1
NGWUTA, AGWU ABRAHAM
PG/M PG/Ph.D/99/27231.
DEVELOPING MAIZE (Zea mays) POPULATIONS
RESISTANT TO STEM BORERS FOR SOUTHEASTERN
NIGERIA.
Crop Science
A THESIS SUBMITTED TO THE DEPARTMENT OF CROP SCIENCE,
UNIVERSITY OF NIGERIA, NSUKKA,
Webmaster
2009
UNIVERSITY OF NIGERIA
2
DEVELOPING MAIZE (Zea mays) POPULATIONS RESISTANT TO STEM
BORERS FOR SOUTHEASTERN NIGERIA.
BY
NGWUTA, AGWU ABRAHAM
B. Agric. (Hons.), M. Sc. Plant Breed. & Genetics (Nig).
PG/Ph.D/99/27231.
A THESIS SUBMITTED TO
THE DEPARTMENT OF CROP SCIENCE,
UNIVERSITY OF NIGERIA, NSUKKA,
IN FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF
DOCTOR OF PHILOSOPHY IN PLANT BREEDING AND GENETICS.
MARCH, 2009
3
CERTIFICATION
Abraham Agwu Ngwuta, a postgraduate student in the Department Crop
Science with Reg. No. PG/Ph.D/99/27231 has satisfactorily completed the
requirement for research study for the degree of Doctor of Philosophy in Plant
Breeding and Genetics.
The study contained in this thesis is original and has not been submitted in part or
in full for any other diploma or degree of this or any other University.
Ngwuta, Abraham Agwu
(Student)
Prof. I. U. Obi
(Supervisor)
Prof. K. P. Baiyeri
(Head of Department)
Dr. S. O. Ajala
(Co-Supervisor)
4
DEDICATION
This study is dedicated to the ALMIGHTY GOD who has given me the grace
to undertake this study, been with me at every stage and ensured its completion. To
HIM be the glory and praise. Amen.
5
LIST OF CO-AUTHOURED CONFERENCE PAPERS ON MAIZE
I. Ngwuta, A.A., S.O. Ajala, I.U. Obi and E.E. Ene-Obong (2001). Potential sources
of resistance to stem borers in local maize populations of south-eastern
Nigeria. African Crop Science Conference Proceedings. Vol. 5 pp 23-28.
2. Ngwuta, A.A., I.U. Obi, S.O. Ajala and E.E. Ene-Obong (2005). Yield stability
test for cultivated open pollinated maize (Zea mays L.) genotypes grown in
southeastern Nigeria. Proceedings of the 39th
Annual Conference of the
Agricultural Society of Nigeria (ASN), pp 217-220.
6
ACKNOWLEDGEMENT
To God is the glory for bringing me this far, sustaining my hands and granting me
the completion of this study. I live to testify to the faithfulness of The Almighty God who
is worthy to be praised.
I am grateful to my supervisor, Professor (Chief) I. U. Obi (OBI-
DIORAMA I) for his patience and the kind way he guided me. I am highly indebted to
Dr. S. O. Ajala, my IITA supervisor, for co-supervising and kindly providing the facilities
for the conduct of the field trials. The financial contribution of Dr. Abebe Menkir and the
invaluable lectures of Prof. M. A. B. Fakorede are highly appreciated.
I appreciate the contributions of Prof. E. E. Ene-Obong and Dr. R. E. C. Mba who
first drew my attention to the stem borer damage in southeastern Nigeria and assisted me
in securing the contacts for the conduct of the study at the International Institute of
Tropical Agriculture (IITA), Ibadan.
The contributions of my teachers, Prof. J. E. Asiegbu, Prof. B. N. Mbah and Prof.
M. I. Uguru are highly appreciated in the able way they taught me. I recall with delight
the pieces of advice and encouragements of Prof. K. P. Baiyeri and Dr. Agwu E. Agwu.
May the Almighty God bless you.
I heartily acknowledge the contributions of Messers; Bamiyo Anjorin, Olumide
Ibikunle, King Adepoju, Emmanuel Itegboje, Layi Adeniyi, Femi Ohunakin and Ayo
Alimi, all staff of the Maize Improvement Programme of IITA, Ibadan. I appreciate your
guidance in the rudiments of proper field technique and data collection.
I salute Dr. Alexander Chukwunweike Odiyi (Big Brother) for his assistance,
care and provisions during the period of the work and beyond. God shall continue to
bless you. I wish to use this opportunity to thank Dr. Ayo Salami, Dr. and Mrs.
Kaka Meseka. Mac Ittah, Dr. Lum Fontem, Buky Bamkefa etc., members of the
7
International Association of Research Scholars and Fellows of IITA (IARSAF-
IITA), Ibadan. I fondly remember the sleepless nights spent together in the
Computer Training Laboratory, the hybridisation of ideas and disagreements
among members. I also give my thanks to Mrs. Chinyere Woods, the Manager of the
Computer Training Laboratory, IITA, for the able way she guided me in the
efficient use of the computers in the management of my data.
I acknowledge with gratitude the assistance of Dr. Yemi Olojede for giving
me the MSTAT software and Prof.. C. O. Muoneke for guiding me on the use of the
software to do diallel analysis. My gratitude also goes to my senior colleagues and
friends, Prof. D. I. Osuigwe, Dr. Alex Awurum, Dr. Emma Opara, Dr. B. A. C.
Agugo, Prof. (Mrs.) G. Onuoha (Ada Zion), Dr. I. J. Ogoke, Dr. M. Agu and Dr. I. I.
Ibeabuchi for being there for me at difficult moments. I appreciate the contributions
of Uzoamaka. K. Agwatu for assisting in organising the local maize cultivars.
I appreciate the assistance and encouragement of my “Big Sister” and her
husband, Prof. & Mrs. S. N. Agwu. My appreciation also goes to Engr. & Mrs.
Josiah Ngwuta, Pastor & Mrs. Emmanuel Nome, Mr. & Mrs. Aja Ngwuta, Mr. &
Mrs. Ogbonnaya Nwani, Pastor & Mrs. Mike Oko and Mr. Ogbonnaya Ngwuta for
their prayers, support and patience while the study lasted.
I am highly indebted to my neighbours, Elder & Elder (Mrs.) Uduma Ikpa and
their children: Amarachi, Chidinma, Nnenna, Kelechi and Ochea. I appreciate the
fatherly love shown my family and me especially the periods I was away to Ibadan
to attend to my study. I pray God to reward you bountifully. I thank Elder Mrs.
Ukoha and Elder E. Udonwu for their prayers and encouragement.
I sincerely appreciate the patience and prayers of my parents who have eagerly
awaited the completion of this study. I thank you for charting this course for me and
8
laying the foundation for moral and academic excellence. I promise you to keep the good
name of the family. My appreciation also goes to Chief Abel Okike and his amiable wife
Lolo Ezinne Okike, for their prayers and encouragements.
Finally, I warmly appreciate the support of my wife, Madam Nneka Ngwuta, who
has patiently stood by me through thick and thin as a result of the exigencies of the
research study. I thank God Almighty for my three kids born to the family while this
study lasted. They depict the goodness of the Lord. To my first son, Nnanna Faithful
Emmanuel “Abidemi”, born 5th
September 2002 (the day I was in Akure planting my last
trial), I promise to make up to you for being absent at your birth. And the second,
Chidiebube Ogonna Johnson who prefers to sit on my laps each time I want to work on
this project at home; I hope that when you are old enough you will understand. However,
if that is your form of encouragement, I appreciate it. To the baby girl, Gift Nmesoma
Grace, born on my birthday (June 15th
2009) we love you.
9
TABLE OF CONTENTS
Title Page------------------------------------------------------------------------------------ i
Certification--------------------------------------------------------------------------------- ii
Dedication----------------------------------------------------------------------------------- iii
List of Co-authoured Conference Papers on Maize Improvement ------------------- iv
Acknowledgement-------------------------------------------------------------------------- v
Table of Contents--------------------------------------------------------------------------- viii
List of Tables-------------------------------------------------------------------------------- x
List of Figures------------------------------------------------------------------------------- xiii
List of Appendices-------------------------------------------------------------------------- xiv
Abstract-------------------------------------------------------------------------------------- xv
INTRODUCTION-------------------------------------------------------------------------- 1
LITERATURE REVIEW----------------------------------------------------------------- 4
Host plant resistance (HPR)--------------------------------------------------- 10
Screening and Selection of Promising Genotypes ------------------------ 10
Mating designs and combining abilities ----------------------------------- 12
Heterosis or hybrid vigour----------------------------------------------------- 14
MATERIAL AND METHODS----------------------------------------------------------- 15
Experiment 1: Field evaluation of local maize germplasm for
resistance to stem borers in four environments-----------------------------
18
Cultural practices--------------------------------------------------------------- 20
Data collection------------------------------------------------------------------ 21
Statistical analysis-------------------------------------------------------------- 23
10
Estimation of genetic variance components--------------------------------- 23
Heritability estimates----------------------------------------------------------- 24
Experiment 2: Diallel evaluation to obtain information on combining
ability and heterosis of selected genotypes and generate reciprocal
populations for further improvement.----------------------------------------
26
Statistical analysis--------------------------------------------------------------- 27
Estimation of heterosis---------------------------------------------------------- 28
RESULTS------------------------------------------------------------------------------------ 29
Multivariate and cluster analyses: using data from artificially infested
plots.------------------------------------------------------------------------------ 34
Multivariate and cluster analyses: using data from naturally infested
plot.------------------------------------------------------------------------------- 45
Multivariate and cluster analyses: using data from non-infested plots
from Ibadan and Ikenne locations, combined.----------------------------- 53
The combining ability and heterotic effects for agronomic attributes and
stem borer damage parameters in the 11 selected genotypes.--------------
61
DISCUSSION------------------------------------------------------------------------------- 85
SUMMARY AND CONCLUSION --------------------------------------------------- 91
REFERENCES --------------------------------------------------------------------------- 93
APPENDICES----------------------------------------------------------------------------- 106
11
LIST OF TABLES
Table
Title
Page
1 Codes, collection sites and entry (genotypes) numbers for the local open-
pollinated and improved maize genotypes evaluated as potential sources of
resistance to S. calamistis and E. saccharina.
19
2 Form of analysis of variance showing sources of variation, degrees of
freedom and mean squares only, for the evaluation of 212 maize genotypes
in four different environments.
25
3 Combined analysis of variance for agronomic and stem borer damage
parameters of 212 maize genotypes evaluated in two (2) environments
30
4 Combined analysis of variance for agronomic and stem borer damage
attributes of 212 maize genotypes evaluated in four (4) environments
31
5 Variance components and heritability (broad sense) estimates of some
agronomic and stem borer damage parameters of the 212 maize genotypes
32
6 Primary data on the evaluation of 212 maize genotypes in two (2) to four (4)
environments
35
7 Plant characteristics, their ranks and rank summation index of the 212 Maize
genotypes evaluated under artificially infested, naturally infested and non-infested
field conditions
36
8 Correlation matrix of some agronomic and damage parameters of the 212
maize genotypes obtained using the data from artificially infested plots
38
9 Discriminant score coefficient matrix of the variables for the first four
canonical discriminant variables obtained for the 212 genotypes using data
from the artificially infested plot
39
10 Distribution of the 212 genotypes in clusters according to their similarities
determined using the data set from artificially infested plots
42
11 Number of genotypes, collection sites and site code of the 212 maize
genotypes shown according to the clusters obtained from the cluster analysis
using data from the artificially infested plots
43
12 Cluster means of agronomic and damage parameters of 212 maize genotypes
obtained from the cluster analysis using data from the artificially infested
plots
44
13 Correlation matrix of some agronomic and damage parameters of the maize
genotypes obtained using data from the naturally infested plots
46
12
14 Discriminant score coefficient matrix of the variables for the first four
canonical discriminant functions obtained for 212 maize genotypes using
data from the naturally infested plots
47
15 Distribution of the 212 maize genotypes in clusters according to their
similarities obtained using data from the naturally infested plots
50
16 Number of genotypes, collection sites and site codes of the 212 maize
genotypes shown according to the clusters obtained from the cluster analysis
using data from naturally infested plots
51
17 Cluster means of agronomic and damage parameters of 212 maize genotypes
obtained from the cluster analysis using data from the naturally infested plots
52
18 Correlation matrix of agronomic and borer damage parameters of the 209
local maize cultivars and three (3) improved maize genotypes obtained using
the noninfested data set
54
19 Discriminant score coefficient matrix of the variables for the four canonical
discriminant functions obtained using the noninfested data sets combined
from Ibadan and Ikenne in southwestern Nigeria
55
20 Distribution of the 212 maize genotypes into clusters according to their
similarities using the noninfested data sets from Ibadan and Ikenne in
southwestern Nigeria
58
21 Number of genotypes, collection sites and site codes shown according to the
clusters obtained
59
22 Cluster means of agronomic and damage parameters of 212 maize genotypes
for the non-infested data sets combined from Ibadan and Ikenne in
southwestern Nigeria
60
23 Combined analysis of variance for stem borer damage parameters of 11
open-pollinated maize genotypes and their crosses evaluated in four (4)
environments
62
24 Combined analysis of variance for agronomic traits of 11 open-pollinated
maize genotypes and their crosses evaluated in seven (7) environments
63
25 Combined analysis of variance for post flowering and agronomic traits
of 11 open-pollinated maize genotypes and their crosses evaluated in
seven (7) environments
64
26 Variance components, heritability (broad sense) and coefficient of variation
estimates of some agronomic and damage parameters obtained from eleven
open-pollinated maize genotypes and their crosses
66
13
27 Means of genotypes (on diagonal italics underlined) crosses (above
diagonal) and SCA effects (below diagonal), GCA effects and array
means for the 11 genotypes evaluated for resistance traits to S. calamistis
and E. saccharina
67
28 Means of genotypes (on diagonal italics underlined) crosses (above
diagonal) and SCA effects (below diagonal), GCA effects and array
means for the 11 genotypes evaluated for two (2) agronomic traits
69
29 Means of genotypes (on diagonal italics underlined) crosses (above
diagonal) and SCA effects (below diagonal), GCA effects and array
means for the 11 genotypes evaluated for three (3) agronomic traits
70
30 Means of genotypes (on diagonal italics underlined) crosses (above
diagonal) and SCA effects (below diagonal), GCA effects and array
means for the 11 genotypes evaluated for three (3) agronomic traits
72
31 Means of genotypes (on diagonal italics underlined) crosses (above
diagonal) SCA effects (below diagonal), GCA effects and array means
for the 11 genotypes evaluated for three (3) agronomic traits.
74
32 Means of genotypes (on diagonal italics underlined) crosses (above
diagonal) SCA effects (below diagonal), GCA effects and array means
for the 11 genotypes evaluated for three (3) yield attributes
78
33 Better-parent (above diagonal) and mid-parent heterosis (below
diagonal) of open-pollinated maze crosses for resistance to maize
damage parameters
79
34 Better-parent (above diagonal) and Mid-parent heterosis (below
diagonal) of open-pollinated maze crosses for flowering and agronomic
traits
80
35 Better-parent (above diagonal) and Mid-parent heterosis (below
diagonal) of open-pollinated maze crosses for three (3) agronomic traits
81
36 Better-parent (above diagonal) and Mid-parent heterosis (below
diagonal) of open-pollinated maze crosses for three (3) agronomic traits
82
37 Better-parent (above diagonal) and Mid-parent heterosis (below
diagonal) of open-pollinated maze crosses for three (3) agronomic traits
83
38 Better-parent (above diagonal) and Mid-parent heterosis (below
diagonal) of open-pollinated maze crosses for three (3) agronomic traits
84
14
LIST OF FIGURES
Figures Title
Page
1 Plot of canonical discriminant variables of the cluster analysis of the 212
maize populations using the artificially infested field data set from Ibadan,
Nigeria.
40
2 Plot of canonical discriminant variables of the cluster analysis Of the 212
maize genotypes using the naturally infested field data set from Egbema,
Nigeria.
48
3 Plot of canonical discriminant variables of the cluster analysis of the 212
maize populations using the non-infested field data sets Combined from
Ibadan and Ikenne in Southwestern Nigeria.
56
15
LIST OF APPENDICES
Appendices Title Page
1 Plant characteristics, their ranks and rank summation index of the 212
maize genotypes evaluated under artificially, naturally and non-
infested field plot conditions.
107
2 Cluster Analysis for data set from artificially infested maize plots.
113
3 Form of combined analysis of variance with degrees of freedom, only
used for the eleven open pollinated maize genotypes and their crosses
evaluated in seven (7) environments.
118
16
ABSTRACT
Development of maize populations resistant to stem borers depends largely on the
existence of useful genes or alleles, which can combine to confer resistance to progenies.
Such genes are often available in areas of stress, having been responsible for the survival
of such crops over the years. Pink stem borer, Sesamia calamistis (Hampson, Noctuidae)
and sugarcane borer, Eldana saccharina (Walker, Pyralidae) are endemic in southeastern
Nigeria. Damages caused by the larvae of these moths are more prevalent during the
second planting season (August-November). Genetic diversity for a range of agronomic
and resistance attributes within 209 local maize collections from southeastern Nigeria and
3 improved check varieties were investigated in field trials in randomised complete block
design (RCBD) with two replications across three environments. Data collected from the
evaluations were subjected to both uni- and multivariate statistics. Furthermore, four traits
namely, leaf feeding, ear damage, shoot breakage and yield were used from across three
environments to construct a selection index. The multivariate analysis on the plant
attributes, using canonical discriminant analysis, revealed the agronomic and borer
damage parameters that contributed significantly to the total variation observed in
different environments. Out of the four canonical discriminant functions obtained, two
had significant (P=0.05) eigenvalues accounting for over 98 % of the total variation. The
first canonical function was mainly associated with yield while the second was associated
with the borer damage attributes. Rank summation index (RSI) used to rank the entries
for resistance to stem borers identified 11 genotypes representing top 5 % of the total as
resistant. In the second experiment the 11 genotypes and their hybrids, made in a diallel
fashion were evaluated for agronomic and borer damage attributes in seven environments
in RCBD with three replications. Data collected were subjected to analysis of variance
and those found significant (P=0.05) were further subjected to diallel analysis using
17
Griffing‟s method 2 model 1 for fixed effects. Significant GCA and SCA effects were
obtained for most of the traits studied in the various environments and in the pooled
environment thus indicating that additive and non-additive gene effects were involved in
the expressions of the traits studied. However, in a few cases, only GCA or SCA was
important thus indicating the relative importance of the genetic component of the
variance. The assessment of the agronomic and borer damage attributes of the parents and
the crosses indicate that the variety crosses were not superior to the parents in most of the
traits. The significant differences observed between the parents and the crosses for dead
heart and leaf feeding damage parameters is suggestive of the occurrence of exploitable
heterosis for the development of genotypes that are resistant to stem borer attack.
Genotypes SE NG-33, SE NG-65 and TZBR Syn W had high negative GCA values for
dead heart while SE NG-62, SE NG-148, TZBR Syn W and TZBR ELD 3 C2 had the
high negative GCA values for leaf feeding damage. For ear damage, SE NG-65, SE NG-
67, SE NG-119, SE NG-148 and AMA TZBR-W-C1 had high negative GCA estimates.
Genotypes SE NG-33, SE NG-62, SE NG-65, SE NG-77, SE NG-106 and SE NG-119
had the highest positive GCA effects for grain yield. The nine genotypes selected formed
two heterotic pools: Group A comprised SE NG-33, SE NG-77, SE NG-106, SE NG-148
and TZBR Syn W while Group B included SE NG-62, SE NG-119, AMA TZBR-W-C1
and TZBR ELD 3 C2. Average yield of the grouped genotypes crossed in all possible
combinations was 1.06 t ha-1
showing 5 % yield increase. Furthermore, the best five
yielding crosses namely; SE NG-33 x TZBR ELD 3 C2, SE NG-62 x SE NG-77, SE NG-
62 x SE NG-106, SE NG-106 x TZBR ELD 3 C2 and TZBR Syn W x TZBR ELD 3 C2,
selected may be used as population crosses or in the formation of composite varieties.
18
INTRODUCTION
Maize (Zea mays L.) is the third most important cereal in the world after wheat
and rice. In Nigeria, maize is popular and widely grown essentially because it matures
during the “hunger period” and can be prepared in a variety of ways. In southern Nigeria,
maize is a major component of the cropping system serving as hunger breaker while other
crops are yet to mature.
In the rain forest zone of southern Nigeria, two crops of maize are possible per
year due to the bi-modal rainfall pattern of the zone. The first season crop can be planted
from mid March to first week of April while the second season planting is from mid
August to early September. The maize produced in the early season is quickly consumed
to avoid damage due to high humidity related diseases and pests. Storage is best with late
maize during the onset of the dry season. Unfortunately, late season maize production is
seriously limited by the activities of stem borers (Obi, 1991). The pink stem borer (S.
calamistis (Hampson)) and the sugarcane stem borer (E. saccharina (Walker)) are the two
stem borer species of economic importance in Southeastern Nigeria (Harris 1962; Appert,
1970; Bowden, 1976). The activities of the larva on the maize plants result in leaf feeding
and stem tunneling, which in turn lead to reduced translocation of nutrients and
assimilates, death of young plants (dead heart), lodging of older plants and direct damage
to maize ears (Usua, 1968; Ezueh, 1978; Bosque-Perez and Mereck, 1990). All these
damage activities tend to cause yield reduction and crop failure. Yield loss of between 10
to 100 % have been reported for stem borer attack in this region (Usua, 1968)
Control measures advocated for stem borers include direct use of insecticides,
cultural control practices especially inter-cropping, early planting and good sanitation
including burning of crop residue and the use of host plant resistance (HPR) (Lawani,
19
1982). Host plant resistance when strategically deployed in appropriate cropping system
is both cost effective and environmentally safe. Therefore, it is often regarded as the hub
in any integrated pest management (IPM) intervention for stem borer control (Teetes,
1985; Kogan, 1982; Belloti, 1990).
