them surviving otherwise they will migrate to the border of the cities and create so many problems.

PAGE 213

Comment: Ravi Singh Personal view on subsidy: Removing subsidy withou t providing options for poor farmers is likely to have severe effects on the livelihood of people. Social economic factors influencing the us e of recycled wheat and maize by small scale farmers in Nakuru A.C. Ndiema Question: Desalegn Debelo What is the status and contribution of decentra lized seed system to overcome the problem of recycled seed? Seed multiplied by small scale farm ers for their own use and for the farmers in their village? Answer: The practice was workable before liberalization of the industry. They are more seed breeders, multipliers and stockiest than before. The issue in question now is trusting the source of the seed. Question: Patrick Ooro What is the projected implication on the quality of seed with particular reference to recycled maize seed? Recycled seed maize has reduction in yields to the tune of 20% loss. In subsequent recycling this loss is said to go upto 50% because maize is a cross pollinated crop. Question: J.K. Macharia Liberalization in seed industry is suppose to ma ke seeds cheaper. How come the seed has become much more expensive? Answer: Seed liberalization has only caused confusion among the faring community giving the seed merchants room to exploit ignorant farmers. Comment: Prof. Njoka We need quality seed if the accompanying technol ogy has to have the benefit required to the farmers. SESSION NO. 3 Survey of National Enemies of Rust Wheat Aphids (RWA) in KenyaMacharia Question: C.A. Kuwite Since the parasites and parasitoids come in late during the season what is their population like in late planted wheat?

PAGE 214

Answer: The population trends are just the same, i.e. the population of predators and parasites tends to lag behind RWA population, thus they are unable to offer adequate control as they peak when RWA has already caused damage to the crop. Question: J.K. Macharia What % age of predators/plant do you consider sufficient to manage RWA biologically? Answer: At the moment, only preliminary studies have been conducted to document the predators that attack RWA. The next phase will be used to dete rmine the bio-efficacy of identified biological agents (natural enemies). Comment: In contrast to other aphids, one RWA per plant can cause symptoms, hence biological control may be complicated. Answer: RWA management requires IPM control strategies (biological control being one) and the future lies on development of host plant resistance strategi es. No single control strategy is effective. Question: How do you expect the effect of natural enemies on the host plant when the number of RWA population is decreased? Answer: The natural enemies are host specific and will only attack aphids but not the host plants supporting RWA, hence no effect in the increase or decrease of RWA populations. Comment: M. Kinyua Study the aphid instead of crop physiology i.e. how the crop behave. However, we decide to study the aphid as well. Current challenges in the management of wheat diseases Ravi Singh Question: How do you select the parents. Is the molecular marker technique for the *** Question: Pathak Clarify the relationship between horizontal/verti cal resistance and slow rusting/major genes? Answer: They are synonymous terms. Horizontal resist ance is synonymous to slow rusting, whereas vertical resistance is synonymous to major genes.

PAGE 215

Question: Pathak New stem rust races have appeared after 15 years in East Africa region. Is it due to change in the environment or evolution of new virulent race? Answer: New races have been evolving from time to time in the region. The evolution and migration of the Sv31 resistant race is a very significant case as seve ral cultivars have become susceptible to this race. Question: Could we screen wild relatives of wheat also to tap the resistant genes? Answer: Yes, it can be done but care must be taken that wild species do not become weed in the field. Comment: Owuoche We should incorporate Sr genes into Lr and Yr germplasm that have shown resistance. Answer: That is a good suggestion and that is one of th e things we plan to carry out in the future. Question: After the 1998 stem rust episode in Uganda *** Answer: The resistant lines were used in crossing progr amme. Advanced generation populations need to come back to the region to identify which lines i nherited resistance to the Sr31 virulent race. Comment: Wolfgang Pfeiffer Fastest way to get Sr resistant varieties to farmers will be using major genes combined with double haploid methods. Then minor genes in phase II varieties. Answer: Cost of making double haploid is still a limiting fact or for its wide use. Also if two crop seasons are grown per year, then traditional breeding scheme is fast enough. Evidence of New Russian Wheat Bi otypes in KenyaJ. Malinga Question: What capacity is there to deploy predators to control aphid epidemics in *** of the large ** of land that are under wheat growing? Answer: We look forward to identifying effective natural en emies, which can then be reared and released on time as RWA attacks the wheat crop usually at the vulnerable growth stages (seedling stage).

PAGE 216

Evidence of RWA Biotypes in KenyaMalinga J. Comment: M Kinyua Study the aphid instead of crop physiology. I.e. how the crop behave. However we decided to study the aphid as well. Evidence of new Russian Wheat Aphid Biotypes in KenyaM. Kinyua Question: Differences in the AFLPs may not necessarily indicat e differences in the effects of the pest. Do you have preliminary observation on the correlati on between the diversity of the pest and the effect on wheat? Answer: Yes, we had previously evaluated the wheat lines in the greenhouse against 3 chores collected from 3 regions (Eldoret, Timau and Njoro) and obser ved that there are differences in virulence of Timau against Njoro and Eldoret shores. This promot ed the story. The future is to chores from the two regions and screen them against specific lin es while evaluating their DNA. The data will be used genotyping of the aphid using one of th e current (programmes e.g. power maker). Physiological Races and Virulence Comment: Temesgen Kebede Some of the large scale state farms use fungicid es to control stem rust. Few large scale private farmers also utilize fungicides to control stem rust. Question: John Muchile What is the short term solution to stem rust pandemic since breeding programs for resistance/tolerance take long? Answer: We do have strategies to screen the resistant va rieties so as to that identify susceptible genotypes were taken to the shuffling in the long run. Comment: The report on Sr31 as resistance is based on the samples collected in 2001 and 2002. The scenario has been changed since 2003. Wheat rust in India. M. Prashar Question: Pathak Why not to test the segregating material rather that advance prevaliased variety of wheat for resistance of wheat rusts?

