Nonpoint Source Pollution Impacts of Agricultural Land Use

Open Access
Publisher Website
Google Scholar
Abstract
The amounts of soluble and particulate phosphorus (P) and nitrogen (N) transported in runoff from agricultural land were predicted using relationships describing the physical and chemical processes involved in their transport. These values were compared with amounts measured in runoff from nine unfertilized grassed, six fertilized wheat, and five fertilized mixed crop watersheds in Oklahoma and Texas over a 10-year period. An estimate of the bioavailability of sediment-bound phosphorus (0.1 M sodium hydroxide extractable) was also included in the analysis. Measured and predicted soluble phosphorus (R2=0.97), particulate phosphorus (R2=0.93), bioavailable phosphorus (R2=0.98), and total nitrogen concentrations (R2=0.93) in runoff from all 20 watersheds in 801 events were not significantly different (5 percent level), with an average standard error of 11,13,13, and 16 percent, respectively, for the predicted value. This was the case for a wide range in measured concentrations (8–9132, 10–88076, and 4–3596 μg/L, and 0.3–254.0 mg/L for soluble, particulate, and bioavailable phosphorus, and total nitrogen, respectively). Predictive relationships may be improved, particularly for low flow events of low soil loss (<50 kg/ha/yr), by making the relationship constants a function of soil loss or land use factors affecting soil loss. Land-use impacts on nonpoint source transport of phosphorus and nitrogen were evident, with losses increasing in the order: grass, mixed-crop, and wheat. The average loss of phosphorus and nitrogen (1.3 and 5.1 kg/ha/yr, respectively) was not of major importance from an agronomic standpoint, when compared to fertilizer input (8 kg phosphorus and 41 kgN/ha/yr). However, soluble phosphorus concentrations were consistently above values associated with accelerated eutrophication (10 μg/L), even in unfertilized cases.