Modeling Phosphorus Transport in Agroecosystems

Joining Users, Developers, and Scientists

Abstracts for Oral Presentations

STIRRING UP PHOSPHORUS AND THE PUBLIC: CORNELL’S LAKE SOURCE COOLING PROJECT

Nelson G. Hairston, Jr., Department of Ecology and Evolutionary Biology, Cornell University

As a part of eliminating the use of CFCs for central air conditioning on campus, Cornell’s facilities engineers proposed to take advantage of the university’s situation near Cayuga Lake by using its cold deep water as a natural cooling resource. They envisioned pumping lake water from depth, where it is continuously 4 °C, through a heat-exchanger at the shore that would remove heat from a closed loop circulating water from the campus central air conditioning system. The lake water would then be returned to the lake near shore in shallow water. The project was predicted to reduce energy use for central air conditioning by 80%. In 1994, I chaired a University faculty technical advisory committee formed to review the potential environmental impacts of this “Lake Source Cooling” project. Chief among our concerns was the movement of phosphorus from the bottom waters of the lake to the illuminated surface waters in summer as a potential cause of algal blooms. Based on a comparison of this artificial internal P load with the existing external loads and the current trophic state of the lake, we concluded that the impacts of LSC would be short-lived and so small as to be difficult or impossible to detect. A local citizens group, however, became extremely concerned, was unconvinced by our analysis, and hired their as a consultant a well-known limnologist from another state. He wrote a report suggesting that the amount of P returned to the surface could “make the situation worse” for blue-green algae, “result in taste and odor problems in drinking water, [and] dead pets and livestock that drank the water,” “produce odiferous compounds ... whose removal from drinking water would require …millions of dollars,” and create “ luridly patterned dying blooms ... [that] could create such an odor that they would clear the beaches and houses for 300 m back from the shoreline.” The State nevertheless permitted the LSC project, and in 2000 the campus cooling system came on line. It has been running continuously for the past 6 years with extensive monitoring of the lake ecosystem. Thus, from a scientific perspective, we have initial data, two radically different hypotheses as to the impact of the P moved by LSC on the lake, and we ran a 58-million-dollar experiment that tested these alternate hypotheses. I will let you know the out come. [Hint: I’m still in town.]

Back to Conference Program

A NATIONAL ASSESSMENT OF THE EFFECTS OF CONSERVATION PRACTICES FOR CROPLAND

Robert L. Kellogg, Natural Resources Conservation Service, Beltsville, MD.

The National Assessment component of the Conservation Effects Assessment Project (CEAP) will provide estimates of the environmental benefits of conservation practices for reporting at the national and regional level and will assess the potential for existing conservation programs and future alternatives to meet the Nation’s environmental and conservation goals. For cropland, a subset of 20,000 National Resources Inventory (NRI) sample points has been selected to serve as "representative fields." USDA developed and implemented a new farmer survey to collect the information needed at the selected NRI sample points to run the field-level process model APEX. APEX is used to estimate field-level effects attributable to conservation practices--reductions in nitrogen, phosphorus, pesticide, and soil loss from farm fields as well as soil quality enhancement. Model output from APEX is used as an input to the SWAT/HUMUS model to assess off-site benefits for water quality--reductions in in-stream concentrations of sediment, nutrients, and pesticides attributable to implementation of conservation practices.

Back to Conference Program

MODELING PHOSPHORUS DYNAMICS IN WATERSHEDS WITH HSPF

Anthony S. Donigian, Jr. and Brian R. Bicknell,
AQUA TERRA Consultants, Mountain View, CA

The Hydrological Simulation Program - Fortran (HSPF) is a comprehensive package for simulation of watershed hydrology and water quality for conventional and toxic organic pollutants. HSPF can simulate the hydrologic, and associated water quality, processes on pervious and impervious land surfaces and in streams and well-mixed impoundments. HSPF incorporates the watershed-scale ARM and NPS models into a basin-scale analysis framework that includes fate and transport in one-dimensional stream channels. It is a comprehensive model of watershed hydrology and water quality that allows the integrated simulation of land and soil contaminant runoff processes with in-stream hydraulic and sediment-chemical interactions. It has been applied to complex watersheds consisting of urban, agricultural, forest, and wetland components. This paper focuses on the P modeling capabilities in both the land and water phases, and discusses application to the Chesapeake Bay Watershed.

