![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Use of Regional Percent Crop Area Factors in Refined Drinking Water Assessments
US EPA Office of Pesticide Programs (OPP)
Environmental Fate and Effects Division (EFED)
Water Quality Technical Team (WQTT)
July 25, 2003
On this Page
- Background
- Regional PCA Factors
- Use of Regional PCAs in refining estimates of drinking water exposure
- Risk Characterization and Regional Default PCAs
- References
Since 1999, EFED has used Percent Crop Area (PCA) factors in FQPA drinking water exposure assessments to adjust for the cropped area in a drinking watershed. PCA factors, representing the maximum potential cropped area nationally, are available for four major crops (corn, soybean, cotton, wheat) and for all agriculture as a default. For the Organophosphate Cumulative Risk Assessment (OP CRA), EFED used a regional assessment of potential drinking water exposure that incorporated regional PCA factors. EFED used the peer-reviewed procedures accepted by the SAP (OPP SAP, 1999) to develop default (all-agriculture) regional PCA factors. Regional PCAs offer one method for incorporating regional considerations into refined drinking water exposure assessments. Regional default PCAs may be particularly useful for assessments involving minor crops grown predominantly outside the corn/soybean/wheat belt of the midwestern US. Regional default PCAs represent an interim step in moving from national PCAs developed from large watersheds (8-digit Hydrologic Unit Codes) to PCAs developed from actual drinking watersheds.
Background
EFED currently provides estimated drinking water concentrations (DWECs) to the Health Effects Division (HED) for use in its human dietary risk assessment. The potential for pesticide exposure from drinking water can vary widely across the country. Different regions in a pesticide use area will vary in vulnerability to surface water contamination due to variations in soil, weather, and agricultural practices. To account for these differences and to better ensure that the DWECs are sufficiently protective to "cover" the entire use area of a particular pesticide, EFED often simulates application to multiple crops in different regions, depending on registered uses of a specific pesticide. In addition, the crops to which the pesticide is applied will be planted more intensely in particular regions than in others.
Development of Percent Crop Area Factors
In 1999, EFED proposed using Percent Crop Area (PCA) factors in pesticide exposure assessments to account for the fact that a watershed of sufficient size to supply a drinking water source is not likely to be devoted entirely to growing crops. The PCA factors were derived by determining what portion of each watershed (represented by nationally-available 8-digit HUCs; see the section Representing Watersheds below) in the continental United States was devoted to growing crops in general, and what portion of each 8-digit HUC was used to grow certain individual crops. OPP's Science Advisory Panel (SAP) reviewed the proposal in May, 1999, and recommended using PCA factors for four major crops - corn, wheat, cotton and soybeans - based on comparisons with surface water monitoring (OPP SAP, 1999). Given limitations of available watershed scales and crop coverages, the SAP did not believe that individual PCAs were yet appropriate for other individual crops due to concerns with scale, uncertainty of crop distribution within watersheds, and a lack of surface water monitoring with which to ensure that PCA-adjusted residue values in drinking water would still be protective. The SAP recommended that a default PCA be applied to all but the four crops listed above.
EFED instituted the use of PCA factors with the November,1999, FQPA science policy guidance document Estimating the Drinking Water Component of a Dietary Exposure Assessment (OPP, 1999) and further documented in the October, 2000, science policy document, Drinking Water Screening Level Assessment (OPP, 2000). This document identified the maximum PCA for each of the four major crops for any 8-digit HUC in the country, as well as the 8-digit HUC with the greatest overall percent cropped area for use as a default (see Table 1).
| CROP | MAXIMUM PERCENT CROP AREA (as a decimal) |
STATE |
|---|---|---|
| Corn | 0.46 | Illinois, Iowa |
| Soybeans | 0.41 | Missouri |
| Wheat | 0.56 | North Dakota |
| Cotton | 0.20 | Mississippi |
| Default | 0.87 | Iowa (national default) |
EFED's current default PCA of 0.87 is based on the most heavily-cropped 8-digit HUC in the entire United States. The 8-digit HUC with the PCA of 0.87 is located in the Midwest.
Representing Watersheds
Ideally, PCA factors would be derived for watersheds delineated for the more than 6,000 surface water intakes in the US. While efforts are underway to do this, the delineation project is not yet ready for use. In the meantime, EFED used 8-digit HUC watersheds to derive a national coverage of watersheds.
"Hydrologic unit boundaries define the areal extent of surface water drainage to a point" (USDA NRCS, 2003). Watersheds can be defined at different scales, from large basins, such as the entire drainage area of the Mississippi River, to small subwatersheds that feed a first-order stream. The USGS developed a hierarchical system of classifying watershed boundaries based on hydrologic units (Seaber et al, 1987; USDA NRCS, 2003). Figure 1 illustrates this classification system. National coverage of hydrologic unit boundaries is available down to the 4th level, 8-digit subbasins (HUC-8).