Whenever good sources of resistance for desirable traits are identified, appropriate
breeding methods, such as recurrent selection, can be employed to increase the frequency
of such desirable genes in order to further increase productivity of such crop. Crop
improvements depend mainly on the availability of genetic variability. Such variability
can be obtained through introduction, selection from available variation, generated
through mutation or through the use of biotechnological tools to obtain desired genes for
desirable traits. Conventional method of developing resistant varieties involves the
identification and use of resistant germplasm in breeding programmes. In looking for
resistant sources, one approach is to search for germplasm in areas where stresses are
prevalent. This approach can identify genotypes with resistance to local stresses
including diseases and insect pests that are also adapted to local ecological problems such
as low soil pH, low soil nutrient and root and stalk lodging (Fajemisin et al., 1985; Kim et
al., 1985; Eberhart et al., 1991).
Maize is not native to Southern Nigeria therefore, all the maize varieties grown in
this region must have been improved varieties introduced in not too distant past and
maintained by the farmers over the years. Usually, farmers‟ selections of seeds for the
next crop represent a form of mass selection for tolerance to environmental stresses such
as insect pests, plant diseases, drought etc. Evidence of exploitable genes for resistance
to maize stem borers is available in literature (Ajala et al., 1995 and Ngwuta et al., 2001).
At IITA, some sources of resistance to S. calamistis and E. saccharina have been
identified and used to form TZBR populations (Bosque-Perez et al., 1989, Kling and
20
Bosque-Perez, 1995). In the course of developing resistant populations, efforts were
aimed at breeding for resistance to these borers separately (Kling and Bosque-Perez,
1995). Some workers (Williams and Davis, 1984; Smith et al., 1989; Wiseman and Davis,
1990; and Mihm, 1995) have noted that the best strategy to a successful host plant
resistance programme is the development of multiple insect resistant varieties. This
approach is currently being used at IITA to develop genotypes with resistance to both
Sesamia calamistis and Eldana saccharina (Schultess and Ajala, 1997; Ajala et al.,
2002). The aim of this study was therefore to identify potential sources of multiple
resistances to stem borers of interest and to generate genetically broad based reciprocal
populations for further improvement efforts. Reciprocal populations have the advantages
of complimenting each other for maximizing heterosis either as varietal crosses or in
inbreds extracted from them and for continuous improvement of the two populations.
The objectives of this study were to:
i. evaluate local and a few improved populations from Southeastern Nigeria for
agronomic traits and stem borer damage parameters
ii. investigate the major characters responsible for the variations among the
maize genotypes assembled, and group them into homogenous subsets so that
representative genotypes can be selected for further studies
iii. investigate the combining ability and heterosis for agronomic attributes and
stem borer damage parameters in the selected genotypes, and
iv. identify the heteroic groups that can be used in inter-population improvement
schemes for the development of high yielding varieties or hybrids.
21
LITERATURE REVIEW
Maize (Zea mays L.) originated from central and south America (Jeffreys, 1967;
Butchinson, 1974; Brucher, 1989) and is presently cultivated in virtually all parts of the
world. It has the highest yield potential compared to other cereal crops (Rehm and Espig,
1991) and ranks third as the most important after wheat and rice (FAO, 1968). The
leading producers of maize are the America producing 185 million tons per year with
United States of America alone producing 125 million tons/year. This is closely followed
by Asia with 107 million tons/year; Europe with 66 million tons while Africa produces
only 30 million tons/year (Rehm and Espig, 1991). Furthermore, based on maize
utilization attributes, cultivated maize varieties have been classified into the following
three groups. Flint maize are, considered the most important for human consumption
while dent maize which gives the highest grain yield are very useful for feed and
industrial uses. The third group, floury maize have soft endosperm and are suitable for
starch production. Other important maize types for more specialized uses are the popcorn
and sweet maize (ref). Whatever the mode of introduction, available record indicates that
maize was reported for the first time in West Africa in 1498 (Van Eignattan, 1965) and
has been in the diet of Nigerians till date. Gradually, the crop has grown from subsistence
to commercial crop on which many agro-based industries depend for raw material.
The rain forest zone of Southeastern Nigeria has long duration of rainfall (April-
November) that is characterized by bimodal distribution. Two crops of maize are
therefore possible per year due to this bimodal distribution because each of the seasons
can fully support maize growth. Nonetheless, storage is best with late season maize
during the onset of the dry season. Unfortunately, late season maize production is
seriously limited by the activities of stem borers. Several workers have reported on stem
22
borers as one of the most serious constraints to maize production in the West and Central
Africa (SAFGRAD, 1967; Fajemisin et al., 1985; Bosque-Perez and Mereck, 1990; Obi,
1990; Shanower et al., 1991; Guonou et al., 1994; Kling and Bosque-Perez, 1995;
Kouame, 1995; Phiri, 1995; Cardwell et al., 1997; Schulthess et al., 1997).Severe crop
yield losses of between 10-100% have been reported (Usua 1968). Reports from other
studies conducted elsewhere have reported stem borer damage of similar magnitude
(Sarup, 1980; Smith et al., 1989; Ajala and Saxena, 1993; Monitor, 2001).
Adeyemi et al., (1966) reported that in West and Central Africa, S. calamistis and
E. saccharina tend to be a problem in areas with bimodal rainfall and during the second
cropping season, only. These areas include the forest-savanna transition and the mid-
altitude zones. The pink stem borer (S. calamistis (Hampson)) and the sugarcane stem
borer (E. saccharina (Walker)) are the most damaging and widespread species of
economic importance in Southeastern Nigeria (Harris, 1962; Appert, 1970; Bowden,
1976). S. calamistis and E. saccharina are polyphagous insects that feed on maize,
sugarcane, sorghum and other grasses. S. calamistis breed throughout the year and has no
resting stage (Harris, 1962). The adults that emerge at the beginning of the cropping
season are smaller and less fecund than those emerging later in the year (Bowden, 1976).
The combined effect of smaller number of less fecund adults results in lower incidence of
Sesamia species in the first season maize crop. The insect growth and the population
increase until it peaks around August-September, when the second maize crops are being
grown. E. saccharina is far more abundant in second season maize crops (Girling, 1980),
having laid its eggs on the senescing parts of maize of the early season.
Damage caused by stem borers is complex because, depending on species, the
insects occupy various parts of the plant and attack the plant at all stages of growth. The
S. calamistis attack the plant during the early stage while the E. saccharina attack during
23
the later stage of the plant development (flowering). E. saccharina lays its eggs on
senescent plant parts, thus preferring tasseling or post tasseling stages. This allows for a
secondary generation only on plants that remain in the field after grain maturity (Sekloka,
1996). Adult females of the S. calamistis lay their eggs between the leaf sheaths and leaf
whorl of the plants. The eggs hatch into early instars which feed on maize leaves. Later
instars migrate to the stem and other parts of the plant where they continue feeding.
Larvae of E. saccharina attacking at about the time of anthesis feed on tassels, leaf
sheaths, collars, midribs and developing ears. The early feeding activities of the larva on
the maize plants lead to leaf feeding and stem tunneling damage. Dead heart occurs when
the meristematic area of the plant is damaged while the overall effect of the tunneling can
lead to lodging of older plants. Attacks during the later growth stages result in damage to
maize ears (Usua, 1968; Ezueh, 1978; Bosque-Perez and Mereck, 1990). All these
damage activities usually lead to various levels of yield reduction and sometimes a total
crop failure. Various workers (Eignattan, 1965; Usua, 1968; Endrody-Younga, 1968;
Girling, 1980; Bosque-Perez and Mereck, 1990; Gounou et al., 1994; Setamou et al.,
1995) have reported yield losses ranging from 10 to 100 %. Ajala and Saxena (1994) also
reported yield reduction of 34-43 % resulting from a combination of stem tunneling and
damage inflicted on the maize ears by the spotted stem borer specie (Chilo partellus
(Swinehoe).
Control measures advocated for stem borers include direct use of insecticides,
cultural and biological measures and the use of host plant resistance (HPR). Chemical
pesticides are in general popular with farmers (Warni and Kuria, 1983) because of their
quick and effective action in control of insect pests. However, the risks to human lives
and to the environment are so great that there is urgent need to change to crop protection
techniques which are less dependent on chemicals. In 1983, there were reports of 2
24
million cases of poisoning by pesticides, 40,000 of which were fatal (Sengooba, 1992).
In Nigeria, the use of chemical pesticides is expensive and not very convenient
particularly with the subsistence farmers, due to the problems of timing of application and
the cost of the insecticides. Farmers have also used as part of their traditional farm
practices, different control measures such as burning of plant residues after harvest, crop
rotation, early planting weeding and inter-planting of different crop species. Although
these cultural control measures have potentials in reducing crop pests, other factors in
farming may limit their effectiveness. For example, resource poor farmers may not be
able to plant their crops all at the same time; those crops planted later may be exposed to
a build up of pests from the earlier plantings, thereby offsetting the effects of cultural
control measures in use. Cultural control measures are inexpensive and in reality rarely
achieves very high levels of control, especially on small-scale farms (Sengooba, 1992)
Biological control is both sustainable and inexpensive and presents no health
hazards or environmental pollution. However, the major shortcoming is the lengthy
detailed ecological studies that are required before a successful package can be
developed. Thus various workers (Wilson, 1988; Jago, 1992; Dickson and Lucas, 1997)
have suggested the use of integrated pest management (IPM) strategy, a multidisciplinary
approach to crop protection where all available methods of reducing the pest population
on a given production system are integrated to achieve optimum economic benefit with
minimal ecological impact.
Host plant resistance has been defined by various workers (Wardle and Burckle,
1923; Mumford, 1931; Painter, 1951; Beck, 1965; Herzog and Fonder-Bark, 1985) as
plant‟s inherent qualities or attributes that render it unsuitable as food or shelter for insect
pests. Resistance to insects can occur naturally in cultivated crops or developed through
organized plant breeding. According to Wiseman (1985), resistance can be classified as
25
to intensities: immunity, high, moderate and low resistance; or by types viz: vertical
(specific) or horizontal (general) resistance (Van Der Plank 1963, 1968). In vertical
resistance the level of resistance offered by a particular host genotype is against a specific
insect biotype whereas in horizontal resistance, the level of resistance offered by the host
genotype is against all insect biotypes. An immune genotype is one which a specific
insect will not damage or use under any known condition. High resistance genotype is
one, which possesses attributes that result in small or minor damage by a specific insect
under a given set of conditions. Moderate resistance or intermediate level of resistance is
one that may result from (a) a mixture of phenotypically high and low resistant plants (b)
plants homozygous for genes which under a given environmental condition produce an
intermediate level of injury or (c) a single clone which is heterozygous for incomplete
dominance for high resistance. Low level of plant resistance indicates attributes
possessed by the genotype that result in less damage or infestation by an insect than the
average for the crop while susceptible genotypes are those on which average or more than
average damage is inflicted by an insect species.
Knowledge of the mechanism of resistance and the interaction between host plant
and pest is important in the use of HPR in pest management because any resistance
mechanism that disrupts the normal behaviour pattern of pests or cause them to wander
around more can expose them to natural enemies or drop off the plant and die (Teetes,
1985). Several authors have described the mechanisms or components of resistance to
include: (1) preference/non-preference (Painter, 1951, 1968) or antixenosis (Kogan and
Orthman, 1978), (2) antibiosis and (3) tolerance. The first two mechanisms of response
involve the responses of insects to plant structural and chemical characteristics
determining their establishment on the plants, while the third involves the response of
plants to insect attack. Plant characters may influence the behavioural responses in two
26
ways: (1) by providing sensory stimuli, and (2) by providing mechanical barriers
example, hardness of tissues, hairiness, silica, lignin etc. DIMBOA, a chemical substance
found in some maize varieties, has been reported (Rojanaridched and Gracen, 1983) to
confer resistance to leaf feeding, sheath and collar feeding caused by European Corn
Borer (Ostrinia nubilis). Mihm (1985) reported that maysin, a chemical substance, found
in the silk of some maize varieties confer resistance to maize against ear worm.
Generally, a tolerant plant shows an ability to grow and reproduce itself or to repair injury
to a marked degree in spite of supporting a population of insects approximately equal to
that damaging a susceptible genotype (Painter, 1951). Wiseman (1985) observed that
resistant genotypes will possess combinations and/or varying levels of the resistance
mechanisms and, that a genotype which is non-preferred does not require the same level
of antibiosis or tolerance that a more preferred genotype must possess. Therefore,
different genotypes may possess the same levels of resistance with different mechanisms
of resistance and/or levels of resistance components. He further suggested that it is
possible to increase the genotype‟s resistance level when only one mechanism is present,
simply by adding another mechanism of resistance. The advantages of HPR as a single
strategy in pest control have been reviewed by many researchers (Maxwell and Jennings,
1980; Smith, 1989) and as a major component in integrated pest management (Teetes,
1985). Host plant resistance is the most practical and environmentally safe method of
pest control on crop fields and when strategically deployed in appropriate cropping
system is both cost effective and environmentally safe. Therefore, it is often regarded as
the hub in any integrated crop or pest management efforts to control stem borer damage.
27
Host Plant Resistance (HPR)
Development of maize varieties resistant to stem borers involves the identification
and the use of resistant sources in breeding programmes. In choosing the germplasm to
initiate a HPR programme, several options are available:
i. Choose materials that have been selected for resistance to the pest species
of concern by other improvement programmes in similar climatic zones
(Chiang and Huden, 1976).
ii. Genotypes known to be resistant to pests, which are related to the species
of concern, can be used.
iii. Genotypes, which are resistant to pests unrelated to species of concern,
have been found to be useful sources (Guthrie et al., 1982; Omolo, 1983).
iv. Use of locally available adapted germplasm (Ajala et al., 1995; Ngwuta et
al., 2001)
v. Screening of exotic germplasm from various breeding programmes and
from germplasm banks.
vi. In cases where resistant materials are not available from the sources
mentioned above, wild relatives of the crop can be screened and identified
resistant sources used to make wide cross for backcrossing...
Screening and Selection of Promising Genotypes
Useful genotypes that can serve as parents of improved varieties need to be
identified from evaluations conducted to discriminate between groups of contending
entries. Several criteria using crop yield and yield components have been used to select
promising genotypes (Diz et al. (1994); Graviois (1998); Guler et al. (2001); Mohammed
28
et al. (2003) and Rabiei et al. (2004); Topal et al. (2004)). Ahmad et al. (1991) used early
maturity while Bridge (2000) and Singh et al. (2004) used crop resistance to insect
damage to select superior genotypes. The use of single trait (univariate) method is not
very desirable due multi colinearity, pleiotropy and linkage. Selection of superior parents
requires that a number of attributes be considered together. Therefore, multivariate
statistical techniques are appropriate for the analysis of aggregates of variables. With the
help of computer-aided software packages it is easy to screen a large number of
genotypes and make selections based on any number of useful traits. Several workers
have reported the use of such procedures as Principal Component Analysis (Falowo,
1982; Ogunbodede, 1997), Canonical Discriminant Analysis (Daoyu and Lawes, 2000),
Correlation and Path-coefficient analyses (Ofori, 1996; Sukchain and Sidhu, 1992), Path
and Cluster analyses (Kozak et al., 2007; Shipley, 2002; Kozak and Kang 2006; Vargas et
al., 2007). The main purpose of discriminant analysis is to find a linear combination of
the measured traits that maximises differences among the pre-existing populations and to
sort the genotypes into their appropriate group with minimum error (Daoyu and Lawes,
2000). The conditions under which the multivariate analyses procedures must be used
include; (i) the traits must show significant differences among the genotypes and (ii) there
must be the presence of significant relationships among traits (Falowo, 1982).
The CANDISC procedure is a dimension reduction technique related to principal
component analysis and canonical correlation. Given two or more groups of observations
with measurements on several quantitative variables, canonical discriminant analysis
derives a linear combination of the variables that has the highest possible multiple
correlation with the groups. The maximum multiple correlation is called the first
canonical correlation or variable while the coefficients of the linear combination are the
canonical coefficients or weights. The canonical variables are essentially used for
29
discriminating among the clusters, and the first two canonical variables are plotted to
show cluster membership.
Mating designs and combining abilities
To obtain information about a base population or to characterize the combining
ability of genotypes is of primary interest of breeders and methods that permit more
accurate assessment of the potentials of germplasms have also been developed. These
include diallel analysis (Jinks 1954, 1956; Hayman 1954a, b; 1957), generation means
analysis (Mather, 1949), and the North Carolina mating designs (Comstock and
Robinson, 1948; 1952). Generation means analysis allows for the separation of gene
effects into additive, dominance and epistasis, thereby enabling the researcher to identify
which of these effects is most important in the inheritance of the trait under consideration.
Each of the North Carolina designs (NCD I, II, III) allows the estimation of additive and
dominance variances. The three designs utilize full- and half-sib progenies that are
produced from biparental mating.
Diallel cross analysis provides the basis for generating preliminary information for
heterotic patterns among population crosses (Hallauer and Miranda Filho, 1981) and aid
in selecting genotypes for further improvement efforts. Diallel crossing schemes and
analyses have been developed for parents that range from inbred lines to broad genetic
base varieties (Gilbert, 1958). Evaluation of diallels may involve one of the four methods
as described by Griffing (1956b); Method I, consists of; F1s, the reciprocals and parents,
Method II includes the F1s and parents, Method III consists of F1s and reciprocals, while
Method IV involves the F1s only. The analysis and interpretation of diallel mating design
vary depending on whether the parents are the reference genotypes or random genotypes
from some reference population. Griffing (1956a) discussed the diallel analysis in detail
30
as well as the analysis of variance where the parents are the genotypes under
consideration (Model I) and where the parents are a sample of genotypes from a reference
population (Model II). Model I estimates apply only to the genotypes included and
cannot be extended to some hypothetical reference population, whereas Model II
estimates are interpreted relative to some reference population from which the genotypes
included are an unselected sample (Hallauer and Miranda, 1995). The Model I analysis
gives information on the hybrids that can be useful for the selection of parents that have
good general combining ability (gca) in a series of crosses and a good specific combining
ability (sca) for specific pairs of parents while Model II in addition to giving gca and sca
provides information regarding the types of gene action. Hallauer and Miranda (1988)
stated that diallel analysis that includes the parents is frequently used for open-pollinated,
synthetic and composite varieties. And these results provide useful information on the
variety performance itself as well as the variety crosses. According to Rood and Major
(1981), the diallel analysis proposed by Hayman (1954a) is particularly useful for
determining the mode of inheritance, whereas that of Griffing (1956a) is more useful to
the applied maize breeder because gca and sca can be calculated for genotypes and
hybrids. Therefore, the performance of specific genotypes can be evaluated. General
combining ability (gca) measures the mean performance of a genotype in while sca is the
deviation from the expected sum of the gca estimates of the two parents involved in a
cross. Therefore, gca measures mainly the additive effects of genes while sca is a measure
of dominance. Several workers have used diallel analysis to study quantitative inheritance
of different traits in maize (Genter and Eberhart, 1974; Yaounes and Andrew, 1978;
Williams et al., 1978; Cantrell and Geadelmann, 1981; Albergoni et al., 1983). All the
workers reported a preponderance of gca, indicating that additive genetic effect was more
important than the non-additive genetic effect (sca) for expressing the various traits
31
studied. Studies by Green (1948), Grissom and Kalton (1956) and Leffel and Hanson
(1961) suggest that combining ability is heritable. Lonnquist (1951) and Stangland et al.
(1963) provided empirical data to show that combining ability can be modified by
recurrent selection. Some workers (Rojas and Sprague, 1952; Matzinger et al., 1959;
Lonnquist and Gardner, 1961; Miller and Marani, 1963; Fatunla, 1973; Younes and
Andrew, 1978) have highlighted forms of the analysis of variance involving the
interaction of gca and sca with location and/or year while Ajala et al., (1986)
elaborated on the procedure for calculating the sum of squares for the interaction terms.
In most cases, the interactions of gca and sca with environments (locations and/or years)
were small. These results suggest that the genotypes in each experiment were consistent
in the expression of combining ability across the environments used.
Heterosis or Hybrid Vigour
Large increases in productivity have been achieved in maize by taking advantage
of heterosis occurring in the first generation of crosses between inbred lines or between
varieties. The development and growing of hybrid maize can justifiably be called one of
plant breeders‟ greatest accomplishments and this success has established a platform that
is being used for other crop species. Heterosis or hybrid vigour is defined as the
superiority of a hybrid over the mean of its parents (mid-parent heterosis) or over the
better parent (better-parent heterosis) with respect to the trait under consideration (Liang
et al., 1972). Several workers have attempted to explain heterosis on either the basis of
genetic or physiological manifestations (Whaley, 1964). Dominance and overdominance
gene-action hypotheses have been advanced in support of the genetic basis (Jinks, 1955),
while nutritional factors and/or growth substances and enzyme systems have been
proposed to explain the physiological basis (Whaley, 1964; Patanothai and Atkins, 1971).
32
Consequently, hybrids are deemed to have increased rate of juvenile growth due to
increased production of growth substances, better utilization of nutritional factors and
more enzyme systems. Thus, heterosis is believed to result from one or more of the
following genetic situations (Moll et al., 1964; Hayes and Foster, 1976; Mushonga et al.,
1997):
(i) Dominance: the accumulated action of favourable dominant genes ;
(ii) Interaction of non-allelic genes or epistasis: complimentary interaction of
additive, dominant and recessive genes at different loci; and
(iii) Intra-locus or intra-allelic interaction or overdominance: favourable
interaction between two alleles at the same locus.
The dominance hypothesis assumes that heterosis is due to the accumulation of
favourable dominant genes as a result masking the deleterious recessive genes in hybrids.
The overdominance hypothesis postulates that the heterozygous condition per se is
responsible for heterosis (Jinks, 1955) and that the heterozygous state of the loci confers a
physiological advantage upon the hybrid. Jinks (1955) proposed analyses which show
that:
(i) Wherever overdominance occurs, non-allelic interaction also occurs,
(ii) If allelic interaction is omitted by removing crosses showing the
interaction, a reduction in the apparent degree of overdominance occurs,
and
(iii) Specific combining ability is always associated with the presence of non-
allelic interaction, whereas general combining ability results from simple
dominance.
