Modeling phosphorus losses from soils amended with cattle manure and chemical fertilizers

Authors: WANG, ZHAOZHI - Agriculture And Agri-Food Canada; ZHANG, T. - Agriculture And Agri-Food Canada; Vadas, Peter; QI, Z. - McGill University - Canada; WELLEN, C. - Ryerson University

Submitted to: Science of the Total Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/11/2018
Publication Date: 10/15/2018
Publication URL: https://handle.nal.usda.gov/10113/5979900
Citation: Wang, Z., Zhang, T.Q., Vadas, P.A., Qi, Z.M., Wellen, C. 2018. Modeling phosphorus losses from soils amended with cattle manure and chemical fertilizers. Science of the Total Environment. 639:580-587.

Interpretive Summary: Manure and fertilizer can be a major source of phosphorus loss in surface runoff, but few models simulate this phosphorus loss process, which can limit their use for management and policy recommendations. We tested the SurPhos model for its ability to simulate manure and fertilizer phosphorus loss using data from field plots in the Lake Erie watershed. SurPhos acceptably simulated soil phosphorus dynamics and dissolved phosphorus loss for both solid and liquid cattle manure, as well as inorganic fertilizer. Results show SurPhos can be reliably used to quantify different sources of phosphorus loss and soil P changes. This allows for better assessment of the effectiveness of different P management practices to reduce in phosphorous losses.

Technical Abstract: While applied manure/fertilizer is an important source of P loss in surface runoff, few models simulate the direct transfer of phosphorus (P) from soil-surface-applied manure/fertilizer to runoff. The SurPhos model was tested with 2008-2010 growing season, daily surface runoff data from clay loam experimental plots subject to different manure/fertilizer applications. Model performance was evaluated on the basis of the coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE), percent bias (PBIAS), and the ratio of the root mean square error to the standard deviation of observed values (RSR). The model offered an acceptable performance in simulating soil labile P dynamics (R2 = 0.75, NSE = 0.55, PBIAS = 10.43%, and RSR = 0.67) and dissolved reactive P (DRP) loss in surface runoff (R2 = 0.74 and NSE = 0.69) for both solid and liquid cattle manure, as well as inorganic fertilizer. Simulated direct P loss in surface runoff from solid and liquid cattle manure accounted for 39% and 40% of total growing season DRP losses in surface runoff. To compensate for the unavailability of daily surface runoff observations under snow melt condition, the whole four years’ (2008-2011) daily surface runoff predicted by EPIC was used as SurPhos input. The accuracy of simulated DRP loss in surface runoff under the different manure/fertilizer treatments was acceptable (R2 = 0.55 and NSE = 0.50). For the solid cattle manure treatment, of all annual DRP losses, 19% were derived directly from the manure. Beyond offering a reliable prediction of manure/fertilizer P loss in surface runoff, SurPhos quantified different sources of DRP loss and dynamic labile P in soil, allowing a better critical assessment of different P management measures’ effectiveness in mitigating DRP losses.