Skip to main content

Table 8 The estimated residual (\(\sigma _{e}^{2}\)), trial (\(\sigma _{t}^{2}\)), replicates within trials (\(\sigma _{r}^{2}\)) and blocks within replicates within trials (\(\sigma _{b}^{2}\)) variance components, genetic variance (\(\sigma _{s}^{2}\)), heritability computed by Method M5 (H2.M5) and predictive accuracy computed by Methods M5 (PA.M5) and M7 (PA.M7), using the classical (CLS) and robust (ROB) approaches, for the maize dataset

From: Robust estimation of heritability and predictive accuracy in plant breeding: evaluation using simulation and empirical data

Stage

Parm †

CLS r

ROB r

CLS qc

ROB qc

CLS r*

ROB r*

CLS qc*

ROB qc*

1

\(\sigma _{e}^{2}\)

352.4962

56.0533

47.8927

49.1810

46.5067

48.6339

45.0147

47.9068

 

\(\sigma _{t}^{2}\)

7.5289

9.7426

6.3445

10.5256

9.1892

12.0945

8.7991

12.2285

 

\(\sigma _{r}^{2}\)

3.4126

3.7156

3.4067

2.9704

3.2630

3.1115

3.6797

3.0902

 

\(\sigma _{b}^{2}\)

0.0000

4.8448

7.9319

6.6856

8.0854

6.5774

7.7984

6.4579

 

\(\sigma _{s}^{2}\)

0.0008

0.0067

0.0074

0.0058

0.0096

0.0071

0.0085

0.0062

3

H2.M5

0.2573

0.8020

0.8328

0.8013

0.8634

0.8231

0.8504

0.8099

 

H2.OK

0.0219

0.2732

0.3139

0.2713

0.3671

0.3035

0.3420

0.2824

 

PA.M5

0.5073

0.8960

0.9126

0.8951

0.9292

0.9077

0.9222

0.8999

 

PA.M7

0.4705

0.8184

0.8338

0.8178

0.8486

0.8287

0.8426

0.8221

  1. † Parm=Parameter; r refers to the raw dataset, i.e, the original dataset before quality control;
  2. qc refers to dataset after quality control; r* refers to the cleaned raw dataset;
  3. qc* refers to the cleaned ’after quality control’ dataset