Figure 1. National hydrologic unit classification system (USDA NRCS, 2003).
The HUC-8 subbasins are generally larger than 700 square miles in size (Seaber et al, 1987). The SAP noted that watersheds draining into drinking-water reservoirs are generally smaller than the HUC-8 subbasins, and may not provide reasonable estimates of actual PCAs for smaller watersheds (OPP SAP, 1999). In the OP CRA, OPP noted that cropping intensity is variable and smaller watersheds capable of supporting drinking water supplies may have PCAs much different than that represented by the subbasin. An example is the Willamette River Valley (OR), where the HUC-8 PCA in the valley was less than 30%, while USGS reported watersheds in the valley with up to 99% agriculture (OPP, 2002).
Regional PCA Factors
The boundaries for the Regional PCAs correspond to the broadest hydrologic units defined by the USGS. These hydrologic units, classified by the USGS as "regions", divide the United States into 21 major hydrologic areas based on surface topography (Figure 1). Eighteen regions cover the continental United States. These major basins "contain either the drainage area of a major river, such as the Missouri region, or the combined drainage areas of a series of rivers, such as the Texas- Gulf region, which includes a number of rivers draining into the Gulf of Mexico (Seaber, et al, 1987). The natural hydrologic boundaries of these regions extend into Canada and Mexico, but the regions themselves are truncated at the borders for mapping purposes.
Because PCAs are developed for 8-digit HUCs, subunits of the major basins, the region scale (HUC-2) is a logical basis upon which to aggregate regional PCA factors. Figure 2 overlays the boundaries of these major basins on the map of PCA factors originally developed by EFED for the SAP in 1999.
Figure 2: Total agriculture PCA (default PCA) by 8-digit HUCs with major basin overlay. See Table 2 for the legend for basin numbers and associated PCA.
To determine the regional PCA we used the same methodology and rationale used to identify the national-scale PCAs. The largest HUC-8 scale PCA within each Major basin was selected to represent a regional default PCA. The default regional PCA represents the HUC-8-scale watershed which is the most intensively cropped within its respective major basin. We determined regional default PCAs for all agricultural crops which serves as a default PCA for all other crops. Table 2 shows the maximum default PCA factors for each of the 18 major basins.
Table 2: Maximum Percent Crop Area factor (PCA) for each Major Basins
| Basin | Basin Name | Default PCA |
|---|---|---|
| 01 | New England | 14 |
| 02 | Mid Atlantic | 46 |
| 03 | South Atlantic - Gulf | 38 |
| Basin | Basin Name | Default PCA |
|---|---|---|
| 04 | Great Lakes | 77 |
| 05 | Ohio | 82 |
| 06 | Tennessee | 38 |
| 07 | Upper Mississippi | 85 |
| 08 | Lower Mississippi | 85 |
| 09 | Souris - Red - Rainy | 83 |
| 10 | Missouri | 87 |
| 11 | Arkansas - White - Red | 80 |
| 12 | Texas Gulf | 67 |
| 13 | Rio Grande | 28 |
| Basin | Basin Name | Default PCA |
|---|---|---|
| 14 | Upper Colorado | 7 |
| 15 | Lower Colorado | 11 |
| 16 | Great Basin | 28 |
| 17 | Pacific Northwest | 63 |
| 18 | California | 56 |
1 See Figure 1 for basin locations.
Use of Regional PCAs in refining estimates of drinking water exposure
Screening-level estimates of DWECs, developed using FIRST, or refined using PRZM-EXAMS should initially be conducted at a national scale and using national-scale percent cropped area values (NPCAs) listed in Table 1.
If dietary risks are found to be exceeded as a result of the national scale assessment, risk assessors can relatively easily use regional PCAs as an initial refinement of pesticide exposure across the United States or the pesticide use area. This refinement may take the form of added characterization of potential exposure or can be used to adjust DWECs. The most appropriate use of regional PCAs can be best determined in discussions between risk assessors and risk managers. Options are:
Using regional PCA factors instead of uniformly applying the maximum national values may obviate the need for further refinement of the drinking water exposure assessment, by a method that has already passed outside peer review.
Using regional default PCA factors to refine DWECs for uses with limited regional extent, such as Section 18 emergency exemptions, Section 24(c) Special Local Need requests, or crops (such as citrus) with limited, well defined growing regions.
Using regional PCA factors to better distinguish which crop-chemical scenarios in a pesticide use area might lead to the greatest surface-water contamination. For instance, a minor crop grown in the South Atlantic-Gulf basin might have a maximum use rate much higher than the predominant use on corn in the Midwest. However, after application of regional PCA factor to the minor use scenario, the DWEC from the Midwest corn scenario would appear to be the more appropriate to use in the human dietary risk assessment.
