Attainable yield and soil texture as drivers of maize response to nitrogen: A synthesis analysis for Argentina

Autor: Cristian Alvarez, Josefina Molino, Juan Manuel Orcellet, P. A. Calviño, Ma. Mercedes Zubillaga, Ma. Guadalupe Tellería, Agustín Pagani, Gustavo Nestor Ferraris, Julia Ester Capurro, Nahuel Ignacio Reussi Calvo, Miguel Boxler, Pablo Barbieri, R.J.M. Melchiori, Ariel Angeli, Luis Alberto Ventimiglia, Fernando O. García, Gabriel Espósito, Hernán Sainz-Rozas, Carolina Alvarez, Juan Manuel Pautasso, Matías Saks, Martín Díaz-Zorita, Santiago Díaz-Valdéz, Fernando Salvagiotti, Vicente Jorge Gudelj, Héctor Carta, Juan Pablo Ioele, Matías Redel, Manuel Ferrari, Mirian Barraco, Sergio Nestor Rillo, Ignacio A. Ciampitti, Hernán Eduardo Echeverría, Angel Berardo, Jose L. Zorzín, Flavio H. Gutiérrez-Boem, Helena Rimski-Korsakov, Sebastian Gambaudo, Adrian A. Correndo, Octavio Pedro Caviglia
Rok vydání: 2021
Předmět:
Zdroj: Field Crops Research. 273:108299
ISSN: 0378-4290
DOI: 10.1016/j.fcr.2021.108299
Popis: The most widely used approach for prescribing fertilizer nitrogen (N) recommendations in maize (Zea Mays L.) in Argentina is based on the relationship between grain yield and the available N (kg N ha−1), calculated as the sum of pre-plant soil NO3--N at 0−60 cm depth (PPNT) plus fertilizer N (Nf). However, combining covariates related to crop N demand and soil N supply at a large national scale remains unexplored for this model. The aim of this work was to identify yield response patterns associated to yield environment (crop N demand driver) and soil texture (soil N supply driver). A database of 788 experiments (1980−2016) was gathered and analyzed combining quadratic-plateau regression models with bootstrapping to address expected values and variability on response parameters and derived quantities. The database was divided into three groups according to soil texture (fine, medium and coarse) and five groups based on the empirical distribution of maximum observed yields (from Very-Low = 13.1 Mg ha−1) resulting in fifteen groups. The best model included both, attainable yield environment and soil texture. The yield environment mainly modified the agronomic optimum available N (AONav), with an expected increase rate of ca. 21.4 kg N Mg attainable yield−1, regardless of the soil texture. In Very-Low yield environments, AONav was characterized by a high level of uncertainty, related to a poor fit of the N response model. To a lesser extent, soil texture modified the response curvature but not the AONav, mainly by modifying the response rate to N (Fine > Medium > Coarse), and the N use efficiencies. Considering hypothetical PPNT levels from 40 to 120 kg N ha−1, the expected agronomic efficiency (AENf) at the AONav varied from 7 to 31, and 9–29 kg yield response kg fertilizer N (Nf)−1, for Low and Very-High yield environments, respectively. Similarly, the expected partial factor productivity (PFPNf) at the AONav ranged from 62 to 158, and 55–99 kg yield kg Nf−1, for the same yield environments. These results highlight the importance of combining attainable yield environment and soil texture metadata for refining N fertilizer recommendations. Acknowledging the still low N fertilizer use in Argentina, space exists to safely increasing N fertilizer rates, steering the historical soil N mining profile to a more sustainable agro-environmental scenario in the Pampas.
Databáze: OpenAIRE
Externí odkaz: