Table 10. Estimated prediction accuracy (correlation between predicted and adjusted grain yield, averaged over 50 TRN-TST partitions, cycle 2011-2012 (W2), CV2.

Models/Environmentsa		Correlation (SE)		Change%b	Numberc
Single environment
0iFN		0.473(0.052)		−	−
5iBH		0.414(0.063)		−	−
5iBN		0.510(0.052)		−	−
5iFN		0.448(0.054)		−	−
Pairs of environments		Interaction Model	Across-Env Model
0iFN	Cor = 0.34	0.512(0.042)	0.436(0.052)	8.3%; 17.5%	50; 47
5iBH		0.451(0.056)	0.386(0.054)	8.9%; 16.8%	50; 48
0iFN	Cor=-0.05	0.467(0.043)	0.271(0.060)	−1.2%; 72.2%	13; 50
5iBN		0.502(0.048)	0.279(0.056)	−1.6%; 80.0%	6; 50
0iFN	Cor = 0.31	0.520(0.043)	0.406(0.056)	9.8%; 27.9%	50; 49
5iFN		0.502(0.050)	0.412(0.060)	12.0%; 21.9%	50; 49
5iBH	Cor = 0.33	0.477(0.053)	0.426(0.061)	15.2%; 11.8%	48; 47
5iBN		0.546(0.047)	0.483(0.051)	7.2%; 13.2%	47; 49
5iBH	Cor = 0.41	0.500(0.049)	0.465(0.050)	20.9%; 7.7%	50; 45
5iFN		0.520(0.051)	0.490(0.055)	16.0%; 6.1%	50; 40
5iBN	Cor = 0.41	0.558(0.050)	0.541(0.052)	9.6%; 3.2%	49; 43
5iFN		0.501(0.052)	0.490(0.055)	11.8%; 2.2%	50; 37
All environments
0iFN		0.513(0.044)	0.298(0.051)	8.5%; 72.6%	45; 50
5iBH		0.536(0.050)	0.481(0.047)	29.6%; 11.5%	50; 49
5iBN		0.531(0.044)	0.401(0.044)	4.1%; 32.4%	37; 50
5iFN		0.561(0.046)	0.523(0.049)	25.3%; 7.4%	50; 43

TRN-TST, training-testing.

^aEnvironments are described by a sequence of codes: 0i, 2i, and 5i denote the number of irrigation cycles; B/F denotes whether the planting system was “bed” (B) or “flat” (F); N/H denotes whether planting date was normal (N) or late (H, simulating heat); Z indicates no tillage.

^bChange in prediction accuracy of the M×E model relative to the prediction accuracy of the single-environment (before semicolon) and relative to the prediction accuracy of the across-environment model (after semicolon).

^cNumber of partitions (of 50) for which the M×E model gave greater accuracy than the single-environment (before semicolon) and the across-environment model (after semicolon).