HSPF offers a comprehensive system for modeling P transport and reactions in soil and water as part of a complete watershed simulation. The optional soil/land surface methods range from simple water- and soil-associated runoff from pervious and impervious areas to detailed P-cycling and transport/leaching in the soil profile. Soil P processes include sorption/desorption, mineralization, immobilization, plant uptake by optional first-order and yield-based methods, and transport to streams with flow and sediment. P and other nutrient inputs to soil can be specified from adjacent land areas, atmospheric deposition, and user-defined applications (e.g. fertilizer, manure) of various P species to four separate soil layers. Losses of P to streams and water bodies are modeled using several optional methods including a simple nonreactive species, or divided into dissolved, sorbed, and organic forms. P dynamics in the water column of free-flowing rivers and lakes/reservoirs include sorption/desorption and transport (deposition/scour) with three size fractions of sediment, and algal cycling that considers both water-column phytoplankton and bed-attached algae. Sources of P to the stream include point sources and atmospheric deposition in addition to the nonpoint contributions. We also describe a large-scale application of nutrient (N and P) modeling with HSPF in the Chesapeake Bay Watershed, which provides loadings to the Chesapeake Bay Water Quality Model.

Back to Conference Program

THE EPIC/APEX PHOSPHORUS COMPONENT

J. R. Williams

The Environmental Policy Integrated Climate (EPIC) model was developed in the early 1980's to assess the effect of erosion on productivity for use in the 1985 National RCA analysis. Since then the model has been expanded and refined to allow simulation of many processes important in agricultural management. The major components in EPIC are weather simulation, hydrology, erosion-sedimentation, nutrient cycling, pesticide fate, crop growth, soil temperature, tillage, economics, and plant environment control. The APEX model was developed to extend the EPIC model capabilities to whole farms and small watersheds. In addition to the EPIC functions, APEX has components for routing water, sediment, nutrients, and pesticides across complex landscapes and channel systems to the watershed outlet. APEX also has groundwater and reservoir components. A watershed can be subdivided as much as necessary to assure that each subarea is relatively homogeneous in terms of soil, land use, management, and weather.

The EPIC and APEX models have been described previously. Only the phosphorus component and other closely related components are described here. The phosphorus model simulates soluble P loss in surface runoff based on the GLEAMS model approach of partitioning pesticides into the solution and sediment phases. A loading function based on the enrichment ratio concept is used to estimate organic P loss with sediment. The P mineralization model is a modification of the PAPRAN mineralization model. The model considers two sources of mineralization: the fresh organic P pool, associated with crop residue and microbial biomass, and the stable organic P pool, associated with the soil humus. The mineralization rate is a function of soil water and temperature and of the C:N and C:P ratios. Mineral P is transferred among three pools: labile, active mineral, and stable mineral. Fertilizer P is labile (available for plant use) at application but may be quickly transferred to the active mineral pool. Flow between the labile and active mineral pools is governed by an equilibrium equation based on the P sorption coefficient.

Recent concern over water quality problems caused by P runoff from feedlots and manure application fields lead to the development of a manure erosion equation for APEX. Depending on the amount of manure cover of the soil the erosion varies from essentially all manure to a combination of manure and soil. Since manure is considered residue, a heavy cover in a feedlot may completely eliminate soil erosion but create the potential for severe manure erosion. Soil erosion potential is also very low in manure application fields with a good grass cover but manure erosion can be high. This situation leads to under estimates organic P using the enrichment ratio approach because the soil erosion rates are near zero. The new manure erosion equation overcomes this problem by providing direct estimates of organic P losses. To better estimate soluble P losses from manure the recently developed equations of Vadas and Sharpley have been installed in APEX. These equations are being tested and compared with previous results and observed data.

