Table 3. Correlation between observed and predicted phenotypic values determined with cross validation for different traits in the maize, wheat, and sugar beet data sets.

Trait-Environment		Heteroscedastic Marker Variances
	BLUP	RMLV	RRWA $\left(h_{\text{p}}^2\right)$	RMLA	BL	HEM
Crossa et al. (2010), 284 maize lines (264 lines, GY)
MFL-WW	0.36	0.28	0.35 (0.8)	0.38	0.36	0.35
MFL-SS	0.45	0.28	0.38 (0.8)	0.39	0.45	0.44
FFL-WW	0.31	0.27	0.32 (0.8)	0.31	0.31	0.32
FFL-SS	0.51	0.35	0.46 (0.8)	0.47	0.48	0.50
ASI-WW	0.51	0.35	0.50 (0.8)	0.52	0.51	0.47
ASI-SS	0.51	0.35	0.44 (0.8)	0.46	0.50	0.45
GY-WW	0.54	0.36	0.46 (0.9)	0.50	0.54	0.52
GY-SS	0.43	0.19	0.34 (0.9)	0.37	0.43	0.35
Pérez-Rodríguez et al. (2012), 306 wheat lines
GY-average	0.65	0.54	0.66 (0.8)	0.66	0.63	0.63
DTH-average	0.59	0.41	0.57 (0.9)	0.60	0.58	0.55
Hofheinz et al. (2012), 310 sugar beet lines
SC	0.83	0.78	0.80 (0.9)	0.80	0.83	0.82
ML	0.85	0.82	0.84 (0.4)	0.86	0.86	0.85

For the RRWA approach, the preliminary heritability estimates $h_{\text{p}}^2$ are given in brackets. BLUP, best linear unbiased prediction; RMLV, modification of the restricted maximum likelihood procedure that yields heteroscedastic variances; RRWA, ridge regression with weighing factors according to analysis of variance components; RMLA, estimation of the error and genetic variance components with restricted maximum likelihood and partitioning according to analysis of variance components; BL, Bayesian LASSO; HEM, heteroscedastic effects model; GY, grain yield; MFL, male flowering; WW, well-watered; SS, severe drought stress; FFL, female flowering; ASI, anthesis-silking interval; DTH, days to heading; SC, sugar content; ML, molasses loss.