PAGE 217

Answer: Yes, The rust laboratory is already testing the segregating lines/advance lines from all over the country. Question: Solomon Gelalcha Would you please share with us your view/experience of programme in the rust control strategy of Indian wheat research programme. Answer: The gene development strategy is not as such su ccessful because of the migratory nature of the rust pathogen. Mover over, the agronomic and qua lity aspect should be considered in the control strategy. Question: Desalegn Debelo Can you share with us the information of puccinia pathways identified in India? Answer: Puccinia pathway indicates that brown and black rusts spread from South to Central India and from Northern hills of the country, yellow and brow n rusts are disseminated down to the plains of North India. SESSION NO. 4 Grain yield, water use and water use effici ency as affected by moisture under rainout shelterP.A. Ooro Comment: Prof. E.M. Njoka Your result should include the environmental effect of rainshelter. Otherwise you would have to give your result in relative terms. Answer: The environmental effect of rainshelter in term s of RH and PAR have been addressed hence will be included. Comment: P.K. Kimurto The rainshelter effects on the performance of the genotype has been established to be negligible or non-significant from previous studies done on the site. Question: Solomon Gelalcha What is your criteria of categorifying your e nvironmental as marginal; is moisture the only criteria? What is the day to maturity of the varieties used in the experiment.

PAGE 218

Answer: Moisture, temperature, soil type etc. can be u sed as a criteria but we used moisture as main criteria. The varieties used are early maturing ones. Question: A.M. Kibe The crop canopy environment must have been aff ected under the rain shelter thus reducing the ET. How did you account for it in your calculation. Your formula for TWU was not really what was used because there was no precipitation. Answer: The out-rain shelters tarpolein was lifted and fa ded in such a way as to allow air movements within the rain shelter. In the TWU formula, R is supposed to be replaced by 1 (irrigation water applied) R(run-off) and D (drainage) was considered to be negligible and therefore was not part of the computations. Evaluation of Monitor and its tank-mix partners for weed control in wheat D.O.K. Amadi Question: Prof. E.M. Njoka You are dealing with different populations and the yield dealing equations are affected by population. How do you account on this? Answer: The different populations are with regard to the different wheat cultivars used in the trials. Our main focus was to target weeds whose response to herbicide treatments may at time be neutral to plant density. Genotype, nutrient and utilization interactio n in wheat grown in the marginal and acid soils environment in KenyaJ. Kamwaga Comment: Prof. E.M. Njoka You cannot attribute your effects of N if the interaction of ** are not indicated. Hence your conclusions were not complete. Answer: The idea here is to test the cultiv ars adaptability to low fertility and also at high fertility situation and find out if there is any cultivar and fertilizer interaction. A cultivar that is efficient at** N ** low fertility and respond well to append N would be preferable. Question: Do you consider the N in grain (protein content) while studying the rates in relation to N use efficiency? Answer: N increases in the grain if you increase the N use efficiency to a certain level.

PAGE 219

Evaluation of impaired wheat varieties under different management levels in Eastern Wallaga Highlands Tolera Abera Question: Owuoche Did you cost land preparation? Answer: Yes indeed. Impact of irrigation frequency and farm yard manure on salt affected soil and wheat production in Dongola areaElmoiez Fadul Question: Solomon Gelalcha Is the 7 days interval economical? (It seems so frequent). Answer: We applied the same amount of water by the end of season for each treatment so economically we used 7 days as recommendation where the output of 7 is high. Comment: Prof. E.M. Njoka In your irrigation interval, care should be taken to put into account the effects of added organic matter (FYM) because of this sponge. Answer: We calculate the ET per day by Jensen and Haize equation and estimate the crop factor then due to loss of water by irrigation, multiply by 70% in this formula. Qimm = Ke ETP x 100 then introduce the plot through partial flame regard less of Ei treated with FYM or no. SESSION NO. 5 Agronomic and economic evaluation of b reak crops and management practices on the grain yield of wheat at Shamb o, Western Oramiya-Tolera Abera Question: Kassa Getu I wonder why you compared barley with field pea as break crops? Answer: The intention was to stimulate the farmers practi ce and demonstrate the difference to the farmers for sustainable wheat production in the area.

PAGE 220

Question: Ravi P.Singh Whether you are able to communicate to polic y makers that improved cultivars and better management practices increase production and is profitable? Answer: The communication has no problem but farmers tend to use their own varieties and management practices. This is due to other socio-economi cs constraints such as cash shortage and credit unavailability to purchase the inputs. Question: M. Kinyua What efforts have been made apart from resear ch to make farmers realize this efforts. Answer: The research results are well communicated to farm ers through on-farms de monstrations and field days organization. So that the farmers see th e research results and compare the traditional and improved practices. Question: Millions of demonstration plots have been plan ted in Ethiopia with modern management by Global 2000. What are the limitations of adoption given the rates of adoption are low? Answer: The main limiting factors for technology adoption are financial and social constraints. Timely availability of technologies and the purchasing power of the farmers limited to the adoption. But now better than ever. Effect of Nitrogen fertilizer levels a nd varieties on Gluten contentBemnet Question: Nitrogen levels did not have an effect on quality of dur *** Comment: Tadesse Dessalegn In Ethiopia, processing industries are willing to work with the researchers and wheat producers and buy all the produce if it m eets their quality standards. Question: C.M. Ndirangu Since farmers are paid on the basis of quantity (ra ther than quality), what is the impact of the study at farm level? Answer: There is a promise from local pasta industries to pay premium price for quality product produced by farmers.

