  SEQ CHAPTER \h \r 1 

   UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON, D.C.  20460

OFFICE OF

PREVENTION, PESTICIDES, AND

TOXIC SUBSTANCES

DP Barcodes: D347778, 

PC Code: 090601

MEMORANDUM							 December 19, 2007

SUBJECT:  Summary Evaluation of Recently Submitted FMC Avian Studies

TO:		Jude Andreasen, Risk Manager Reviewer

		Susan Lewis, Risk Manager

Special Review and Reregistration Division (7505P)

FROM:	Christopher J. Salice, Ph.D., Biologist

Environmental Risk Branch IV 

Environmental Fate and Effects Division (7507C)

THROUGH:	  SEQ CHAPTER \h \r 1 Elizabeth Behl, Branch Chief

		Edward Odenkirchen, Senior Biologist

		Environmental Risk Branch IV

		Environmental Fate and Effects Division (7507C)

Background:

Although the carbofuran IRED was signed in 2006, the registrant, FMC,
submitted four studies on the response of avian species to carbofuran
exposure between May and July 2007 for OPP consideration.  FMC’s
intent was to provide the Agency with additional information for use in
probabilistic models to reduce some uncertainties OPP had identified
regarding the estimates of the magnitude carbofuran-induced effects on
birds in agro-ecosystems.  There are no available EPA guidelines for the
submitted studies.  One study addressed the inhibition and recovery of
brain cholinesterase activity following dosing with carbofuran.  Another
study was intended to provide insight into the avoidance behavior of
birds when exposed to carbofuran in feed.  The last two studies were
submitted to provide information on the effect of a food matrix on the
acute toxicity of carbofuran.  EFED has completed scientific review of
the submitted studies and has classified all four studies as
SUPPLEMENTAL.

Brain Acetylcholinesterase Inhibition and Recovery in Bobwhite Quail

Study 1: Brewer, L.W., Stafford, J.M., and Moore, D. 2007. 
Determination of the Time Course of Brain Cholinesterase (ChE) Activity
Depression and Recovery in Northern Bobwhite (Colinus virginianus)
Following Scheduled Oral Dosing with Furadan 4F. Springborn Smithers
Laboratories, 2900 Quakenbush Rd., Snow Camp, North Carolina 27349.
Springborn Smithers Study Number 282.4110 and FMC Study Number
A2007-6201. FMC Corporation, P.O. Box 8, Route 1 and Plainsboro Rd.
Princeton, NJ 08543.  June 7, 2007.  Unpublished report. MRID 47107601

The objective of this study was to determine the time course of brain
cholinesterase (ChE) activity depression and recovery in northern
bobwhite quail (Colinus virginianus) following exposure to carbofuran
via scheduled oral dosing with Furadan 4F.  The experimental protocol
was not based on any procedures or guidelines specified by the EPA or
the OECD; guidelines for this type of study do not currently exist.  The
following are key aspects of the protocol:

Ten replicate groups, containing 10 birds each (5 males and 5 females),
were established for dosing at each of the three dose level (0.75, 1.50
and 3.0 mg a.i./kg body weight; 300 birds total). 

A single control group of 10 birds received no dosing with the test
substance, but was administered an equal dose of distilled water.  

The treatment replicates were assigned to respective time steps, which
indicate the amount of time allowed to pass after dosing before a given
bird was sampled for brain ChE analysis.  The time steps for sampling
were 5 min., 15 min., 30 min., 60 min., 90 min., 120 min., 150 min., 180
min., 240 min., and 360 min.  

Within a given time-step group and dose level, birds were dosed at
5-minute intervals to allow time between birds for euthanization and
head removal (and freezing) after dosing.  

Once all dosing was complete, frozen brains were shipped and analyses
for total ChE conducted. 

 The percent survival of bobwhite quail following dosing until
euthanization of the control, 0.75, 1.5 and 3.0 mg a.i./kg treatment
groups was 100, 100, 99 and 63%, respectively.  Thirty-seven of the 100
bobwhite dosed with 3.0 mg a.i./kg body weight died while the remaining
birds at this dose level survived.  All the deaths occurred between the
30- and 360-minute time-step.

