Brief description of results on genomic selection of CIMMYT maize in Africa (Yoseph Beyene et al.) Several populations each with 200 F2 x tester individuals genotyped with 200 SNPs and evaluated in DROUGHT and OPTIMUM environments. C0 200 F2 x tester individuals genotyped with 200 SNPs and evaluated in DROUGHT and OPTIMUM environments. Phenotypic selection of the best plants to form C1 Cross the C1, genotyped them, and make predictions to form the C2 plants. Cross the C2, genotyped them, and make predictions to form the C3 plants. Evaluated all cycles of selection together with checks, parents, pedigree, etc. in a field experiment.

Table 1. Mean grain yield performance (Mg ha-1) at cycle 0 (C0), cycle 1 (C1), cycle 2 (C2), cycle 3 (C3), F1, pedigree-selected lines and commercial checks for the eight bi-parental populations evaluated in four managed drought stress environments, and average grain yield gains per selection cycle including and excluding C0. Entries Across eight populations Three DTMA populations Five WEMA populations GY (Mg ha-1) AD (days) PH (cm) C0 2.286 63.910 179.900 2.410 63.670 189.160 2.212 64.060 174.300 C1 2.420 64.080 181.900 2.319 63.940 188.320 2.482 64.160 178.100 C2 2.438 64.410 184.600 2.378 64.830 193.340 2.474 179.400 C3 2.593 64.100 182.200 2.613 64.720 193.220 2.581 63.730 175.600 Pedigree 2.417 64.400 181.700 - F1 2.394 63.930 Parents 2.361 64.000 178.400 2.186 64.130 183.030 2.431 63.950 176.600 Checks 2.180 64.310 180.100 2.191 181.030 2.169 64.700 179.300 LSD0.05 0.219 0.700 5.260 0.342 1.130 6.900 0.257 0.850 6.320 Average gain per cycle (C0, C1, C2, C3) 0.094 0.090 0.960 0.069 0.404 1.720 0.110 -0.105 0.520 Average gain per cycle (C1, C2, C3) 0.086 0.010 0.150 0.151 0.390 2.450 0.050 -0.215 -1.250

Table 1. Mean performance for grain yield (Mg ha-1), anthesis date (days), and plant height (cm) at cycle 0 (C0), cycle 1 (C1), cycle 2 (C2), cycle 3 (C3), hybrid (F1), and pedigree-selected lines for all eight bi-parental populations evaluated in four managed drought stress environments, and average grain yield, anthesis days and plant height per selection cycle including and excluding C0.   CML440/CML504 CML441 /CML444 CML444/Malawi 6x1008 6x1016 6x1017 6x1020 6x1028 ------------------------------------------------------Grain yield (Mg/ha)------------------------------------------------- Cycle0 2.171 2.517 2.540 2.283 2.261 2.532 2.141 1.844 Cycle1 2.592 2.237 2.129 2.697 2.583 2.384 2.317 2.429 Cycle2 2.185 2.333 2.562 2.471 2.573 2.965 2.263 2.098 Cycle3 2.622 2.387 2.883 2.754 2.918 2.375 2.734 2.127 F1 - 2.28 2.957 2.120 2.436 2.177 Parents 2.133 2.248 2.181 2.635 2.464 2.446 2.4 2.214 Pedigree 2.47 2.423 2.456 2.327 2.413 LSD0.05 0.832 0.831 0.536 0.631 0.516 0.392 0.498 0.430 Average gain per cycle (C0, C1, C2, C3) 0.095 -0.040 0.146 0.119 0.192 0.011 0.173 0.052 Average gain per cycle (C1, C2, C3) 0.015 0.048 0.377 0.029 0.168 -0.005 0.208 -0.150

About 2.5 cycles of genomic selection in 1 year with an average increase of 0.069 Mg ha-1 per cycle under DROUGHT. This means about 0.173 Mg ha-1 of gains per year.

About 2.5 cycles of genomic selection in 1 year with an average increase of 0.0495 Mg ha-1 per cycle under DROUGHT. This means about 0.124 Mg ha-1 of gains per year.