Beal (1880) was the first to study heterosis in crosses between open pollinated
varieties of maize and reported a 51% superiority in yield over the parents and
33
consequently proposed the use of this procedure as a method of increasing yield in
maize. Later, the striking research result of inbreeding depression and heterosis among
inbred lines diverted attention to the use of hybrids resulting from inbred lines. However,
in the 1950s the studies and use of variety crosses received renewed interest because of
the development of quantitative genetics and recurrent selection procedures for the
improvement of breeding populations on the premise of using two base populations and
selecting for general and specific combining abilities. Nevertheless, a number of useful
heterotic patterns among open pollinated maize varieties have been reported (Miranda
Filho and Vencosky, 1984; Crossa and Gardner, 1987; Khalifa and Drolsom, 1988;
Mungoma and Pollak, 1988; Misevic, 1989; Misevic et al., 1989; Ajala, 1992; Ajala,
1993). Open-pollinated varieties have also been used as important sources for the
development of inbred lines for the development of hybrids and synthetic varieties.
Estimates of both mid-parent (MP) and better-parent (BP) heterosis for several
traits have been reported in different maize populations (Lonnquist and Gardner, 1961;
Paterniani and Lonnquist, 1963; De Leon and Lonnquist, 1978). Most of these workers
have reported substantial heterosis in inter-varietal crosses of maize and concluded that
heterosis increases with genetic divergence. Wellhausen (1978) recommended the
production of two diverse breeding populations based on the heterotic patterns of the
selected populations, which will maximize the population-cross performance.
Any observed differences in the expression of trait by individuals of the same
species, whether they are controlled by genetic or environmental factors, can be described
in terms of heritability concept. Heritability estimates can be made in a broad or narrow
sense. Broad sense heritability gives an estimate of the proportion of the total variation
(phenotypic) that is due to genetic effect While, narrow sense heritability is defined as the
proportion of the total variability that can be transmitted from parent to offspring
34
(Obilana and Fakorede, 1981; Becker, 1984; Hallauer and Miranda, 1988). Thus, while
the broad sense heritability gives a measure of the total genetic variation consisting of the
transmissible (additive) and non-transmissible (non-additive) values, the narrow sense
heritability gives a measure of the total variation that is transmissible. Narrow sense
heritability therefore determines the degree of resemblance between relatives and
expresses its predictive advantages, that is, the reliability of the phenotypic value as a
measure of the breeding value (Obilana and Fakorede, 1981).
35
MATERIALS AND METHODS
Two hundred and twelve maize genotypes (212) used in this study comprised of
209 local open-pollinated cultivars and three advanced genotypes from International
Institute of Tropical Agriculture (IITA), Ibadan Nigeria. To collect the local maize
cultivars, students of Michael Okpara University of Agriculture, Umudike, Abia State
normally residing in maize producing areas of Southeastern Nigeria were requested to
collect at least one-kilogramme samples from their respective locations. The collection
site, site codes and genotype names of the germplasm are explained in Table 1. The
genotypes were numbered according to the locations where they were collected and given
codes A to Z and AA to AI for ease of grouping. Genotypes 1 to 209 were collected from
the various locations in southeastern Nigeria while genotypes 210 (TZBR-W 1), 211
(AMA TZBR-W-C-1) and 212 (TZBR Eld 3 C2), which were at different stages of
improvement were collected from IITA, Ibadan, Nigeria. Genotype 209 from Ikwuano
was collected after the serial numbering of the genotypes was concluded thus, the
disparity in the genotype numbers from Ikwuano. The highest number of genotypes (65)
was collected from Ikwuano followed by Umuahia North (21), Bende (16) and Umuahia
South (15) all from Abia state. Other locations however contributed 1 to 7 genotypes. The
collected genotypes were evaluated in two stages as follows:
36
Table 1: Codes, collection sites and accession numbers for the local open-pollinated and improved maize genotypes evaluated as potential sources of resistance to S. calamistis and E. saccharina.
Code Collection sites Entry numbers of maize genotypes
A Ikwuano 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 38 39 40 41 42 43 44
45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 209
B Umuahia North 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85
C Umuahia South 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
D Uzuakoli 101
E Bende 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
F Isiala Ngwa 118 119 120 121 122 123
G Ohafia 124 125 126 127 128
H Aba North 129
I Aba south 130 131
J Osisioma 132 133 134
K Arochukwu 135 136 137 138 139
L Isuikwuato 140 141 142 143
M Nnochi 144
N Ukwa West 145 146 147 148
O Owerri North 149 150 151 152
P Obowo 153 154 155
Q Aboh Mbaise 156 157 158 159 160 161 162
R Ehime Mbano 163 164 165 166 167
S Ihite Uboma 168
T Egbu 169
U Ahiazu 170 171 172 173 174 175
V Mbaitoli 176 177 178 179 180
W Ezinihite 181 182 183 184 185 186
X Orlu 187 188 189
Y Nkanu 190
Z Udi 191 192 193
AA Enugu South 194
AB Nsukka 195 196
AC Awgu 197
AD Ovoko 198
AE Ohaozara 199 200 201 202 203
AF Afikpo 204 205
AG Ezzamgbo 206
AH Abakaliki 207 208
AI IITA-Ibadan 210 (TZBR-W-1) 211 (AMA TZBR-W-C-1) 212 (TZBR Eld 3 C2)
TZBR syn W = Tropical Zea Borer Resistant, synthetic white; AMA TZBR-W-C-1= Amakama Tropical Zea Borer Resistant, white cycle 1 and,
TZBR Eld 3 C2 = Tropical Zea Borer resistant (Eldana resistant) cycle 2
37
Experiment 1: Field evaluation of local maize germplasm for resistance to stem
borers in four environments.
Cultural practices
The genotypes from Southeastern Nigeria numbered SE NG-1 to SE NG-
209 (SE NG = Southeastern Nigeria) and the three improved ones from IITA, Ibadan
were evaluated during the planting seasons of 2001, in four environments, one of which is
artificial infestation with egg masses of S. calamistis at Ibadan. The other three
environments include Egbema, a stem borer endemic location in Southeastern Nigeria, a
non-infested trial at Ibadan and another one planted at Ikenne. The trials at Ibadan were
established on 22nd
May 2001, while those of Ikenne and Egbema were planted on 25th
and 29th
May 2001, respectively. Both Ibadan and Ikenne are located in the southwest of
Nigeria and are separated by a distance of 80 km. Egbema on the other hand is in Imo
State and is approximately 800 km south of Ibadan. The experiments were laid out in a
Randomized Complete Block Design (RCBD) with two replications.
At each site, the land was ploughed and harrowed. The plots were measured and
marked out. Row lengths were 5 m in Egbema and Ikenne while the inter-row spacing
was 0.75 m. In Ibadan, row length was 6.5 m which was divided into two equal halves of
3 m and each with a space of 0.5 m in the middle. One half of the plot was artificially
infested with egg mass of S. calamistis while the other half constituted the non-infested
plot. Row spacing was also at 0.75m apart. In all the plots two seeds were planted at an
intra-row spacing of 0.25 m. Thinning was done at three weeks after planting (3WAP)
leaving one plant per stand to give a population density of 53,333 plants ha-1
.
Fertilizer was applied in two splits in each trial. A basal application with 60 kg of
N, 30 kg of P2O5 and 30 kg of K2O was given immediately after thinning. The second
application was made with 30 kg of N ha-1
. Weeds were controlled with a pre-emergence
38
application of Primextra herbicide (4 l ha-1
). Further weed control on the plots was done
manually using hand hoes at 6 and 9 WAP. At 11 WAP, Gramoxone alone was applied to
ensure weed control at crop maturity stage. All genotypes of the infested trial at Ibadan
were artificially infested with egg masses of S. calamistis containing about 40 eggs at
black head stage introduced into the leaf sheath at 3 weeks after emergence (WAE).
Data collection
The following data were collected in the three locations for some agronomic
characteristics:
(a) Plant stand: Number of plants that established per plot.
(b) Days to 50 % silking: The numbers of days from planting to 50 % of the plants
in a plot to have their silk emerge from the sheaf.
(c) Plant height: Plant height was measured (in cm) from the base of the plant to
the base of the flag leaf, obtained from 5 plants taken at random in each plot.
(d) Plant aspect: Rating was done on a 1-9 scale based on overall plant appeal,
where 1= excellent plant type and 9= poor plant type.
(e) Ear aspect: Rating was done on a 1-9 scale based on overall ear appeal, where
1= excellent ear filling and 9= poor filling.
(f) Stalk lodging: The percentage of plants with stalks broken below the ear.
(g) Root lodging: the percentage of plants that are fallen or leaning (from the base
of the plant) beyond 45 0
.
(h) Husk cover rating: Rating on a 1-9 scale, where 1= very tight husk and well
extended beyond the tip and 9= well exposed ear tip.
(i) Plants at harvest: The total number of plants surviving up to the time of
harvest in each plot.
39
(j) Ears harvested: The total number of ears harvested per plot.
(k) Grain moisture: This was determined using a portable moisture-testing meter
for grains (Dickey-John model) expressed in percentages.
(l) Field weight (FWT): The weight, in kg, of all the de-husked ears per plot in
each plot.
(m) Grain yield (t ha-1
) adjusted to 15 % moisture and based on 80 % shelling
percentage was calculated as;
Grain yield = FWT
x (100 - Moisture %) x Plot size
1000 85 x 10, 000
Data on stem borer damage parameters were collected at Ibadan and Egbema for the
artificially and naturally infested trials, respectively. These included:
(a) Leaf feeding: Rating on a scale of 1 to 9, where 1= no visible leaf injury or a small
amount of pin or shot-hole type on a few leaves, and 9= most leaves having long
lesions (Davis and Williams, 1989).
(b) Dead heart: Counts were made of the number of plants showing the symptom in a
plot and expressed in percentage as a proportion of the number for the plot.
(c) Stem tunneling: the actual length of stem tunneled by the borer, expressed as a
percentage of the plant height. Data were taken from 5 plants selected at random
within each plot.
(d) Ear damage: the proportion of the ears damaged per plot, expressed in percentage.
(e) Root lodging: the proportion of the plants per plot broken at the ground level,
expressed in percentage.
(f) Stalk lodging: the proportion of the plants per plot broken below the ear,
expressed in percentage.
40
Statistical analysis
Analysis of variance (ANOVA) was computed separately for the genotypes in
each location and then combined across environments. This was carried out using the
Restricted Maximum Likelihood (REML) procedure (SAS, 1999) for the agronomic traits
and stem borer damage parameters. The statistical model for the combined ANOVA was:
Yijk = EiRj + Gk + GEij + eijk
Where: overall mean
i = 1, 2, 3, 4 (environments)
j = 1, 2 (replications)
k = 1, 2, 3…212 (genotypes)
Ei = effect of the ith
environment
Gk = the effect of the kth
genotype
GEik = the interaction effect between the kth
genotype and the ith
environment
eijk = error (residual)
Data which were obtained as percentages were transformed using the square root
transformation procedure as described by Obi (2002) before performing the ANOVA.
The environments and genotypes were considered random in determining the expected
mean squares for the analysis. The form of ANOVA is given in Table 2. Multivariate
analyses using the cluster and canonical discriminant functions were made using SAS
(1999) procedure to group the entires.
Estimation of genetic variance components
Using the expected mean squares for ANOVA the variance components were
calculated as suggested by Hallauer and Miranda (1988):
41
Genetic variance (
g) =
M3 – M2
re
G x E interaction variance (
ge) =
M2 - M1
r
Error variance (
e) = M1
Mean square for GxE = M2
Mean square for genotypes = M3
Heritability estimates
Broad Sense Heritability (BSH) gives an estimate of the proportion of the total
variation (phenotypic) that is due to genetic effect and calculated as follows:
BSH (h2) =
g
ph
where,
g = genetic variance,
ph = total or phenotypic variance and
e = environmental variance.
=
g
g+
e
42
Table 2: Form of analysis of variance showing sources of variation, degrees of freedom
and mean squares only, for the evaluation of 212 open-pollinated maize genotypes
(populations) in four different environments
Sources of variation Degrees of Mean Expected
freedom. squares mean squares
Blocks in environments e(r-1)
Environment (E) e-1
Genotypes (G) g-1 M3
e+ r
ge + re
g
GxE (g-1)(e-1) M2
e+ r
ge
Error e(r-1)(g-1) M1
e
e = number of environments; r = number of replications; g = number of genotypes
43
A Rank Summation Index (RSI) method of Mulumba and Mock (1978) was then
used to rank the genotypes for their overall performance with respect to damage caused
by borers and agronomic performance using data across all environments. To obtain the
RSI, genotypes were first ranked for each parameter (in this case, 1 = best and 212 =
poorest) and parameter ranks summed to generate overall performance of each genotype.
Thus, the lower the RSI of any genotype, the greater is its resistance and the better is its
agronomic performance. Using the RSI of the damage parameters and including
important agronomic attributes as a selection index, selection among the cluster groups
was made as proposed by Ajala et al. (1995). Thus eleven genotypes representing top 5 %
of the entries were made based on a combination of low pest damage, high yield and
other desirable agronomic traits. The eleven selected genotypes were then used to
generate a set of diallel crosses for further studies.
Experiment 2: Diallel evaluation to obtain information on combining ability and
heterosis of selected genotypes and generate reciprocal base populations for further
improvement.
The diallel crosses were generated during the off-season (December 2001 to
March 2002 using irrigation facilities at the International Institute of Tropical
Agriculture, Ibadan. A set of F1s (55 crosses) and the 11 parents were evaluated for
agronomic attributes and resistance to stem borer damage parameters in a total of seven
environments. The trials were established in the environments as follows:
(1) Artificially infested, early season planting at Ibadan - 2nd
April 2002,
(2) Non-infested, early season planting at Ibadan - 2nd
April 2002,
(3) Non-infested, early season planting at Ikenne - 24th
April 2002
44
(4) Non-infested, late season planting at Ikenne - 16th
September 2002
(5) Artificially infested, late season planting at Ibadan - 11th
September 2002,
(6) Non-infested, late season planting at Ibadan – 11th
September 2002.
(7) Late season planting at Akure, 7th
September 2002.
The agronomic practices and field layout are as in Experiment 1, this time with
three replications. Data were collected on the agronomic and borer resistance parameters
as described in Experiment 1.
Statistical analysis
Analysis of variance (ANOVA) was computed separately for the genotypes in
each environment and then combined across the seven (7) environments. This was carried
out using the Restricted Maximum Likelihood (REML) procedure (SAS, 1999) for the
agronomic traits and stem borer damage parameters. Where genotypic effect was found
significant, the variable was submitted for further analysis to partition the general and
specific combining effects. Estimates of GCA (gi) and SCA (si) effects were calculated
using the Diallel-SAS05: A comprehensive programme for Griffing‟s and Gardner-
Eberhart (GEAN II) analyses (Zhang et al., 2005). Consequently, gca and sca effects
were estimated by Griffing‟s Method 2, Model I, for fixed effects for the parents and a set
of their crosses. The model for the randomized complete block analysis for a set of the
parents and the crosses (genotypes) is:
Xijkl = µ + bk + vij + (bv)ijk +eijkl,
where, Xijkl = observed value of the experimental unit,
µ = overall mean,
bk = effect of the kth
block (replicate)
vij = effect of the ijth
cross
45
(bv)ijk = genotype x block interaction
eijkl = error effect.
Where the genotypic effect was found significant, vij was further partitioned into general
and specific combining ability as: vij = gi + gj + sij
where, gi = gca effect of the ith
parent,
gj = gca effect of the jth
parent,
sij = sca of the ijth
cross, i ≠ j
The Standard error (S.E.) to test gca and sca effects were:
S.E. (gi – gj) = differences between gca effect,
S.E. (sij – skl) = sca effect of two crosses with no genotype in common, and
S.E. (sij – sik) = sca effect of two crosses with one common genotype.
The calculations of the genotypic variance components and heritability estimates
were done as in Experiment 1 except that the mean squares used were those of the hybrids
as described by Becker (1984).
Estimation of Heterosis
Mid-parent (MP) and better-parent (BP) heterosis were calculated as described by
Liang et al., (1972) and Uguru (2004):
__ __
F1 MP X
100 or
F1 BP X
100
MP 1 BP 1
Where F1 = mean of F1 cross
MP = mid parental value
BP = mean of better parent
46
RESULTS
Highly significant (P = 0.01) differences were obtained in almost all cases for all
the traits for both environments and genotypes. Furthermore, genotype-by-environment
interaction effects were significant for all traits except plant stand, leaf feeding, and plant
and ear heights (Tables 3 and 4). Genotypic variance (
G) estimates for the traits studied
were generally less than the environmental variances except for plant and ear aspects that
had 0.02 and 0.00, respectively (Table 5). Generally, genotypes by environment (
GxE)
variance estimates were the lowest and in some cases negligible, but the estimates were
relatively high for stalk lodging and grain moisture percent with values of 1.47 and 5.79,
respectively. Broad sense heritability estimates for plant and ear aspects and stalk lodging
were high with values higher than 60 % in each case while days to 50 % silking and ear
damage rating had moderate (35-59 %) estimates. Heritability estimates for other traits
were low having values of < 25 % (Table 5).
47
Table 3: Combined analysis of variance for agronomic and stem borer damage parameters of
212 maize genotypes evaluated in two (2) environments.
Source of
variation
Degrees of
freedom
Leaf
feeding
Plant
aspect
Root
lodging
Stalk
lodging
Ear
aspect
Blocks/ Env. 2 1.4 3.6** 7.4 7.5 0.2
Env (E) 1 221.1** 10.9** 213.0** 1042.0** 0.3**
Genotype (G) 211 0.7* 2.1** 6.0** 14.5** 2.1**
G x E 211 0.7 0.9** 4.8** 10.7** 0.8**
Error 422 0.6 0.8 3.5 7.3 0.6
*, ** = significant at P = 0.05 and 0.01, respectively
48
*, ** = significant at P = 0.05 and (P=0.01), respectively
Table 4: Combined analysis of variance for agronomic and stem borer damage attributes of 212 maize genotypes evaluated in four (4) environments.
Source of
variation df.
Plant
stand
Days to
50 % silk
Plant height
(cm)
Ear height
(cm)
Husk
cover
rating
Plants at
harvest
Field
weight
(kg)
Ears
harvested
Ear
damage
Grain
moisture
(%)
Grain yield
(t ha-1
) x
1000
Blocks/Env 4 66.2** 303.5** 15510.7**
16433.2**
1.8**
44.6**
6.6** 29.9** 0.8 270.3** 5.0**
Env (E) 3 7163.1** 6279.0** 539024.8** 233989.9** 130.9** 7825.9** 205.6** 6247.3** 184.5** 7282.8** 395.1**
Genotype (G) 211 10.9** 90.3** 2109.0**
1665.6**
1.3**
13.3**
0.9** 27.0** 3.5** 54.8** 2.1**
G x E 633 6.9 6.7** 424.7
224.3
0.5** 7.6*
0.2** 9.8** 1.2* 39.4** 0.5**
Error 844 6.9 5.5 442.1 239.5 0.4 6.7 0.1 8.1 0.9 26.6 0.3
49
Table 5: Variance components and heritability estimates of some agronomic and stem borer damage
parameters of the 212 maize genotypes.
Source of variation σ2G σ
2E σ
2GxE σ
2e σ
2Ph h
2 bs (%)
Plant stand 0.48 16.87 0.00 7.00 17.35 2.77
Days to 50 % silking 10.44 14.79 0.00 5.50 25.23 41.38
Leaf feeding 0.01 0.52 0.06 0.60 0.59 1.69
Plant height (cm) 204.20 1270.00 0.00 442.10 1474.20 13.85
Ear height (cm) 173.60 551.20 0.00 239.50 724.80 23.95
Plant aspect 9.31 0.02 0.01 3.50 9.34 99.68
Root lodging (%) 0.28 0.49 0.65 7.30 1.42 19.72
Stalk lodging (%) 1.08 2.43 1.47 0.50 4.98 21.69
Husk cover rating 0.10 0.31 0.06 0.40 0.47 21.28
Plants at harvest 0.71 18.44 0.35 6.80 19.50 3.64
Field weight (kg) 0.09 0.48 0.02 0.10 0.59 15.25
Ears harvested 2.15 14.71 0.84 8.10 17.70 12.15
Ear aspect 0.31 0.00 0.11 0.80 0.42 73.81
Ear damage 0.3 0.43 0.07 0.90 0.83 7.50
Moisture (%) 1.92 17.08 5.79 26.70 24.79 7.75
Grain yield (t ha-1
) 0.03 0.93 0.06 20.30 1.02 2.94
σ2
G = genetic variance; σ2
E = environmental variance (within families); σ2
GxE = genetic by environment
interaction variance; σ2e = experimental error; σ
2ph = total (phenotypic) variance and h
2bs = broad sense
heritability
50
Overall means for borer damage parameters and agronomic characteristics of the
genotypes evaluated in different environments (Table 6) show that in general, the proportion
of plants per plot that suffered dead heart was low (2.0 %). Leaf feeding rating was also low
with an average value of 3.0. Mean ear damage obtained for the genotypes was 4.2 with a
range of 2.4 to 4.8. On the average, genotypes attained 50 % silking in 63 days but with a
range of 55 to 74 days, indicating that the maize genotypes grown in southeastern Nigeria are
mainly late maturing (126 days). Average height of the genotypes was 202.3 cm ranging
from, 146.2 to 253.9 cm. Range of plant aspect rating was 2.5 to 6.5 resulting in a moderate
average rating of 4.8. Percentage of plants which lodged at the root (root lodging) was low
(1.9 %) showing that many genotypes posses the ability to withstand rain storm. Similarly,
percentage of the plants which lodged below the ear (stalk lodging) was average of 4.9 %
ranging from zero to 10.5 %. Husk cover ratings for the genotypes averaged 3.1 with a range
of 2.0 to 4.3 indicating that the ears are tightly covered. With respect to number of plants at
harvest, 13 plants per plot survived up to harvest, one less the number (plant stand) which
survived the seedling stage. Mean field weight of the ears harvested, per plot was 1.1 kg,
ranging from 0.48 to 2.53 kg. Ear aspect rating was moderate (4.8) with a range of 2.8 to 7.0.
