The U.S. Geological Survey and the Bureau of Land and Water Conservation of the Pennsylvania Department of Environmental Protection conducted a cooperative study to determine the effects of manure application and antecedent soil-phosphorus concentrations on the transport of phosphorus from the soil of a typical farm site in Lancaster County, Pa., from September 1992 to March 1995. The relation between concentrations of soil phosphorus and phosphorus transport needs to be identified because excessive phosphorus concentrations in surface-water bodies promote eutrophication.

The objective of the study was to quantify and determine the significance of chemical, physical, and hydrologic factors that affected phosphorus transport. Three study plots less than 1 acre in size were tilled and planted in silage corn. Phosphorus in the form of liquid swine and dairy manure was injected to a depth of 6-8 inches on two of the three study plots in May 1993 and May 1994. Plot 1 received no inputs of phosphorus from manure while plots 2 and 3 received an average of 56 and 126 kilograms of phosphorus per acre, respectively, from the two manure applications. No other fertilizer was applied to any of the study plots. From March 30, 1993, through December 31, 1993, and March 10, 1994, through August 31, 1994 (the study period), phosphorus and selected cations were measured in precipitation, manure, soil, surface runoff, subsurface flow (at 18 inches below land surface), and corn plants before harvest. All storm events that yielded surface runoff and subsurface flow were sampled. Surface runoff was analyzed for dissolved (filtered through a 0.45-micron filter) and total concentrations. Subsurface flow was only analyzed for dissolved constituents. Laboratory soil-flask experiments and geochemical modeling were conducted to determine the maximum phosphate retention capacity of sampled soils after manure applications and primary mineralogic controls in the soils that affect phosphate equilibrium processes.

Physical characteristics, such as particle-size distributions in soil, the suspended sediment and particle-size distribution in surface runoff, and surface topography, were quantified. Hydrologic characteristics, such as precipitation intensity and duration, volumes of surface runoff, and infiltration rates of soil, were also monitored during the study period. Volumes of surface runoff differed by plot.

Volumes of surface runoff measured during the study period from plots 1 (0.43 acres), 2 (0.23 acres), and 3 (0.28 acres) were 350,000, 350,000, and 750,000 liters per acre, respectively. About 90 percent of the volume of surface runoff occurred after October 1993 because of the lack of intense precipitation from March 30, 1993, through November 30, 1993. For any one precipitation amount, volumes of surface runoff increased with an increase in the maximum intensity of precipitation and decreased with an increase in storm duration. The significantly higher volume of surface runoff for plot 3 relative to plots 1 and 2 was probably caused by lower infiltration rates on plot 3.

Soil concentrations of plant-available phosphorus (PAP) for each study plot were high (31-60 parts per million) to excessive (greater than 60 parts per million) for each depth interval (0-6, 6-12, and 12- 24 inches) and sampling period except for some samples collected at depths of 12-24 inches. The high levels of PAP before manure applications made it difficult to detect any changes in the concentration of soil PAP caused by manure applications. Manure applications to the study area prior to this study resulted in relatively high concentrations of soil PAP; however, the manure applications to plot 3 during the study period did cause an increase in the soil concentration of PAP after the second manure application. The percentages of total phosphorus in plant-available and inorganic forms were about 5 and 80 percent, respectively, in the 0-24--inch depth interval of soil on the study plots. Concentrations of total phosphorus on sand, silt, and clay particles from soil were 700, 1,000, and 3,400 parts per million, respectively. About 70 percent of the total mass of phosphorus in soil to a depth of 24 inches was associated with silt and clay particles.

Soil-flask experiments indicated that soils from the study plots were not saturated with respect to phosphorus. Soils had the capacity to retain 694 to 1,160 milligrams of phosphorus per kilogram of soil. The measured retention capacity probably exceeded the actual retention capacity of soil because laboratory conditions optimized the contact time between soil and test solutions.

Geochemical modeling indicated that the primary mineralogical controls on the concentration of dissolved phosphorus in surface runoff and subsurface flow were aluminum and iron oxides and strengite (if it exists). Aluminum and iron oxides bind phosphate in solution and strengite is an iron-phosphate mineral. The mineralization of organic phosphorus into dissolved inorganic forms could also supply phosphorus to surface runoff and subsurface flow.

Phosphorus inputs to the plots during the study period were from precipitation and manure. Phosphorus inputs from precipitation were negligible. The loads of phosphorus to the plots from manure applications in May 1993 and May 1994 were 112 and 251 kilograms per acre for plots 2 and 3, respectively; about 60 percent of the load occurred in 1994.

Phosphorus outputs in surface runoff differed between study plots. The cumulative yields of total phosphorus during the study period for plots 1, 2, and 3 were 1.12, 1.24, and 1.69 kilograms per acre, respectively. Differences between plots were primarily evident for dissolved yields of phosphorus. The percentage of the total phosphorus output in surface runoff that was in the dissolved phase varied from 6 percent for plot 1 to 26 percent for plot 3.

The cumulative yields of dissolved phosphorus from plots 2 and 3 were 135 and 500 percent greater, respectively, than the dissolved yield from plot 1. Even though volumes of surface runoff were different on the plots, the primary cause of the difference between plots in the yield of dissolved phosphorus in surface runoff was differences in the concentration of dissolved phosphorus. After the second manure application, concentrations of dissolved phosphorus in surface runoff on plots 2 and 3 were significantly higher than the concentration for plot 1.

An increase in the concentration of dissolved phosphorus in subsurface flow from plots 2 and 3 was measured after manure applications. The mean concentrations of dissolved phosphorus in subsurface flow after the first manure application were 0.29, 0.57, and 1.45 milligrams per liter of phosphorus for plots 1, 2, and 3, respectively.

The loss of dissolved phosphorus in surface runoff was related to the soil concentration of PAP. The model relating dissolved phosphorus in surface runoff to soil PAP indicated that concentrations of dissolved phosphorus in surface runoff would exceed 0.1 milligram per liter if soil concentrations of PAP exceeded 9 parts per million; this PAP concentration was exceeded by each study plot. Over 50 percent of the variation of dissolved phosphorus in surface runoff was explained by soil concentrations of PAP in the 0-6-inch depth interval.

The loss of suspended phosphorus in surface runoff was primarily affected by the particle-size distribution of suspended sediment in surface runoff. Surface runoff was enriched with fines relative to the soil matrix. Generally, over 90 percent of sediment in runoff was comprised of silt and clay particles; only 50-60 percent of particle sizes from the intact soil matrix were in the silt- to clay-size range. Concentrations of suspended phosphorus in surface runoff were not significantly related to soil concentrations of total phosphorus in the 0-6-inch depth interval.

Concentrations of dissolved phosphorus in subsurface flow were also related to soil concentrations of PAP. The relation indicated that dissolved concentrations of phosphorus in subsurface flow would exceed 0.1 milligram per liter if soil concentrations of PAP in the 0-6-inch depth interval of soil were greater than 49 parts per million; this PAP concentration was exceeded by each study plot.

The significant relation of high concentrations of dissolved phosphorus in water to soil concentrations of PAP indicated that soils with comparable concentrations of soil PAP would be potential sources of dissolved phosphorus to surface water and subsurface water tables. The percentage of the total phosphorus lost from a system in the dissolved form increased as soil concentrations of PAP increased. This indicates that best-management practices to reduce phosphorus losses from this system not only need to target suspended forms of phosphorus but also dissolved forms. Practices aimed at reducing the loss of dissolved phosphorus from the system increase in importance with an increase in soil concentrations of PAP.