Using regional PCAs in association with a pesticide use map to distinguish between possible exposure for the same crop in different regions of the country and to focus spatial modeling refinements in locations for which crop-chemical scenarios are likely to result in the greatest exposure. For instance, PRZM-EXAMS output from a Maine potato scenario might be greater than that from an Idaho potato scenario if the same PCA were applied to both. However, application of the regional default PCAs (14% for Maine and 63% for Idaho) might lead the risk assessor to use the results of the Idaho scenario instead.
Risk Characterization and Regional Default PCAs
Although PCA factors based on 8-digit HUCs have been peer-reviewed by the SAP, it is imperative that risk assessors characterize the assumptions and implications of employing regional default PCAs. While certain uncertainties, such as those of scale and location of cropped areas within a watershed, are common to those connected with use of the national default PCA, some uncertainties are magnified by a regional assessment. For instance:
Urban and turf uses are not included in the PCA factors, which account for agricultural crops only. This could be expected to have a minor effect on the national default PCA of 0.87. However, turf uses may well co-occur with agriculturally cropped land in many Florida watersheds, for instance. The risk assessor should consider if a regional PCA is still useful in scenarios where non-crop uses of a chemical could potentially be used in the same watershed as the agricultural use being modeled. If so, the resultant uncertainty in the regional PCA should be characterized.
Regional assessment introduces national border effects. Basins along the border with Mexico (such as the Rio Grande basin) and Canada (such as the Great Lakes basin) do not include those portions of the watershed outside of the United States. Depending on the intensity of agriculture across the border, the regional PCAs might be underestimating or overestimating the fraction of land which could potentially contribute to pesticide contamination of drinking water.
While watersheds the size of the Great Basins provide logical boundaries for separating transport of pesticides in runoff on the field scale, adjacent basins are in fact hydrologically connected. For instance, the Lower Mississippi Basin receives inflow from the Upper Mississippi, and the Lower Colorado from the Upper Colorado. However, dilution of pesticide residues in such large rivers means that they should add minimal additional residues to those simulated by PRZM-EXAMS to runoff from a small watershed to the Index Reservoir.
Assessment of drinking-water exposure to pesticides on a sub-national scale will help risk managers to focus needed mitigation actions on regions where the greatest exposure is likely to occur. Furthermore, understanding the spatial variability of PCAs can be useful in prioritizing locations for development of future PRZM-EXAMS modeling scenarios. As a more robust set of standard scenarios is established for crops across the country, more routine use of regional PCAs may be possible with better security in the protectiveness of resulting DWEC values.
References
OPP SAP. 1999. Proposed Methods for Determining Watershed-derived Percent Crop Areas and Considerations for Applying Crop Area Adjustments to Surface Water Screening Models. May, 1999, Science Advisory Panel report
[Background Document available at
http://www.epa.gov/scipoly/sap/meetings/1999/may/pca_sap.pdf;
SAP Report available at
http://www.epa.gov/scipoly/sap/meetings/1999/may/final.pdf].
OPP. 1999. Estimating the Drinking Water Component of a Dietary Exposure Assessment. FQPA Science Policy Document. Federal Register Notice: November 10, 1999 (Volume 64, number 217). Electronic copy available at http://www.epa.gov/pesticides/trac/science/.
OPP. 2000. Drinking Water Screening Level Assessment. Part B: Applying a Percent Crop Area Adjustment to Tier 2 Surface Water Model Estimates for Pesticide Drinking Water Exposure Assessments. FQPA Science Policy Document Public Comment Draft September 1, 2000. Federal Register: October 11, 2000 (volume 65, number 197). Electronic copy available at http://www.epa.gov/pesticides/trac/science/.
OPP. 2002. Organophosphate Pesticides: Revised OP Cumulative Risk Assessment . Revised assessment released June 10, 2002. Electronic copy available at http://www.epa.gov/pesticides/cumulative/rra-op/.
Seaber, P.R., F.P. Kapinos, and G.L. Knapp. 1987. Hydrologic Unit Maps. USGS Water-Supply Paper 2294. Second printing 1994. Electronic copy available at
http://pubs.usgs.gov/wsp/wsp2294/pdf/wsp_2294_a.pdf.
USDA Natural Resources Conservation Service (NRCS). 2003. Watershed Boundary Dataset (WBD). Web site
http://www.ncgc.nrcs.usda.gov/products/datasets/watershed/. Last visit: 7/25/03.
Steps Toward Regional Assessment: What We Can Do Now. October, 2002 Issue Paper presented to the EFED Water Quality Tech Team by Kevin Costello.