PAGE 221

Question: All varieties in your study are tall and less respons ive to N particularly in terms of yield. When will semi-dwarf varieties be available. Answer: Since 2002, high yielding quality meeting semi-dwarf varieties have been released at Debre Zeit, Sinana and Adet Research Centre. SESSION NO. 6 Monitoring of bread wheat cultivars a nd advanced lines for their resistance to yellow rust-Temesgen Kebede Question: Ravi Singh How has the resistance to stem rust seen until 2003 behaved during 2004? Answer: We have yet summarized the data for 2004 in the mean time we will include this data. Question: M. Kinyua How has it managed the crop improvement programme as far as disease management is concerned in Ethiopia. Answer: There has been loose collaboration among breeders and pathologists and the data generated from rust (wheat) monitoring materials were less utilized However, this will be improved in the future in planning the crossing programme. Evolution of Kenyan wheat lines fo r bread making quality E. Kimani Question: Nyirigira Need of more clarification because you said in the talk that in wheat grain, the protein quality is due to genotype while the protein quantity is due to weather. Answer: Protein quality is a genotypic trait while the pr otein quantity varies with environment due to climatic conditions and soil fertility. Question: C.A. Kuwite Are you able to look for varieties that take shorter time for maximum water absorption dough development so that cooks will not take longtime cooking/baking?

PAGE 222

Answer: Varieties with shorter DDT are available and are pr eferred in baking. Bakers prefer varieties with higher water absorption as it relates to amount of loaf baked. Question: You told us that these is positive correlation betw een DDT and loaf volume. What would happen if you shorten DT by using commercialized fermenters (yeast)? Answer: Yeast is already used for the bread leavening a constant amount for all the samples DDT is an effect/a result of protein amount in the flour. Question: Tadesse Dessalegn Do you consider national quality standards in your breeding process? Answer: There are standards in the country used by th e researchers as well as the factories millers and bakers. Question: Kibe Why was Heroe used as the control? Answer: K. Heroe was used as the check in the field trials. It is also a released variety as compared to the others that were yet to be released. Its p oorer performance however was probably due to the higher altitude though its r ecommended in lower altitude. Question: P. Njau What is the practical applicability of heterosi s in wheat breeding? Can you comment on the rate of depression of the varietal characters measured? Do you think there might be a correlation between combining ability and choice of parent for top and back crossing. Answer: High hybrid vigor expected from crossing than simple selection. The rate of breeding depression is very minimum. Yes, there is correlation betw een the combining ability of the parents and their performance which leads to utilization of the pa rental lines in breeding programme (crossing). Question: Ravi Singh Is the hybrid vigour seen in your experiment is due to data taken on plant basis or due to the heat stressed environment of the experimental site? Answer: May be because the data is taken on plant basis or some other thing. Anyway, the heat stress on the parental lines may encourage the hybrids to perform better than the parental lines did.

PAGE 223

Question: Wolfgang The high parent heterosis you have shown is much higher compared to reports in the literature for wheat. This can not be explained by heterosis for yield components other factors? Answer: There may be other factor but I cannot expl ain relating to this specific experiment. Question: P.Njau Is it possible to fix the hybrid vigour in wheat ** may be use of DH technique? Answer: I dont think it is necessary to go to fixation of heterosis in wheat because wheat is self-pollinated crops and the extent of inbreeding depression observed was very minimum (insignificant). Performance of wheat genotype s in Western KenyaA.W. Kore Question: The grain yield per ha you reported is too low to be research managed plots. Why so? Answer: The low yields are possibly due to the high temp eratures which hastened maturity and hence did not give enough time for assimilation, we hope to try other genotypes that can tolerate heat and also give other agronomic packages. Comment: W. W. Wagoire The trial is commendable. The climate is rather hostile and with time foliar diseases will be a problem. Please keep up and lets get more information. These comments are in view of similar work due in Uganda under similar conditions. Question: It looks like termites will become a major constr aint in wheat production, what measures have you put in place to ensure sustainable production which is termite free? Answer: We would need to do further work with entomologist on this. Comment: M. Kinyua It is a little bit hard to take control of small scale farmers, however it tends to provide (wheat) a better option to farmers who are ** to maize and rice in the Western regions. Comment: Owuoche Wheat is grown by farmers in Kisii and we need to be aggressive to help them improve the crop. Question: Was 2 tonnes the highest from test material?

PAGE 224

Answer: We believe that it is possible to get higher yields than this if adaptable cu ltivers are used and also proper agronomic practices are brought in place. Question: D.O.K. Amadi How was the behaviour of weaver birds and quelea with the wheat crop in your trials? Answer: During the trial period, these birds were not a pr oblem probably because wheat is new to the area. Question: Due to recipient poverty in Western Kenya, can we say that wheat can be used as an alternative crop for poverty alleviation to improve livelihoods? Answer: For now, we can recommend that wheat can be used to supplement what is currently being grown rather than being used as an alternative crop. Question: Owuoche Probably you could try to plant wheat in higher areas of Western part of Kenya? Answer: Some effort has been made in the higher a ltitudes as well and the results are commendable. Question: Nyirigira Did you, in your observations see crop pest and diseases reactions? Answer: The main pest observed were the termites while some insignificant leaf rust was also noticed during the trial period. Quantification of the value of improved wh eat production options in South Western Uganda-William Wagoire Question: D.O.K. Amadi Hand weeding currently seems to be the only op tion available for weed control in wheat in Uganda. Comment on the possible use of herbicides for weed control on your what farm. Answer: The cost of herbicide and their application woul d not be appealing to small-scale farmers. The hand weeding uses family labour which in most cases they never consider.

