  SEQ CHAPTER \h \r 1 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON D.C., 20460



OFFICE OF PREVENTION, PESTICIDES AND TOXIC SUBSTANCES

MEMORANDUM	

SUBJECT:	Estimates of Crop Irrigation for Endothall Registration (DP#
369862, 369863)

FROM:                Bill Phillips, II, Ph.D., Agronomist

Biological Analysis Branch 

Derek Berwald, Economist

		Economic Analysis Branch

		Biological and Economic Analysis Division (7503P)

THRU:	               Timothy Kiely, Chief

		Economic Analysis Branch

		Arnet Jones, Chief

		Biological Analysis Branch

Biological and Economic Analysis Division (7503P)

TO:		Sidney Jackson, Risk Manager Reviewer

		Barbara Madden, Risk Manager

		Risk Integration Minor Use Emergency Response Branch 

		Registration Division (7505P)

Product Review Panel:  October 7, 2009   

Summary

EPA is evaluating a proposed use of the aquatic herbicide endothall for
weed control in irrigation canals.  This memo provides estimates for the
percentage of crops that may be exposed to endothall  that EPA is
reasonably certain will not be exceeded. Crops will be exposed to
endothall by irrigation water, so the maximum percent that could be
exposed is the share of production from irrigated land.  

Estimates, summarized in Table 1, are derived using two methods.  Where
data on irrigated and dryland production are available, the estimates
are based on the share of production from irrigated land, which is the
maximum percent of the crop that could be exposed to endothall. 
Estimates using the share of production that is irrigated are provided
for barley, corn, dry edible beans, peanuts, oats, rice, sorghum,
soybeans, sugarbeets, sugarcane, and wheat.  For crops of concern that
do not have available data on irrigated and dryland production, BEAD
uses the share of production grown in 17 western states where irrigation
is used extensively and endothall use will be concentrated. Estimates
based on production in the 17 western states are provided for apples,
grapes, oats, green peas, wheat, watermelon, and strawberries.  

In addition, BEAD reviewed the protocol for the residue trials and
concludes that the resulting data are conservative.  The trials used
applications of treated irrigation water applied through overhead
sprinklers, which would result in high exposure to the edible portion of
the crop.  Many fields are irrigated with drip lines or a flood system,
which would result in less of the treated water, and thus less
pesticide, deposited on the crop.  In the registrant’s field trials
plant samples were taken for residue testing from crops irrigated
overhead using water that was treated the day of harvest; this is not
common practice under actual production conditions.  



Table 1.  Estimated Percent Crop Treated with Endothall for Dietary Risk
Assessment

	Share of Irrigated

Production	Share of 

Production

 in West 1	Estimated Percent Crop Treated for Dietary Risk Assessment

Apple	n/a	65	65

	Fresh market	n/a	78	78

	Processing	n/a	44	44

	Juice	n/a	49	49

	Canned	n/a	14	14

Barley for grain	36	-	36

Corn for grain	19	-	19

Dry Beans 2	32	-	32

Grape	n/a	94	94

	Fresh market	n/a	99	99

	Processing	n/a	94	94

Green Peas	n/a	11	11

Oats for grain	7	37	7

Peanut for nuts	42	-	42

Rice	100	-	100

Sorghum for grain	15	-	15

Soybean for beans	9	-	9

Strawberry	n/a	91	91

	Fresh market	n/a	89	89

	Processing	n/a	100	100

Sugarbeet for sugar	37	-	37

Sugarcane for sugar	54	-	54

Watermelon	n/a	39	39

Wheat for grain	14	70	14

Source:	USDA 2009a, 2009b, 2009c, 2009d, EPA calculations.

1	Arizona, California, Colorado, Idaho, Kansas, Montana, Nebraska,
Nevada, New Mexico, North Dakota, Oklahoma, Oregon, South Dakota, Texas,
Utah, Washington, and Wyoming.  Estimates are only provided for certain
crops.



Background

Endothall is an herbicide registered for control of plants in water
bodies.  It is currently registered for weed control in irrigation
canals, but with the requirement that water cannot be released from the
system or applied to crops for seven days after treatment.  

The Agency is currently evaluating a petition for tolerances on many
irrigated crops to allow registration of endothall for use in irrigation
canals to control aquatic weeds without the requirement that the treated
water be held or contained prior to being used to irrigate a crop. 
Because this is a new use pattern for endothall, dietary risks must be
estimated and analyzed.  Normal methods for estimating the percent crop
treated for a new pesticide are inappropriate in this case, because the
pesticide is not directly applied to the crop.  Exposure occurs via
irrigation water.    