Grain moisture content obtained at harvest was 23.1 %, ranging from 12.3 to 29.8 %. This
range indicates that some genotypes with low moisture contents may be early maturing
having matured long before harvest. Grain yield, considered one of the most important traits
in pest resistant breeding programmes, gave a mean of 1.6 t ha-1
with a range of 0.6 t ha-1
3.7 t
ha-1
. Most of the top yielders were improved checks from IITA, Ibadan and a few locally
collected genotypes from southeastern Nigeria. The coefficients of variation for grain yield, a
measure of the relative levels of variability for traits, obtained in this study showed that
moderate values of about 25 % were obtained in almost all cases. Root lodging had the
51
highest coefficient of variability value of 99.4 %. CVs for other traits were much lower
ranging from 55.4 % for stalk lodging to 3.8 % for days to silking (Table 6).
Rankings of the 212 open-pollinated maize genotypes (Table 7) for stem borer
resistance parameters namely; ear damage rating, stalk lodging and leaf feeding, and grain
yield showed that genotype AMA TZBR-W-C-1 (from IITA-Ibadan) had the best overall
resistance levels with a Rank Summation Index (RSI) value of 91. This was followed by
genotypes SE NG-77 and SE NG-67 (from Umuahia North), SE NG-62 (from Ikwuano), SE
NG-148 (from Ukwa West), SE NG-106 (from Bende), SE NG-119 (from Isiala Ngwa), SE
NG-33 (from Ikwuano) and SE NG-65 (from Umuahia North), in that order. Selection of the
top 5 % (eleven genotypes in all) included genotypes TZBR Syn W and TZBR Eld 3 C2 both
improved genotypes from IITA, Ibadan. Genotype SE NG-32 (from Ikwuano) was the worst
having scored an RSI value of 1441 (Appendix 1).
Multivariate and cluster analyses: using data from artificially infested plots.
When data for 13 plant characteristics namely: plant stand, days to 50 % silking, leaf
feeding damage, plant height, ear height, stalk lodging, husk cover rating, plants at harvest,
field weight, ears harvested, ear damage, percentage grain moisture and grain yield, were
subjected to both uni- and multivariate analyses, eight (8) traits namely, days to 50 % silking,
leaf feeding damage, plant height, stalk lodging, plants at harvest, ears harvested, ear damage
and grain yield, showed significant differences among the genotypes.
52
Table 6: Primary data on the evaluation of 212 maize genotypes in two (2) to four (4) environments
Plant trait Mean Minimum Maximum Range Cv % LSD(0.05)
Dead heart (%) 2.04 0.00 6.50 6.50 80.00 3.30
Leaf feeding 2.81 2.00 4.50 2.50 26.30 1.00
Ear damage 4.20 2.40 4.80 2.40 22.80 1.00
Plant stand 14.82 10.75 17.00 6.25 17.90 5.30
Days to 50 % silking 62.51 55.38 73.75 18.38 3.80 5.10
Plant height (cm) 202.32 146.15 253.90 107.75 10.40 44.49
Ear height (cm) 109.80 66.25 164.84 98.59 14.10 34.89
Plant aspect 4.75 2.50 6.50 4.00 19.41 1.30
Root lodging (%) 1.88 0.00 6.50 6.50 99.38 2.60
Stalk lodging (%) 4.86 0.00 10.50 10.50 55.42 3.70
Husk cover 3.09 2.00 4.25 2.25 19.00 1.18
Plants at harvest 13.41 9.75 16.13 6.38 19.40 5.10
Field weight (kg) 1.12 0.48 2.53 2.05 31.00 0.78
Ears harvested 9.81 4.25 14.25 10.00 28.90 5.49
Ear aspect 4.85 2.75 7.00 4.25 16.40 1.10
Grain moisture (%) 23.06 12.29 29.88 17.59 22.40 10.19
Grain yield (t ha -1
) 1.63 0.62 3.67 3.05 31.20 0.98
53
Table 7: Plant characteristics, their ranks and rank summation index of the 212 Maize genotypes evaluated under artificially infested,
natural infested and non-infested field conditions.
Artificially infested________ Naturally infested_______ _____Non-infested______
Genotype Ear Stalk Leaf Yield Ear Leaf Yield Ear Grain yield RSI*
damage lodging damage (t ha -1 ) damage damage (t ha -
1 ) damage (t ha -
1 )
AMA TZBR-W-C1 3 (2)+ 2 (33) 3 (18) 1.61 (4) 2.00 (1) 2.00 (7) 1.64 (2) 3.00 (11) 1.97 (13) 91
SE NG-77 3 (2) 1 (10) 3 (18) 0.92 (41) 2.50 (12) 2.00 (7) 1.46 (8) 3.00 (11) 1.96 (15) 124
SE NG-67 3 (2) 0 (1) 3 (18) 0.80 (56) 2.00 (1) 2.00 (7) 1.16 (33) 3.00 (11) 2.17 (9) 138
SE NG-62 3 (2) 0 (1) 3 (18) 1.11 (23) 2.00 (1) 2.50 (112) 1.30 (16) 2.50 (3) 2.58 (4) 180
SE NG-148 3 (2) 3 (57) 3 (18) 1.12 (22) 3.00 (38) 2.00 (7) 1.34 (10) 2.75 (6) 1.51 (33) 193
SE NG-106 3 (2) 5 (116) 3 (18) 1.20 (17) 2.50 (12) 2.00 (7) 1.32 (14) 2.75 (6) 2.30 (6) 198
SE NG-119 2 (1) 1 (10) 4 (136) 1.42 (9) 3.00 (38) 2.00 (7) 1.49 (6) 2.25 (1) 2.74 (3) 211
SE NG-33 4 (26) 1 (10) 3 (18) 1.25 (15) 3.50 (93) 1.50 (1) 1.56 (4) 3.25 (21) 1.65 (23) 211
SE NG-65 4 (26) 1 (10) 3 (18) 1.41 (10) 2.50 (12) 2.00 (7) 0.99 (73) 3.75 (47) 1.60 (26) 229
TZBR-W-1 3 (2) 2 (33) 3 (18) 1.20 (18) 2.50 (12) 2.50 (112) 1.20 (28) 3.00 (11) 1.90 (17) 251
TZBR ELD 3 C2 3 (2) 1 (10) 3 (18) 0.56 (109) 2.00 (1) 2.50 (112) 1.59 (3) 2.50 (3) 2.01 (12) 270
*RSI = Rank summation index, +values in parenthesis represent performance ranking
54
Simple correlations among the eight traits were significant (P=0.05) for all the traits
except for Stalk lodging with ear damage, leaf feeding and grain yield (Table 8).
However, significant relationships were obtained involving these traits and stalk lodging.
Thus, indicating the possibility of a form of indirect relationships existing among these
traits. Since the genotypes differed in these traits that were significantly correlated with
one another, it was therefore justified to adopt the canonical discriminant analysis as
recommended by Falowo (1982).
Discriminant analysis is a measure of the effectiveness of each function
(eigenvalue) in discriminating between groups and the proportion of the discriminant
power accounted for by each function. The first two discriminant functions (CAN 1 and
CAN 2) had significant (P= 0.05) correlations and accounted for 97.8 % and 1.1 % of the
total variation, respectively resulting in 98.9 % when combined (Table 9). The score
coefficients (weights) loadings on CAN 1 function was dominated by plant height (2.92)
while CAN 2 was associated with leaf feeding (0.61) and ear damage (0.51). Therefore,
CAN 1 may conveniently be regarded as „height‟ function and CAN 2 as borer damage
function. The plot of the first two discriminant functions shows five clusters of the
genotypes, identified by numbers 1 to 5 (Figure 1). Thus, the bi-plot shows the
distribution of the genotypes along the axes as tall versus short (Y-axis) and resistant
versus susceptible (X-axis).
55
Table 8: Correlation matrix of some agronomic and damage parameters of the 212 maize
genotypes obtained using the data from artificially infested plots.
Plant
characteristics
Plant
height
Plants at
harvest
Ears
harvested
Ear
damage
Stalk
lodging
Leaf
feeding
Plants at harvest 0.327**
Ears harvested 0.307** 0.597**
Ear damage -0.216** -0.145** -0.262**
Stalk lodging 0.208** 0.512** 0.307** 0.070
Leaf feeding -0.320** -0.127** -0.269** 0.199** 0.000
Yield 0.348** 0.466** 0.735** -0.551** 0.077 -0.314**
* , ** = Significant at P = 0.05 and 0.01, respectively.
56
Table 9: Discriminant score coefficient matrix of the variables for the first four canonical
discriminant variables obtained for the 212 genotypes using data from the artificially
infested plot.
Plant characteristics CAN 1 CAN 2 CAN 3 CAN 4
Leaf feeding 0.11 0.61 -0.54 0.65
Plant height (cm) 2.92 0.11 -0.35 0.17
Stalk lodging -0.20 -0.61 0.11 0.15
Plants at harvest -0.05 0.12 0.91 0.23
Ears harvested 0.07 0.41 0.47 0.14
Ear damage 0.12 0.51 -0.18 -0.43
Grain yield (t ha-1
) -0.19 -0.86 -0.75 -0.11
Eigenvalue 7.40 0.08 0.07 0.01
57
Percentage variance 97.80 1.13 0.95 0.12
Canonical correlation 0.94** 0..28* 0.25
0.09
*,** = Significant at P = 0.05 and 0.01, respectively.
CAN 1, CAN 2, CAN 3 and CAN 4 = Canonical disscriminant functions 1, 2, 3 and 4,
respectively
58
Can1
7.5
2
2 2
2 2
2 2
5.0 2 2 2
2 2 5 5 5
5 5 5 5 5 5
5 5 5 5555 5
5 5 5 5 5 5
2.5 5 55 5 5 5 5 5
5 5 5 5 5 5 5 5 5
5 5 5 555 5 5 5 5 5 5
5 5 55555 55 555 55 5 5 5 5
5 51 5 5 1 5 5 1 1
0.0 1 1 51 1 11 1 1
1 1 1 11 1 1 11 1
1 1 1 1 1 1 1 1 1 1
11 1 11 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 11 1 11 1
-2.5 1 1 11 1 11 1 11 1
3 1 331 3 3 333
3 3 3 3 3
33 33 3 3
3 3 3 3 3 3 3
-5.0
3 3
4
4
-7.5
4
4
-10.0
-3 -2 -1 0 1 2 3
Can2
59
Summary of the grouping of the 212 maize genotypes by the Cluster procedure
(SAS, 1999) in relation to their sites of collection is presented in Table 10. The local
genotypes from southeastern Nigeria were distributed in all the groups whereas the three
(3) improved genotypes fell within Clusters III and V (Table 11). Entry 210 from IITA,
Ibadan was grouped together with 86 local genotypes while the other improved
genotypes, entries 211 and 212, were grouped in cluster III along with 26 local entries.
Group means for plant height shows that Clusters II and V contain tall genotypes with
heights ranging from 192.2 to 217.6 cm; Clusters I and III genotypes are of medium
height with values ranging from 151.8 to 168.8 cm, while Cluster IV were made up of
relatively short genotypes of about 124.0 cm (Table 12). With regards damage
parameters, Clusters II and III had the lowest estimates of 2.7 to 3.5 for leaf feeding and
3.1 to 4.2 for ear damage. Grain yield, a very important attribute that reflect the combined
effects of other measured traits, including the damage parameters, shows that Clusters II
and V genotypes are high yielding, with grain yield of >1.00 t ha-1
while, Clusters I and
III were medium and Cluster IV low yielding.
60
Table 10: Distribution of the 212 genotypes in clusters according to their similarities determined using the data set from artificially infested plots.
Site Cluster I Cluster II Cluster III Cluster IV Cluster V .
code A 1 6 7 10 11 13 15 22 23 2 40 3 17 19 4 8 9 12 16 18 20
25 26 28 31 32 33 34 35 37 63 21 42 43 5 24 27 29 30 36 38
47 48 49 51 52 53 54 55 56 44 39 41 45 46 50 59
57 58 62 209 60 61 64
B 67 73 79 82 85 68 71 77 83 65 69 70 72 74 75
76 78 80 81 84
C 87 88 89 90 95 98 100 93 96 99 86 91 92 94 97
D 101
E 102 110 114 115 105 108 104 116 106 107 111 112 113 117
F 118 119 121 122 120 103 123
G 127 128 126 124 125
H 129
I 131 130
J 132 133 134
K 140 143 135 136 138 139
L 140 142
M 144
N 148 146 147 145
O 150 152 151 149
P 153 155
Q 158 159 162 161 156 157 160
R 163 166 165 164 167
S 168
T 169
U 170 171 172 173 174 175
V 180 177 176 178 179
W 183 181 182 184 185 186
X 187 188 189
Y 190
Z 193 191 192
AA 194
AB 195 196
AC 197
AD 198
AE 200 201 199 202 203
AF 204 205
AG 206
AH 208 207
AI 211 212 210
Total 80 13 28 4 87
61
Table 11: Number of genotypes, collection sites and site code of the 212 maize genotypes shown according to the clusters obtained from the cluster analysis using data from
the artificially infested plots.
Groups No. of genotypes ________Site codes________ _____________________________Collection sites____________________________
Cluster I 80 A B C D E Ikwuano Umuahia North Umuahia South Uzuakoli Bende
F G J K N Isiala Ngwa Ohafia Osisioma Arochukwu Ukwa West
O P Q R T Owerri North Obowo Abo Mbaise Ehime Mbano Egbu
U V Y Z AB Ahiazu Mbaitoli Nkanu Udi Nsukka
AE Ohaozara
Cluster II 13 A B E F K Ikwuano Umuahia North Bende Isiala Ngwa Arochukwu
O P R AH Owerri North Obowo Ehime Mbano Abakaliki
Cluster II 28 A B C E G Ikwuano Umuahia North Umuahia South Bende Ohafia
I N O Q V Aba South Ukwa West Owerri North Abo Mbaise Mbaitoli
W AC AF AG AI Mbaitoli Awgu Afikpo Ezzamgbo IITA-Ibadan
Cluster IV 4 A F L Ikwuano Isiala Ngwa Isuikwuato
Cluster V 87 A B C E F Ikwuano Umuahia North Umuahia South Bende Isiala Ngwa
G H I J K Ohafia Aba North Aba South Osisioma Arochukwu
L M N Q R Isuikwuato Nnochi Ukwa West Abo Mbaise Ehime Mbano
S U V W X Ihite Uboma Ahiazu Mbaitoli Mbaitoli Orlu
Z AA AB AD AE Udi Enugu South Nsukka Ovoko Ohaozara
AH AI Abakaliki IITA-Ibadan
62
Table 12: Cluster means of agronomic and damage parameters of 212 maize genotypes obtained
from the cluster analysis using data from the artificially infested plots.
Plant characters Cluster I Cluster II Cluster III Cluster IV Cluster V
Leaf feeding 3.4 ±0.7 2. 7 ±0.6 3.5 ±0.8 4.6 ±1.2 3.3 ±0.6
Plant height (cm) 168.8 ±6.7 217.6 ±6.7 151.8 ±5.2 125.0 ±9.7 192.2 ±8.0
Stalk lodging (%) 5.0 ±0.8 4.4 ±1.0 5.3 ±1.1 5.1 ±1.1 4.8 ±0.8
Plants at harvest 9.1 ±1.9 9.6 ±2.3 8.0 ±2.1 5.6 ±3.0 9.9 ±1.7
Ears harvested 6.4 ±2.0 7.5 ±2.0 5.6 ±2.1 3.8 ±3.0 7.1 ±1.7
Ear damage 4.3 ±1.9 4. 2 ±2.0 3.1 ±1.7 3.1 ±1.4 4.9 ±1.9
Grain yield (t ha-1
) 0.9 ±0.4 1.5 ±0.7 0.8 ±0.4 0.5 ±0.2 1.1 ±0.5
Values= cluster means ± standard deviation
63
Multivariate and cluster analyses with data from natural infested plot
Eighteen traits made up of both agronomic and borer damage parameters were
subjected to uni- and multivariate analyses (SAS, 1999) but only ten of then contributed
significantly in clustering the genotypes into the cluster groups. Correlation studies
among these traits revealed significant relationships amongst them (Table 13). Similar to
earlier observations with artificially infested data set, yield was significantly and
negatively correlated with days to 50 % silking, stalk lodging, plant and ear aspects, ear
damage and dead heart while it was positively correlated with plant height, plants at
harvest and number of ears harvested, however, the coefficients of determination in most
cases were too low to be of any practical consideration. The few cases where R2 were
high (>16 %) included correlation of yield with number of ears harvested, plant aspect,
plants at harvest and ear aspect. Furthermore, number of ears harvested was highly
correlated with plants at harvest and plant aspect (Table 13).
Discriminant analysis revealed that only CAN 1 and CAN 2 had significant (P=
0.05) correlations, accounted for 97.3 % and 2.1 % respectively resulting in a toal of 99.4
% of the total variations existing among the genotypes (Table 14). CAN 1 was mainly
associated with grain yield, while CAN 2 was loaded with number of ears harvested and
plant height. The bi-plot of the first two discriminant functions (Figure 2) showed the
distribution of the clusters along the perpendicular axes. The discriminant score
coefficient for yield was highest, -3.06 and is the main determinant of CAN 1 function.
Due to the negative sign of the coefficient of yield, all genotypes loaded above zero on
the Y-axis were identified as low yielding whereas the ones below were high.
64
Table 13: Correlation matrix of some agronomic and damage parameters of the maize genotypes obtained using data from the
natural infested plots.
Plant
characteristics
Days to 50
% silking
Plant
height
Stalk
lodging
Plant
aspect
Plants at
harvest
Ears
harvested
Ear
aspect
Ear
damage
Dead
heart (%)
Plant height 0.001
Stalk lodging 0.067 0.215**
Plant aspect 0.093 -0.065 0.317**
Plants at harvest -0.180** 0.129** -0.061 -0.302**
Ears harvested -0.29 0.028 -0.348** -0.588** 0.539**
Ear aspect 0.185** -0.184** 0.147** 0.313** -0.175** -0.277**
Ear damage -0.148** -0.086 0.075 0.101* 0.067 0.077 0.306**
Dead heart (%) 0.013 0.099* 0.184** 0.221** -0.472** -0.421** 0.092 -0.084
Yield (t ha -!) -0.351** 0.216** -0.249** -0.570** 0.409** 0.677** -0.439** -0.095* -0.146**
* , ** = Significant at P = 0.05 and 0.01, respectively.
65
Table 14: Discriminant score coefficient matrix of the variables for the first four
canonical discriminant functions obtained for 212 maize genotypes using data
from the natural infested plots.
Plant characters CAN 1 CAN 2 CAN 3 CAN 4
Dead heart (%) -0.08 0.23 -0.09 0.08
Ear damage 0.10 0.14 -0.02 0.48
Days to 50 % silking 0.18 -0.08 -0.08 0.30
Plant height (cm) -0.04 0.61 -0.16 0.08
Stalk lodging -0.13 0.33 0.43 -0.33
Plant aspect 0.16 0.13 0.19 0.53
Plants at harvest -0.02 -0.18 0.67 0.57
Ears harvested 0.06 1.16 -0.08 0.18
Ear aspect -0.05 0.30 0.41 -1.01
Yield (t ha-1
) -3.06 -0.29 0.11 -0.48
Eigenvalue 8.84 0.19 0.04 0.01
Percentage variance 97.26 2.14 0.47 0.13
Canonical correlation 0.95** 0.55 *
0.09
0.04
*, ** = Significant at P = 0.05 and 0.01, resectively.
CAN 1, CAN 2, CAN 3 and CAN 4 = Canonical discriminant functions 1, 2, 3 and 4, respectively
66
Can1
7.5
4 4
4
4 4 4
4 4 4 4
5.0 4
4 4 4
1 1 4 4
4 1
1 11 1 1 1
1 1 1
2.5 11 1 11 1 1 1
1 1 1 1 1 1 1
1 11 1 111 1 11 1 1 1
1 11 11 1 111 11 1 1
3 1 1 1 3 13 3 3
3 3 3 3 33 3 333 3 3 3 3
0.0 3 33 3 3 33 3333 3
33 3333 3 33 3 3
- 3 3 3 333 33 3
3.0 3 33 3 3 3 3 3 3
3 3 3 33 3 3 33
3 3 3 3 3 3 3
-2.5 3 3 3 3 3
5 5
5 5 5 5 5
5 55 5 5 5
5 5
5 5 5 5
-5.0 5 5
5
5 5 5
5
5 5 5 5
-7.5
2
2
2
-10.0
2
-12.5
-3 -2 -1 0 1 2 3
67
The second axis of CAN 2 distributed the clusters on the basis of their number of ears
harvested and plant height. Clusters II and V are located on the lower half of Y-axis
while Clusters I and IV are found in the upper half of the axis. Cluster III contained
genotypes which were found around the intercept of the two discriminant functions.
Clustering of the genotypes by the multivariate procedures summarised in Table 15
produced five distinct clusters with two improved genotypes appearing in Cluster II
alongside two local genotypes. The third improved genotype was located in Cluster V
with 29 locals. Cluster III with 89 genotypes contains the highest number followed by
Cluster I with 71, both of which were all local genotypes (Table 16). Clusters I and III
contain genotypes with average yield of 0.8 t ha-1
whereas Clusters II and V were made
up of genotypes producing relatively high yield of 1.5 to 1.7 t ha-1
. Generally, the ratings
for the damage parameters namely; stalk lodging, ear damage and dead heart, were
lowest for Cluster II which had the highest yield (Table 17).
68
Table 15: Distribution of the 212 maize genotypes in clusters according to their similarities obtained using data from the natural infested plots.
Site
code Cluster I Cluster II Cluster III Cluster IV Cluster V
A 1 5 6 7 8 10 2 3 9 11 12 20 26 4 15 17 16 18 25 30
13 14 19 21 23 24 33 27 29 32 34 35 36 37 22 28 51 42 49 58 62
31 38 40 43 44 47 41 45 46 53 54 55 56
48 50 60 57 59 63 64 65 69 71
B 70 74 76 78 81 85 72 75 79 80 83 84 73 82 209 67 68 77 80
86 87 89 88
C 91 92 93 94 95 96 90 99 97
98 100
D 101
E 102 110 113 114 115 103 104 107 108 109 112 117 105 106 111 116
F 121 118 120 122 119 123
G 124 125 128 126 127
H 129
V 180 178 179 180 176 177
W 181 186 183 184 185 182
X 189 188 187
Y 190
Z 192 193 191
AA 194 195
AB 196
AC 197
AD 198
AE 199 201 202 203 200
AF 204 205
AG 206
AH 207 208
AI 211 212 210
Total 71 4 89 18 30
69
Table 16: Number of genotypes, collection sites and site codes of the 212 maize genotypes shown according to the clusters obtained from the cluster
analysis using data from natural infested plots.
