planning rice breeding programs for impact multi-environment trials: design and analysis
TRANSCRIPT
Planning rice breeding programs for impact
Multi-environment trials:
design and analysis
IRRI: Planning breeding Programs for Impact
Introduction:Problem of individual trials?
Multi-environment trials (METs) used to predict performance in farmers fields
Its predictive power = low
SO
IRRI: Planning breeding Programs for Impact
IntroductionProblem of METs?
Must be planned carefully to ensure they are predictive and efficient
very expensive and require much coordination and time
SO
IRRI: Planning breeding Programs for Impact
Learning objectives
• To clarify the purpose of variety trials
• To introduce linear models for multi-environment trials (MET’s)
• To describe the structure of the analysis of variance for MET’s
• To model the variance of a cultivar mean estimated from a MET
• To examine the effect of replication within and across sites and years on measures of precision
IRRI: Planning breeding Programs for Impact
To predict performance:
• Off-station
• In the future
WS 2002 WS 2003 +
Purpose of MET’s
IRRI: Planning breeding Programs for Impact
0 Yield (t/ha) 6Single trial
0 Yield (t/ha) 6Mean of 3 trials
MET’s reduce SEM for cultivars
IRRI: Planning breeding Programs for Impact
Simplest MET model considers trials “environments”
Yijkl = M + Ei + R(E)j(i) + Gk + GEik + eijkl [7.1]
The genotype x environment model
Where:• M = mean of all plots
• Ei = effect of trial i
• R(E)j(i) = effect of rep j in trial I• Gk = effect of genotype k
• GEik = interation of genotype k and trial i
• eijkl = plot residual
IRRI: Planning breeding Programs for Impact
The genotype x environment model
Trials and reps are random factorsThey sample the TPE
We do not select varieties for specific trials or reps
Genotypes are fixed factorsWe are interested in the performance of the
specific lines in the trial
IRRI: Planning breeding Programs for Impact
The genotype x environment model
The GE interaction is a random factor
Interactions of fixed and random factors are always random
Random interactions with genotypes are part of the error variance for genotype means
IRRI: Planning breeding Programs for Impact
Single trial: Yijk = μ + Rj + Gi + ek(j)
GE model: Yijkl = M + Ei + R(E)j(i) + Gk + GEik + eijkl
Relationship between GE model and single-trial model:
IRRI: Planning breeding Programs for Impact
ANOVA for GLY model
Source Mean Square EMS
Environments (E)
Replicates within E
Genotypes MSG σ2e + rσ2
GE + reσ2G
G x E MSGE σ2e + rσ2
GE
Error (Plot Residuals)
MSe σ2e
IRRI: Planning breeding Programs for Impact
Variance of a cultivar mean
Where: • e = number of trials• r = number of reps per trial
σ2Y = σ2
GE/e + σ2e/re [7.2]
IRRI: Planning breeding Programs for Impact
Estimating σ²G, σ²GE and σ²e
σ2e = MSerror
σ2GE = (MSGE – Mserror)/r
σ2G = (MSG – MSGE)/re
IRRI: Planning breeding Programs for Impact
σ2e = .45 (t/ha)2
σ2GE = 0.30 (t/ha)2
Hypothetical values:
σ2Y = σ2
GE/e + σ2e/re [7.2]
Example: modeling the LSD for a MET program using GE model
Example: modeling the LSD for a MET program using GE model
Number of sites Nr of reps/site SEM t/ha LSD
1 1 .87 2.61
2 .72 2.16
4 .64 1.92
2 1 .61 1.83
2 .51 1.53
4 .45 1.35
5 1 .39 1.08
2 .32 0.96
4 .29 0.87
10 1 .27 0.810.69 2 .23
4 .20 0.60
Table 1. The effect of trial and replicate number on the standard deviation of a cultivar mean: genotype x environment model
IRRI: Planning breeding Programs for Impact
The “real” SEM (with GE component estimated separately) for a single trial is:
SEM = (σ2GE/e + σ2
e/re)0.5
= ((0.3/1) + (0.45/4)) 0.5
= 0.64 t/ha
The “apparent” SEM (with GE and G components confounded) for a single trial is:
SEM = (σ2e/r)0.5
= (0.45/4) 0.5
= 0.35
IRRI: Planning breeding Programs for Impact
Yijklm = M + Yi + Sj + YSij + R(YS)k(ij)+ Gl + GYil + GSjl + GYSijl + eijklm
Yijkl = M + Ei + R(E)j(i) + Gk + GEik + eijkl
σ2Y = σ2
GY/y + σ2GS/s + σ2
GYS/ys + σ2e/rys
The genotype x site x year model
A more realistic MET model subdivides the “environment” factor into “years” and “sites”:
SourceMean
squareEMS
Years (Y)
Sites (S)
Y x S
Replicates within Y x S
Genotypes (G) MSG σ2e + rσ2
GYS + rsσ2GY+ ryσ2
GS+ rysσ2G
G x S MSGS σ2e + rσ2
GYS + ryσ2GS
G x Y MSGY σ2e + rσ2
GYS + rsσ2GY
G x Y x S MSGYS σ2e + rσ2
GYS
Plot residuals MSe σ2e
ANOVA for GSY model
IRRI: Planning breeding Programs for Impact
Estimating σ2GY , σ
2GS , σ
2GY S, and σ2
e
σ2e = MSerror
σ2GYS = (MSGYS – MSerror)/r
σ2GY = (MSGY – MSGYS)/rs
σ2GS = (MSGS – MSGYS)/ry
σ2G = (2MSG - MSGS – MSGY)/2rsy
IRRI: Planning breeding Programs for Impact
Example: Modeling the LSD for a MET program using the GSY model
For NE Thailand OYT:
σ2e = 0.440 (t/ha)2
σ2GS = 0.003 (t/ha)2
σ2GY = 0.049 (t/ha)2
σ2GYS = 0.259 (t/ha)2
(Cooper et al., 1999)
IRRI: Planning breeding Programs for Impact
Number of sites
Number of years
Number of replicates/site LSD (t ha-1)
1 1 1 2.45
2 2.06
4 1.85
2 1 1.79
2 1.52
4 1.37
5 1 1 1.10
2 0.93
4 0.83
2 1 0.81
2 0.69
4 0.62
Example: Modeling the LSD for a MET program using the GSY model
IRRI: Planning breeding Programs for Impact
Conclusions from error modeling exercise?