PAGE 225

Question: Since production of wheat from your presentation at small scale level percentage farm size of upto 0.25ha percentage land, should it be promot ed or there are other alternative crops. Is it economical to grow wheat in Uganda. Answer: In South Western Uganda wheat is one of the fe w alternative cash crops while in Eastern Uganda, wheat price is more stable than maize which is the alternative crop. Otherwise at national level, wheat is not a high priority. Question: M. Kinyua Whether recommendations have f iltered through to the farms, what has been the impact? Answer: Yes, the elite lines UW400 and UW309 have been adopted and are being demanded for by the farmers (before they are released). Paper Title: Allelism of resistance genesC.A. Kuwite Question: Owuoche Why did you not include cross of RxS? Answer: Our aim was to study whether the resistance genes were allelic in these cultivars. Including such a cross would not have served our purpose. Stability analysis and participatory evalua tion of bread wheat varietiesKassa Getu Question: How were the small holder groups organized? How did you gain their support. Did you have to offer them incentives before or after the tests? Answer: The farmers were organized by NGOs. Had it not been organized that way, it would have been difficult for us to deal with farmers in different districts. For the first year, we give farmers incentives. Once they were sensitized they started to produce variety of their choice at their own cost and were benefited from their produce. Question: M. Kinyua Did you find any contradictions in the way they ranked the varieties and factors on the way you know them as a researcher.

PAGE 226

Answer: Yes, farmers do have some times a different inter est from what researchers think. However, these variations are area specific, and so is the variety release. Question: J.K. Macharia Why was the farmers used for vari ety evaluation not gender sensitive? Answer: Although the number of females were very few, female farmers have attended the variety evaluation during the mini-workshop. Due to diffe rent social reasons, females usually do not come to meetings. Consumption use of water, water and nitrogen use efficiency by wheat in relation to irrigation and nitrogen A. Kibe Comment: Pivot can be used in combination with drip irrigation tubes to increase water productivity. Answer: Yes its possible. Question: Kinyua What are the recommendations for the local farmers. Answer: Local wheat farmers should consider irrigating their wheat crop. The large scale farmers are able to harness the water (rain) and pump it to a higher elevation and then supply it through gravity. Yields can be increased to over 6 tons/ha. Small sc ale farmers can also grow wheat in small plots (1/4 acre) under intensive management (water and N inputs) together with other pulse (chick pea) and oil (mustard linseeds) crops for subsiste nce use. Harvesting can be managed manually (sickles). Water can be used judiciously so as to supply just sufficient amounts to achieve maximum potential (genetic) of the crop. On farm trials need to be done in order to evaluate water nutrient management optio ns and economic evaluation. Comment: John Muchile Centre pivot irrigation starting in Kenya are lik ely to expand, are the farmers likely to suffer salinization as in India. Answer: Salinization harzards normally occur when wate r is applied in excess quantities under conditions of high evaporative demand. Sprinkler irriga tion tends to be a conservative method (though expensive in the initial stage). It tends to avoid salinity problems, especially when just sufficient amounts of water are applied.

PAGE 227

Comment: M.C. Mahagayu There has to be some home made solutions to t ackle the issue of irrigation for the case of small scale resource from farmers. Given yield stability of bread wheat genotypes in favourable and stressed environment in Ethiopia-Debelo Comment: Solomon Gelalcha The varieties we have in Ethiopia basically s hould have come either from CIMMYT/Mexico or somewhere else. But we are trying to improve the long existing obsolete varieties through crossing. Question: Ravi Singh How many lines survived with two stem rust race? Answer: From international nursery materials only four su rvived. Most of the commercial varieties were also out by the stem rust. But there are varieties which maintained their resistance, though not preferred by the farmers for their low yield potential. Question: M. Kinyua Introductions were doing better than the Ethiopian materials why? What could you think is the reason. Answer: Ethiopian crosses are single crosses. They are good in disease resistant than in yield potential. Seedling and adult plant resistance of wheat var ieties to local stem rust isolates in EthiopiaEmebet Fekadu Question: Where did you carry out the tests on adult plant resistance, was it in the field or greenhouse? Answer: It was in the greenhouse at five leaf state assessment that it repr esents adult plants of the crop. Question: R.P. Singh Have you tested the wheat lines w ith the new race of stem rust? Answer: We did not test those materials with the new varieties. Integrating of other small cereals in wheat producti on systems in improving livelihoods-Wolfgang P.

PAGE 228

Question: M. Kinyua How do you identify high yielding varieties in your environment. How do you compare environments taking interactions in mind. Answer: Testing under both stress and non stress to identif y input responsive and/or efficient genotypes. Wide testing with use of a common set of standa rds to allow for comparing different experiment. Question: C.M. Ndirangu Are there any non-phonological anatomical or ph ysiological characteristics that are associated with drought tolerance in wheat? Answer: List of traits was presented. Currently CTD and NO VI in the near future offer the largest potential routine use in applied breeding characterization of progenitors for e.g. carbon isotope discrimination is recommended. Question: Ruth Wanyera Among the biotic stresses, you mentioned nematodes, Fusarium ep, take all and Septaria sp, what happens to stem rust in this type of environment? Answer: Rust diseases are in general the major di sease problems under dryland production. Disease epidemics are less frequent under drought. Howeve r we need resistance to all prevalent diseases in varieties to combat epidemics in water years. Question: Owuoche Data from drought tolerant experiments are ofte n characterized by high CV. How do you handle such data? Answer: Uniform area within field for experiment. Larger plot size, more replications, eliminate masking factors, spatial. Experimental design. Question: Is insect resistance not a ** in developing drought tolerant germplasm. This dryland evnironment is where the insects ** best. Answer: Absolutely insect resistance is very important un der drought and often the major biotic constraint. Question: Tadesse Dessalegn What will be the success of using large vs fe w crosses? Regional programmes could not handle large number of segregating population if they go for large number of crosses.

