PROCESSING UNITS AND STEPSNHD+ catchment used as the spatial unit of analysis

• There are a large number of catchments (~2.6 million).• Each has a single water body (the stream reach) and the outflow at the

catchment outlet reflects the direct runoff from the entire catchment.• The entire US land area is accounted for.• NHD+ catchments comprise a range of areas highly relevant to farming

practices at the local scale.• Generally small and of a scale relevant to local farming activities• 90% are 2.4 mi2 (620 ha) or smaller• 50% are <= 0.56 mi2 (145 ha)

RESULTSBACKGROUND AND OBJECTIVESPesticides are used on numerous agricultural crops across the US to control pestsand improve food yield and quality. The study conducted by the PyrethroidWorking Group characterized the potential for pyrethroids to enter flowingsurface waters based on a spatially explicit analysis of crop proximity to surfacewaters using multi-year data on diverse agricultural production patterns.

Standard exposure calculations in the USEPA EFED regulatory risk assessmentframework assume that 100% of the area around the water body is cropped andtreated, and therefore subject to drift and runoff entry. However, at the real worldscale at which growers operate, these conservative simplifying assumptions arenot realistic in most use areas. This study used government spatial cropping andhydrology data to estimate catchment-scale composition by crop Percent CropArea (PCA) and crop proximity to flowing water bodies.

The resulting datasets provide a useful set of metrics across various croppingpatterns which can be applied to pesticide exposure assessments that may needspatially-explicit refinements to crop-water interactions. Because the cropproximity results are linked to the NHD+ framework, including these specificattributes into other NHD+-based analyses is extremely efficient.

SPATIAL AND TEMPORAL VARIATIONS OF NATIONAL CROPPING PATTERNS FOR HIGHER-TIER PESTICIDE EXPOSURE ASSESSMENTChristopher M. Holmes1, Joshua Amos1, Paul Hendley2, Russell Jones3, Scott Jackson4

1 Waterborne Environmental, Inc.; 2 Phasera Ltd.; 3 Bayer CropScience; 4 Valent USA

Intersect with crop

DATASETSUSDA NASS Cropland Data Layer (CDL) 2008-2012• Derived from satellite and ground training data• Produced annually for entire country• Over 50 crop classes

USGS National Hydrography Plus (NHDPlus) • Contains over 2.6 million stream segments and

associated drainage areas• Provides framework for national analysis• Allows results to be linked to other data using

same framework (e.g., USGS StreamCat with over 500 environmental & ecological metrics)

USDA Census of Agriculture (2012)• Accurate quantification of crop harvested acres• Assist in defining appropriate cropping regions

PWG CDL Crop Group

US EPA Scenario Crop

of Interest

Selected Spatial Extent

2012 USDA Census

Harvested (acres) (crop

group)

CDL 2012 (acres) (crop

group)

5-yr Composite

(acres) (crop group)

Hybrid (acres)

CDL 2012 % of 2012

Survey

5-yr composite % of 2012

Survey

Hybrid %

Year(s) selected

% of National

Acres Repres-ented

Alfalfa Alfalfa National 16,710,820 16,165,805 38,176,095 97% 228% 2012 100%Citrus Citrus FL 539,908 976,906 1,872,681 181% 347% 2012 61.3%Corn Corn National 94,816,833 95,651,409 194,904,058 101% 206% 2012 100%

Cotton Cotton National 9,384,080 13,451,958 25,452,190 143% 271% 2012 100%Grapes Grape NY 39,216 28,879 67,778 74% 173% Composite 3%

Grass Seed Grass Seed OR 420,767 420,1222 920,297 99% 218% 20132 39.1%Peanuts Peanuts GA, FL, AL 1,146,206 1,368,424 3,214,760 119% 280% 2012 70.7%

Soybeans Soybeans National 76,104,385 75,243,102 175,900,965 99% 231% 2012 100%

Sunflower Sunflower National 1,876,890 1,594,947 6,584,195 2,676,850 85% 351% 143% Hybrid1 100%

Sweet Corn Sweet Corn National 571,611 301,398 1,129,230 53% 198% Composite 100%

Tree Fruit Apples WA, NY, MI, NC, PA 384,237 557,432 1,075,285 585,389 145% 280% 152% Hybrid1 56%

Tree Nuts Almonds CA 1,496,610 1,704,659 3,129,560 114% 209% 2012 100%Pecans GA, TX, NM 330,314 322,999 726,851 453,758 98% 220% 137% Hybrid1 57.1%

