Parameterization of AquaCrop model for vining pea biomass and yield predictions and assessing impacts of irrigation strategies considering various sowing dates

Abstract

The AquaCrop model was parameterized and evaluated using data relative to supplemental irrigated vining pea for industry using observations in farmers’ fields located in the Ribatejo region, Portugal. Data refers to field observations relative to leaf area index (LAI), soil water content, biomass and final yield relative to two contrasting rainfall years, 2011 and 2012, a wet and a dry year, respectively. LAI data were used to parameterize the canopy cover (CC, %) curve. Results showed that after parameterization of the CC curve the respective root mean square errors (RMSE) were smaller than 15 % of the fraction of ground cover by the crop. The model performance relative to the soil water balance simulation revealed a slight over-estimation bias for the wet year but estimation errors were small, with RMSE representing <5 % of the average soil water content observations. Good accuracy for biomass and yield predictions was obtained, with deviations between predictions and observations ranging from 1 to 7 % of the average observed biomass and up to 11 % of the average observed yield. Overall, results have shown that AquaCrop is appropriate for vining pea biomass and yield estimation. Therefore, the model was explored to assess impacts of sowing dates and alternative supplemental irrigation strategies on water use and yield. Results have shown that sowing by early January is appropriate in terms of improved water use and yield and that a very mild deficit irrigation schedule may be adopted without significantly affecting yields.