PAGE 229

Answer: Initially the number of crosses can be larger sin ce F2 can be planted in solid seeded plots under drought conditions with best performing populations planted in the following cycle for individual pant selection. Ideally under higher moisture c onditions. Parallel planting of F2 under drought observation and higher moisture scenarios is an op tion if resources are available. Number of crosses are so far your programme. Question: Patrick Ooro Materials from Australia especially for drought stress have been around with tiller inhibiting gene. Is it the same trend with CIMMYT? If No, what is your comment. Answer: For most of our scient countries we need high tilling to ensure stand un der drought and input responsiveness in better years. Often stands are poor due to poor seed or poor land preparation and planting conditions. Further yi eld from secondary tillers if damage from stress occurs (e.g. Frost) can enhance risk efficiency similarly, unsynchronized tillering can expand period of water extraction. SESSION NO. 7 Control of Russian Wheat Aphids (RWA) in wheat using systematic insecticides in KenyaM. Macharia Comment: Singh Use of Gaucho is now banned in France. Answer: There are alternative such as cruiser which are effective against RWA. Besides, Kenya farmers can apply foliar systemic herbicides on condition the herbicides is applied on detecting of the initial symptoms of RWA infestation (i.e. rolled leaves resembling spring onions). Question: D.O.K. Amadi Yield loss attributable to BYDV in wheat in Kenya is about 47%. Would this percentage loss be true irrespective of the time of infestation. Answer: Yield losses will vary depending on the stage of crop at infestation time, prevailing weather conditions. Comment: Abbas Elsir There is no serious effect of RWA on wheat produc tion in Sudan, however it has serious effect on legumes especially *** if its sowing is delayed.

PAGE 230

Position of Russian Wheat Aphids in RwandaM. Macharia. Comment: Nyirigira Since RWA is not a problem to wheat production in Rwanda they should be no reason to make input in RWA research. Question: What does each country think about Russian Wh eat Aphid incorporating it in their wheat programmes? Answer: Varied opinion in the E. Africa region. Uganda: ** are different although they perceive it to be a problem. Ethiopia: RWA is a sporadic pest and becomes *** during shortage of rainfall. Kenya feelings: With the current development of vi rulent biotypes as in Kenya, it is necessary to input I some strategic research, as it is just a question of time before we are in RWA crisis. Comment: Research budgets in most of the ECA countries ar e limiting to address what would be considered strategic research and proposition of research i ssues is varied. Otherwise the RWA, from the presentations made, is a real threat. Reducing the threat of RWA on wheat and barleyMacharia & Migui Question: Ravi Singh Movement of yellow rust from East Africa to S outh Africa did not occur. Yellow rust in South Africa entered from Southern France through human error. Do you think RWA entered in Kenya from South Africa? Answer: Its possible the RWA may have come from the South, as there were very strong winds blowing into Kenya from the South. Improvement of yield in the drought tol erant wheat varieties of KenyaP. Njau Question: Ravi Singh Can you compare the cost of producing DH verses single seed descend technique? Answer: In Kenya most of breeding work is manual and only single season per year. This makes it 8 years to develop homozygous lines at a cost of 5 dolla rs per day for three months per year. In case of DH technique the initial cost is expensive but it is very cost effective and that we only need the

PAGE 231

media and maintenance of the greenhouse. Most of the work is done in the field and so cost reduced. We expect the cost of DH production to be the inbreeding (segregating) population. Question: Why did you use F2 instead of e.g. F3 seed after screening for disease resistance? Answer: Here we were dealing with very specific char acter (drought) and both parents were commercially adapted so, we expected to get a recommencement in F2 plants. Question: Kassa Getu Is any one of DN gene families sequenced? Answer: Yes, some of them have been sequenced particul arly DW2 and DW4 but others have not yet been sequenced. It is important to note them as a ra nge of DW genes 1-9 already tagged but not all have been sequenced. Production of bread making quality of Et hiopian wheat cultivars using direct and indirect quality traitsTadesse Dessalegn Comment: Ravi Singh Often it is mentioned that grain yield and qua lity have negative relationship. However, my comment is that genes that contribute to quality infact do not have negative effect on quality. The reason why yield and quality often do not go togeth er is due to the probability of combining yield and quality genes together. Yield progress usually is followed by quality in subsequent breeding cycles. So progress in quality follows progress in yield. Question: E. Kimani Explain the negative correlation between protei n content and mixograph development time (NB in some literature they are positively correlated) Answer: The negative correlation might be due to the low protein percentage of the materials as protein has a confounding effect on many parameters, th erefore, materials should have optimum protein for optimum results or optimum mixing time. Comment: Wolfgang In 50% of your lines you found HMW GLU B1 7 + 9 which would indicate the presence of the 1B. 1R translocation. To breed for higher quality do you plan to reduce the proportion of 1B and 1R?

PAGE 232

SESSION NO. 8 Integrating of other small cereals in wheat producti on systems in improving livelihoodsWolfgang Pfieffer Comment: M. Kinyua We have selections of triticals at KARI-Njoro in small samples. We have two releases which are recognized by ministry. They are also in some cases very good in threshing than wheat. Question: Kibe Considering its threshing problems how would our farmers overcome it? How can we get tritical varieties? There are small machines (hand driven) available. It would however be important to look at the end utilization value of triticale in comparison to other cereals e.g. it doesnt shatter, its nutritional value is higher, its drought resistant. There are certain varieties that are easier to thresh than wheat. We have a few lines in KAR I-Njoro. Two are released varieties and we can multiply seed if required. CIMMYT can provide the seed through NPBRC KARI-Njoro too. Question: Kibe Considering its taste. How would our farmers in E. Africa adapt it? Answer: Triticale is rich in lysine and tryplophan. It woul d be good if we were to consider it for blending purposes in order to improve nutrition. It can be made into biscuits too. It has been used to make flour mixtures for making bread (chapatti). Question: P. Ooro Zero/reduced tillage has worked extremely with re gard to large scale farmer s in Kenya. For their small scale counterparts, the major limitation is getting the right planters for zero/reduced tillage systems. Answer: Agree, machine for smallholders has been deve loped. Examples from Asia indicate options through machine sharing, renting of machines from government agencies etc. Given that large farms in Kenya have resources to acquire plante rs. They may serve as pilot projects with an adoption of technologies in a sub sequent phase by smallholders and local machine development or evaluation of existing planters for small scale farming. Question: H.G. Mwangi What is the potential of minimum/conservati on tillage in this ECASDA region with similar environments like Mexico e.g. 500mm and soils as a way of conserving environment and resources.