Vegetables / Ground Fruit

Lettuce CA 1,101,487 625,473 1,367,047 57% 124% Composite 72.0%

Onion CA " " " " " " 26.5%Tomato CA " " " " " " 74.2%Pepper FL 239,277 57,657 140,655 24% 59% 2012 16.6%

Strawberry FL " " " " " " 16.9%

Potato CO, ID, ME, WA, WI 1,779,460 1,519,791 4,595,914 85% 258% 2012 60.0%

Potato ME 112,508 123,434 357,025 110% 317% 2012 5.3%Wheat, Spring Wheat, Spring ND, MT, MN,

SD, OR, WA, ID 12,065,566 13,872,640 16,561,447 115% 137% 2012 99.2%

Wheat (all) Wheat National 49,038,649 55,399,808 134,930,046 113% 275% 2012 100%

1A hybrid of 2012 and 5-yr composite was used. Sunflower: ND & SD were 2012, all other states composite. Pecans: GA was 2012 and TX, NM composite. Apples: PA was composite, all other states 2012.2Because it was processed at a later date, 2013 CDL was available and used for grass seed in Oregon.

Summary of spatial extent and year of CDL for each crop based on national examination of CDL and USDA Census of Agriculture acreages. Values in black were selected for use.

Tier II scenarios estimate the runoff, erosion and chemical mass leaving a 10-ha field entering a 1-ha pond, 2 m deep. It is assumed the entire drainage area for the pond is 100% cropped and treated. Drift that occurs assumes a wind blowing directly towards the pond for every application in a season. The crop is planted and treated every year for 30 years.

Repeat for all crops for each of 5 years of CDL

REFINE TIER II SCENARIO EECS WITH CROP-SPECIFIC PCA PROBABILITY DISTRIBUTIONS

Percent Crop Area values by year50th %ile

75th %ile

90th %ile

95th %ile

99th %ile

Max

Cotton 2008 1.2 6.3 18.2 29.5 58.9 100.0Cotton 2009 1.0 5.9 17.4 28.4 56.0 100.0Cotton 2010 0.46 4.0 16.3 28.0 56.5 100.0Cotton 2011 0.76 7.0 22.8 35.9 64.7 100.0Cotton 2012 0.85 6.4 19.7 31.4 61.2 100.0

Cotton 5-Yr Composite 0.87 7.4 28.1 44.5 76.5 100.0

Zone 25th

50th

75th

90th

95th

99th

Max 0-10m Zone 0.000 0.000 0.000 3.2 12.0 51.4 100.0

10-50m Zone 0.000 0.000 0.513 6.0 15.9 52.8 100.0 50-200m Zone 0.000 0.220 2.8 14.1 26.4 59.0 100.0 10-200m Zone 0.000 0.219 2.7 13.6 25.5 57.8 100.0

Entire

0.121 0.847 6.4 19.7 31.4 61.2 100.0

Proximity Zone Percent Crop Area

Cotton PCA percentiles of different proximity zones (2012)

Cotton PCA percentiles for 10-200m Proximity Zone across five years (2008 – 2012)

To address the assumption of 100% cropping, the distribution of PCA values for each crop was applied to the 30 years of annual maxima EECs using a probability approach. • Potential runoff & erosion loading in each

catchment assumed to be proportional to PCA • PCA distributions binned into 10 groups

(0-25, 50, 75, 87.5, 90, 93, 96, 98, 99, 100%) and used highest PCA in each group

Comparison of PCA percentiles for 10-200m Proximity Zone across multiple crops

Refined EEC distributions after applying PCA probabilities to MS and TX cotton scenarios

SELECTION OF SPATIAL EXTENT AND YEAR OF CDL USING CENSUS OF AGRICULTURE

• PCA fraction for each group was multiplied by the crop-specific maximum annual EEC for each year• Scaled the contribution of each modified yearly maximum EEC in the 10 groups of 30-yr

distributions by group’s occurrence probability (i.e., their percentile range; e.g., an occurrence probability of 0.25 would be assigned for the group defined as 50th to 75th percentile)

• Results in distribution of 300 water-body-year EECs (30 yrs x 10 PCA groups x group probability)• Retains the maximum EECs in Tier II scenario, simply reduces likelihood of occurrence• Results show assumptions used in EPA tier II exposure estimation under FIFRA overestimate the

exposures that will occur from pesticide use in real world catchments due to PCA and proximity

USEPA STANDARD TIER II AQUATIC EXPOSURE

For more information, contact holmesc@waterborne-env.com

10-ha field, 100% cropped and treated

Wind is always toward the pond

Buffer streams by selected distances in GIS