Groups Number of
genotypes ----------- Site codes -------- -------------------------------------------Collection sites ---------------------------------------
Cluster I 71 A B C D E Ikwuano Umuahia North Umuahia South Uzuakoli Bende
F G I J K Isiala Ngwa Ohafia Aba south Osisioma Arochukwu
L O P Q R Isuikwuato Owerri North Obowo Aboh Mbaise Ehime Mbano
U V W X Z Ahiazu Mbaitoli Ezinihite Orlu Udi
AE AF AH Ohaozara Afikpo Abakaliki
Cluster III 4 A S AI Ikwuano Ihite Uboma IITA-Ibadan
Cluster III 89 A B C E F Ikwuano Umuahia North Umuahia South Bende Isiala Ngwa
G H J K L Ohafia Aba North Osisioma Arochukwu Isuikwuato
M N O P Q Nnochi Ukwa West Owerri North Obowo Aboh Mbaise
R U V W X Ehime Mbano Ahiazu Mbaitoli Ezinihite Orlu
Y AA AC AD AE Nkanu Enugu South Awgu Ovoko Ohaozara
AG Ezzamgbo
Cluster IV 18 A B C E F Ikwuano Umuahia North Bende Isiala Ngwa Aba south
I K N O Q
Umuahia
South Arochukwu Ukwa West Owerri North Aboh Mbaise
U Z AA Ahiazu Udi Enugu South
Cluster V 30 A B E F I Ikwuano Umuahia North Bende Isiala Ngwa Aba south
N Q R V W Ukwa West Aboh Mbaise Ehime Mbano Mbaitoli Ezinihite
X Z AB AE AI Orlu Udi Nsukka Ohaozara IITA-Ibadan
70
Table 17: Cluster means of agronomic and damage parameters of 212 maize genotypes
obtained from the cluster analysis using data from the naturally infested plots.
Plant
characteristics Cluster I Cluster II Cluster III Cluster IV Cluster V
Days to 50 % silk 67.1 ± 3.5 64.3 ± 2.1 66.0 ± 3.2 68.7 ± 3.7 65.3 ± 2.2
Plant height (cm) 194.2 ± 20.3 189.4 ± 13.8 201.3 ± 27.5 183.0 ± 22.7 196.8 ± 19.8
Stalk lodging (%) 4.2 ± 1.5 1.4 ± 1.0 3.7 ± 1.5 4.0 ± 1.4 3.0 ± 1.2
Plant aspect 5.2 ± 0.5 3.1 ± 0.3 4.8 ± 0.6 5.6 ± 0.6 4.1 ± 0.6
Plants at harvest 10.0 ± 1.3 11.6 ± 0.3 10.5 ± 0.8 9.1 ± 1.9 11.4 ± 0.8
Ears harvested 6.1 ± 1.3 10.6 ± 0.4 7.6 ± 1.6 4.2 ± 0.7 9.0 ± 0.6
Ear aspect 5.2 ± 0.6 3.4 ± 0.5 4.8 ± 0.5 5.2 ± 0.5 4.2 ± 0.7
Ear damage 3.5 ± 0.7 2.6 ± 0.7 3.6 ± 0.8 3.4 ± 1.1 3.2 ± 0.7
Dead heart (%) 2.4 ± 1.6 0.6 ± 0.5 2.0 ± 1.4 2.9 ± 1.5 1.2 ± 0.8
Grain yield (t ha-1
) 0.8 ± 0.1 1.7 ± 0.1 0.8 ± 0.1 0.5 ± 0.1 1.5 ± 0.18
Values = cluster means and standard deviations
71
Multivariate and cluster analyses: using data from non-infested plots from
Ibadan and Ikenne locations, combined.
Data from two non-infested environments of Ibadan and Ikenne, were tested
for homogeneity of variances using Bartlett‟s test and because they showed non
significant effects for the traits, they were pooled for the multivariate analysis.
Correlation coefficients among the traits were significant (Table 18) thus supporting
the use of multivariate analysis in grouping the large number of genotypes with the
aim of selecting few representative genotypes which could be used for further
development efforts. The score coefficients among the plant characteristics
accounting for variations among the genotypes which were loaded on the first two (2)
discriminant functions, CAN 1 and CAN 2, accounted for 95.4 % and 3.1 % of the
total variation individually and 98.5 % when combined (Table 19). CAN 1 was
associated with grain yield while CAN 2 was associated mainly with days to 50 %
silking which was negatively associated with number of ears harvested. The plot of
CAN 1 and CAN 2 shows the relative distribution of the clusters (Figure 3).
72
Table 18: Correlation matrix of agronomic and borer damage parameters of the 209 local maize
cultivars and three (3) improved maize genotypes obtained using the uninfested data set.
Plant characteristics
Days to 50 %
silking
Plant
height (cm)
Plants
harvested
Ears
harvested Ear damage
Plant height (cm) 0.140**
Plants harvested -0.004 0.138**
Ears harvested -0.014 0.123* 0.515**
Ear damage 0.160** -0.184** -0.133** -0.191**
Grain yield (t ha-1
) -0.102* 0.228** 0.415** 0.730** -0.487**
*, * = Significant at P = 0.05 and P = 0.01., respectively
73
Table 19: Discriminant score coefficient matrix of the variables for the four canonical
discriminant functions obtained using the noninfested data sets combined from Ibadan
and Ikenne in southwestern Nigeria.
Plant characters CAN 1 CAN 2 CAN 3 CAN 4
Days to 50 % silking 0.17 0.84 0.23 0.41
Plant height (cm) -0.15 -0.36 -0.53 0.31
Plants at harvest 0.04 0.17 0.05 0.76
Ears harvested -0.34 -1.25 0.32 0.18
Ear damage -0.12 0.23 0.94 0.07
Grain yield (t ha-1
) 3.19 1.20 0.54 -0.35
Eigenvalue 8.03 0.26 0.03 0.01
Percentage variance 95.42 3.13 0.35 0.09
Canonical correlation 0.94** 0.43** 0.11 0.06
*, * = Significant at P = 0.05 and P = 0.01., respectively
CAN 1, CAN 2, CAN 3 and CAN 4 = Canonical discriminant functions 1, 2, 3 and 4, respectively
74
4
75
Genotypes belonging to Clusters I, III, and IV were located above the zero axis of
CAN 1 while, Cluster V genotypes were distributed among the four partitions created
by the two discriminant functions. Cluster IV members were located on the right side
of CAN 2 axis whereas Clusters I, II, III and V had their genotypes distributed on
both sides of CAN 2 axis. The cluster members as identified by the multivariate
procedure are summarised including the site of collection and the number of entries in
Table 20. The highest number of local genotypes, 134 in all, was contained in Cluster
V. The improved genotypes were grouped in Clusters I and III that had 37 and 8
entries, respectively (Table 21). Genotypes belonging to Clusters I, III and IV were
characterized by low estimates for ear damage of 1.3 to 1.8 and high grain yield of 1.6
to 2.9 t ha-1
(Table 22). Clusters II members had the lowest yields of 0.6 t ha-1
while
having the highest score for ear damage and the lowest number of plants and ears at
harvest. On the other hand, genotypes belonging to Cluster IV had the highest values
for grain yield, number ears harvested and number of plants at harvest.
76
77
78
79
80
The combining ability and heterotic effects for agronomic attributes and stem borer
damage parameters in the eleven (11) selected genotypes.
Highly significant differences among environments and genotypes were
obtained for all traits except grain moisture (Tables 23 to 25). A breakdown of the
significant genotype mean squares revealed that both parents and crosses were generally
significant except for days to 50 % silking, root lodging, stalk lodging, husk cover rating
and number of plants at harvest where effect due to parents were not significant (Tables
24 and 25). Apart from dead heart and plant aspect, gca effects were significant for all
traits while sca effects were significant for most of the traits except ear damage rating.
Differences between the parental genotypes and the crosses, a measure of average effect
of heterosis, were significant only for dead heart and leaf feeding damage (Table 23).
Genotype x environment interaction mean squares were significant for plant stand, days
to 50 % silking, ear height, plant aspect stalk lodging, number of plants at harvest, field
weight, number of ears harvested, ear aspect and grain yield (Tables 24 and 25) but not
for dead heart, leaf feeding, ear damage (Table 23). All other parameters were also not
significant (Tables 24 and 25). Except for the damage parameters of dead heart, leaf
feeding, ear damage and husk cover rating, the interactions of the main effects with
environment were significant for most of the traits (Tables 24 and 25)
81
Table 23: Combined analysis of variance for stem borer damage parameters of 11 open-
pollinated maize genotypes and their crosses evaluated in three (4) environments.
Source of Degree of Dead Leaf Ear
variation freedom heart (%) feeding damage
Environment (E) 3 2.3* 42.9** 359.5**
Blocks in environments 6 1.9* 18.2** 163.1**
Genotype 65 1.7** 2.6** 38.9*
Parents 10 2.6** 2.4* 37.7**
Crosses 54 1.3* 2.6** 39.7*
GCA 10 1.2 4.5** 86.8*
SCA 44 1.3* 2.2** 29.0
Parents x Crosses 1 12.1* 6.4** 4.7
Genotype x E 195 0.8 0.9 23.4
Parents x E 30 0.6 1.1 11.6
Crosses x E 162 0.8 0.9 25.8
GCA x E 30 1.0 1.0 33.6
SCA x E 132 0.8 0.9 24.0
Parents-Crosses x E 3 0.7 0.1 9.5
Error 520 0.9 1.1 27.0
*, ** significant at the 0.05 and 0.01 levels of probability, respectively
82
Table 24: Combined analysis of variance for agronomic traits of 11 open-pollinated maize genotypes and their
crosses evaluated in three (7) environments. Source of Degree of Plant Days to Plant Ear Plant
variation freedom stand 50 % silking height (cm) height (cm) aspect
Environment (E) 6 3508.7** 921.5** 111398.0** 37727.6** 44.5**
Blocks in environments 14 95.1** 56.7** 3708.3** 2222.2** 1.8**
Genotype 65 103.7** 14.1** 2185.6** 1116.6** 1.2**
Parents 10 56.7** 6.7 3040.4** 1454.7** 1.4**
Crosses 54 114.2** 15.2** 2039.9** 1063.6** 1.1**
GCA 10 193.3** 36.2** 2573.1** 1234.8** 1.1
SCA 44 96.2** 10.4** 1918.8** 1024.7** 1.1**
Parents x Crosses 1 8.4 30.4 1502.2 601.7 1.0
Genotype x E 390 20.5** 5.1** 589.4 312.8** 0.5**
Parents x E 60 18.6* 6.6** 537.2* 265.2* 0.5
Crosses x E 324 20.8** 4.7** 600.7** 324.4** 0.5**
GCA x E 60 21.9** 5.6** 868.2** 567.5** 0.7**
SCA x E 264 20.6** 4.5** 539.9** 269.2** 0.5*
Parents-Crosses x E 6 21.4 8.8** 501.5 165.1 1.4**
Error 910 12.8 3.1 360.6 195.3 0.4
*, ** significant at the 0.05 and 0.01 levels of probability, respectively
83
Table 25: Combined analysis of variance for post flowering and agronomic traits of 11 open-pollinated maize genotypes and their crosses
evaluated in three (7) environments.
Source of Degree of Root Stalk Husk Plants at Field Ears Ear Grain Grain
variation. freedom lodging lodging cover harvest weight (kg) harvested aspect moisture (%) yield (t ha -1
)
Environment (E) 6 3673.7** 12333.3** 126.3** 2195.3** 169.6** 3513.8** 36.7** 2156.8** 35.5**
Blocks in environments 14 206.8** 343.3** 0.3 87.8** 5.6** 71.2** 5.4** 326.6** 1.0**
Genotype 65 56.1** 193.1** 0.7** 83.4** 3.6** 89.8** 7.7** 13.8 0.8**
Parents 10 57.3 110.4 0.6 25.6 5.3** 51.4* 10.7** 49.7 1.2**
Crosses 54 55.2** 198.9** 0.7** 95.4** 3.3** 97.8** 7.2** 5.2 0.7**
GCA 10 97.7** 360.6* 0.9* 147.1** 5.1** 165.2** 9.8** 8.7 1.1**
SCA 44 45.5* 162.1** 0.7** 83.6** 3.0** 82.5** 6.6** 4.4 0.6**
Parents x Crosses 1 94.6 704.0 0.0 18.1 0.4 41.1 1.0 118.3 0.03
Genotype x E 390 32.8 97.5** 0.3 19.4** 0.8** 23.2** 2.2** 12.5 0.2**
Parents x E 60 33.0 63.6 0.2 22.8** 0.9** 24.4** 1.5** 43.8** 0.2**
Crosses x E 324 31.5 99.5** 0.3 18.5** 0.8** 23.1** 2.3** 6.2 0.2**
GCA x E 60 33.5 141.5** 0.3 19.9** 1.0** 24.7** 3.3** 6.9 0.2**
SCA x E 264 31.1 90.0** 0.3 18.1** 0.8** 22.8** 2.1** 6.1 0.2**
Parents-Crosses x E 6 103.1** 324.2** 0.1 40.0** 0.3 17.3 6.5** 39.2** 0.1
Error 910 1.3 2.3 0.4 11.3 0.3 23.1 0.9 11.4 0.1
*, ** significant at P = 0.05 and 0.01, respectively
84
Genetic variance estimates (σ2
G) obtained for most plant characteristics were
lower than the environmental variance (σ2
E) except for dead heart and ear aspect, while
variances due to genotype x environment interaction (σ2
GxE) were low in nearly all the
traits studied. Broad sense heritability estimates obtained for the traits were also,
generally low ranging from 0.0 % for grain moisture to 14.1 % for ear aspect.
Coefficient of variation estimates calculated for the plant traits ranged from 3 % for days
to 50 % silking to 178 % for dead heart (Table 26)
All the genotypes had very low (< 2 %) levels of dead heart, leaf feeding damage
ratings of between 2 and 5, and a range of 4 to 10 % ear damage. Similarly, the crosses
had low levels (< 2 %) of dead heart, with similar values for both leaf feeding and ear
damage as for the parents. Array means for the three parameters were less variable with
ranges of: 0.0 to 0.6 for dead heart, 2.5 to 3.5 for leaf feeding and 6.5 to 8.5 for ear
damage (Table 27).
Genotypes coded SE NG-65, SE NG-148, TZBR Syn W and AMA TZBR-W-C1
had significant negative gca effects for deadheart while SE NG-67, SE NG-77 and SE
NG-106 also had negative gca effects for leaf feeding. Negative gca effects for ear
damage rating were also obtained for SE NG-65, SE NG-67, and AMA TZBR-W-C1
whereas SE NG-62, SE NG-106 and TZBR ELD 3 C2 had significant positive gca for ear
damage (Table 27). Significant negative sca effects for dead heart were obtained for; SE
NG-62 x SE NG-33, SE NG-148 x TZBR Syn W, SE NG-67 x SE NG-65 and , SE
NG-119 x TZBR Syn W, SE NG-33 x TZBR ELD 3 C2, SE NG-119 x AMA TZBR-W-
C1 and SE NG-106 crossed to SE NG-77 and SE NG-1196. For the leaf feeding damage,
significant sca effects were obtained for TZBR Syn W x AMA TZBR-W-C1, AMA
TZBR-W-C1 x SE NG-106 and SE NG-148, SE NG-119 x SE NG-65 and SE NG-77, SE
NG-67 x SE NG-62 and SE NG-148 (Table 27).
85
86
87
Number of plants per plot recorded for the parental genotypes ranged from 16 to
22 while days to 50 % silking ranged from 60 to 61 days. Similarly, the number of plants
per plot and days to 50 % silking obtained for the crosses were from 13 to 22 plants and
60 to 61 days, respectively (Table 28). Array means for the two traits ranged from 18 to
20 for plant stand and 59 to 61 days for days to anthesis. SE NG-77, SE NG-106 and
SE NG-148 had significant positive gca effects for plant stand while SE NG-77, SE NG-
106, SE NG-148, and AMA TZBR-W-C1 had significant negative gca estimates for days
to 50 % silking. All other gca effects were either small or not in the desired direction..
Significant positive sca effects for plant stand were obtained for TZBR Syn W crossed to
four locals (SE NG-33, SE NG-67, SE NG-106 and SE NG-119) for AMA TZBR-W-C1
crossed to three (SE NG-65, SE NG-67 and SE NG-148) and for TZBR ELD 3 C2
crossed to two (SE NG-65 and SE NG-119). In addition, SE NG-77 crossed to SE NG-
119 and SE NG-148;, SE NG-62 crossed to SE NG-65 and SE NG-119, and SE NG-33
x SE NG-148 were also significant for the same trait (Table 28). For days to 50 %
silking, significant negative sca effects were obtained for TZBR ELD 3 C2 crossed to SE
NG-67, SE NG-148 and TZBR Syn W, SE NG-119 crossed to SE NG-148 and SE
NG-106, SE NG-33 crossed to SE NG-62 and SE NG-148, SE NG-65 crossed to SE
NG-77 and SE NG-67 x AMA TZBR-W-C1 (Table 28).
The parental genotypes were generally tall (1.8 to 2.1 m) with ear height ranging
from 0.9 to 1.1 m. Rating for plant aspect was from 3 to 4. Values obtained for crosses
were 1.6 to 2.1 m for plant height, 0.8 to 1.1 m for ear height and rating for plant aspect
was from 2.7 to 4.1. Array means for the three parameters were less variable ranging
from 1.9 to 2.1 m for plant height, 0.98 to 1.05 m for ear height and 3.08 to 3.29 for plant
aspect (Table 29).
88
89
90
SE NG-67, SE NG-77, SE NG-106, SE NG-119, SE NG-148, TZBR Syn W and
TZBR ELD 3 C2 had significant positive gca effects for plant and ear heights while SE
NG-77, SE NG-148, TZBR Syn W and TZBR ELD 3 C2 had negative gca effects for the
same traits. SE NG-62, SE NG-65 and AMA TZBR-W-C1 had negative gca effect for
one of the two traits (plant and ear heights). Significant sca effects for plant height were
obtained for SE NG-67 crossed to SE NG-33, SE NG-106 and AMA TZBR-W-C1, SE
NG-106 crossed to SE NG-119 and TZBR Syn W, SE NG-65 crossed to SE NG-119
and TZBR ELD 3 C2, TZBR Syn W crossed to SE NG-33 and SE NG-77 and SE NG-62
x SE NG-148. Similar observations were made for ear height. For plant aspect,
significant negative sca effects were obtained for SE NG-148 crossed to SE NG-65, SE
NG-67 and SE NG-106, AMA TZBR-W-C1 crossed to SE NG-65, SE NG-67 and SE
NG-77, SE NG-106 crossed to SE NG-62 and TZBR ELD 3 C2 and SE NG-33 x SE
NG-119 (Table 29).
Root lodging for the parental genotypes ranged from 5 to 11 % while stalk
lodging was from 4 to 11 %. Husk cover rating among the genotypes was low (< 3). For
the crosses, ranges of 5 to 13 % and 4 to 17 % were obtained for root lodging and stalk
lodging, respectively. Husk cover rating for the crosses was from 1.7 to 2.4. Array means
were less variable with 6 to 8 % for root lodging, 8 to 12 % for stalk lodging and 1.9 to
2.1 for husk cover rating (Table 30).
SE NG-33, SE NG-62, SE NG-106 and TZBR Syn W had negative gca effect for
root lodging, stalk lodging and husk cover rating while SE NG-77 had gca estimates in
the opposite direction. SE NG-33, SE NG-62, SE NG-65, SE NG-119, AMA TZBR-W-
C1 and TZBR ELD 3 C2 had significant gca in the three parameters, but SE NG-67, SE
NG-106, SE NG-148 and TZBR Syn W had significant gca only in two parameters and
SE NG-77 had only in one.
91
92
Significant negative sca effects for root lodging were obtained for SE NG-62 crossed to
SE NG-33 and AMA TZBR-W-C1, SE NG-65 x SE NG-148 and SE NG-67 x AMA
TZBR-W-C1. For stalk lodging, significant sca effects were obtained for the following
crosses namely SE NG-33 x SE NG-148, SE NG-62 x SE NG-65, SE NG-67 x AMA
TZBR-W-C1 and SE NG-106 x TZBR Syn W. Significant negative sca effects for husk
cover rating were obtained for SE NG-33 crossed to SE NG-65, SE NG-67 and SE NG-
106 and SE NG-148 crossed to SE NG-65 and SE NG-67 (Table 30).
At harvest, the parental genotypes had a range of 17 to 21 plants per plot while
the number of ears harvested was from 14 to 19 ears per plot. Weight of maize harvested
per plot among the genotypes ranged from 1.5 to 2.7 kg per plot. Similarly, the crosses
had 14 to 22 plants per plot with 12 to 19 ears and 1.1 to 2.8 kg of maize at harvest.
Array means were 18 to 20 for plants at harvest, 15 to 17 for number of ears at harvest
and 1.9 to 2.3 kg per plot (Table 31). SE NG-33 and SE NG-65 had significant positive
gca effects for the three traits while SE NG-62, SE NG-67, SE NG-106 and SE NG-119
had significant positive gca effects for two of the traits. SE NG-77 and AMA TZBR-W-
C1 had significant positive gca in one while the other gca effects were either negative or
not significant (Table32) Significant positive sca effects for plants at harvest were
obtained for SE NG-33 crossed to SE NG-62, SE NG-77, SE NG-119 and AMA
TZBR-W-C1, SE NG-65 crossed to SE NG-67 and SE NG-148, SE NG-77 crossed to
SE NG-62, SE NG-67, SE NG-106 and AMA TZBR-W-C1, SE NG-62 crossed to SE
NG-106 and SE NG-148 and TZBR ELD 3 C2 crossed to SE NG-148, TZBR Syn W
and AMA TZBR-W-C1.