PAGE 233

Answer: It will be a pre-requisit for sustainable wheat production in the region due to numerous benefits (soil organic matter, pH water holding, nutrient supply). A direct transfer of the technology developed in Mexico may not be possible. Adaptiv e research needed with available inputs such as herbicides, machinery and varieties needed. Question: Ruth Wanyera Please explain what you mean by soil health in the cropping systems? Answer The different small grains have different susceptib ility to soil born pathogens e.g. take all. Their inclusion in a crop rotation will decrease the i noculum load in the soil and hence increase soil health for the following cereal crop (example oats). Question: Durum wheat in Ethiopia is grown in high potential areas unlike other countries which produce in marginal. With the current quality demand and in the country and high grain protein content in moisture areas there is a need to push DK in lo w lands. What is your strategy of including drought tolerant cultivars in the international nurseries? Answer: The situation for Ethiopia is different from othe r countries and had plans to make separate tables. Drought tolerant durums are included in interna tional nurseries but all are semi-dwarfs, mainly directed to terminal drought environments. For th e more specific conditions of Ethiopia we have developed germplasm with height short interm ediate and tall from crosses between (Ethiopian landraces and Ethiopian varieties) and CIMMYT hi gh yielding dwarfs. This germplasm is targeted to a range of agro-ecologies in Ethiopia. The current status of stem rust in whea t production in KenyaRuth Wanyera Question: W. Wagoire You report reoccurrence of stem rust in Kenya in 1996. Is there a possibility with the reported occurrence in Uganda in 1998? Answer: We may not know, since there was no follow up until year 2002, but chances are that it could be the Ugandan race. Question: W. Wagoire Do you have data on yearly basis that you study in for 2002 004? And if not we need to follow up the occurrences every season so as to est ablish the magnitude of the problem. Answer: Yes, there is data for each year, but it was not s hown in the presentation. It is true that Sr was observed in 1996, there was no follow up to coll ect the spares and carry on the identification.

PAGE 234

Comment: R. Singh Kenyan collections from Kenya were sent to USA where preliminary results are indicting that the Ugandan and Kenyan stem rust are likely to be the same. Responses of physiological traits to drou ght tolerance in bread wheat under tropical conditionsP.K. Kimurto Comment: Kibe The physiological traits are evident in all growing crops. They however vary by genotype. We should aim at determining the farmer friendl y traits that are evident on the plant. Comment: This can be done by relating the physiological response at different stages of crop growth and relate them to DM, growth, yield attributes and gr ain yield. This way, we will be able to see the sensitivity of genotype to environmental stress, and therefore, determine the most dependable physiological parameter. Comment: P. Njau It is possible to develop varieties with the tr aits (physiological) but this will depend on the correlation of the traits and their adap tive effect on yield and kennel weight. Question: Patrick Ooro Noting that the traits studied appeared to have been associated with genotypes differently. Is it possible to combine these traits on a single ideotype. May be breeders can also comment. Answer: Yes, its possible to get ideotype. But care shoul d be taken not to combine traits that will be antagonistic to each other and finally reduce yield. The breeders should take this challenge and us it in their breeding programmes. Cell membrane stability as a measure of drought tolerance in Ethiopian bread and durum wheat genotypesA. Zemede Question: Wolfgang The region significantly differ in soil fertility, ho wever average yields were similar. Data on 1000 grain weight was 26g for the better soil and 4g for the poor soil location. There must have been severe stress during grain fill can you please elaborate? Answer: Between the two locations the big difference is the distribution of rainfall pattern, so at ** district during the grain filling stage of the crop occurred the moisture stress.

PAGE 235

Question: Enlighten us on why both sites P was not applied i.e. 92.0 and 41.0 with much higher MRR (than those recommended) was not recommended? Answer: Because the interaction of the two nutrients treat ments gave the highest net benefit and highest agronomic grain yield. The response of bread wheat to nitrogen and phosphorous fertilizer at different agro-ecologies of North-West ern EthiopiaA. Zemede Question: Abbas Elsir I noticed that the MRR at 92N level and zero phosphorus is higher as 138N and 46P taking into consideration it is cost effective and labour saving. In the sense that if you take the cost labour into consideration the MRR of the high fertilizer level will be reduced. Answer: 138N and 46P fertilizer rate gave highest net be nefit and within the range of acceptable MRR, (100%). About labour cost, we used broadcasting fert ilizer application, so it is not as such higher than the control treatment, however, the comment is well accepted. Comment: Z. Nyakwara There is no labour which is free and land should be quantified. Whether owned or hired, hence to arrive at the MRR all variables making up to (total variable costs) should be checked. (Quantified). Comment: The comment is well accepted. Evaluation of wheat ( T.aestivium ) double haploid for resistance in the greenhouseJ. Malinga Comment: Wolfgang It is critical in this kind of yield loss studies to eliminate all confounding factors. Example if we want to determine the effect of stem rust and the stem rust resistant genotype is susceptible to yellow rust with high disease incidence of both Sr and Yr. The effect Jr of the Sr resistant genotype could not be determined and be masked Bx and Yr. Comment: Kibe In order for use to have better conclusive results it might be better to pool a parameter across all sites, and regress the dependable against the i ndependent variables. Carry out regression (and multiple regression) analyses. This woul d give us better conclusive results.

