Simulation and calibration of extended hortsyst model in predicting macronutrients in tomato (Solanum lycopersicum L.)

Abstract

The production potential of high crop yields is latent in Mexico, but it must be done in moderation so as not to continue increasing pollution and save production costs with optimal fertilizer use. The HortSyst dynamic model which predicts dry matter production (DMP) was calibrated and evaluated using data from the summer to fall 2018, actualized crop parameters and a time step of 5 min, for tomato grown from transplant until the beginning of the reproductive phase in a plastic greenhouse located at the central part of Mexico. The HortSyst model performance without calibration as predicting DMP had an efficiency (EF) of 0.91 and a radiation use efficiency (RUE) of 4.86 g MJ-1 m-2 PAR, failing to simulate the magnitude of the fluctuation between measurements, compared with the calibrated that failing to simulate the fluctuation pattern through measurements with EF 0.99 and RUE 5.89 g MJ-1 m-2 PAR. In order to modeling macronutrients (N, P, K, Ca and Mg) concentrations, Gaussian curves and a polynomial curve using as an input cumulative relative development rate (RDRc) were used, which resulted in a coefficient of determination (R2 ) of 0.99 for N, P, Ca and Mg models and R2 of 0.86 for K model. To determine macronutrients uptake, the concentration of the nutrients was multiplied by the weight produced, resulting in an EF greater than 0.98 in the simulations for all macronutrients.