Volume 24, Number 6-7, September-November 2004
Crop model STICS (Simulateur mulTIdisciplinaire pour les Cultures Standard)
Page(s) 367 - 381
Agronomie 24 (2004) 367-381
DOI: 10.1051/agro:2004032

Methodology of adaptation of the STICS model to a new crop: spring linseed (Linum usitatissimum, L.)

Francis Fléneta, Pierre Villonb and Françoise Rugetc

a  AlternatecH, Section Agro-Transfert, Chaussée Brunehaut, Estrées-Mons, BP 136, 80203 Peronne Cedex, France
b  Université de Technologie de Compiègne, Département G.S.M, BP 20529, 60205 Compiègne Cedex, France
c  Unité Climat, Sol et Environnement, INRA, Domaine Saint Paul, Site Agroparc, 84914 Avignon Cedex 9, France

(Received 5 June 2003; accepted 10 May 2004)

Abstract - The STICS simulation model was adapted for linseed. An original procedure was used. Firstly, options were selected from among the possibilities available in STICS to simulate the processes of crops. Secondly, the model was calibrated following six steps: 1. gathering of information, 2. the use of parameters from the literature or from other models, 3. the use of STICS parameters for other crops if there is an analogy with linseed, 4. the use of the experimental data to determine parameters which can be measured or calculated, 5. the use of the experimental data to determine parameters by testing a range of values, and 6. the checking of consistency between the parameters and their physical or biological meaning. After adaptation to linseed, the simulations of leaf area, biomass, water consumption, plant nitrogen content, seed number and seed yield were in good agreement with the measurements used for calibration. Thirdly, the results of calculations by STICS were compared with measurements not used for calibration. There was little difference between calculations and measurements of leaf area, biomass, plant nitrogen content and seed number, while seed yield was overestimated because of diseases and lodging, which are not taken into account by the model. However, the differences in seed yield between treatments were properly simulated. This work was a first step towards developing a model to improve linseed crop management. To this end, modifications are needed to account for all yield limitations.

Key words: linseed / modeling / calibration / validation / STICS

Corresponding author: Francis Flénet

© INRA, EDP Sciences 2004