93
94
With respect number of ears harvested, significant positive sca effects were obtained for
SE NG-77 crossed to SE NG-33, SE NG-67, SE NG-106, SE NG-119 and AMA
TZBR-W-C1, SE NG-106 crossed to SE NG-33, SE NG-62, SE NG-65, SE NG-148
and TZBR ELD 3 C2, SE NG-148 crossed to SE NG-65 and SE NG-67, AMA TZBR-
W-C1 crossed to SE NG-67 and TZBR ELD 3 C2, SE NG-33 x SE NG-119 and TZBR
Syn W x TZBR ELD 3 C2. Also, significant positive sca effects for number of ears
harvested were obtained for SE NG-33 crossed to SE NG-106, SE NG-119, AMA
TZBR-W-C1 and TZBR ELD 3 C2, SE NG-65 crossed to SE NG-67, SE NG-148 and
AMA TZBR-W-C1, SE NG-77 crossed to SE NG-67, SE NG-106, SE NG-119 and
AMA TZBR-W-C1, SE NG-106 crossed to SE NG-62, SE NG-148 and TZBR ELD 3
C2, TZBR ELD 3 C2 crossed to TZBR Syn W and AMA TZBR-W-C1 and SE NG-67 x
SE NG-148 (Table 31).
Ear aspect rating for the parents ranged from 2.8 to 5.1 while grain moisture
ranged from 18 to 24 %. Grain yield obtained for the parents ranged from 0.66 to 1.33 t
ha -1
. Means for the crosses ranged from 2.7 to 5.7 for ear aspect, 17.2 to 19.8 % for
grain moisture and 0.63 to 1.36 t ha -1
for grain yield. Array means were again, less
variable ranging from 3.3 to 4.2 for ear aspect rating, 18.1 to 18.7 % for grain moisture
and 0.9 to 1.1 t ha -1
for grain yield (Table 32). SE NG-33, SE NG-62, SE NG-65, SE
NG-77, SE NG-106, SE NG-119 and TZBR Syn W had significant negative gca effects
for ear aspect. For grain yield, SE NG-33, SE NG-62, SE NG-65, SE NG-77, SE NG-106
and SE NG-119 had significant positive gca effects while all the other genotypes had gca
effects in the opposite direction. Significant sca effects for grain yield were obtained for
SE NG-33 crossed to SE NG-106, SE NG-119 and SE NG-148; SE NG-67 crossed to SE
NG-148, AMA TZBR-W-C1 and TZBR ELD 3 C2, SE NG-106 x SE NG-148 and TZBR
Syn W x TZBR ELD 3 C2. For grain yield, significant positive sca effects were obtained
95
for SE NG-77 crossed to SE NG-33, SE NG-67, SE NG-106, SE NG-119 and AMA
TZBR-W-C1, SE NG-106 crossed to SE NG-33, SE NG-62, SE NG-65, SE NG-148
and TZBR ELD 3 C2, SE NG-148 crossed to SE NG-65 and SE NG-67, TZBR Syn W
crossed to SE NG-67 and TZBR ELD 3 C2, SE NG-33 x SE NG-119 and AMA TZBR-
W-C1 x TZBR ELD 3 C2 (Table 32).
Both mid- and better parent heterosis for dead heart, leaf feeding and ear damage
were generally negative with values of -31.6 % and -46.0 % for dead heart; -6.5 and -
12.5 % for leaf feeding and 4.3 and -7.6 % for ear damage for better- and mid-parent
heterosis respectively (Table 33). In general low heterotic estimates were obtained for
other traits.except in some crosses namely; TZBR Syn W x SE NG-106, TZBR Syn W x
SE NG-33 and SE NG-148 x SE NG-33 for plant stand (Table 34), SE NG-148 x SE
NG-119, TZBR ELD 3 C2 x SE NG-148 and TZBR ELD 3 C2 x SE NG-148 for plant
and ear heights and TZBR Syn W x TZBR ELD 3 C2, TZBR Syn W x SE NG-62, SE
NG-62 x SE NG-106 (Table 35), SE NG-148 x SE NG-65, SE NG-77 x SE NG-62,
TZBR ELD 3 C2 x SE NG-62, SE NG-106 x SE NG-62 for root lodging, SE NG-77 x SE
NG-65, SE NG-119 x SE NG-67 for stalk lodging and TZBR Syn W x SE NG-65, SE
NG-77 x SE NG-65, TZBR Syn W x SE NG-119, for husk cover rating (Table 36)
Others include, SE NG-33 x AMA TZBR-W-C1, TZBR Syn W x SE NG-106, SE NG-
148 x AMA TZBR-W-C1 for number of plants at harvest, SE NG-62 x TZBR ELD 3 C2,
SE NG-62 x SE NG-106, SE NG-106 x TZBR ELD 3 C2, TZBR Syn W x TZBR ELD 3
C2 for field weight, SE NG-33 x SE NG-148, SE NG-62 x SE NG-106, SE NG-106 x
TZBR Syn W for number of ears harvested (Table 37) and SE NG-62 x TZBR ELD 3
C2, SE NG-62 x AMA TZBR-W-C1, SE NG-62 x SE NG-77, SE NG-67 x 69 x TZBR
ELD 3 C2 for ear aspect rating, SE NG-119 x AMA TZBR-W-C1, SE NG-106 x AMA
TZBR-W-C1, SE NG-62 x AMA TZBR-W-C1, TZBR Syn W x AMA TZBR-W-C1 for
96
grain moisture content and SE NG-62 x TZBR ELD 3 C2, SE NG-62 x SE NG-106, SE
NG-106 x TZBR ELD 3 C2, TZBR Syn W x TZBR ELD 3 C2 for grain yield (Table 38).
97
98
99
100
101
102
103
104
DISCUSSION
In searching for resistance genes to be particular stress factor, oe approach will be
to systematically search in areas where the stress occur (Omolo, 1983; Ampofo and
Saxena, 1984; Ampofo et al., 1986b; ICIPE, 1991; Ajala et al., 1995).. Farmers‟ seeds
saved represent a form of mass selection, thus such materials may be carrying specific
adaptive genetic factors which enable them to survive in such challenged environments.
Therefore, the highly significant differences obtained from the evaluation of 212
genotypes grown in southeast Nigeria indicate that there are variations among the
genotypes from which useful selections can be practiced for stem borer resistance and
improved agronomic performances. Knowledge and nature of genotype by environment
interaction are important to the breeder in taking decisions concerning selection methods
and testing procedures for the crop. Coefficient of variation (CV %), is a measure of the
reliability of the data obtained. Nonetheless, high CVs are common in areas where
results obtained are variable due to inherent variability in infestation, sites and other
factors of the environment resulting in non-uniformity of genotypes. Highly significant
genotypic variance in all the traits is indicative of the magnitude of variation that exists
among the genotypes providing the opportunity for selection. Significant GxE
interaction, as obtained in this study indicate the specific adaptive nature of the local
genotypes evaluated, or the particular preferences of the people in a specific locality,
therefore developing maize genotypes for the southeastern region of Nigeria will
necessarily benefit from two approaches, i) will be to improve on locally adapted
varieties based on information generated from participatory identification of locally
desired traits, and ii) is to generate a broad based widely adapted population from which
experimental varieties can be generated as desired for different areas.
105
Clustering helps to reduce the size of data generated from evaluations into
manageable groups based on the discriminating variables thus permitting selection of the
best entities within any desired group. The canonical discriminant technique revealed
that out of the 23 parameters assessed only 11 variables comprising both resistant and
agronomic traits, were important in discriminating the genotypes. Among the
discriminating variables, grain yield was more important than all others put together.
Under borer infestation, the borer damage parameters were second to grain yield in
determining which genotype belonged to which cluster. Generally, the local varieties
were highly variable resulting in their being distributed in all the groups. Thus,
genotypes distributed alongside the improved genotypes can be considered as good as
the improved for both agronomic and stem borer attributes. Furthermore, using the rank
summation index method, the best local genotypes were distributed in the same groups
where the improved genotypes were located whereas; the worst two genotypes were
located in a different cluster. The identification of the best genotypes within the
supposedly best groups supports the usefulness of a selection index, in this case RSI for
selection purposes (Ngwuta, et al., 2001). That a large proportion of the local genotypes
evaluated were collected from Abia State could be related to the fact that the State has an
array of agricultural research stations namely Michael Okpara University of Agriculture,
the National Cereals Research Institute, National Seed Service (NSS) and the National
Root Crops Research Institute. Collaborative research links between these and other
research centers across the country could have resulted in the introduction of diverse
maize germplasm to the region. The eight local genotypes that were finally selected for
further studies were from within 30 km radius of the National Cereals Research Institute
station at Amakama, National Cereal Research Institute had the national mandate for
maize until mid 1990s when the research mandates for different institutes in Nigeria
106
were re-distributed. Extensive evaluations of maize genotypes at this station could have
resulted in inadvertent distribution of improved genotypes around the station from on-
farm and other form of farmer participatory testing. Continuous cultivation and use of
remnant harvest as seed every year could have led to the re-classification of such
genotypes as local. In addition, the high level of interaction between this station, IITA
and other research institutes in Nigeria enabled the formation of AMA TZBR-W-C-1,
adjudged the best in this study. AMA TZBR-W-C-1 was formed through a shuttle
breeding approach involving the identification of resistant indivulas from a number of
open pollinated varieties evaluated at Amakama, selfing these individuals and
reconfirming their level of stem borer resistance under artificially infestation with
Sesamia and Eldana at Ibadan before recombination. Two cycles of selection using this
approach led to the development of the variety hence the name AMA from Amakam
(IITA Project 4-Improving maize-grain legume systems in West and Central Annual
Report 2000).
The range of values obtained for most of the agronomic traits and stem borer
damage parameters revealed the possibility of identifying genotypes with desirable traits
among the local open-pollinated varieties from southeastern Nigeria. Low estimates for
leaf feeding damage and dead heart suggests that most of the genotypes possess
resistance factors for damages caused by early instars. The low damage may have been
as a result of the bores finding the plants unsuitable for food and therefore may have
wandered away without causing much damage to the plants, starve to death or get killed.
Barrow (1985; 1987) attributed similar results to either the insects getting killed or
repelled and secondly, to the impairment of insect development. Resistance to attack and
damage by stem borers at seedling stage is desirable in the southeastern Nigeria where
the borer species have three or more overlapping generations per year or cropping cycle.
107
Mihm (1985) had erlier advocated the breeding of maize varieties that possess resistance
characteristics for a long time or in key plant parts in order to prevent, or minimize,
feeding damage and yield losses.
Comparison of parents with crosses signify the presence of considerable average
heterosis or vigour in resultant crosses In cases where there were no apparent differences
between the parental means and the crosses means, it was still possible to identify a few
crosses which were outstanding in their performances. Significant parent vs crosses
mean squares obtained for deadheart and leaf feeding damage revealed that vigour per se
play a significant role in resistance and thus suggest the use of hybrid varieties in
controlling stem borer attack in the region. Significant variances due to gca and sca
implies that both additive and non-additive (particularly dominance) variances are
important in controlling the expression of most of the agronomic and stem borer damage
parameters in maize. Dead heart, a stem borer damage parameter of the seedling stage
showed significant variance for sca but not for gca further confirming that hybrid
breeding approach should be considered in breeding for resistance to the trait. Similar
observation was reported for improved maize populations (Ajala, et al. 2008).. A high
general combiner is characterized by its better breeding value when crossed with a
number of other parents. Depending upon the character of interest, the magnitude and
direction of gca will determine its usefulness. Furthermore, per se performance of the
parent is also considered together with gca since the former offers reliability to gca as a
guide to selection of parent. Significant differences among crosses for sca has been
associated with genetic diversity between genotypes and consequently used to group
genotypes into heterotic pools (Cruz and Vencovsky, 1989; Pswarayi and Vivek, 2004).
Estimated values of sca provides important information about the hybrid performance as
related to its parents, showing the importance of non-additive interactions due to major
108
or minor gene effects when in a particular hybrid combination. Cruz and Vencovsky
(1989) reported that parents with high positive gca and showing high sca in crosses with
some other genotypes could be related to the allelic frequencies at the loci with some
dominance effects. Workers at the international maize and wheat improvement center
(CIMMYT) and the International Institute of Tropical Agriculture have used the
measures of GCA and SCA effects to establish heterotic patterns among genotypes.
(Beck et al., 1990; Crossa et al., 1990 and Vasal et al., 1992; Badu-Apraku et al, 2008,
Ajala et al, 2008; Dhiliwayo et al 2009). Han et al., (1991), Vasal et al., (1992) and
Gama et al., (1995) reported that on the average, crosses produced by crossing
genetically diverse populations or lines generate better SCA effects in the desired
direction..
The overall aim of this study was to use existing local genotypes to develop
better performing newer genotypes for the south east. It is widely recognized that hybrids
are more productive than open pollinated varieties. Trend in agricultural development
efforts is therefore directed towards the development and use of hybrids. Towards this
end, identification of individual populations that will perform well in crosses or the
development of heterotic pools from which hybrids can be generated at will is essential.
Consequently, five parameters comprising percentage dead heart, leaf feeding damage,
ear damage, days to anthesis and grain yield were considered in the formation of a
heterotic pair from the selcted maize populations in this study. In general crosses
involving SE NG-65 and SE NG-67 had poor combining abilities with very low grain
yield values and were thus eliminated. The remaining nine genotypes can be grouped as
follows based on their combining abilities and grain yield. Group A comprising SE NG-
33, SE NG-77, SE NG-106, SE NG-148 and TZBR Syn W while Group B included SE
NG-62, SE NG-119, AMA TZBR-W-C1 and TZBR ELD 3 C2. Furthermore, the best five
109
yielding crosses namely SE NG-33 x TZBR ELD 3 C2, SE NG-62 x SE NG-77, SE NG-
62 x SE NG-106, SE NG-106 x TZBR ELD 3 C2 and TZBR Syn W x TZBR ELD 3 C2
may be used as population crosses or in the formation of composite varieties. This gave
rise to mean yield of 1.26 t ha-1
representing a 42 % yield advantage over the mean of the
parents (0.89 t ha-1
).
110
SUMMARY AND CONCLUSION
A total of 212 open-pollinated maize genotypes made up of 209 locally collected
cultivars from southeast Nigeria and three improved populations with known levels of
resistance to stem borers from IITA were evaluated in field trials from 2000 to 2002 in
two experiments..
In the first experiment, the entries were evaluated in three locations in a total of 4
environments namely Egbema, a stem borer hotspot location in southeastern Nigeria, and
in Ikenne and Ibadan, both in the southwestern part of Nigeria. The two Ibadan trials
were under artificially infestation with stem borers and under non-infested condition,
respectively.. Data on agronomic and borer damage parameters were collected and
subjected to uni- and multi-variate analyses. A total of five discriminant groups or
clusters were formed in each case. The groupings were essentially based on their
differences in yield and borer damage attributes namely cob damage, leaf feeding
damage and stalk breakage. Using the rank summation index involving the yield and
borer damage parameters, 11 populations were selected representing the top 5 % of the
entries. Among the entries selected were the three advanced genotypes and eight local
maize cultivars.
In the second experiment, the selected genotypes were crossed in a diallel fashion
and evaluated in field trials along side the parents to give a total of 66 entries. The 66
genotypes were evaluated in seven environments in a randomized complete block design
with three replications. Data collected were subjected to analysis of variance and only
those found significant were subjected to diallel analysis for general and specific
combining ability using the Diallel-SAS05: A comprehensive programme for Griffing
and Gardner-Eberhart (GEAN II) analysis (Zang et al., 2005) outlined in Griffing‟s
method 2 model 1 of 1956 for fixed effects. Significant GCA and SCA effects were
111
obtained for most of the traits studied in the various environments and in the pooled
environment thus indicating that additive and non-additive gene effects were involved in
the expressions of the traits studied. However, in a few cases, only GCA or SCA was
important thus indicating the relative importance of the genetic component of the
variance.
The assessment of the agronomic and borer damage attributes of the parents and
the crosses indicate that the variety crosses were not superior to the parents in most of
the traits. The significant differences observed between the parents and the crosses for
dead heart and leaf feeding damage parameters is suggestive of the occurrence of
exploitable heterosis for the development of genotypes that are resistant to stem borer
attack.
Genotypes SE NG-33, SE NG-65 and TZBR Syn W had high negative gca values
for dead heart while SE NG-62, SE NG-148, TZBR Syn W and TZBR ELD 3 C2 had the
high negative gca values for leaf feeding damage. For ear damage, SE NG-65, SE NG-
67, SE NG-119, SE NG-148 and AMA TZBR-W-C 1 had high negative gca estimates.
Genotypes SE NG-33, SE NG-62, SE NG-65, SE NG-77, SE NG-106 and SE NG-119
had the highest positive gca effects for grain yield.
The nine genotypes selected formed two heterotic pools: Group A comprised SE
NG-33, SE NG-77, SE NG-106, SE NG-148 and TZBR Syn W while Group B included
SE NG-62, SE NG-119, AMA TZBR-W-C1 and TZBR ELD 3 C2. Average yield of the
grouped genotypes crossed in all possible combinations was 1.06 t ha-1
showing 5 %
yield increase. Furthermore, the best five yielding crosses (SE NG-33 x TZBR ELD 3
C2, SE NG-62 x SE NG-77, SE NG-62 x SE NG-106, SE NG-106 x TZBR ELD 3 C2
and TZBR Syn W x TZBR ELD 3 C2) selected may be used as population crosses or in
the formation of composite varieties.
112
REFERENCES
Adeyemi, S. A. O., J. Donelly, and J. A. Adetoyinbo (1966) Studies on chemical control
of stem borers of maize. Nigerian Agricultural Journal. 3: 61-66.
Aguiar, A. M., L. A. Carlini-Garcia; A. R. Da Silva; M. F. Santos; A. A. F. Garcia and
C. L. De Souza Jr. (2003). Combining ability of inbred lines of maize and
stability of their respective single crosses. Ssientia Agricola. 60 (1): 83-89.
Ahmad, Q., M.A. Rana and S. U. H. Siddiqui (1991). Sunflower seed yield as influenced
by some agronomic and seed characters. Euphytica 56, 137–142.
Ajala, S. O. (1992). The potential of some tropical maize genotypes as sources of high-
yielding inbred lines. Discovery and Innovation. 4: 79-83.
Ajala, S. O. (1993). Populations diallel among maize genotypes with varying levels of
resistance to the spotted stem-borer, (Chilo patellus SWINEHOE). Maydica
38:39-45.
Ajala, S. O. and K. N. Saxena (1993). Assessing tolerance to Chilo partellus (Swinehoe)
attack in maize.Ann. plant resistance to insects Newsletter. 19: 34.
Ajala, S. O., L. M. Smith II, and A. Odulaja (1995). Potential of Kenyan Local maize
(Zea mays L) germplasm as a source of resistance to the spotted stem borer chilo
partellus (Swinhoe). Trop. Agric. (Trinidad), 72 (4): 297 - 302.
Ajala, S.O. (1992). Combining ability for maturity and agronomic traits in some tropical
maize (Zea mays L.) populations. Tropical Agric. 69: 29-34.
Ajala, S.O. and K. N. Saxena (1994). Interrelationship among Chilo partellus
(Swinehoe) Damage parameters and Their Contribution to Grain Yield Reduction
in Maize (Zea mays L). Applied Entomology, 29(4): 469 - 476.
Ajala, S.O., J.G. kling, F. Schulthess, K. Cardwell and A. Odiyi (2002). Progress in
breeding for resistance to Maize stem borers Sesamia calamistis and Eldana
saccharina in West and Central Africa. In Integrated approaches to higher maize
productivity in the new millennium. Proceedings of the seventh Eastern and
Southern Africa regional maize conference (5-11 February, 2002), Nairobi
Kenya. Pp 539.
Ajala, S. O., A. C. Odiyi, C. The, G. Olaoye (2008). Population crop diallel o maize
genotypes with varying levels of resistance to the pink stalk borer (Sesamia
calamistis: Hampson) and sugarcane borer (Eldana saccharina: Walker). Maydica
53: pp 79-86.
Allard, R. W. (1960). Principles of plant breeding. John Wiley and Sons Inc. New York.
Pp 458.
113
Ampofo, J. K. O. (1986). Maize stalk borer (Lepidoptera: Pyralidae) damage and plant
resistance. Environ. Entomol. 15: 12 - 119.
Ampofo, J. K. O., Z. T. Dabrowski and E. O. Omolo. (1986a)Registration ICZI-CM (GP
145) and 1CZ2-CM) (GP 146) maize germplasm lines Crop Sci. 26: 650.
Ampofo, J.K.O., K.N. Sexena, J.G. Kibuka, E.O. Nyangiri (1986b). Evaluation of some
maize genotypes for resistance to the stem borer Chilo partellus (Swinhoe) in
Western kenya. Maydica 31: 379 - 389.
Anderson, E. and W. L. Brown (1952). Origin of maize and its genetic significance. In:
Heterosis. Pp 124-128. Iowa State College Press.
Appert, J. (1970). Insects harmful to maize in Africa and Madagascar. Madagascar
Institute of Agronomic Research, 71 pp.
Barrow, M.R. (1987) The effect of first generation Maize stalk borer Busseola fusca
(Fuller) (Lepidoptera: Noctuidae), on yield of different maize genotypes. J. Ent.
Soc. South Africa 50(2): 291 - 298.
Bdliya, P. M. and S. Burris (1988). Diallel analysis of tolerance of drying injury in seed
corn. Crop Sci. 28: 935-938.
Beal, W. J. (1880). Re. Michigan Board Agric. P 287-288. Cited in Hallauer, A. R. and J.
B. Miranda (1988). Quantitative genetics in maize breeding. 2nd
edition. IOWA
State Univ. Press Ames.
Beck, D. L., S. K. Vasal and J. Crossa (1990). Heterosis and combing ability of
CIMMYT‟s tropical early and intermediate maturity maize (Zea mays L.)
germplasm. Maydica 35: 273-278.
Beck, S. D. (1965). Resistance of plants to insects. A Rev. Ent. 10: 207-232.
Becker, W. A. (1984). Manual of quantitative genetics (4th
edition). Academic
Enterprises. Pullman, Wash., USA.
Belloti, A. (1990). A review of control strategy is for important cassava pests in the
Americas. In: S. K. Han and F. E. Caveness (eds.) Integrated pest management
for tropical roots and tuber crops. IITA, Ibadan Nigeria, pp 58-65.
