Accuracy and simultaneous selection gains for N-stress tolerance and N-use efficiency in maize tropical lines

• Published in: Scientia agricola Vol. 74; no. 6; pp. 481 - 488
• Main Authors: Mendonça, Leandro de Freitas, Granato, Ítalo Stefanine Correia, Alves, Filipe Couto, Morais, Pedro Patric Pinho, Vidotti, Miriam Suzane, Fritsche-Neto, Roberto
• Format: Journal Article
• Language: English
• Published: São Paulo - Escola Superior de Agricultura "Luiz de Queiroz" 01.12.2017; Universidade de São Paulo
• Subjects: abiotic stress; AGRICULTURE, MULTIDISCIPLINARY; correlation between traits; mixed models; winter maize; correlation between traits; winter maize; mixed models; abiotic stress
• ISSN: 0103-9016; 1678-992X; 0103-9016
• DOI: 10.1590/1678-992x-2016-0313

ABSTRACT Maize plants can be N-use efficient or N-stress tolerant. The first have high yields in favorable environments but is drastically affected under stress conditions; whereas the second show satisfactory yields in stressful environments but only moderate ones under optimal conditions. In this context, our aim was to assess the possibility of selecting tropical maize lines that are simultaneously N-stress tolerant and N-use efficient and check for differences between simultaneous selection statistical methods. Sixty-four tropical maize lines were evaluated for Nitrogen Agronomic Efficiency (NAE) and Low Nitrogen Tolerance (LNTI) response indices and two per se selection indices, Low Nitrogen Agronomic Efficiency (LNAE) and Harmonic Mean of Relative Performance (HMRP). We performed eight selection scenarios: LNAE; HMRP; Additive index; Mulamba-Mock index; and Independent culling levels. The last three was predicted by REML/BLUP single-trait and multi-trait using genotypic values of NAE and LNTI. The REML/BLUP multi-trait analysis was superior to the single-trait analysis due to high unfavorable correlation between NAE and LNTI. However, the accuracy and genotypic determination coefficient of NAE and LNTI were too low. Thus, neither single- nor multi-trait analysis achieved a good result for simultaneous selection nor N-use efficiency nor N-stress tolerance. LNAE obtained satisfactorily accurate values and genotypic determination coefficient, but its performance in selection gain was worse than HMRP, particularly in terms of N-use efficiency. Therefore, because of the superior performance in accuracy, genotypic determination coefficient and selection, HMRP was considered the best simultaneous selection methodology of the scenarios tested for N-use efficiency and N-stress tolerance.