Submitted to: American Water Resources Association Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 9/1/2005
Publication Date: 9/21/2005
Citation: Hively, W.D. 2005. Characterization of phosphorus loading source areas in a monitored dairy farm landscape, five years after implementation of best management practice. In: American Water Resources association Conference Proceedings, New York City Watershed Science and Technical Conference, September 21-22, 2005, Fishkill, New York. p. 65.
Interpretive Summary: Limiting phosphorus (P) losses from the agricultural landscape, often acheived through implementation of best management practices (BMPs), is important because this nutrient is widely recognized as a primary cause of eutrophication of inland waters. Successful implementation of BMPs, however, relies upon clear identification of P loading source areas, which is often complicated due to spatial and temporal variation in precipitation, soil moisture, extent of saturated areas, flowpaths, and farm management. The objective of this project was to identify and characterize P loading source areas throughout a 160-ha agricultural watershed located in the headwaters of the Cannonsville Reservoir basin, Delaware County, New York. The watershed contains a third generation, family-run dairy farm maintaining approximately 80 milking cows and 35 replacement heifers. Two years of field data collection, including grabsampling of surface runoff and collection of soil and stream sediment samples, provided a basis to assess P contributions from near-barn concentrated source areas, agricultural fields, and forested areas. Seventy-five grabsamples of surface runoff and tile drain flow were collected at diverse landscape positions (roadways, stream channels, culverts, diversion ditches, tile drain outlets, barnyards, and field areas) on five dates between June 2000 and May 2001. Sampling conditions included small rainfall events, snowmelt events, and baseflow. Observed runoff concentrations of dissolved P were elevated at near-barn locations, roadways, and tile drains; were moderate at intensively managed field locations; and were low at extensively managed and forested upper-watershed locations. Results of extensive soil testing portrayed increasing soil P contents with increasing with proximity to the barn and frequency of manure application. Stream sediment samples exhibited low P contents in upper stream reaches, with increasing P enrichment in the lower watershed and downstream from the barnyard. Although less than two percent of the entire study watershed is occupied by impermeable areas, and only a third of that area receives manure, nutrient loading from critical impermeable source areas is potentially a significantly source of stream water P, particularly in the summertime when short-duration precipitation events produce runoff primarily from the near stream landscape. Overall, results illustrated the importance of reducing runoff from high-P near-barn source areas, the need to plan and manage P application to agricultural fields and pastures so as to avoid excessive accumulation of P, and the beneficial role that upper-hillside forested land plays in providing clean water to dilute high concentration runoff from near-barn, frequently manured areas.
Technical Abstract: Two years of field data collection provided the basis for characterization of phosphorus (P) loading source areas throughout a 160-ha monitored dairy farm watershed located in the headwaters of Cannonsville Reservoir, a New York City (NYC) drinking water supply. Best management practices (BMPs) implemented on the farm using the NYC Watershed Agricultural Program’s Whole Farm Planning process had resulted in a 43% reduction in watershed TDP loading over a four-year monitoring period (1), but at the time of sampling it was evident that several P loading problem areas remained. Soil and water sampling was therefore used to characterize P loading sources and processes throughout the landscape. Grabsamples (n=75) of surface runoff and stream flow were collected during baseflow, rainfall, and snowmelt. Observed runoff concentrations of total dissolved phosphorus (TDP) were elevated at near-barn locations, roadways, and tile drains; were moderate at intensively managed field locations; and were low at extensively managed and forested upper-watershed locations. Soil samples (n=78) were collected throughout the watershed at field and non-field locations. Soil test phosphorus (STP), which rainfall simulation had shown to correlate with TDP concentrations in surface runoff (2), generally increased with proximity to the barn and with frequency of manure application. Only near-barnyard soils exhibited excessive levels of STP. Stream sediment samples (n=41) exhibited low STP in upper stream reaches, with increasing P enrichment in the lower watershed and downstream from the barnyard. Sample results assisted in the development of P extraction coefficients that were used in distributed hydrological modeling of TDP loading. Additionally, field observations were used to characterize the magnitude of P load contributions from ten specific source areas including roadways, barnyards, barnyard filter areas, stream crossings, wetlands, and forests. A second round of BMPs, implemented in 2002-2003, have since corrected many of the identified P loading problem areas.