Endothall will be used to control weeds in irrigation systems, not as
treatment for the water in the systems.  Most of the water that goes
through a canal to irrigate crops will not be treated.  Instead,
endothall will be periodically injected into irrigation water to keep
weed populations down.  Because the water in the canal is rapidly turned
over, after several hours most if not all of the endothall will no
longer be in the canal systems.  In addition, the use directions for
endothall restrict applications to irrigation canals to once every seven
days, and a maximum of six times per year.  

Due to the rate of endothall used, the number of applications, and the
timing of applications the residue trials for endothall seem to be
conservative to BEAD.  This supposition is supported given the protocol
used where crops were overhead irrigated six times with an inch of water
containing the maximum concentration of endothall.   After the final
treatment the crop was harvested the same day, resulting in maximum
endothall residue concentrations on the crop.  From one of the
registrants field protocols:  

“[d]uplicate control and treated samples of each commodity were
harvested from the respective tests.  Samples of field corn forage,
sorghum forage and wheat forage and hay were harvested 0 days after the
second or third application (0 DAT).  Samples of sweet corn forage,
kennels plus cob with husks removed (K+CWHR) and stover, field corn
grain and stover, sorghum grain and stover, and wheat grain and straw
were harvested following the sixth application at 0 DAT (or at 1 DAT in
one wheat test).” 

Under actual field production conditions, crop harvest is conducted only
after a specified number of days after treatment. Due to standard
agronomic practices this duration is extended when, for example, wheat
is left to dry in the fields for some time before harvest, to reduce the
moisture content.  While in other cases crops such as grapes are not
overhead irrigated when fruit are present, because additional moisture
on the bunch can lead to fungal diseases.  There is an additional
consideration that should be given here and that is that much of the
crops irrigated from surface water sources are done so by using some
type of delivery system other than overhead sprinklers, systems such as
flood or drip irrigation.

Method for Estimating Percent Crop Treated

For some crops, the estimates in this memo are based on yield and
acreage data for irrigated and non-irrigated crops from the 2007 Census
of Agriculture (USDA 2009d).  The estimates provided will be an upper
bound on the percentage of crops that may have been irrigated with water
from canals that were treated with endothall, because some of the
irrigated crops are not grown in areas where endothall will be used.  

Other crops that are important to the dietary risk assessment do not
have data on irrigated and non-irrigated production available.  For
these crops, this memo provides estimates of the amount of the crop
grown in 17 western states where endothall use is expected.  These are
states where substantial amounts of water are transported through open
canal systems to be used for irrigation.  It is these systems where weed
control is important, and endothall is likely to be used.      

Estimates from the 2007 Census of Agriculture

For select crops, the 2007 Census of Agriculture (USDA 2009d) provides
estimates of the acres and yield from farms with irrigation, farms
without irrigation, and farms where some of the crop is irrigated. 
Using this information we provide estimates of the share of production
irrigated for those crops where the data are available. These crops are
barley, corn, dry edible beans, oats, peanuts for nuts, rice, sorghum,
soybeans, sugarbeets, sugarcane, and wheat.   

Table 2 shows the estimated percent of the crop that has been irrigated
for the food crops for which data are available.  To estimate the
production from the irrigated acres, the first step is to estimate the
crop production from irrigated land.  Using the data in Table 2, we
multiply the irrigated yield by the irrigated acreage, which is the
acreage from farms where the entire crop is irrigated plus the irrigated
acreage from partially irrigated farms.  To get an estimate of the share
of the crop irrigated, we divide this by an estimate of total
production.  To estimate total production from the information in Table
2, we multiply the yield by the acreage for each type of farm, and add
these together.  Taking wheat as an example, estimated production from
irrigated farmland is (1,806,902 + 1,557,177) × 80.3 = 270,135,544
bushels, which is the numerator for our percentage calculation.  Total
estimated production is 1,806,902 × 80.3 + (1,557,177 + 3,703,599) ×
42.7 + 43,865,291 × 37.0 = 1,992,745,133 bushels.  Dividing estimates
for production from irrigated land by the estimate of total production
yields an estimate of the share of production that has been irrigated,
in this case 14%.  The estimates for all the food crops with available
data in the census are in the last column of Table 2.  

