Phenotypic plasticity of maize grain yield and related secondary traits: Differences between inbreds and hybrids in response to contrasting water and nitrogen regimes

Abstract

Correct characterization of heritability and phenotypic plasticity (PP) is critical for breeding purposes. The latter refers to the variation range of a trait in response to changes in the environment and has been assessed as the difference between percentiles 10th (P10) and 90th (P90) of each trait, which does not reflect below the median (P50) variations to the same extent as the above the median ones. This inconsistency may affect the classification of stable (low PP) or plastic (high PP) given to traits, as well as their relative ranking and PP relationship with heritability. The objectives of current research were to evaluate corrected PP (PPC) variation in grain yield (GY) and related secondary traits among contrasting maize genotypic groups (inbreds and hybrids) grown under contrasting water regimes (WR) and nitrogen (N) availabilities. The relationship between PPC and broad-sense heritability (H2) was also assessed. Field experiments were conducted during three (N) or seven (WR) growing seasons at two mid-latitude environments of Argentina. Measured traits were days and thermal time to 50% anthesis (A50 and TTA) and to 50% silking (S50 and TTS), the anthesis-silking interval in days (ASID) and in TT (ASITT), plant height (Ph), prolificacy (Pr), GY, kernel numbers (KN), and kernel weight (KW). Values for percentiles 10th (P10), 50th (P50) and 90th (P90) of each trait were identified for each treatment combination. P50 was set to 1, and values obtained for P10 and P90 were expressed as ratios with P50. P10 was corrected (P10C= P50 – P50/P10) to reflect the below P50 variations to the same extent as those above P50. Corrected PP was estimated as PPC= P90 - P10C. P90 values of all traits corresponded to non-stressed plots whereas P10C values corresponded to stressed plots, except for ASIs (opposite trend). A large plasticity (PPC>mean PPC) was usually verified for ASIs, GY and KN. Mean inbreds PPC had a larger variation than mean hybrids PPC (+19% for WR and +29% for N), except for GY under contrasting WR (hybrids > inbreds). A common trend across all evaluated traits was the markedly larger effect on PPC of P10C than of P90, in agreement with the predominant representation of stressful conditions by the former and in contrast to previous studies where no correction was performed on P10. Our results demonstrated the lack of relationship between H2 and PP and improved current knowledge about the importance of environment modulation on PP of most expansion-related and production traits, highlighting the relevance of the evaluated resource (water or N) as well as of the genotypic group (hybrids or inbreds) on final phenotype expression.