184-15 Genotype x Location and Genotype x Country Interaction for Maize Hybrid Yield in the Central American Lowland Tropics: Implications for Breeding and Testing Programs.

See more from this Division: C01 Crop Breeding & Genetics
See more from this Session: Tools for Evaluating and/or Enhancing Genetic Progress
Tuesday, November 2, 2010: 11:45 AM
Long Beach Convention Center, Room 101A, First Floor
Share |

Felix San Vicente1, Gary Atlin1, Biswanath Das1, Roman Gordon2, Hector Deras3 and Oscar Cruz4, (1)CIMMYT, Texcoco, Mexico
(2)IDIAP, Los Santos, Panama
(3)CENTA, San Andres, El Salvador
(4)DICTA, Tegucigalpa, Honduras
Maize is the staple in Guatemala, Honduras, El Salvador, and Nicaragua, with annual consumption averaging over 80 kg person-1.  Yields average less than 2 t ha-1 (although higher in El Salvador).  Increasing maize yield is critical to the food security and incomes of the poor.  National variety release agencies treat each country separately, and give little consideration to data from neighboring countries.   Because national research resources are limited,  release decisions are made based on low-precision results from only a few locations in each country. However, CIMMYT and large seed companies treat the region as a single target population of environments (TPE).  If genotype x country (GC) interaction is small, hybrid release and recommendation decisions in each country will be more accurate if made on the basis of trials conducted at many locations across the region rather than on data from a few trials.  To test this hypothesis, yield data from advanced CIMMYT-coordinated hybrid trials conducted in 2007-2009, were analyzed using a model that partitioned the genotype x location (GL) interaction into GC and genotype x location (country) (GL(C))components.  GL(C) was always much larger than the GC component, and most tests of genotype x country interaction were not significant.  Genetic correlations among hybrid means in different countries were close to 1.0.  These results indicate that greater gains from selection in each country will be achieved by treating the region as a single TPE.  The best new private-sector hybrids evaluated across 19 locations out-yielded the regional check H-59, released approximately 20 years ago, by 1.5 t ha-1 in 2009, while the highest-yielding CIMMYT three-way cross out-yielded H-59 by over 1 t ha-1.  Integrated breeding and testing programs for the region are developing broadly-adapted hybrids with the potential to double yields if seed and fertilizer are made accessible to smallholders.
See more from this Division: C01 Crop Breeding & Genetics
See more from this Session: Tools for Evaluating and/or Enhancing Genetic Progress