Note that the data available in the Census of Agriculture only provides
overall yields for farms where only part of the crop is irrigated,
although it does provide total production and irrigated and dryland
acreage data.  For this analysis, BEAD assumed the yield from the
irrigated acreage on partially irrigated farms to be equal to the
average yield per acre from irrigated farms.  This is a conservative
assumption, because the implied non-irrigated yield is substantially
lower than from dryland farms in all but one case, which tends to
inflate the estimates of the share of crop irrigated.  

These estimates are based on only one year of data.  However, it is
unlikely that the amount of acreage irrigated will change a large amount
from one year to the next, because irrigation equipment is an expensive
capital investment.  

Note that these are estimates of the amount of the crop irrigated, not
the amount of the crop that may have endothall residues.  Not all of the
crop that is irrigated will be grown in areas where endothall is likely
to be used, and even in the West, where endothall use for irrigation is
expected, not all systems will necessarily be using endothall, a new
method of control.  Therefore, the share of crops that are actually
exposed to endothall are likely to be lower than the estimates of
irrigated acreage in Table 2.    



Table 2.  Share of Crop Production Irrigated	 	 	 	 	 	 	 	 	 
 	 	 

	Entire crop irrigated

Part of crop irrigated

None of crop irrigated

EPA Estimated Share of Production Irrigateda

	 	 	Average Yield Per Acre

 	Acres	Acres not	Average Yield Per Acre



Average Yield Per Acre



Crop	Farms	Acres

	Farms	irrigated	irrigated

	Farms	Acres



	Barley for grain (bushels)	3,566	605,340	100.0

675	136,135	155,262	53.3

15,607	2,625,220	49.9

36%

Corn for grain (bushels)	17,927	6,103,769	180.0

20,984	7,053,000	6,435,486	150.0

308,849	66,656,287	144.3

19%

Dry edible beans, excluding limas (cwt)	2,275	317,193	22.6

181	36,310	50,432	16.5

3,780	1,051,614	15.9

32%

Oats for grain (bushels)	1,304	67,948	83.7

316	10,266	15,560	63.4

40,938	1,415,375	58.1

7%

Peanuts for nuts (pounds)	1,301	251,564	3722.5

1,080	167,225	186,723	3234.1

3,801	595,052	2725.8

42%

Rice (cwt)	6,084	2,758,792	72.0

-	-	-	-

-	-	-

100%

Sorghum for grain (bushels)	2,092	443,599	86.9

2,391	401,615	670,030	74.7

21,759	5,254,590	69.2

15%

Soybeans for beans (bushels)	7,007	2,175,069	45.3

13,326	3,062,006	3,396,716	40.8

258,777	55,282,030	40.2

9%

Sugarbeets for sugar (tons)	1,535	387,224	30.0

60	7,451	20,424	24.2

2,427	838,718	23.4

37%

Sugarcane for sugar (tons)	210	431,796	38.7

9	3,409	3,373	33.3

473	408,088	34.7

54%

Wheat for grain, all (bushels)	7,695	1,806,902	80.3	 	7,518	1,557,177
3,703,599	42.7	 	145,597	43,865,291	37.0	 	14%

Source:  USDA 2009d and EPA Calculations

a Assumes irrigated acres on farms where only part of the crop is
irrigated obtain the same yield as acres on farms where the entire crop
is irrigated





Crop Estimates 

Endothall residues on some crops are of particular concern to EPA based
on residue trials and dietary habits.  These crops are peas, watermelon,
strawberries, apples, oats, grapes and wheat.  For these crops BEAD
provides estimates of the proportion of the crop that is grown in the 17
western states, which are the states expected to be users of endothall
in irrigation water.  According to the Farm and Ranch Irrigation Survey
(USDA 2004), almost 70% of the irrigated acreage in the United States is
in the 17 western states, and about 87% of the irrigation water applied
to crops is in these western states, which correspond to the states
served by Bureau of Reclamation water projects.  

The share of the crop grown in the 17 western states can be used as a
crude measure of the share of the crop that may have come in contact
with endothall.  These estimates are the best that can be made with the
available data, and they likely overestimate the crop that will
encounter endothall treated water, because not all production in the
western states will be irrigated with water from canals that have been
treated with endothall. Not all irrigation systems will use endothall,
some crop production is irrigated with wells that do not require weed
control, and some crops such as wheat and oats may be farmed without
irrigation, even in the arid West.   When possible (wheat and oats), the
estimates provided  in Table 2 above, based on the 2007 Census of
Agriculture, are preferable.  

