|Yakirevich, Alex - BEN GURION UNIV, ISRAEL|
Submitted to: Computers in Agriculture
Publication Type: Abstract Only
Publication Acceptance Date: May 3, 2009
Publication Date: N/A
Technical Abstract: Quantitative evaluation of the effect of field manure application on bacterial concentrations in creeks adjacent to the field requires developing microbial transport models. Reliable testing of such models with bacteria monitoring data requires a better understanding and estimation of the uncertainty in the monitoring data caused by either spatial and/or temporal variability of bacterial concentrations. The objectives of this work were: (i) to develop and calibrate an event-based model describing bacterial transport with runoff water at a field scale; and (ii) to characterize the associated uncertainties both in the input and the model predictions. Experiments were carried out at the ARS Beltsville experimental watershed site (OPE3) in Maryland. Bovine manure was collected from the adjacent USDA dairy facilities, accumulated and stored for 5 months before the application. Fecal coliforms (FC) content in the manure, after applied to the field, varied within six orders of magnitude, and the distribution was highly asymmetrical. Manure was broadcast with 60 ton/ha applicatioin rate on the 3.6 hectare experimental field equipped with soil moisture sensors. The top soil texture is mostly sandy loam and loamy sand. Water content and fecal coliform (FC) contents were measured in the upper 10-cm soil layer before and after manure applications. Both runoff volume and FC concentrations in runoff water were monitored using flume equipped with a refrigerated sampler. Standard meteorological data were measured in the vicinity of the experimental field. A total of 6 runoff events were recorded from May 2 through July 8, 2004. Fecal coliform concentations were highest in the runoff water collected 7 days after the manure application and then declined with time. No clear relationships between FC content and runoff flow rates was observed. A bacterial transport add-on module was developed for the event-based kinematic runoff and erosion model (KINEROS2). The bacterial transport module describes convective-dispersive overland transport and accounts for bacteria release from manure, reversible attachment-detachment to soil, and surface straining of infiltrating bacteria. The model was successfully calibrated with our runoff experimental data and adequately described the FC transport with runoff water. Furthermore, model simulations showed that the spatial variation of soil infiltration and the variation of bacteria concentrations in manure were major sources of uncertainty in predictions of bacteria transport.