• σ2GS was very small in this case
little evidence of specific adaptation to sites
• σ2GSY was very large in this case
much random variation in cultivar performance from site to site and year to year
• σ2e very large, methods to reduce plot error are needed
• σ2GYS was very large compared to σ2
GY and σ2GS
sites and years are equivalent for testing
IRRI: Planning breeding Programs for Impact
Deciding whether to divide a TPE
• If TPE = large and diverse, it may be worthwhile to divide it into sets of more homogeneous sites
• If no pre-existing hypothesis about how to group
environments, use cluster, AMMI, or pattern analysis
• If there is a hypothesis that can be formed based on geography, soil type, management system, etc, group trials according to this fixed factor
IRRI: Planning breeding Programs for Impact
Environments can be grouped into subregions:
Yijklm = M + Si + Ej(Si) + R(E(S))k(ij)+ Gl + GSil + GE(S)lij + eijklm
Yijkl = M + Ei + R(E)j(i) + Gk + GEik + eijkl
• Subregions are fixed
• Trials within subregions are random
• If GS interaction term is not significant, subdivision is unnecessary, and could be harmful
The genotype x subregion model
IRRI: Planning breeding Programs for Impact
SourceMean
squareEMS
Subregions (S)
Locations within subregions (L(S))
Replicates within L(S)
Genotypes (G) MSG σ2e + rσ2
GL(S) + rlσ2GS+ rlsσ2
G
G x S MSGS σ2e + rσ2
GL(S) + rlσ2GS
G x L(S) MSGL(S) σ2e + rσ2
GL(S)
Plot residuals MSe σ2e
Expected mean squares for ANOVA of the genotype x subregion model for testing fixed groupings of sites
IRRI: Planning breeding Programs for Impact
Example: Are central and southern Laos separate breeding targets?
Should breeders and agronomists in Laos consider central and southern regions as separate TPE for RL rice?
22 traditional varieties tested in 4-rep trials at
3 sites in central region, 3 in south in WS 2004
Source df MS F
Subregions (S) 1 5459785
Locations within subregions (L(S))
4 17284169
Replicates within L(S)
18 292059
Genotypes (G) 21 3644949 4.77**
G x S 21 764412 0.76
G x L(S) 84 1006974 6.58**
Plot residuals 378 153101
ANOVA testing hypothesis: central & southern regions of Laos = separate RL breeding targets
22 TVs tested in WS 2004
IRRI: Planning breeding Programs for Impact
Are central and southern Laos separate breeding targets?
Genotype x subregion interaction is not significant when tested against variation among locations within subregions
Subdivision is therefore not needed
Subdivision might even be harmful, because it would reduce replication within each subregion
IRRI: Planning breeding Programs for Impact
Can anyone briefly clarify the purpose of variety trials?
When should you divide a TPE?
IRRI: Planning breeding Programs for Impact
Summary 1
• Purpose of a variety trial is to predict future performance in the TPE
• Random GEI interaction is large, and reduces precision with which cultivar means can be estimated
• Variance component estimates for the GLY model can be used to study resource allocation in testing programs
• Within homogeneous TPE, the GSY variance usually the largest. If so, strategies that emphasize testing over several sites or several years likely equally successful
IRRI: Planning breeding Programs for Impact
Summary 2
• Little benefit from including more than 3 replicates (and often more than 2) in a MET
• Standard errors and LSD’s estimated from single sites are unrealistically low because they do not take into account random GEI
• Fixed-subregion hypotheses allow a hypothesis about the existence of genotype x subregion interaction to be tested against genotype x trial within subregion interaction