Modelling the growth, development and yield of Triticum durum Desf under the changes of climatic conditions in north-eastern Europe.

How agricultural ecosystems adapt to climate change is one of the most important issues facing agronomists at the turn of the century. Understanding agricultural ecosystem responses requires assessing the relative shift in climatic constraints on crop production at regional scales such as the temperate zone. In this work we propose an approach to modeling the growth, development and yield of Triticum durum Desf. under the climatic conditions of north-eastern Poland. The model implements 13 non-measurable parameters, including climate conditions, agronomic factors, physiological processes, biophysical parameters, yield components and biological yield (latent variables), which are described by 33 measurable predictors as well as grain and straw yield (manifest variables). The agronomic factors latent variable was correlated with nitrogen fertilization and sowing density, and biological yield was correlated with grain yield and straw yield. An analysis of the model parameters revealed that a one unit increase in agronomic factors increased biological yield by 0.575. In turn, biological yield was most effectively determined by climate conditions (score of 60-62) and biophysical parameters (score of 60-67) in the 2nd node detectable stage and at the end of heading. The modeled configuration of latent and manifest variables was responsible for less than 70% of potential biological yield, which indicates that the growth and development of durum wheat in north-eastern Europe can be further optimized to achieve high and stable yields. The proposed model accounts for local climate conditions and physiological processes in plants, and it can be implemented to optimize agronomic practices in the cultivation of durum wheat and, consequently, to expand the area under T. durum to regions with a temperate climate.