Crop Science UNIVERSITY OF NIGERIA AGWU ABRAHAM.pdfCrop Science A THESIS SUBMITTED TO THE DEPARTMENT...

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


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


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



means for the 11 genotypes evaluated for two (2) agronomic traits

69



means for the 11 genotypes evaluated for three (3) agronomic traits

70



means for the 11 genotypes evaluated for three (3) agronomic traits

72


diagonal) SCA effects (below diagonal), GCA effects and array means

for the 11 genotypes evaluated for three (3) agronomic traits.

74


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


diagonal) of open-pollinated maze crosses for three (3) agronomic traits

81



82



83



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


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


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

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.

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).

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).

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.

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.

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).

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.

http://www.crop/

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

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

Crop Science UNIVERSITY OF NIGERIA AGWU ABRAHAM.pdfCrop Science A THESIS SUBMITTED TO THE DEPARTMENT...

Documents

Transcript of Crop Science UNIVERSITY OF NIGERIA AGWU ABRAHAM.pdfCrop Science A THESIS SUBMITTED TO THE DEPARTMENT...