Back to Conference Program

SWAT WATERSHED SCALE PHOSPHORUS PROCESS MODELING CAPABILITIES

Jeff Arnold and Colleen Green, USDA-ARS

Phosphorus is an essential crop nutrient that is simulated in models due to its impact on water quality and crop yield. Models can avoid some of the tribulations associated with field studies; however a model’s ability to adequately simulate P is dependent on the usage of quality data for calibration purposes and the correct representation of P model processes and its fractionation. Properly modeling P is difficult due to problems associated with chemical analyses to the availability of P data, including sediment attached-P transport, dissolution, and availability. Also, channel P processes and sources are not well understood and consequently channel models need further testing and refinement. An obvious need exists to improve our understanding of P processes. Phosphorus, in its various forms, should be simulated adequately so that the results are defensible to policy makers and environmental managers.

Back to Conference Program

MODELING BUFFER PROCESSES WITH THE RIPARIAN ECOSYSTEM MANAGEMENT MODEL (REMM)

David Bosch, Richard Lowrance, Jennifer Gilbert, and Randy Williams

In many landscapes, riparian buffer zones can be effective conservation practices for mitigating nonpoint source pollution. Their potential use as a best management practice has been limited because of the lack of a design procedure that can quantify their effectiveness for different climatic, soil, and vegetation conditions. The Riparian Ecosystem Management Model (REMM) was developed for researchers and natural resource agencies as a modeling tool that can help to quantify the water quality benefits of riparian buffers under varying site conditions. Processes simulated in REMM include surface and subsurface hydrology; erosion, sediment transport, and deposition; carbon, nitrogen, and phosphorous transport, removal, and cycling; and vegetation growth. Phosphorous (P) is simulated in REMM as a combination of organic and inorganic forms. The organic pools include those in litter residue and in humus. Simulation of inorganic P is based upon the processes within the EPIC model, defined by labile, active, and stable forms of P. The model operates on a daily time step. Management options such as vegetation type, buffer configuration, and biomass harvesting can be simulated. Recent work with the REMM model has focused upon parameter sensitivity and application for optimizing buffer sizes based upon potential P loading and soil P pools.

Back to Conference Program

COMPLEXITIES OF P MODELLING AT THE CATCHMENT SCALE –
OR WHY MODEL P?

Paul Quinn
Newcastle University School of Civil Engineering and Geosciences
http://www.ncl.ac.uk/wrgi/TOPCAT/

Modelling P is vital to our understanding and capability to resolve environmental problems. Equally P and sediment loss management should be part of new vision to create a vibrant and dynamic rural environment. The problem is not with our understanding of P processes and our ability to measure P, it is in how we apply our models across scale to resolve P problems. It can be argued that models based on knowledge created at the plot or laboratory scale should only be applied at that scale, and attempts to apply them at the catchment scale is either wrong or must at least change the nature of the application. Catchment scale modelling is in fact the ‘policy scale’ and therefore the models used at this scale should be appropriate to our knowledge of processes at this scale (including the uncertainty) and be tailored closely to the policy needs of the catchment. Further, the results of any modelling and decision support tools must be taken up by those who control the rural landscape, i.e. the people. P modellers must embrace the ‘applied’ element their work now and work closely with communities to resolve the many multi-faceted problems arising from agriculture. Modelling tools and management tools, given the gravity of the P problem, are probably best situated to take a lead role in rural land use change. Therefore, it is proposed that a suite of scale appropriate P models are needed within catchments, that are useable or defensible to the people who will either create future change (planners and policy makers) or be expected to change practices (i.e. farmers).

