Maize breeding effects on grain yield genetic progress and its contribution to global yield gain in Argentina

Abstract

Context or problem: Maize production in Argentina has increased in recent years, following the global gain (GG) in grain yield (GY). The GG in GY depends on genetic progress (GP), which requires frequent quantification to detect potential plateaus and variations in its contribution to the GG. In this sense, hybrid adoption in Argentina shifted from double- and three-way to single cross (F1) hybrids in the 1990s, thereby increasing the level of heterosis of released hybrids. Since heterosis increase may have had a greater impact on estimates of maize GG than those based on a single cross type, GP based exclusively on F1 hybrids and its contribution to GG could be lower than that including different types of crosses. Objective or research question: The main objectives of this work were to (i) analyze grain yield GP, (ii) dissect grain yield GP into the corresponding trends in its physiological determinants and numerical components, and (iii) estimate the relative contribution of GP to GG in the main maize producing region of Argentina for the period marked by significant changes in heterosis level (i.e. from 1960s to 1990s) and the period of massive adoption of F1 hybrids (from 1990s onwards). Methods: We used a dataset obtained at the core of the mentioned region (INTA Pergamino; 33°56’S, 60°34’W) from era-decade experiments (ERA) including 24 hybrids released between 1965 and 2016 and grown with no abiotic or biotic restrictions. We quantified GY, its numeric components (KN: kernel number m−2, KW: individual kernel weight) and its physiological determinants (BT: total shoot biomass at maturity, HI: harvest index). Results: A GP of 0.84 % y−1 was computed for GY from 1965 to 1993, and of 0.51 % y−1 thereafter. We also detected genetic progress for KN (0.53 % y−1) and BT (0.26 % y−1) but not for KW. A GP of 0.66 % y−1 was computed for HI up to 1993, which plateaued thereafter. A 40 % contribution of GP to GG was estimated for the Pergamino site from 1965 to 1993, predominantly driven by changes in heterosis level. The contribution dropped to 32 % from 1993 onwards, when F1 hybrids were massively adopted by the farmers. This contribution was much smaller (e.g. 19 %) in areas less representative of the target population of environments of dominant breeding programs. Conclusions: Our findings underline that the relative contribution of GP to GY improvement at the regional level varies markedly depending upon the period included in the analysis and the environment used for evaluation. The decline observed at less representative locations of the main target population of environments may guide breeders in their decision for developing new programs.