The 17 western states are Arizona, California, Colorado, Idaho, Kansas,
Montana, Nebraska, Nevada, New Mexico, North Dakota, Oklahoma, Oregon,
South Dakota, Texas, Utah, Washington, and Wyoming.  It is in the West
that large scale water delivery systems are important to agriculture;
these are the systems that will use endothall for irrigation, as they
are the systems that currently use acrolein or xylene, aquatic
herbicides for which endothall may substitute.  Agriculture in the East,
where it is less arid, does not require large scale water transfers. 
BEAD does expect to see endothall use in the East, but primarily for
drainage systems rather than irrigation water delivery systems where
water will be applied to crops. 

Where available, estimates of the crop grown in the 17 western states
are provided for different end-uses of the particular crop.  This
information can be useful for dietary risk estimates, which are based on
consumption of specific food products.  The estimates of the amount
grown in the 17 western states are provided in Table 3.   

Peas

Only two western states are important producers of green peas for
processing, which can appear in baby foods.  In 2008, about 103,000 tons
were produced in Washington State, and about 33,000 tons in Oregon. 
Production in these two states was about 11% of US production of almost
412,000 tons (USDA 2009a).  

Watermelon

Four western states produce enough watermelon to show up in the USDA
data, Arizona, California, Oklahoma, and Texas.  California and Texas
produced about 6,100,000 cwt. and 6,000,000 cwt. in 2008, respectively. 
Arizona produced almost 2,900,000 cwt., and Oklahoma only produced about
360,000 cwt.  Together, the almost 15,400,000 cwt. of watermelon
represented about 39% of US production of about 39,600,000 cwt. in 2008
(USDA 2009a).  

Strawberries

California is the major strawberry producing state in the United States.
 In 2008, California produced almost 23,000 cwt. of strawberries, and
Oregon another 238 cwt., the two together totaling about 91% of the
total US production.  About 89% of the fresh market strawberries come
from the western states of California, Washington and Oregon, and almost
all the processing strawberries come from the same three states,
accounting for almost all strawberries used for processing (USDA 2009b).
 

Apple

Available data on apple production covers the production for the western
states of Arizona, California, Idaho, Oregon, Utah, and Washington, with
Washington being by far the largest producer.  Combined these states
accounted for about 65% of total US apple production in 2008, with over
59% of the total coming from Washington alone (USDA 2009b).  

Oats

Data on oat production is available for 13 of the seventeen western
states, with South Dakota being the largest producer in the West, and
3rd in the country, with about 8.8 million bushels.  In all, the western
states produce about 37% of the nation’s oats (USDA 2009c).  

Grapes

California was the largest producer of grapes in 2008, accounting for
over 89% of total production.  The 17 western states combined accounted
for about 94% of total production.  Looking at the end uses of these
grapes, the picture is similar.  About 99% of fresh grapes were from the
western states in 2008, and about 94% of the processed grapes.  Of the
grapes used for processing, California was the most important producer
of wine grapes (about 99%) and seems to account for all dried grapes
(USDA 2009b). 

Wheat

Wheat is widely grown in western states, with all 17 states appearing in
the available production data.  Together, the 17 western states account
for about 70% of the total wheat production.  

Table 3.  Share of Production of Specific Crops in the 17 Western
States, 2007

	US Production	Western States Production	Western States Share

Green Peas (Tons)	411,780	43,666	11%

Watermelon (1,000 cwt.)	39,551	15,387	39%

Strawberries (1,000 cwt.)	25,317	22,913	91%

Fresh Market 	20,911	18,653	89%

Processing 	4,406	4,404	100%

Apple (Million Pounds)	9,769	6,394	65%

Fresh Market 	6,304	4,911	78%

Processing 	3,372	1,500	44%

Juice 	1,534	750	49%

Canned 	1,191	170	14%

Oats (1,000 Bushels)	88,635	32,565	37%

Grapes (Tons)	7,303,260	6,887,000	94%

Fresh Market 	985,200	976,050	99%

Processing 	6,304,350	5,944,650	94%

Wheat (1,000 Bushels)	2,499,524	1,752,810	70%

Source:  USDA 2009a, 2009b, 2009c, EPA Calculations

The 17 Western states are Arizona, California, Colorado, Idaho, Kansas,
Montana, Nebraska, Nevada, New Mexico, North Dakota, Oklahoma, Oregon,
South Dakota, Texas, Utah, Washington, and Wyoming.



