Submitted to: Soil Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 11, 2004
Publication Date: June 1, 2004
Repository URL: http://www.ars.usda.gov/SP2UserFiles/Place/53102000/pdf_pubs/P1836.pdf
Citation: Vaughan, P.J., Shouse, P.J., Suarez, D.L., Goldberg, S.R., Ayars, J.E. 2004. Boron transport within an agricultural field: uniform flow versus mobile-immobile model simulations. Soil Science. 169(6):401-412. Interpretive Summary: Boron is a naturally occurring element in soils and is particularly concentrated in soils of the west side of the San Joaquin Valley in California. Boron is an essential nutrient for plants but can become toxic to plants at relatively low concentrations and sensitive plants, such as melons grown in this area, can be adversely affected. The objective of this work was to determine how best to model the vertical movement of boron in the soil of an irrigated agricultural field located in this area. Two different computer models were tested, comparing measured boron concentrations in the field soils over a two-year period with the models' predictions. Results indicate that, for a period of two years or less, the predictions of the simpler model, the uniform flow model, are as good as those of the mobile-immobile water model. Theoretical arguments suggest that the mobile-immobile model may be more accurate over longer periods. Further work is needed to determine whether this is the case for periods longer than two years. This work has importance in the evaluation of potential negative consequences of irrigating with lower quality water that could result in soil boron concentrations that are toxic to certain plants.
Technical Abstract: The transport of boron in soil is important to agriculture because boron concentrations in soil water are beneficial to plants only over a limited range (0.37 to 1.39 mmol L-1 for tolerant crops). Irrigation water in the San Joaquin Valley, California commonly has elevated B concentrations and soil water B can reach phytotoxic levels due to the concentrating effects of transpiration and soil surface evaporation. Because the constant capacitance model successfully computed B speciation in soil water and on mineral surfaces, it was incorporated into a multicomponent solute transport code and a two-year field test of the model was performed for 43 sites within a 65 ha field in the San Joaquin Valley. The model successfully predicted the adsorbed B (SOB(OH)3-) concentration with a median scaled root mean square error (SRMSE) of 11% for 43 sites. The median SRMSE for prediction of total B was 36% and 46% for solution B. Higher SRMSE for solution B may be caused by lack of detail in specifying the lower boundary condition. A steady increase in SRMSE from east to west in the field, the same trend as the seven tile drains, raises the possibility that the bottom boundary conditions could have an unknown systematic variation in this direction. A mobile-immobile water transport model failed to exhibit a significant improvement over the standard uniform flow model (UFM) so the simpler UFM was preferred. The change in total B mass at all sites generated was accurately predicted with a relative error of only 4.1%. This work has potential practical application in studying effect of water management practices on soil B.