PAGE 236

Comment: M. Macharia I feel that the field studies needs to be supported by greenhouse studies through inoculation by using fixed number of aphids/plant. Comment: The inconsistent in the data over the season and locations pints to an ***** Answer: Yes RWA resistance is very limited in bread wheat particularly in spring wheat. So the partial identification of *** is useful. SESSION NO 2: GENERAL DISCUSSION Current challenges in the management of wheat diseasesRavi Singh Question: Macharia With the high turnover of scientists, there is n eed to have the RWW more frequently 2-3 years. Answer: Ravi The recommendation was taken to account. Noted that previously it was span 2 years. Michire: Gaucho (seed dresser) ban in France. Producted use h as been suspended in a specific sector in France, this was seen to have effect in s unflower sector (lowered production). Trials are underway to see effect of Gaucho. But is being used in other sectors. Kinyua (ECAMAW Chairperson): There is need to discuss recommendations of the region to complete the workshop. Meet every 2-3 yrs because of high turnover of scientists. Comment: Funding of wheat from ECAMAW ended in 2001 Success has been due to personal initiatives. Hope for funding in Africa livelihood programme. There is no ASARECA funding. 1/3 ECAMAW funding. Macharia: It is a way of dissemination. As stakeholde rs we should try to meet more frequently. Comment: Strategy of getting the funds. C. Kuwite: With the production and import data, we can use this to get the government to give more money.

PAGE 237

Comment: Impact government because the imports show the demand. They should give the money for more production. Pandey gave ideas to approach the gove rnment as a region to influence them to act. Why does the national co-ordinator find it hard to influence the Minister. Wagoire: In Uganda, its difficult to get a meeting with th e Minister on such an issue because of protocol. Thus using the regional office is easier to get report from ministers. Debelo: RWW is important for exchanging of ideas. A pproach the government through the ECAMAW, ASARECA etc. Abbas: In Sudan, financial difficulty. Politicians dont recognize wheat re search. They would react more from CGs. Nyirigira: Talks between scientists and government hasnt bore any fruit. CGs could make more impact; airing out imports and the possibility of more production in wheat and the problems of production. C.Kuwite: Access to Minister is not probable. The Mini ster of Regional Co-operation (who opened the meeting) could be a good channel to air th e wheat production constraint in the region. Prashar: In India field days invite politicians to enhan ce awareness. Invite them oftenly to make them aware. This makes them sensitive to the issue. Can CG centres talk to the government? CIMMYT. This should be possible thus the formation of Regional programme. Ravi. Wofgang: Should be done jointly with regional and CG centres. Owuoche: Liberalization has caused harm. Levies of wheat research were erased. Suggestion: Impose levy on import wheat for research. Michire: Parliamentary select committee have meetings with scientists so they can influence the parliament.

PAGE 238

Wagoire: Have regional document of the findings a nd what is to be done, constraints? Wasike Lusike If wheat scientists have a quantifiable document e.g. in the next 15 years. SR will reduce wheat production and increase imports past this on media, it would get in. Solomon: In Ethiopia, there are questions on impact of r esearch on poverty allevia tion. The approach is important. Thus show the support from the govern ment and why there has been a drawback. Kibe: Effectiveness can be achieved through educational system university. Try to channel funds for textbooks for use by the lecturers for teaching students to obtain right information for our region e.g. agronomic strategies, variety deve lopment, etc. Source of dissemination.

PAGE 239

12th Regional Wheat Workshop For Easte rn, Central and Southern Africa Participants E-mail Addresses No. Name Country/ Institution Position E-mail address 1 Emebet Fekadu Ethiopia Pathologist f_emebet@yahoo.com 2 Bemnet Gashawbeza Ethiopia Durum wheat co-ordinator bemnet_mehret@yahoo.com 3 Desalegn Debelo Ethiopia National Bread Wheat Research Co-ordinator ddebelo@yahoo.com desalegnd@freemail.et 4 Tesmegen Kebede Ethiopia Crop protection Division Head/Pathologist temkebede@yahoo.com 5 Kassa Getu Ethiopia Plant Biotechnologist kassagetu@yahoo.com 6 Solomon Gelalcha Ethiopia Plant breeder sgelalcha@yahoo.com 7 Minale Liben Ethiopia Agronomist minaleliben@yahoo.com 8 Tadesse Dessalegn Ethiopia Wheat Breeder Tadesseyf@hotmail.com Tadesseyf@freemail.et 9 Alemayehu Zemende Ethiopia Breeder in Durum wheat azemede@yahoo.com 10 Elmoiez M. Fadul Sudan Soil Scientist fadulen@hotmail.com 11 Abbas Elsir M. Elamin Sudan Socio-economist abbaselsir@yahoo.com 12 William Wamala Wagoire Uganda Breeder/National Wheat Co-ordinator wwagoire@infocom.co.ug 13 Antoine ruzindana Nyirigira Rwanda Agr onomist Researcher in wheat and maize programme a_nyirigira@yahoo.fr 15 C.A. Kuwite Tanzania Pathologist ckuwite@yahoo.co.uk 16 M. Prashar India/ICAR Pathologist dwrfd@hotmail.com 17 Dr. M. Kinyua Kenya Breeder mgkinyua@africaonline.co.ke 18 Dr. T.C. Riungu Kenya Breeder tcriungu@yahoo.com 19 Dr. A. Wangai Kenya Pathologist/Virologist wangai@africaonline.co.ke 20 Dr. J. Owuoche Kenya Wheat breeder and genetist Owuoche@yahoo.com 21 M. Macharia Kenya Entomologist karinjr@africaonline.co.ke 22 Samuel Migui Kenya Entomologist sam.migui@csiro.au 23 Njau P.N. Kenya Plant breeder njaupnn@yahoo.com 24 Wanyera R. Kenya Plant pathologist cepnjr@wananchi.com