Suggested Federal Register Language 

EPA is establishing tolerances on multiple commodities to support the
application of the aquatic herbicide endothall to be used in irrigation
canals without a holding period.    

For a new agricultural pesticide use, EPA typically estimates percent
crop treated by comparison with the amount of use of other pesticides
for the same crop or site.  That approach is inappropriate for the new
use for endothall, because the use is on irrigation canals rather than
crops and EPA does not have data on the frequency of use of aquatic
herbicides on irrigation canals.    

Instead, EPA has estimated percent crop treated for endothall by
estimating the percent crop irrigated which serves as an upperbound for
crops that may be exposed to endothall in irrigation water.  EPA used
two methods to estimate percent crop irrigated.  The preferred method,
used where data on irrigated production are available, is an estimate of
the share of total production that is irrigated.  Estimates from this
method are provided for barley, corn, dry edible beans, oats, peanuts,
rice, sorghum, soybeans, sugarbeets, sugarcane, and wheat. Where data on
irrigated production are not available, EPA estimated the percent crop
irrigated by determining the percentage of United States production of a
crop that is grown in 17 western states where endothall may be used. 
The 17 western states are Arizona, California, Colorado, Idaho, Kansas,
Montana, Nebraska, Nevada, New Mexico, North Dakota, Oklahoma, Oregon,
South Dakota, Texas, Utah, Washington, and Wyoming.  These states are
the states where large scale water projects predominate, and where other
chemicals are used in canals for weed control.  These types of
irrigation projects are relatively rare in other parts of the country.

Use of these estimates in the exposure assessment is conservative,
because it is the equivalent of assuming 100% of irrigated crops have
irrigated with water from endothall-treated canals.

In fact, even in areas with surface water delivery systems, all
irrigation canals may not be treated with endothall.  Additionally, some
crops, even in the heavily irrigated areas of the West, are not
irrigated, such as dryland grain production.   

References

USDA 2004.   US Department of Agriculture, National Agricultural
Statistics Service, The Census of Agriculture, Farm and Ranch Irrigation
Survey (2003) Volume 3, Special Studies Part 1.  November 2009.  
Available here:
http://www.agcensus.usda.gov/Publications/2002/FRIS/index.asp

USDA 2009a.  US Department of Agriculture, National Agricultural
Statistics Service, Vegetables 2008 Summary.  January 2009.  Available
here:   HYPERLINK
"http://usda.mannlib.cornell.edu/usda/current/VegeSumm/VegeSumm-01-28-20
09.txt"
http://usda.mannlib.cornell.edu/usda/current/VegeSumm/VegeSumm-01-28-200
9.txt 

USDA 2009b.  US Department of Agriculture, National Agricultural
Statistics Service, Noncitrus Fruits and Nuts 2008 Summary.  July 2009. 
Available here:   HYPERLINK
"http://usda.mannlib.cornell.edu/usda/current/NoncFruiNu/NoncFruiNu-07-0
8-2009.pdf"
http://usda.mannlib.cornell.edu/usda/current/NoncFruiNu/NoncFruiNu-07-08
-2009.pdf 

USDA 2009c.  US Department of Agriculture, National Agricultural
Statistics Service, Crop Production 2008 Summary.  January 2009. 
Available here:   HYPERLINK
"http://usda.mannlib.cornell.edu/usda/current/CropProdSu/CropProdSu-01-1
2-2009.txt"
http://usda.mannlib.cornell.edu/usda/current/CropProdSu/CropProdSu-01-12
-2009.txt 

USDA 2009d.   US Department of Agriculture, 2007 Census of Agriculture
United States Summary and State Data Volume 1, Geographic Area Series,
Part 51.   February 2009, updated September 2009.   Available here:
http://www.agcensus.usda.gov/Publications/2007/Full_Report/Volume_1,_Cha
pter_2_US_State_Level/st99_2_032_032.pdf



 Registrant submission: 47520713.  Arsenovic, M. (2008) Endothall
(Hydrothol 191): Magnitude of the Residue on Grain Cereal Group (Except
Rice): Lab Project Number: Z9768. Unpublished study prepared by
Interregional Research Project No. 4.  590 pages.  

 The exception is peanuts, where the implied yield estimate is slightly
higher than the dryland yield.  The difference is less than 3%.  

 PAGE   