RESULT ON RAPID CYCLING SELECTION BASED ON MARKERS ON OTHER BIPARENTAL POPULATIONS

Across 10 biparental populations – OPTIMUM ENVIRONMENTS Grain Yield 5.0 5.5 6.0 6.5 7.0 7.5 Group Cycle3 Cycle2 Cycle1 Cycle0 F1 Parents Pedigree Checks

Three cycles of marker only selection in 1.5 years Across 10 biparental populations – OPTIMUM ENVIRONMENTS Gran yield (ton. ha-1)= 6.6577 +0.1837Cycle Grain Yield (ton ha-1) 6.0 6.5 7.0 7.5 Cycle Cycle0 Cycle1 Cycle2 Cycle3 Three cycles of marker only selection in 1.5 years Overall genetic gain is 1.7- 3.5 times higher than from conventional breeding reported in sub-Saharan Africa.

Across 10 biparental populations – DROUGHT ENVIRONMENTS Grain yieldY 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Group Cycle3 Cycle2 Cycle1 Cycle0 F1 Parents Pedigree Checks

Three cycles of marker only selection in 1.5 years Across 10 biparental populations – DROUGHT ENVIRONMENTS Grain Yield (ton. ha-1) = 2.5025 +0.0455Cycle Grain Yield (ton. ha-1) 2.0 2.5 3.0 Cycle Cycle0 Cycle1 Cycle2 Cycle3 Three cycles of marker only selection in 1.5 years Overall genetic gain is 1.7- 3.5 times higher than from conventional breeding reported in sub-Saharan Africa.

Performance of hybrids developed through marker selection vs conventional breeding   Pop 5 - Managed Drought  GY (t/ha) % P % C C3S2 2.8 24.0 40.4 2.7 18.9 36.3 2.6 18.7 36.2 18.5 36.0 16.4 34.4 Mean Pedigree 2.2 Mean of checks 1.7 LSD 0.7 CV 17.8 Heritability 0.4 No of Locations 1 Pop 5 - Optimum 7.0 15.2 31.4 6.9 14.0 30.4 13.0 29.6 6.6 9.8 27.0 9.6 26.9 6.0 4.8 1.0 8.1 0.6 4   Pop 7 - Managed drought  GY (t/ha) % P % C C3S2 4.1 30.6 43.7 3.7 22.9 37.5 3.5 19.2 34.4 17.9 33.4 17.6 33.2 Mean Pedigree 2.8 Mean of checks 2.3 LSD 1.0 CV 17.3 Heritability 0.2 No of Locations 2 Pop 7- Optimum % P 6.4 8.9 12.1 8.8 12.0 6.3 7.2 10.5 6.2 6.1 9.4 5.9 9.2 5.8 5.6 0.9 7.6 0.6 4

HYBRIDS DERIVED FROM MARKER TESTED IN DROUGHT ENVIRONMENTS SELECTION TESTED IN DROUGHT ENVIRONMENTS

Means of Hybrids, from C3, Parents, F6, Cross, and Check from two biparental populations evaluated in 3-4 DROUGHT and 3-4 OPTIMUM populations MANAGEMENT Group BLUE GY BLUE GY Drought C3 4.008 Optimal 7.997 3.891 7.909 3.801 7.713 3.722 7.669 3.622 7.608 3.556 7.576 3.551 7.550 3.545 3.541 7.541 3.502 7.468 Parent 3.255 7.123 2.466 5.699 Cross 2.688 6.542 F6 3.525 7.325 2.761 7.219 2.781 6.505 2.645 6.885 2.022 6.566 Check 3.567 5.685 2.852 6.359 3.162 6.700 1.656 4.033 2.124 6.016 MANAGEMENT Group BLUE GY BLUE GY Drought C3 3.196 Optimum 7.089 3.119 7.059 2.614 7.037 2.490 6.932 2.415 6.867 2.375 6.840 2.371 6.825 2.365 6.797 2.359 6.795 2.353 6.767 Parent 2.267 6.383 1.963 6.183 Cross 2.307 5.913 F6 2.057 6.187 2.049 7.215 2.730 6.072 2.206 6.415 1.930 6.686 Check 1.871 5.038 2.291 5.827 1.246 6.533 1.376 2.929 2.036 5.347

Grain Yield Mg ha-1 Entry

Grain Yield Mg ha-1 Entry

Fig. 1i Grain Yield Mg ha-1 Entry

Fig. 1j Grain Yield Mg ha-1 Entry

HYBRIDS DERIVED FROM MARKER TESTED IN OPTIMUM ENVIRONMENTS SELECTION TESTED IN OPTIMUM ENVIRONMENTS

CONCLUSIONS RAPID CYCLING OF GENOMIC SELECTION IN MAIZE BIPARENTAL BREEDING POPULATIONS WORK!

http://genomics.cimmyt.org Thanks!! j.crossa@cgiar.org