Bosque-Perez, N. A. and J. H. Mareck (1990). Distribution and species composition of
Lepidopterous maize borers in southern Nigeria. Bulletin of Entomological
Research 80: 363 - 368.
Bosque-Perez, N. A., J. H. Mareck, Z. T. Dabrowski, L. Evrett, S. K. Kim and Y. Efron
(1989). Screening and breeding maize for resistance to S. calamistis and E.
saccharina. P. 163-169. In CIMMYT. 1989. Towards insect resistant maize for
the third world. El Batan Mexico. 9-14 Mar. 1989. CIMMYT, Mexico City.
Bowden, J. (1976). Stem borer ecology and strategy for control. Annals of Biology. 84:
107-134.
114
Bridge, J. (2000). Nematodes of bananas and plantains in Africa: research trends and
management strategies relating to the small-scale farmer. Acta Horticulturae 540:
391–408.
Brucher, H. (1989). Useful plants of Neotropical origin and their world relatives.
Springer. Berlin.
Butchinson, J. B. (ed.), (1974). Evolutionary studies in World Crops. Cambridge Univ.
Press. London.
Cantrell, R. G. and J. L. Geadelmann (1981). Inheritance of husk leaves in maize. Crop
Sci. 21: 541-544.
Cardwell, K. F., F. Shulthess, R. Ndemah and maize yield losses due to pests and
diseases in Cameroon. Agri., Ecosyst. And Environ. 3: 33-47.
Chiang, M. S. and M. Hudon (1976). A study of maize inbred lines for their resistance to
European corn borer, Ostrinia nubilis. Phytoprotection. 57: 36-40.
Cocherham, C. C. (1963). Estimation of genetic variances. Pp 53-94. In: Plauson, W. D.
and H. F. Robinson (eds.) Statistical genetics and Plant breeding. Nat. Acad. Sci.
Res. Council Publ. No. 982, Washington, D. C., USA.
Comstock, R. E. and H. F. (1952). Robinson estimation of average dominance of genes.
In: Heterosis. Gowen, J. W. (ed.), pp 494-516. Iowa State College Press, Ames,
Iowa.
Comstock, R. E. and H. F. Robinson (1948). The components of variance in a population
of bi-parental progenies and their use in estimating the average degree of
dominance. Biometrics 4: 254-266.
Crossa, J. and C. O. Gardner (1987). Integration exotic germplasm for improving
adapted maize populations. Crop Sci. 27: 187-190.
Cruz, C. D. and R. Vencovsky (1989). Comparacao de alguns metodos de analise
dialelica. R. Bras. Genet. 12: 425-438.
Daoyu, Z. and G. S. Lawes (2000). Manova and discriminant analyses of phenotypic
data as guide for parent selection in kiwifruit (Actinidia deliciosa) breeding.
Euphytica 114: 151-157.
Davis, F. M. and W. P. Williams (1989). Methods used to screen for and to determine
mechanisms of resistance to the southwestern corn borer and fall armyworm. In
CIMMYT. 1989. towards Insect Resistant Maize for the Third World. Proc. Int.
Symp. Methodologies for developing Host Plant Resistance to Maize Insect, El
Batán Mexico, 9-14 March 1987. CIMMYT Mexico City, P101-108.
De Leon, J. L. and J. H. Lonnquist, (1978). Heterosis in full-sibs within and between
half-sib families in open-pollinated varieties of maize. Crop Sci. 18: 26-28.
Dickson, C. H. and J. A. Lucas (1977). Plant pathology and plant pathogens. 2nd
edition
Oxford: Blackwell Scientific Publication.
115
Diz, D. A., D.S. Wofford and S. C. Schank (1994). Correlation and path-coefficient
analyses of seed-yield components in pearl millet x elephant grass hybrids.
Theoretical and Applied Genetics 89: 112–115.
Eberhart, S. A., S. K. Kim, J. Mareck, L. L. Darrah and M. M. Goodman (1991). A
comprehensive breeding system for maize improvement in Africa. Pp 175-193. In
N. Q. Ng, P. Perrino, F. Attere and H. Zedan (eds.) Crop Genetic Resource of
Africa. Vol. 11. IITA, IBPGR, UNEP, CNR.
Endrody-Younga, S. (1968). The stem borer, S. botaniphaga Tams and Bowden (Lep.,
Noctuidae) and maize crop in central Ashanti, Ghana. Ghana J. Agricultural Sci.
1: 103-131.
Ezueh, M. I. (1978). An Introductory Economic Entomology for Nigeria. Ethiope
Publishing Corporation, Benin City, Nigeria. 123pp.
Fajemisin, J. M., Y. Efron, S. K. Kim, F. H. Khadri, Z. T. Dabrowski, J. H. Mareck, M.
Bjarnsoon, V. Parkinson, L. A. Everett and A Diallo (1985). Population and
varietal development in maize for tropical Africa through resistance breeding
approach. Pp 385-407. In Brandolini and F. Salami (eds.) Breeding strategies for
maize production improvement in the tropics. FAO and Istitut Agronomico Per
L‟Oltremare, Firenze, Italy.
Fakorede, M. A. B. and S. B. Agbana (1983). Heterotic effects and association of
seedling vigour with mature plant characteristics and grain yield in some tropical
maize genotypes. Maydica 28: 327-338.
Falowo, D. O. (1982). An application of principal component analysis: Response to F3
Amazon cocoa seedlings to insecticide treatments. Nigerian Agricultural Journal.
17/18: 178-188.
FAO, (1968). Food composition Table for Use in Africa. Food and Agriculture
Organisation (FAO), of the United Nations, Nutrition Division. 306 pp.
Fatunla, T. (1973). Evaluation of exotic tomato genotypes in diallel cross. Egypt J.
Genet. Cytol. 2: 316-321.
Gardner, C. O. (1963). Estimates of genetic parameters and their implications in plant
breeding. Pp 225-252. In: Statistical genetics and plant breeding. NAS-NRC. No.
982.
Gardner, C. O. and E. Paterniani (1967). A genetic model used to evaluate the breeding
potential of open-pollinated varieties of corn. Cience
Gardner, C. O. and S. A. Eberhart (1966). Analysis and interpretation of the variety cross
diallel and related populations. Biometrics. 22: 439-451.
Genter, C. F. and S. A. Eberhart (1974). Performance of original and advanced maize
populations and their diallel crosses. Crop Sci. 14: 881-885.
Gilbert, N. E. G. (1958). Diallel cross in plant breeding. Heredity 12: 477-492.
116
Girling, D. J. (1980). E. saccharina as a crop pest in Ghana. Tropical pest management.
26: 152-157.
Gounou, S., .F. Schulthess, T. Shanower, W.N.O. Hammond, H. Braima, A.R. Cudjoe,
R. Adjakloe, K.K. Antwi and I. Olaleye, (1994). Stem Borer and Ear borers of
maize in Nigeria. Plant Health Management Research Monograph No. 4
International Institute of Tropical Agriculture, Ibadan, Nigeria. 25 pp.
Gravois, K. A. (1998). Optimizing selection for rough rice yield, head rice, and total
milled rice. Euphytica 101: 151–156.
Green, J. M. (1948). Inheritance of combining ability in maize hybrids. J. Ame. Soc.
Agron. 40: 58-63.
Griffing, B. (1956a). Concepts of general and specific combining ability in relation to
diallel crossing system. Aust. J. Biol. Sci. 9: 463-493.
Griffing, B. (1956b). A generalized treatment of the use of diallel crosses in quantitative
inheritance. Heredity. 10: 31-50.
Grissom, D. B. and R. R. Kalton (1956). Inheritance of combining ability for forage
traits in brome grass (Bromus inermis Leyss) Agron. J. 48: 289-293.
Guler, M., M. S. Adak and H. Ulukan (2001). Determining relationships among yield
and some yield components using path coefficient analysis in chickpea (Cicer
arietinum L.). European Journal of Agronomy 14: 161–166.
Guthrie, W. D. (1989). Breeding for insect resistance in maize. Plant Breeding Rev.6:
209-243.
Guthrie, W. D., C. T. Tseng, J. Knole and J. L. Jarvis (1982). European corn borer and
maize chlorotic dwarf virus resistance-susceptibility in inbred lines of dwarf
maize. Maydica. 27: 221-233.
Guthrie, W.D., F.F. Dicke and C.R. Neiswander (1960) Leaf and sheath feeding
resistance to the European Corn borer in eight Lines of dent corn, Ohio Agric.
Experimental Station Research bulletin 860. 38 pp.
Hallauer, A. R. and J. B. Miranda (1988). Quantitative genetics in maize breeding. 2nd
edition. IOWA State Univ. Press 468 pp.
Hallauer, A. R., J. B.
Hallauer, A. R., W. A. Russell and K. R. Lamkey (1988). Corn breeding. P. 463-565. In
G. F. Sprague and J. W. Dudley (ed.) Corn and corn improvement. Agronomy
monograph 18, 3rd
edition. ASA. CSSA. Madison, WI.
Han, G. C., S. K. Vasal, D. L. Beck and E. Elias (1991). Combining ability of inbred
lines derived from CIMMYT maize (Zea mays L.) germplasm. Maydica 36: 57-
64).
117
Harris, K. M. (1962). Lepidopterous stem borers of cereals in Nigeria. Bulletin of
Economical Research. 53: 139-171.
Hayes, J. D. and C. A. Foster (1976). Heterosis in self-pollinating crops with particular
references to barley. Pp 239-257. In: Janossy, A. and G. H. Lupton (eds.)
Heterosis in plant breeding. Proceedings of the seventh Congress of EUCARPIA.
Akademiai Kiado, Budapest.
Hayman, B. I. (1954a). The theory and analysis of diallel crosses. Genetics 39: 789-809.
Hayman, B. I. (1954b). The analysis of variance of diallel tables. Biometrics 10: 235-
244.
Hayman, B. I. (1957). Interaction, heterosis and diallel crosses. Genetics 42: 336-355.
Herzog, D. C. and J. E. Funderbark (1985). Plant resistance and cultural practice
interactions with biological control. In: Biological control in agricultural IPM
system. Academic Press.
ICIPE, (1991). Integrated pest management of crop borers for resource-poor farmers in
Africa. – Contribution of plant resistance to insect pests: Achievements and
Highlights 1980-1990. International Center of Insect Physiology and Ecology,
Nairobi, Kenya. 30 pp.
IITA, (1983). International Institute of Tropical Agriculture, Research Highlights for
1983 Ibadan, Nigeria. 172 pp.
Jago, N. D. (1992). Integrated Pest Management in the Sahelian zone, peasant –level
farm environment of North-West Mali 25-32. In: Gibson, R. W. and A.
Sweetmore (eds.) proceedings of a seminar on crop protection for resource-poor
farmers. Pp 168. East Sussex, United Kingdom.
Jeffreys, M. D. W., (10670. Pre-Columbian maize in South Africa. Nature 215: 695-697.
Jinks, J. L. (1954). The analysis of continuous variation in a diallel cross of Nicotiana
rustica. Genetics. 39: 767-788.
Jinks, J. L. (1955). A survey of the genetical basis of heterosis in a variety of diallel
crosses. Heredity 9: 223-238.
Jinks, J. L. (1956). The F2 and backcross generation from a set of diallel crosses.
Heredity 10: 1-30.
Khalifa, I., and N. Drolsom. (1988). Combining ability for European corn borer
resistance and three agronomic traits in maize. maydica 33:247-259.
Khush, G.S. and D.S. Brar (1991). Genetics of resistance to insects in crop plants. in
N.C. Brady (ed.) Advances in Agronomy 45: 223-274.
Kim, S. K., Y. Efron, J. Fajemisin and F. Salamini (eds.). Breeding strategies for maize
production improvement in the tropics. FAO, Rome/Instituto Agronomico per
L‟Oltremare, Florence, Italy.
118
Kling, J.G. and N.A. Bosque-Perez (1995). Progress in Screening and Breeding for
Resistance to the Maize stem Borers E. saccharina and S. calamistis In: D.C.
Jewell, S.R. Waddington, J.K. Ransom and K.V. Pixley eds. Maize Research for
Stress Environments. Proceeding of Fourth Eastern and southern African
Regional Maize Conference held at Harare, Zimbabwe, 28th
- 1st March, 1994.
CIMMYT, Mexico D.F. pp182 - 186
Kogan, M. O. (1982). Plant Resistance in pest management. In: Metcalf, R. A. and W. H.
Lukmann (eds.). Introduction to insect pest management. 2nd
Edition. Wiley-
Interscience, New York. Pp 93-137.
Kogan, M. O. and E. F. Orthman (1978). Antixenosis- a new term proposed to replace
Painter‟s “nonpreference” modality of resistance. ESA Bull. 24:175-176.
Kouame, K. L. (1995). Seasonal abundance of the two maize stem borers S. calamistis
and E. saccharina and bionomics of the Sesamia egg parasitoid, Telenomus
busseolae. Ph. D. Thesis Simin Frazer University pp 137.
Kozak, M. and M. S. KANG (2006). Note on modern path analysis in application to crop
science. Communications in Biometry and Crop Science 1: 32–34.
Lawani, S.M. A review of the effects of various agronomic practices on cereal stem
borer populations. Tropical Pest Management. 28 (3): 266-276.
Lee, C. H. (1974). Reciprocal recurrent selection and genetical study in two varieties of
maize (Zea mays L.). Master of Agriculture Science dissertation. University of
Malaya, Malaysia.
Leffel, R. C. and W. D. Hanson (1961). Early generation testing of diallel crosses of
soybeans. Crop Sci. 1: 169-174.
Liang, G. H., C. R. Reddy and A. D. Dayton (1972). Heterosis, inbreeding depression
and heritability estimates in a systematic series of grain sorghum genotypes.
Crop Sci. 12: 409-411.
Lin, C. S. and M. R. Binns (1991). Genetic properties of four types of stability
parameters. Theor. Appl. Genet. 82: 505-509.
Lonnquist, J. H. (1951). Recurrent selection as a means of modifying combing ability in
corn. Agron. J. 43: 311-315.
Lonnquist, J. H. and C. O. Gardner (1961). Heterosis in interracial crosses in maize and
its implication in breeding procedures. Crop Sci. 1: 179-183.
Manrique, L. A. (1993). Crop production in the tropics. A review J. Plant Nutrition 16:
1485-1516.
Mather, K. (1949). Biometrical Genetics. 1st edition. Methuen Ltd. London, 369 pp.
Matzinger, D. F., G. F. Sprague and C. C. Cockerhan (1959). Diallel crosses of maize in
experiments repeated over locations and years. Agron J. 51: 346-350.
119
Maxwell, F. G. and P. R. Jennings (1980). Breeding plants resistant to insects. John
Wiley and Sons Inc. New York.
McClure, G. W. (1892). Corn crossing. Illinois Agric Exp. Stn. Bull. 290.
Mihm, J. A. (1985). Breeding for host plant resistance to stem-borers. Insect Sci. applic.
Vol. 6, No. 3, pp 369-377.
Mihm, J. A. (1995). Maize pest management-multiple insect and disease resistant
varieties are the key to success. P 170-181. In: D. C. Jewell, S. R. Waddington, J.
K. Ransom and K. V. Pixely (eds). Maize research for stress environments.
Proceedings of the fourth Eastern and Southern African regional maize
conference held at Harare, Zimbabwe. 28 Mar. to 1 Apr. 1994. Mexico, D. F.
Miller, P. A. and A. Marani (1963). Heterosis and combining ability in diallel crosses of
upland cotton, Gossypium hirsutum L. Crop Sci. 3: 441-444.
Milligan, G.W. and Cooper, M.C. (1985). An examination of procedures for
determrining the number of clusters in a data set, Psychometrika 50: 159-179.
Miranda Filho (1981). Quantitative genetics in maize breeding. Iowa State University
Press, Ames Iowa, USA. .
Miranda Filho J. B. and R. Vencovsky (1984). Analysis of diallel crosses among open-
pollinated varieties of maize (Zea mays L.) Maydica 24: 217-234.
Misevic , D., D. A. Maric, D. E. Alexander, J. Dumanovic and S, Patnovic. (1989).
Population cross diallel among high oil populations of Maize. Crop Sci. 29: 613-
617.
Misevic, D. (1989). Heterotic patterns among US Corn Belt Yugoslavian and exotic
maize populations. Maydica 34:353-363.
Mohammad, S. A., B.M. Prasanna and N. N. SINGH (2003). Sequential path model for
determining interrelationships among grain yield and related characters in maize.
Crop Science 43: 1690–1697.
Mohyuddin, A.I. and M.R. Ahique, (1978). An assessment of loss caused by Chilo
partellus to maize in Pakistan, PANS 24: 111 - 113.
Moll, R. H., M. F. Linsey and H. F. Robinson (1964). Estimates of heritability and
degree of dominance in corn. Genetics 49: 411-423.
Monitor, (2001). Biotechnology and development Monitor. The Network University
(TNU), The Netherlands. No. 48, 24 pp.
Morrow, G. E. and F.D. Gardner (1893). Field experiments with corn. Illinois Agric Exp.
Stn. Bull. 25: 173-203.
Mulumba, N.N. and J. J. Mock, (1978). Improvement of yield potential of the Eto Blanco
maize (Zea mays L.) population by breeding for plant traits. Egyptian J. Genet.
Cytol. 7: 40 - 51.
120
Mungoma, C. and L. M. Pollak (1988). Heterotic patterns among ten corn belt and exotic
maize populations. Crop Sci. 28: 500-504.
Mushonga, J. N., S. C. Gupta and L. R House (1997). Combining ability and heterosis
for diastatic activity in grain sorghum. African Crop Science Journal, Vol. 5.
No.2, pp. 99-106.
Nass, L. L., M. Lima, R. Vencovski and P. B. Gallo (2000). Combining ability of maize
inbred lines evaluated in three environments in Brazil. Brazilian J. of Genet. 8:
81-85.
Ngwuta, A. A., S. O. Ajala, I. U. Obi and E. E. Ene-Obong (2001). Potential Sources of
Resistance to Maize Stem Borers (S. calamistis (Hampson) and E. saccharina
(Walker)) in Local Maize Populations of South-eastern Nigeria. African Crop
Science proceedings, Vol. 5. 23-28.
Obi, I. U. (1991). Maize, its agronomy, diseases, pests and food values. Published by
Optimal computer solutions Limited. Enugu, Nigeria 206 pp.
Obi, I. U. (2002). Statistical methods of detecting differences between treatment means
and research methodology issues in laboratory and field experiments. 2nd
ed.
Published by AP Express Publishers Limited. Nsukka, Nigeria. 117pp.
Obilana, A. T. and M. A. B. Fakorede (198). Heritability: A treatise. Samaru J. of Agric.
Research. 1: 72-82.
Ofori, I. (1996). Correlation and path-coefficient analysis of components of seed yield in
bambara groundnut (Vigna subterranea). Euphytica 91: 103-107.
Ogunbodede, B. A. (1997). Multivariate analysis of genetic diversity in Kenef, Hibiscus
cannabinus (L). African Crop Sci. J. 5 N0. 2: 127-133.
Ogunbodede, B.A., S.R. Ajibade and S.A. Olakojo (2000). Heterosis and combining
ability for yield and yield related characters in some Nigerian local varieties of
maize (Zea mays). Moor J. Agricultural Research 1: 37-43.
Omolo, E. O. (1983). Screening of local and exotic maize lines for stem borer resistance
with special reference to Chilo partellus. Insect Sci. Appl. 4: 105-108.
Painter, R. H. (1951). Insect resistance in crop plants. The MacMillan Co., New York
520 pp.
Painter, R. H. (1968). Crops that resist insects provide a way to increase world food
supply. Kans. Agric. Exp. Sta. Bull. 22: 520.
Patanothai, A. and R. F. Atkins (1971). Heterotic response for vegetative growth and
fruiting development in grain sorghum, Sorghum bicolor (L.) Moench. Crop Sci.
11: 839-843.
Paterniani, E. and J. H. Lonnquist (1963). Heterosis in interracial crosses of corn (Zea
mays L.) Crop Sci.3: 504-507.
121
Phiri, G. S. N. (1995). Interaction of the spotted stem borer Chilo partellus (Swinehoe)
with some alternative hosts and its larval parasitoid Cotesia sesamiae (Cameron)
in Malawi. Ph. D. thesis, University of Reading. 347 pp.
Pswarayi, A. and B. Vivek (2004). Combining ability of CIMMYT‟s early maturing
maize (Zea mays L.) germplasm under stress and non-stress conditions and
identification of testers. Available on www.cropscience.org.au.
Rabiei, B., M. Valizadeh, B. Ghareyazie, and M. Moghaddam (2004). Evaluation of
selection indices for improving rice grain shape. Field Crops Research 89: 359–
367.
Rehms, S. and G. Espig (1991). The cultivated plants of the tropics and subtropics.
Priese GmbH Berlin, Germany. 552 pp.
Rojanaridpiched, C. and V.E. Gracen (1983). Mechanisms of resistance to the European
corn borerin maize. A. Pl. resist. Insects Newsl. 9, 33-34.
Rojas, B. and G. F. Sprague (1952). A comparison of variance component in corn yield
trials. III. General and specific combining ability and their interaction with
locations and years. Agron. J. 44: 462-466.
Rood, S. B. and D. J. Major (1981). Diallel analysis of leaf number, leaf development
rate, and plant height of early maturing maize. Crop Sci. 21: 867-873.
SAFGRAD, (1987) Proceeding of Workshop on the Establishment of Semi-Arid Food
Grain Research and Development (SAFGRAD) Research Network (Maize in
Central and Western Africa), Ougadougou, Burkina Faso, March 23rd
- 27th
1997.
321 pp.
Saleh, G. B., D. Abdullah and A. R. Anuar (2002). Performance, heterosis and
heritability in selected tropical maize single, double and three-way cross hybrids.
138: 21-28.
Sanborn, J. W. (1890). Indian Corn. Rep. Maine Dep. Agric. 33: 54-121.
Sarup, P. (1980). Insect pest management in maize. In: Breeding, production and
protection methodologies of maize in India (ed. J. Singh). Pp 193-197. All India
co-ordinated maize improvement project. IARI, New Delhi.
SAS Institute (1999). Statistical analysis system user‟s guide version 8. SAS Institute,
Inc., Cary, N. C.
