The area of vegetable production is growing rapidly world-wide, as are efforts to increase production on existing lands in these labor- and input-intensive systems. Yet information on nutrient losses, greenhouse gas emissions, and input efficiency is lacking. Sustainable intensification of these systems requires knowing how to optimize nutrient and water inputs to improve yields while minimizing negative environmental consequences. This work characterizes soil nitrogen (N) dynamics, nitrate (NO3¯) leaching, greenhouse gas emissions, and crop yield in five diversified vegetable systems spanning a gradient of intensification that is characterized by inputs, tillage and rotational fallow periods. The study systems included a low input organic system (LI), a mechanized, medium scale organic system (CSA), an organic movable high tunnel system (MOV), a conventional system (CONV) and an organic stationary high tunnel system (HT). In a three-year vegetable crop rotation with three systems (LI, HT and CONV), key N loss pathways varied by system; marked N2O and CO2 losses were observed in the LI system and NO3– leaching was greatest in the CONV system. Yield-scaled global warming potential (GWP) was greater in the LI system compared to HT and CONV, driven by greater greenhouse gas flux and lower yields in the LI system. The field data from CONV system were used to calibrate the Root Zone Water Quality Model version 2 (RZWQM2) and HT and LI vegetable systems were used to validate the model. RZWQM2 simulated soil NO3¯-N content reasonably well in crops grown on bare ground and open field (e.g. beet, collard, bean). Despite use of simultaneous heat and water (SHAW) option in RZWQM2 to incorporate the use of plastic mulch, we were not able to successfully simulate NO3¯-N data. The model simulated cumulative N2O emissions from the CONV vegetable system reasonably well, while the model overestimated N2O emissions in HT and LI systems.

Agricultural system models integrate many different processes that cannot all be measured in field experiments and help quantify soil water dynamics, crop evapotranspiration, and crop growth with high temporal resolution. Understanding soil water dynamics and crop evapotranspiration is essential to improve agricultural management of field crops. For example, the interaction between nitrogen application rate and water dynamics is not sufficiently understood. In most cases, model simulations deviate from field measurements, especially when model input parameters are indirectly and unspecifically derived. The extent to which measured soil hydraulic property inputs decrease the discrepancy between measured and simulated soil water status is not well understood. Consequently, this study: (i) investigated thr use of measured soil hydraulic properties as Root Zone Water Quality Model (RZWQM2) inputs compared to indirectly derived inputs; (ii) explored the capability of calibrating measured soil hydraulic property input parameters for one crop and using them for other crops without further calibration; (iii) studied the effect of the nitrogen application rate on the behavior of soil water dynamics and crop evapotranspiration using RZWQM2 under different rainfall amounts. To evaluate the model in different field management conditions, a field experiment with soybean, corn, wheat, and fallow soil was conducted from 2015 – 2017 to collect field data to calibrate and validate the RZWQM2 model. The model presented a satisfactory response to using measured soil hydraulic property inputs and a satisfactory capability to quantify the effect of nitrogen rates on daily crop evapotranspiration, soil water dynamics, and crop growth. With sufficient measurements of soil hydraulic parameters, it was possible to build a RZWQM2 model that produced reasonable results even without calibration.