PAGE 240

25 Mwangi H.G Kenya Soil and crop scientist hgachuna@yahoo.com 27 Z. Nyakwara Kenya Socio-economist znmageto@yahoo.com 28 J. Kamwaga Kenya Agronomist johnsonkamwaga@yahoo.com 29 D.O.K. Amadi Kenya Agronomist dvdamadi@yahoo.com 30 Ndiema A.C. Kenya Socio-economist Achesambu@yahoo.com 31 E. Kimani Kenya Biochemist enyambura12@yahoo.com 32 Ooro P.A. Kenya Wheat agronomist paooro@yahoo.com 33 Antony M. Kibe (Dr.) Kenya Agronomist akmwangi@yahoo.com 34 Malinga J. Kenya Plant breeder/entomologist joycemalinga@yahoo.com 35 Kimurto P.K. Kenya Agronomy/Physiology pkimurto@yahoo.co.uk 36 Victor W. Wasike Kenya Agronomist vwwasike@yahoo.com 37 Clerkson M. Mahagayu Kenya Socio-economist mahagayu75@yahoo.com 38 Dr. C.M. Ndirangu Kenya/Egerton Plant breeder mwangicn2000@yahoo.com 39 J.K. Macharia Kenya/Egerton Plant pathologist josmac60@yahoo.com 40 W.P. Masikonde Kenya/Bayer E.A. Marketer www.bayerea.com 41 Prof. E.N. Njoka Kenya/Dean Faculty of Agric. 42 J.M.E. Muchile Kenya/Bayer Crop Science Product Development Co-ordinator/p athologist jmuchile@bayerea.net 43 W.A.O. Kore Kenya/KARIKibos Breeder wikore2000@yahoo. co.ke 44 John N. Ndungu Kenya, Kari-Njoro Biochemist/chemist Joched2003@yahoo.com 45 Dr. Pfeiffer Wolfgang CIMMYT Principal sc ientist, head breeding (small grains) IAE programme W.PFEIFFER@CGIAR.ORG 46 Dr. Ravi Singh CIMMYT Principal scientist (geneticist/pathologist) r.singh@cgiar.org 47 Dr. Dave Hodson CIMMYT 48 Dr. Prashar M. CIMMYT 49 Dr. Alpha Diallo CIMMYT

PAGE 241

1. Abbas Elsir Mohammed Elamin Agricultural Research cooperation Wad Medani Sudan 2. Elmoiez M. Fadul Faculty of Agriculture University of Khartoum Dongola Research station Dongola Sudan 3. Minale Liben Adet Agricultural Research Centre P. O Box 08 Bahir Dar, Ethiopia 4. Tolera Abera Oromiya Agricultural Research Institute Bako Agricultural Research Centre P. O Box 03, Oromiya Ethiopia 5. Desalegn Debelo, EARO Kulumsa R. C., P. O Box 489 Asella, Ethiopia 6. Tadesse Dessalegn, Adet Agricultural Research Centre P. O Box 08, Bahir Dar Ethiopia 7. Solomon Gelalcha Kulumsa Agricultural Research Centre, P. O Box 489, Asella, Ethiopia 8. Richard V. Ndondi Selian Agricultural Research Institute P. O Box 6024 Arusha Tanzania 9. Catherine Kuwite Selian Agricultural Research Institute P. O Box 6024 Arusha Tanzania 10. William Wamalwa Wagoire Kachwekano Agricultural Research and Development Centre P. O Box 421 Kabale Uganda 11. Alemayehu Zemede Agricultural Research Centre P. o Box 32 Debre-Zeit Ethiopia 12.Kassa Getu Ethiopia Research Organization, Holetta Agricultural Research Centre P. O Box 2003, Addis Ababa, Ethiopia 13. Emebet Fekadu Ethiopian Agricultural Organization, Research Centre P. O Box 37, Ambo Ethiopia 14. Bemnet Gashawbeza Mengesha Ethiopian Agricultural Research Organization, Debre Zeit Agricultural Research Center, P.O. Box 32, Debre Zeit, Ethiopia; 15. Temesgen Kebede Ethiopian Agricultural Research Organization, Kulumsa Research Centre P. O Box 489, Asella Ethiopia

PAGE 242

16. Kinyua M.G KARI Njoro Kenya 17.Nyakwara Z. A KARI-Njoro Kenya 18. Ndiema A. KARI-Njoro Kenya 19. Macharia M. KARI-Njoro Kenya 20. Amadi DOK KARI-Njoro Kenya 21. Ravi Singh CIMMYT 22. Hugo DG CIMMYT 23. Diallo CIMMYT 24. Prashar M. India 25. Kamwaga KARI-Njoro Kenya 26. Shivaji P CIMMYT 27. Kibe M. A. Egerton University Kenya 28. Kore W. KARI-Njoro Kenya 29. Kimani E KARI-Njoro Kenya 30. Abebe Demisse ASARECA 31. Kimurto P. Egerton University Kenya 32. Njau P. KARI-Njoro Kenya 33. Wolfgang P. CIMMYT 34. Migue S. Australia 35. Wanyera R. KARI-Njoro Kenya 36. Malinga J. KARI-Njoro Kenya 37. Wangai A. KARI-Njoro Kenya 38. Riungu T. KARI-Njoro Kenya 39. Owuoche J. KARI-Njoro

PAGE 243

Kenya 40. Ooro P. KARI-Njoro Kenya 41. Mwangi H. KARI-Njoro Kenya 42. Macharia M. G Egerton Unoiversity Kenya 43. Zubeda Mduruma CIMMYT Ethiopia