Back to Conference Program

UNCERTAINTY IN MODEL CALIBRATION, VALIDATION,
AND EVALUATION DATA

R.D. Harmel¹, R.J. Cooper², R.M. Slade³, R.L. Haney1, J.G. Arnold¹
¹ USDA-ARS, Temple, TX
² The Macaulay Institute, Craigiebuckler, Aberdeen, Scotland, UK
³ USGS (retired), Austin, TX

The scientific community has not established an adequate understanding of the uncertainty inherent in measured water quality data, which is introduced by four procedural categories: streamflow measurement, sample collection, sample preservation/storage, and laboratory analysis. Although previous research has produced valuable information on relative differences in procedures within these categories, little information is available that compares the procedural categories or presents the cumulative uncertainty in resulting water quality data. As a result, quality control emphasis is often misdirected, and data uncertainty related to watershed modeling is typically either ignored or accounted for with an arbitrary margin of safety. Faced with the need for scientifically defensible estimates of data uncertainty, the objectives of this research were to: (1) compile selected published information on uncertainty related to measured streamflow and water quality data for small watersheds, (2) use a root mean square error propagation method to compare the uncertainty introduced by each procedural category, and (3) use the error propagation method to determine the cumulative probable uncertainty in measured streamflow, sediment, and nutrient data. Best case, typical, and worst case "data quality" scenarios were examined. Averaged across all constituents, the calculated cumulative probable uncertainty (+/-%) contributed under typical scenarios ranged from 6% to 19% for streamflow measurement, from 4% to 48% for sample collection, from 2% to 16% for sample preservation/storage, and from 5% to 21% for laboratory analysis. Under typical conditions, errors in storm loads ranged from 8% to 104% for dissolved nutrients, from 8% to 110% for total N and P, and from 7% to 53% for TSS. Results indicated that uncertainty can increase substantially under poor measurement conditions and limited quality control effort. This research provides introductory scientific estimates of uncertainty in measured water quality data. The results and procedures presented should assist modelers in quantifying the "quality" of calibration and evaluation data sets, determining model accuracy goals, and evaluating model performance.

Back to Conference Program

QUANTIFYING THE EFFECTIVENESS OF PHOSPHORUS CONTROL BMPS USING MODELING AND MODEL-BASED APPROACHES

M.W. Gitau[1], W. J. Gburek[2], T.L. Veith[3]

[1] Program Associate, Biological and Agricultural Engineering, University of Arkansas, Fayetteville, AR.
[2] Hydrologist, USDA-ARS Pasture Systems and Watershed Management Research Unit, University Park, PA.
[3] Agricultural Engineer, USDA-ARS Pasture Systems and Watershed Management Research Unit, University Park, PA.

There has been a steady shift towards modeling and model-based approaches as primary methods of quantifying watershed-wide BMP effectiveness. The advantages of using models include: 1) multiple BMPs can be studied simultaneously; 2) the impacts of individual BMPs can be determined while also determining the effects of BMP combinations; 3) location-specific responses can be obtained; 4) modeling offers a practical means of analyzing various “what-if” management scenarios. We discuss experiences with the Soil and Water Assessment Tool (SWAT), and a recently developed BMP characterization tool, with focus on a farm-sized watershed within the Cannonsville Reservoir Watershed, New York. Monitoring studies have been carried out at selected locations and at the watershed outlet on this farm, during pre- and post-BMP periods, thus providing data with which to compare model simulations. On average, the BMPs reduced DP losses by 31% and TP losses 21%, based on SWAT simulations. Because most of the fields had a combination of at least two BMPs, it was difficult to separate out individual BMP impacts based on SWAT simulations. For DP, integration of BMP tool efficiencies allowed individual BMP impacts to be incorporated while still maintaining the same level of representation as was obtained using model simulations. As the SWAT model is often used with little or no post-BMP data to verify simulation results, this study served to validate SWAT model suitability for evaluating BMP impacts. The BMP tool was found to suitably complement the model by providing insights into individual BMP impacts, and providing BMP efficiency data where the model was lacking.