Schulthess, F. and S.O. Ajala (1997). Recent advances at IITA in the control of
stemborers in West and Central Africa. In Strategy for sustainable Maize
production in West and Central Africa. Proceeding s of a regional Maize
workshop (21-25) April, 1997. IITA-Cotonou, Benin Republic. Pp 451.
Sekloka, S. (1996). Contribution a l‟etude bioecologique de E. saccharina Walker
(Lepidoptera: Pyralidae) Sur Zea mays L. et les Plantes hotes alternatives.
Memoirs Ing. Agr. UNB, Benin.
122
Sengooba, T. (19931992). Crop protection strategies- their status with resource-poor
farmers. Pp 17-23. In: Gibson, R. W. and A. Sweetmore (eds.) Proceedings of a
seminar on crop protection for resource –poor farmers. Pp 168. East Sussex,
United Kingdom.
Setamou, M. , F. Shulthess, N. A. Bosque-Perez and A. Thomas-Odjo (1995). The effect
of stem and cob borers on maize subjected to different nitrogen and silica
treatments, with special reference to S. calamistis Hampson (Lepidoptera:
Noctuidae). Entomologia Experimentalis et Applicata. 77: 205-210.
Shanower, T. F., F. Schulthess, and S. Gounou, (1991) Distribution and abundance of
some stem and cob borers in Benin Plant Health Management Research
Monograph No.1, International Institute of Tropical Agriculture, Ibadan, Nigeria.
Shipley, B. (2002). Cause and Correlation in Biology. A User’s Guide to Path Analysis,
Structural Equations and Causal Inference. Cambridge, UK: Cambridge
University Press
Singh, B. U., P.G. Padmaja and N. Seetharama (2004). Stability of biochemical
constituents and their relationships with resistance to shoot fly, Atherigona
soccata (Rondani) in seedling sorghum. Euphytica 136: 279–289.
Smith, C. M. (1989). Plant resistance to insects: A fundamental approach. John Wiley
and Sons Inc. N. Y. 286 pp.
Smith, M. E., J. A. Mihm, and D. C. Jewell, (1989). Breeding for multiple resistances to
temperate subtropical and tropical maize insect pests at CIMMYT. In CIMMYT
1989: Towards Insect Resistant Maize for the Third World. Proceeding of the
International symposium on Methodologies for Developing Host Plant Resistance
to Maize Insect. CIMMYT, Mexico, D.F. p. 222 - 234.
Strangland, G. R., W. A. Russell and O. S. Smith (1983). Evaluation of the performance
and combining ability of selected lines derived from improved maize populations
Crop Sci. 23: 647-651.
Sukchain and B. S. Sidhu (1992). Correlation and path-coefficient analysis for
reproductive traits in Guinea grass. Euphytica 60: 57-60.
Sulthess, F., N. A. Bosque-Perez and S. Gounou (1991). Sampling lepidopterous pests on
maize in West Africa. Bull. Of Entomological Research. 81: 279-301.
Teetes, G. L. (1985). Insect resistant sorghums in pest management. Insect Sci. and Its
Application. 6: 443-451.
Topal, A., C. Aydin, N. Akgun and M. Babaoglu (2004). Diallel cross analysis in durum
wheat (Triticum durum Desf.): Identification of best parents for some kernel
physical features. Field Crops Research 87: 1–12.
Troyer. A. F. and A. R. Hallauer (1968). Analysis of diallel set of maize. Crop Sci. 8:
581-584.
123
Uguru, M. I. (2004). Crop Genetics and Breeding. Pub. The comic printers, Onitsha,
Nigeria.
Usua, E. J. (1968). Effects of varying populations of Busiola fusca larvae on the growth
and yield of maize. J. of Econ. Ent. 61: 375-376
Vacaro, E., J. F. B. Neto, D. G. Pegoraro, C. N. Nuss and L. D. H. Conceicao (2002).
Combining ability of 12 maize populations. Pes. Agropec. Bras. 37: (1) 1-10.
Van der Plank, J. E. (1963). Plant Diseases: Epidemics and control. Academic Press,
New York.
Van Eijnattan, C. L. M. (1965). Towards the improvement of maize in Nigeria. Ph. D.
Ph. D. Thesis, Wageningen, The Netherlands.
Vargas, M., J. Crossa, M. P. Reynolds, P. Dhungana, and K.M. Eskridge (2007).
Structural equation modeling for studying genotype x environment interactions of
physiological traits affecting yield in wheat. J. of Agric. Sci., Cambridge 145:
151–161.
Vasal, S. K., G. Srinivasan, S. Pandey, H. S. Cordava, G. C. Han and F. C. Gonzalez
(1992). Heterotic patterns of Ninety-two white tropical CIMMYT maize lines.
Maydica 37: 259-270.
Warui, C. M. and J. N. Kuria (1983). Population, incidence and the control of maize
stalk borers Chilo partellus (Swinehoe), C. orichalcociiellus strand (Pyralidae)
and S. calamistis Hamps. In coast province, Kenya. Insect Sci. and its
application. 4: 11-18.
Webber, R. H. (1901). Loss of vigor in corn from inbreeding . Science 13: 257-258.
Wellhausen, E. J. (1978). Recent developments in maize breeding in the tropics. In:
Walden, D. B. (ed.) Maize Breeding and Genetics. John Wiley, USA.
Whaley, W. G. (1964). Phisiology of gene action in hybrids pp. 98-113. In: J. W. Gowen
(ed.). Heterosis. Hofner publishing company, Inc. New York.
William, R. M. J. and J. S. Brown (1998). Heterosis and combining ability of cottons
originating from different regions of the United States. J. of cotton Sci. 2: 77-84.
Williams, W. P. and F. M. Davis (1984). Registration of Mp 705, Mp 706 and Mp 707
germplasm lins of maize. Cro Sci. 24: 12-17.
Williams, W.P., F. M. Davis and G. E. Scott (1978). Resistance of corn to leaf feeding
damage by the fall armyworm. Crop Sci. 18: 861-863.
Wilson, L. (1988). Biological control: A sustainable solution to crop pest problems in
Africa. In: Proceedings of the inaugural Conference and workshop of the IITA
Biological Control Program Center for Africa. Pp 1-4. International Center for
Tropical Agriculture, Cotonou, People‟s Republic of Benin.
124
Wiseman, B. R. (1985). Types and mechanisms of host plant resistance to insect attack.
Insect Sci. Applic. 6 (3): 239-242.
Wiseman, B. R. and F. M. Davis (1990). Plant resistance to insects attacking corn and
grain sorghum. Florida Entomologist. 73: 446-458.
Younes, M. H. and R. H. Andrew (1978). Productivity and prolificacy in a diallel series
of market sweet corn hybrids. Crop Sci. 18: 224-226.
Zhang, Y. M., S. Kang and K. R. Lamkey (2005). Diallel-SAS: A Comprehensive
Program for Griffing‟s and Gardner-Eberhart Analyses. Agron. J. 97: 1097-1106.
125
APPENDICES
126
127
128
129
130
131
Appendix 2: Cluster Analysis for Data Set from Artificially Infested Maize Plots. The FASTCLUS Procedure Replace=FULL Radius=0 Maxclusters=5 Maxiter=1 Initial Seeds ___________________________________________________________________________ Cluster DYSK PLHT PHARV EHARV __________________________________________________________________________ƒ 1 74.5000000 168.1000000 10.0000000 4.5000000 2 72.0000000 229.7000000 10.0000000 8.5000000 3 63.5000000 140.8000000 8.0000000 2.5000000 4 62.0000000 112.5000000 1.5000000 1.0000000 5 62.5000000 198.7000000 10.5000000 3.5000000 Initial Seeds Cluster S_LODG EDAM LF_DAM YIELD __________________________________________________________________________ƒ 1 6.0000000 6.5000000 4.5000000 0.6206590 2 5.0000000 2.5000000 3.5000000 1.2685165 3 5.0000000 2.0000000 3.5000000 0.4193255 4 4.5000000 1.0000000 5.5000000 0.1586960 5 6.0000000 6.5000000 2.5000000 0.6384885 Criterion Based on Final Seeds = 3.2196 Cluster Summary Maximum Distance RMS Std from Seed Radius Nearest Cluster Frequency Deviation to Observation Exceeded Cluster ________________________________________________________________________ 1 80 3.0017 17.1465 3 2 13 3.0893 15.2257 5 3 28 2.6678 17.5590 1 4 4 3.9158 17.2450 3 5 87 3.3545 17.5720 1 Cluster Summary Distance Between Cluster Cluster Centroids ____________________________ 1 17.0345 2 25.4220 3 17.0345 4 27.3053 5 23.4765 Statistics for Variables Variable Total STD Within STD R-Square RSQ/(1-RSQ) __________________________________________________________________ DYSK 3.89266 3.86578 0.032460 0.033549 PLHT 20.14976 7.15244 0.876389 7.089907 PHARV 2.03901 1.88906 0.157947 0.187574 EHARV 2.06749 1.97252 0.107010 0.119834 S_LODG 0.88333 0.86905 0.050411 0.053087 EDAM 1.95907 1.88195 0.094676 0.104577 LF_DAM 0.72451 0.68589 0.120779 0.137370 YIELD 0.48908 0.45979 0.132939 0.153322 OVER-ALL 7.37377 3.13329 0.822863 4.645358 Pseudo F Statistic = 240.40 Approximate Expected Over-All R-Squared = 0.90010 Cubic Clustering Criterion = -6.682 WARNING: The two values above are invalid for correlated variables. Cluster Means
132
Cluster DYSK PLHT PHARV EHARV __________________________________________________________________________ƒ 1 62.7250000 168.7650000 9.0687500 6.4437500 2 64.4230769 217.5692308 9.6153846 7.5000000 3 63.3928571 151.8375000 8.0357143 5.6250000 4 59.7500000 124.9750000 5.6250000 3.7500000 5 63.6551724 192.1867816 9.9310345 7.1379310 Cluster Means Cluster S_LODG EDAM LF_DAM YIELD __________________________________________________________________________ƒ 1 5.0062500 4.3062500 3.3625000 0.9247487 2 4.4230769 4.1538462 2.6923077 1.4638604 3 5.2500000 3.1428571 3.5000000 0.7831826 4 5.1250000 3.1250000 4.6250000 0.4533950 5 4.7988506 4.9310345 3.2528736 1.1045689 Cluster Standard Deviations Cluster DYSK PLHT PHARV EHARV __________________________________________________________________________ƒ 1 3.757608316 6.733537825 1.878835739 1.990473753 2 3.983539851 6.741461836 2.346574153 1.979057015 3 4.010073295 5.209724259 2.076894813 2.141412652 4 3.523729085 9.654489457 2.954516317 1.892969449 5 3.912049324 7.960565817 1.708685769 1.901177198 Cluster Standard Deviations Cluster S_LODG EDAM LF_DAM YIELD __________________________________________________________________________ƒ 1 0.805415806 1.946261834 0.720297934 0.418166362 2 0.975665453 1.951330907 0.560448538 0.733722105 3 1.084401183 1.676952388 0.793492048 0.447076854 4 1.108677891 1.436140662 1.181453907 0.202513759 5 0.822855812 1.886568827 0.604602326 0.457070939 The FREQ Procedure Table of CLUSTER by ENTRY CLUSTER(Cluster) ENTRY Frequency‚ Percent ‚ Row Pct ‚ Col Pct ‚ 1‚ 2‚ 3‚ 4‚ 5‚ 6‚ Total ________ƒˆ________ˆ________ˆ________ˆ________ˆ________ˆ________ˆ 1 ‚ 1 ‚ 0 ‚ 0 ‚ 0 ‚ 0 ‚ 1 ‚ 80 ‚ 0.47 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.47 ‚ 37.74 ‚ 1.25 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 1.25 ‚ ‚ 100.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 100.00 ‚ ________ƒˆ________ˆ________ˆ________ˆ________ˆ________ˆ________ˆ 2 ‚ 0 ‚ 1 ‚ 0 ‚ 0 ‚ 0 ‚ 0 ‚ 13 ‚ 0.00 ‚ 0.47 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 6.13 ‚ 0.00 ‚ 7.69 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ ‚ 0.00 ‚ 100.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ ________ƒˆ________ˆ________ˆ________ˆ________ˆ________ˆ________ˆ
133
3 ‚ 0 ‚ 0 ‚ 1 ‚ 0 ‚ 0 ‚ 0 ‚ 28 ‚ 0.00 ‚ 0.00 ‚ 0.47 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 13.21 ‚ 0.00 ‚ 0.00 ‚ 3.57 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ ‚ 0.00 ‚ 0.00 ‚ 100.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ ________ƒˆ________ˆ________ˆ________ˆ________ˆ________ˆ________ˆ 4 ‚ 0 ‚ 0 ‚ 0 ‚ 1 ‚ 1 ‚ 0 ‚ 4 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.47 ‚ 0.47 ‚ 0.00 ‚ 1.89 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 25.00 ‚ 25.00 ‚ 0.00 ‚ ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 100.00 ‚ 100.00 ‚ 0.00 ‚ ________ƒˆ________ˆ________ˆ________ˆ________ˆ________ˆ________ˆ 5 ‚ 0 ‚ 0 ‚ 0 ‚ 0 ‚ 0 ‚ 0 ‚ 87 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 41.04 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ 0.00 ‚ ________ƒˆ________ˆ________ˆ________ˆ________ˆ________ˆ________ˆ Total 1 1 1 1 1 1 212 0.47 0.47 0.47 0.47 0.47 0.47 100.00 (Continued) Table of CLUSTER by ENTRY CLUSTER ANALYSIS FOR ART INFESTED MAIZE PLOT (IBAD_INF) 14 Canonical Discriminant Analysis of MAIZE Clusters The CANDISC Procedure Observations 212 DF Total 211 Variables 8 DF Within Classes 207 Classes 5 DF Between Classes 4 Class Level Information Variable CLUSTER Name Frequency Weight Proportion 1 _1 80 80.0000 0.377358 2 _2 13 13.0000 0.061321 3 _3 28 28.0000 0.132075 4 _4 4 4.0000 0.018868 5 _5 87 87.0000 0.410377 CLUSTER ANALYSIS FOR ART INFESTED MAIZE PLOT (IBAD_INF) 15 Canonical Discriminant Analysis of MAIZE Clusters The CANDISC Procedure Univariate Test Statistics F Statistics, Num DF=4, Den DF=207 Total Pooled Between Standard Standard Standard R-Square Variable Deviation Deviation Deviation R-Square / (1-RSq) F Value Pr > F DYSK 3.8927 3.8658 0.7823 0.0325 0.0335 1.74 0.1433 PLHT 20.1498 7.1524 21.0400 0.8764 7.0899 366.90 <.0001 PHARV 2.0390 1.8891 0.9039 0.1579 0.1876 9.71 <.0001 EHARV 2.0675 1.9725 0.7544 0.1070 0.1198 6.20 <.0001 S_LODG 0.8833 0.8691 0.2212 0.0504 0.0531 2.75 0.0294 EDAM 1.9591 1.8820 0.6724 0.0947 0.1046 5.41 0.0004 LF_DAM 0.7245 0.6859 0.2808 0.1208 0.1374 7.11 <.0001 YIELD 0.4891 0.4598 0.1989 0.1329 0.1533 7.93 <.0001 Average R-Square Unweighted 0.1965765 Weighted by Variance 0.8228633 Multivariate Statistics and F Approximations S=4 M=1.5 N=99
134
Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.10136163 19.87 32 739.16 <.0001 Pillai's Trace 1.03564440 8.87 32 812 <.0001 Hotelling-Lawley Trace 7.57080698 47.03 32 512.49 <.0001 Roy's Greatest Root 7.40456873 187.89 8 203 <.0001 NOTE: F Statistic for Roy's Greatest Root is an upper bound. CLUSTER ANALYSIS FOR ART INFESTED MAIZE PLOT (IBAD_INF) 16 Canonical Discriminant Analysis of MAIZE Clusters The CANDISC Procedure Adjusted Approximate Squared Canonical Canonical Standard Canonical Correlation Correlation Error Correlation 1 0.938625 0.935872 0.008191 0.881017 2 0.280749 . 0.063417 0.078820 3 0.258505 . 0.064242 0.066825 4 0.094775 0.008997 0.068224 0.008982 Eigenvalues of Inv(E)*H = CanRsq/(1-CanRsq) Eigenvalue Difference Proportion Cumulative 1 7.4046 7.3190 0.9780 0.9780 2 0.0856 0.0140 0.0113 0.9893 3 0.0716 0.0625 0.0095 0.9988 4 0.0091 0.0012 1.0000 Test of H0: The canonical correlations in the current row and all that follow are zero Likelihood Approximate Ratio F Value Num DF Den DF Pr > F 1 0.10136163 19.87 32 739.16 <.0001 2 0.85190076 1.58 21 577.71 0.0488 3 0.92479297 1.34 12 404 0.1918 4 0.99101768 0.37 5 203 0.8701 CLUSTER ANALYSIS FOR ART INFESTED MAIZE PLOT (IBAD_INF) 17 Canonical Discriminant Analysis of MAIZE Clusters The CANDISC Procedure Total Canonical Structure Variable Can1 Can2 Can3 Can4 DYSK 0.144280 -0.283948 0.210861 0.730558 PLHT 0.997143 -0.067835 0.022753 0.012882 PHARV 0.371001 0.295488 0.667526 -0.051447 EHARV 0.331176 0.072114 0.377594 -0.222470 S_LODG -0.229122 -0.023050 0.220516 0.310949 EDAM 0.248211 0.651152 0.309755 -0.251114 LF_DAM -0.308591 0.553598 -0.366893 0.644352 YIELD 0.380599 -0.240665 -0.025602 -0.281145 Between Canonical Structure Variable Can1 Can2 Can3 Can4 DYSK 0.751663 -0.442469 0.302546 0.384303 PLHT 0.999772 -0.020344 0.006283 0.001304 PHARV 0.876216 0.208739 0.434192 -0.012269 EHARV 0.950252 0.061891 0.298388 -0.064455 S_LODG -0.957852 -0.028822 0.253891 0.131257 EDAM 0.757170 0.594128 0.260236 -0.077347 LF_DAM -0.833452 0.447217 -0.272907 0.175720 YIELD 0.979790 -0.185312 -0.018151 -0.073080 Pooled Within Canonical Structure
135
Variable Can1 Can2 Can3 Can4 DYSK 0.050596 -0.277062 0.207083 0.739368 PLHT 0.978299 -0.185182 0.062515 0.036475 PHARV 0.139459 0.309060 0.702717 -0.055813 EHARV 0.120887 0.073244 0.385996 -0.234363 S_LODG -0.081104 -0.022703 0.218601 0.317660 EDAM 0.089983 0.656829 0.314484 -0.262730 LF_DAM -0.113521 0.566654 -0.377983 0.684092 YIELD 0.140989 -0.248062 -0.026560 -0.300570 CLUSTER ANALYSIS FOR ART INFESTED MAIZE PLOT (IBAD_INF) 18 Canonical Discriminant Analysis of MAIZE Clusters The CANDISC Procedure Total-Sample Standardized Canonical Coefficients Variable Can1 Can2 Can3 Can4 DYSK -0.076774400 -0.301084514 0.296387686 0.592069354 PLHT 2.918741874 0.109841931 -0.351115250 0.167787580 PHARV -0.054210128 0.122665733 0.910202441 0.227900292 EHARV 0.074491568 0.414440718 0.472000695 0.136159089 S_LODG -0.199896101 -0.605105648 0.107931222 0.154962632 EDAM 0.116824348 0.508199982 -0.183383300 -0.426851880 LF_DAM 0.107005504 0.608582491 -0.542197635 0.645844311 YIELD -0.194977911 -0.856099641 -0.745462317 -0.109351635 Pooled Within-Class Standardized Canonical Coefficients Variable Can1 Can2 Can3 Can4 DYSK -0.076244222 -0.299005327 0.294340934 0.587980724 PLHT 1.036048248 0.038989930 -0.124633269 0.059558548 PHARV -0.050223403 0.113644640 0.843264265 0.211140032 EHARV 0.071070020 0.395404619 0.450320750 0.129905027 S_LODG -0.196665547 -0.595326437 0.106186928 0.152458255 EDAM 0.112225459 0.488194260 -0.176164262 -0.410048495 LF_DAM 0.101300424 0.576135451 -0.513289955 0.611410629 YIELD -0.183301577 -0.804831752 -0.700819990 -0.102803066 Raw Canonical Coefficients Variable Can1 Can2 Can3 Can4 DYSK -0.019722869 -0.077346750 0.076140164 0.152098956 PLHT 0.144852412 0.005451276 -0.017425279 0.008327025 PHARV -0.026586498 0.060159465 0.446394364 0.111770087 EHARV 0.036030033 0.200456418 0.228296990 0.065857340 S_LODG -0.226298722 -0.685029046 0.122186964 0.175430363 EDAM 0.059632453 0.259408350 -0.093607165 -0.217884584 LF_DAM 0.147692790 0.839987125 -0.748360394 0.891417210 YIELD -0.398666099 -1.750443944 -1.524226778 -0.223588351 Class Means on Canonical Variables CLUSTER Can1 Can2 Can3 Can4 1 -1.359660603 0.036760407 0.105325224 -0.103885950 2 5.508757592 -0.730599739 -0.507830151 -0.095616849 3 -3.874647605 -0.438911242 0.054647063 0.137824427 4 -7.372509788 0.778790547 -1.611656380 0.025679733 5 2.013093927 0.180819732 0.035543469 0.064276921
136
Appendix 3: Form of combined analysis of variance with degrees of freedom, only used
for the eleven (11) open pollinated maize genotypes and their crosses evaluated in
seven (7) environments.
Source of
Variation D.f. D.f.
Environment (E) e-1 6
Blocks/ E e(1-r) 14
Genotype g-1 65
Parents n-1 10
Crosses n(n-1)/2 54
GCA p-1 10
SCA P(p-3)/2 44
Parents x Crosses 1
Genotype x E (g-1)(e-1) 390
Parents x E (n-1)(e-1) 60
Crosses x E (n(n-1)/2)(e-1) 324
GCA x E (p-1)(e-1) 60
SCA x E (p(p-3)/2)(e-1) 264
Parents-Crosses x E 6
Error e(r-1)(-1) 910
Total ger-1 1385