Improving node number simulation in soybean

Nereu Augusto Streck,Nilson Lemos De Menezes,Felipe Brendler Oliveira,Ana Paula Schwantes,Gizelli Moiano De Paula

doi:10.1590/s0100-204x2009000700002

Nereu Augusto Streck, Nilson Lemos De Menezes + Show 3 more

Open Access

https://doi.org/10.1590/s0100-204x2009000700002

Copy DOI

Abstract

The objective of this study was to improve the simulation of node number in soybean cultivars with determinate stem habits. A nonlinear model considering two approaches to input daily air temperature data (daily mean temperature and daily minimum/maximum air temperatures) was used. The node number on the main stem data of ten soybean cultivars was collected in a three-year field experiment (from 2004/2005 to 2006/2007) at Santa Maria, RS, Brazil. Node number was simulated using the Soydev model, which has a nonlinear temperature response function [f(T)]. The f(T) was calculated using two methods: using daily mean air temperature calculated as the arithmetic average among daily minimum and maximum air temperatures (Soydev tmean); and calculating an f(T) using minimum air temperature and other using maximum air temperature and then averaging the two f(T)s (Soydev tmm). Root mean square error (RMSE) and deviations (simulated minus observed) were used as statistics to evaluate the performance of the two versions of Soydev. Simulations of node number in soybean were better with the Soydev tmm version, with a 0.5 to 1.4 node RMSE. Node number can be simulated for several soybean cultivars using only one set of model coefficients, with a 0.8 to 2.4 node RMSE.

Highlights

The calculation of node appearance rate (NAR) is an important part of simulation models for soybean [Glycine max (L.) Merrill] growth, development and yield (Jones & Laing, 1978; Sinclair, 1986; Sinclair et al, 2005)
Maximum air temperature exceeded the optimum temperature for NAR (31oC) in most weeks during the node appearance phase, mainly in the 2004/2005 growing season, while minimum air temperature was always above minimum temperature for NAR (7.6oC) in all three growing seasons
A lack of association between the performance of a nodes on a stem (NN) simulation model with maturity group for the determinate soybean cultivars used in the present study is consistent with the results reported by Setiyono et al (2007) for semi‐determinate and indeterminate soybean cultivars and indicates that vegetative development is independent of reproductive development in soybean, even though the former overlaps the latter

Summary

Introduction

The calculation of node appearance rate (NAR) is an important part of simulation models for soybean [Glycine max (L.) Merrill] growth, development and yield (Jones & Laing, 1978; Sinclair, 1986; Sinclair et al, 2005). Integrating NAR over time results in the accumulated number of nodes on a stem (NN), which is an excellent measurement for plant development (Streck et al, 2006). The NN is related to the timing of the developmental stages. Fehr & Caviness (1977), vegetative development (VD) is described by assigning a number based on the status of the leaf on a node above it. The VD is coded as V‐stages, so that VC = cotyledonary stage (the edges of the unifoliate leaf pair are not touching), V1 = unifoliate leaf pair is expanded and the edges of the first trifoliolate leaf are not touching, and so forth until Vn = the last main stem node. Other developmental stages in soybean, such as R1 (beginning of flowering), R3 (beginning pod), and R5 (beginning seed) are related to NN, as the vegetative phase

Objectives

Methods

Results

Conclusion