Back to Conference Program

IMPACT OF REAL WATERSHED HYDROLOGY ON MODELING
PHOSPHORUS TRANSPORT

William J. Gburek[1], M. Todd Walter[2], and Tammo S. Steenhuis[2]

[1]USDA-ARS (retired), University Park, PA
[2]Department of Agricultural and Biological Engineering, Cornell University

In the ideal case, modeling phosphorus (P) loss from a watershed would be viewed as first modeling two individual controls on P loss, one characterizing P availability over the landscape (i.e., P source), and the other characterizing locations of surface runoff production, or leaching where significant (i.e., P transport). Then, the interactions between these source and transport components would be modeled to quantify P loss from the watershed. There is a substantial body of laboratory and plot research related to characterizing P source as a function of soils, land use, and land management practices. However, field-based research related to potential for P transport at the watershed scale is sorely lacking. Consequently, most widely applied watershed models addressing watershed hydrology and nutrient dynamics (e.g., the ARS models SWAT and AnnAGNPS) fall short of the ideal, especially related to their characterization of P transport mechanisms.
Models, by definition, are an approximation of reality. Thus, the model developer must always compromise on level of detail included within the various model components. Sometimes these compromises are based on purpose of the model and desired complexity, sometimes on the model developer’s bias, and sometimes, as inferred previously, on availability of information – this latter area will be the focus of this presentation, specifically as related to representing surface runoff dynamics within a comprehensive hydrology-water quality watershed model. The simplistic P transport routines typically employed in currently used watershed models are examined in context of “real” watershed hydrology as controlled by climate, topography, and land use, and strengths, weaknesses, and limitations of the approaches are discussed.

Back to Conference Program

A NEW MODEL FOR SURFACE APPLICATION OF ANIMAL MANURE
AND MANURE PHOSPHORUS TRANSFORMATION AND LOSS IN RUNOFF

Peter Vadas, Bil Gburek, Andrew Sharpley, Peter Kleinman,
Philip Moore, and Miguel Cabrera

Agricultural P transport in runoff is an environmental concern. An important source of P runoff is surface-applied manures, but computer models used to assess P transport rarely simulate P loss from surface manures. We developed a new model to rectify this weakness. The model operates on a daily basis and simulates manure application to soil, dividing manure P into water-extractable and stable pools, while letting some manure P infiltrate into soil if a slurry is applied. Manure dry matter decomposes, and manure stable P pools decompose into water-extractable pools with time. Manure dry matter and P are assimilated into soil to simulate bioturbation and freeze-thaw activity. Some water-extractable P is leached from manure when it rains, and a portion of leached P can be transferred to runoff. Most manure P leached into soil by rain remains in the top 2 cm, while some leaches deeper. This 2 cm soil layer can contribute P to runoff via desorption. Data from field studies in Texas, Pennsylvania, Georgia, and Arkansas USA were used to build and validate the model, confirming that it accurately simulates daily changes in manure dry matter, manure and soil P pools, storm-by-storm P concentrations in runoff, and cumulative P loads in runoff. Therefore, our manure P runoff model represents an important modification for larger models used to assess agricultural P loss. One particular value of the model is that it identifies which sources of P (soil or manure) are dominant contributors of P to runoff. This can help target management practices most effective in mitigating P loss

Back to Conference Program

AN ALTERNATIVE SUBROUTIEN FOR MODELING INORGANIC PHOSPHORUS SORPTION AND DESORPTION ON SOILS

Nathan Nelson, Kansas State University

Excess phosphorus applied to soils with low P adsorption capacities can enter surface water via leaching and subsurface transport, thereby negatively impacting water quality. Computer simulation models can be used to describe the effects of management practices on P leaching losses provided the models are appropriately validated. The objectives of this research were to modify and validate P subroutines in the GLEAMS (Groundwater Loading Effects of Agricultural Management Systems) model to more accurately reflect P sorption and desorption, then use the modified model to determine crop and waste management effects on long-term P leaching losses below the root zone. GLEAMS was modified with the Langmuir equation to partition labile P between adsorbed and solution phases. The modification improved predictions of percolate P concentrations and soil P accumulation in acid sandy soils receiving waste-based P additions. The modification also increased model sensitivity to changes in crop and P management. Phosphorus-based waste applications decreased predicted P leaching by 20 kg ha^-1 yr^-1 compared to N-based waste applications. Eliminating all P applications decreased the predicted P leaching losses by less than 1 kg ha^-1 yr^-1compared to P-based waste application. Results show that P can continue leaching from P saturated soils even in the absence of P additions.

Back to Conference Program