Evaluation of the CSM‐CROPGRO‐Canola Model for Simulating Canola Growth and Yield at West Nipissing in Eastern Canada

Abstract

With increasing demands for renewable energy and dietary vegetable oils, the production of canola has become widespread in recent years. Modeling canola growth and yield is a helpful approach to predict canola responses to various environments, especially under climate change. However, few studies have been performed for predicting growth and yield of canola in Canada. In this study, we evaluated the CSM‐CROPGRO‐Canola model in Decision Support System for Agrotechnology Transfer v4.6 for simulating spring canola at West Nipissing in Eastern Canada. The model was evaluated using plant and soil data collected from field experiments over three growing seasons (2012–2014). The model could predict the observed crop development and successfully mimic the characteristics of canola regarding light absorption and utilization using combinations of leaves and pods. The accumulations of aboveground biomass were satisfactorily simulated in the life cycle under different nitrogen (N) fertilizer application rates, with a normalized RMSE of 19%. The seed yields were successfully predicted with different N application rates except for an underestimation under zero N application. The underestimation of yield under low N rates was possibly related to the deficiency in the simulated N mineralization that could also be associated with inaccurate input soil data. A better simulation of seed yields under low N application was achieved when the soil organic matter module based on the CENTURY model was used in DSSAT v4.6. The calibrated model simulated soil moisture and inorganic N contents satisfactorily, showing a good performance of the CSM‐CROPGRO‐Canola model for the study region.

Simulated Canola Yield Responses to Climate Change and Adaptation in Canada

“…The CSM-CROPGRO-Canola model, included in the Decision Support System for Agrotechnology Transfer (DSSAT v4.6, Hoogenboom et al, 2015;Deligios et al, 2013), has been successfully adapted for simulating canola growth and yield in Canada (Jing et al, 2016). Our objectives for this study were (i) to simulate canola growth and yield under distinct climate conditions across three locations across Canada, (ii) to simulate the responses of canola growth and yield under a set of future climate scenarios developed with a regional climate model, and (iii) to evaluate the potentials of climate change adaptation measures for canola production in Canada.…”

“…Seed yields were manually measured at maturity in 2012 and 2013, and no yield was measured in 2014 due to waterlogging during canola ripening. Details of this experiment can be found in Jing et al (2016). At Brandon and Normandin, N fertilizer at a rate of 90 kg ha -1 was applied at seeding each year.…”

“…It was further adapted for simulating spring canola in eastern Canada with N and water stresses (Jing et al, 2016). The CSM-CROPGRO-Canola model requires soil data; daily weather data, cultivar parameters; and management data such as seeding dates, irrigation, and fertilization to simulate crop growth and yield on a daily step.…”

“…They are categorized as species, ecotype, and cultivar. Using the same methodology of model calibration and evaluation for cultivar InVigor 5440 at West Nipissing as detailed in Jing et al (2016), we calibrated the cultivar coefficients for 5030LL based on field experiment data for 2010 at Normandin. The experimental data at Brandon and the remaining independent data at Normandin for 2011 and 2012 were used for model evaluation.…”

“…In fact, these two cultivars are similar although InVigor 5440 has a slightly higher yield potential than 5030LL. Root-mean squared error (RMSE) and the normalized rootmean squared error (nRMSE), along with model simulation efficiency (EF) and index of agreement (d) used in Jing et al (2016), were applied to evaluate the performance of the model.…”