The strategy and potential utilization of temperate
germplasm for tropical germplasm improvement—a
case study of maize (Zea Mays L.)
W. Wen, T. Guo, V.H. Chavez T., J.
Yan, S. Taba
CIMMYT
Heterotic groups
A, B, C, D……
A
X
B
C
X
D
Heterosis exploitation
Organization of germplasm in genetically divergent heterotic groups is beneficial for a
systematic and optimum exploitation of heterosis (Melchinger 1999).
Well-developed heterotic patterns of mature breeding programs are artificial constructs created
by breeders and enhanced by the process of breeding hybrids (Tracy and Chandler 2006)
• CML: CIMMYT Maize Lines ,Tropical
• GEM: US Germplasm Enhancement of Maize
(GEM) lines, Temperate
• The CMLs originated from 35 mostly broad-based
populations and pools with mixed origins, which
are carefully selected inbred lines with good
general combining ability (GCA) and a significant
number of value-adding traits such as drought
tolerance, nitrogen (N) use efficiency, acid soil
tolerance, resistance to disease and insect pests
(www.cimmyt.org).
• To date, the total number of CMLs held in trust in
CIMMYT genebank is 530.
• The Germplasm Enhancement of Maize project
(GEM) is a cooperative effort of the United States
Department of Agriculture’s Agricultural Research
Service (USDA-ARS), land-grant universities,
private industry, international, and nongovernmental organizations.
• The objective is widening the germplasm base of
commercial hybrid corn in the United States
through introduction and incorporation of novel
and useful germplasm gathered from around the
globe (http://www.public.iastate.edu/~usdagem/)
Features of CMLs
• The lack of structure within CMMYT lines
and lower genetic distance between
current heterotic groups due to the
mixed genetic constitution of CIMMYT
germplasm used to extract the inbred
lines.
Why use temperate germplasm?
 Some favorable alleles may be unique to temperate
germplasm, for instance, the allele of the gene encoding βcarotene hydroxylase (crtRB1), associated with β-carotene
concentration and conversion in maize kernels, which was
detected most recently (Yan et al. 2010)
 Some GEM lines performed well in tropical and subtropical
areas.
 There should be favorably unique alleles or genomic regions
in temperate germplasm that can be helpful in tropical maize
improvement as well as utilization of tropical lines in
temperate areas.
Traditional MAS
Including two steps: identifying QTL and then
estimating their effects.
Weakness of QTL identification methods:
 Biparental populations, not representative, do not have the same level of allelic diversity and
phase as the breeding program as a whole
 Costly
 Additional effort is needed to validate the mapping results
 Biased estimation
Genomic selection
Genomic selection which involves calculating breeding values
for all markers and utilizes all markers in selection without
testing them for their significance may dramatically change
the role of phenotyping especially with the rapid
development of high through-put genotyping platform
(Jannink et al. 2010)
Heffner et al. 2009
Objectives
 to detect the potential of utilizing temperate
germplasm in tropical maize breeding
 to construct a simple model using information of
phenotypic values and molecular markers to predict
single crosses (hybrid) performance
 to compare the efficiency of hybrid performance
prediction by using random markers and selective
markers with association analysis
Material Summary
CML
GEM
Total
F1
Heterotic
Pattern
A
B
A/B
SS
NSS
Number
Description
48
38
8
35
19
148
654
Dent kernel
Flint kernel
Dent/Flint kernel
Stiff stalk
Non-stiff stalk
F1: CML (A) X GEM (SS), CML (B) X GEM (NSS)
CML
F1
GEM
Phenotyping
F1/TL
F1/AF
Range
Inbreds/TL
Mean
Range
Inbreds/AF
Trait
Range
Mean
Mean
Range
DA (day)
76-103
92.11 ±4.66 60-82
71.44 ±4.04 80-103
92.01 ±4.13 61.67-82
71.31 ±3.71
DS (day)
76-106
93.09 ±4.82 61-82
71.01 ±3.74 80-106
92.98 ±4.30 61-80
70.92 ±3.46
PH (cm)
125-275
209.12
±21.48
140-285
228.38
±18.25
155-250
209.91
±17.30
228.29
±14.21
EH (cm)
50-170
112.1 ±18.68 77.5-175
122.76
±17.82
70-150
112.6 ±15.43 85-165
122.33
±14.36
EW (kg)
0.12-2.23
0.87 ±0.37
0.1-2.78
0.85 ±0.46
0.13-1.77
0.92 ±0.34
0.21-2.78
0.95 ±0.49
EWP (kg)
0.04-0.84
0.27 ±0.09
0.13-1.42
0.43 ±0.14
0.13-0.84
0.28 ±0.08
0.21-0.83
0.43 ±0.09
183.33-261.25
Mean
SNPs used for genotyping
Chromosome
1
2
3
4
5
6
7
8
9
10
Unknown
Total
No. of SNPs
217
132
140
134
154
98
77
115
96
83
20
1,266
The frequency of the same specific allele of each SNP in
GEM lines, CMLs and all 148 lines.
Allelic Frequency
0.9
GEM
0.8
CML
0.7
All
0.6
0.5
0.4
0.3
0.2
0.1
0
1
101
201
301
401
501
601
701
801
901 1001 1101 1201
SNP ID
PC2(4.5%)
16
CML
14
12
GEM_NSS
10
GEM_SS
8
6
4
2
0
-2
0
5
10
15
20
25
PC1 (13.5%)
Principal component analysis (PCA) of 148 maize inbred lines based on 1266 SNPs
PC2 (13.5%)
0.100
0.050
0.000
F1 Group 1 NSS
-0.050
F1 Group 2 SS
-0.100
-0.150
-0.250
-0.200
-0.150
-0.100
-0.050
0.000
0.050
0.100
0.150
PC1 (52.9%)
Principal component analysis (PCA) of 654 maize F1 lines based on 872 SNPs
Prediction of hybrid performance
Training sample set: n=430
Testing sample set: n=215
Prediction of hybrid performance
Plant Height
Days to Anthesis
0.8
0.9
0.7
0.8
0.7
0.6
0.6
0.5
A
R
R
0.5
0.4
A
0.4
B
0.3
C
0.2
B
0.3
C
0.2
0.1
0.1
0
0
1266
1184
756
500
250
100
Number of markers
50
25
1266
1184
756
500
250
100
50
25
Number of markers
Marker type A: marker selection referring to association results from inbreds, and starting from
the marker with highest significance.
Marker type B: marker selection referring to association results from F1s, and starting from the
marker with highest significance.
Marker type C: markers selected randomly from the genome-covering 1,266 SNPs
Implication and discussion
• Utilizing temperate lines in tropical breeding
program
In this study, unique alleles harbored within the temperate lines imply the
utilization of temperate lines in tropical breeding program and the
potential of mutual improvement
Clear heterotic patterns of the GEM lines can broaden the genetic divergence
between heterotic groups within CMLs
• The potential and strategy of genomic
selection in maize breeding
The high prediction accuracy indicates the potential and efficiency of genomic
selection based on our model, for predicting hybrid performance.
Replacing a larger number of genome-covering markers with a smaller
number of trait-associated markers may be more effective because the
latter produces less noise
PPB in Oaxaca Project
• Trails of landrace collections, hybrids, CIMMYT
gene pools
• Seed production of hybrids and inter-varietal
hybrids in Central Valley of Oaxaca
• Seed processing and Declared seeds for
distribution
• Low cost seed bags: 500-600 pesos for 20kg.
• Good yield and plant performance of some gene
pools: potential useful gene flow into the local
maize production communities
PPB in Valle Central, Oaxaca
End.
• Thank you very much.