A negative exponential growth model was successfully fit to the recovery
phase for each of the three dosing treatments (p >> 0.05 in the G test).
 Rearrangement of the fitted negative exponential growth models for time
to 50% recovery indicated half-lives of 1.1, 2.9, and 4.4 hours for the
0.75, 1.5 and 3.0 mg a.i./kg bw treatments, respectively.  We conclude
from this study that the rate of ChE recovery is inversely related to
carbofuran dose.  These estimates of half life are based on recovery to
brain ChE activity levels that actually exceed activity levels seen in
control birds; therefore, study results may overestimate half life. 
Estimates of half-life were independently calculated by EPA using Excel
and a logarithmic curve fit; half-life estimates were similar to those
reported by the study-author (see below).  The time to peak ChE
inhibition is approximately 30 minutes and is consistent across all
tested doses.  The mean ChE level at which dosed bobwhite quail died was
1.24 (SD 2.2) (moles/min/g brain tissue.  Birds at the 0.75 and 1.5 mg
ai/kg dose levels showed recovery of ChE activity that exceeded the
control activity levels (> 12 (moles/min/g brain tissue); birds dosed
with 3.0 mg ai/kg, however, did not on average recover to control levels
during the course of the study.  Using the study author’s regression
equation, the estimated time to complete recovery for birds in the high
dose is 10.9 hours.  

Based on our review we conclude that the study is scientifically sound
and provides useful insight into the mechanism of carbofuran ChE
inhibition and subsequent recovery.  However, because there are no
guidelines for avian cholinesterase studies this study is classified
SUPPLEMENTAL.  

Results Synopsis:

Bobwhite Quail endpoints

	Mortality

37% mortality in high dose group (3.0 mg/kg)

1% mortality in medium dose group (1.5 mg/kg)

All deaths occurred between the 30 and 360 minute time steps

	Brain cholinesterase inhibition and recovery

		Time to 50% recovery (1/2 life) (7.4 µmoles/min/g brain)
-study-author determined

High dose (3.0 mg/kg):			4.4 hours

Medium dose (1.5 mg/kg):		2.9 hours

Low dose (0.75 mg/kg):			1.1 hours

		Time to 50% recovery (1/2 life) (7.4 µmoles/min/g brain) -EPA
determined

High dose (3.0 mg/kg):			4.9 hours

Medium dose (1.5 mg/kg):		2.2 hours

Low dose (0.75 mg/kg):			1.1 hours

		Time to 100% recovery (12 µmoles/min/g brain) - based on study-author
analysis results

High dose (3.0 mg/kg):			10.9 hours

Medium dose (1.5 mg/kg):		5.0 hours

Low dose (0.75 mg/kg):			2.9 hours

		Time to 100% recovery (12 µmoles/min/g brain) - EPA determined

High dose (3.0 mg/kg):			19.8 hours

Medium dose (1.5 mg/kg):		5.8 hours

Low dose (0.75 mg/kg):			3.1 hours

 

Avoidance of Carbofuran Treated Feed by Mallard Ducks

Study 2: Stafford, J.M. 2007.  Assessment of Mallard Duck (Anas
platyrhynchos) avoidance to feed containing Furadan 4F. Springborn
Smithers Laboratories, 2900 Quakenbush Rd., Snow Camp, North Carolina
27349. Springborn Smithers Study Number 282.4111 and FMC Study Number
A2007-6202. FMC Corporation, P.O. Box 8, Route 1 and Plainsboro Rd.
Princeton, NJ 08543.  May 15, 2007.  Unpublished report.  MRID 47128701

The objective of this study was to evaluate mallard duck avoidance to
feed treated with Furadan 4F (active ingredient, carbofuran).  The study
employed a two-feed choice (treated and non-treated diet) test design
conducted with six concentrations of active ingredient (a.i.) mixed into
the diet of six respective treatment groups.

The experimental protocol was not based on any procedures or guidelines
specified by the EPA or the OECD; guidelines for this type of study do
not currently exist.  The following are key aspects of the protocol:   

Six treatment groups, each containing 10 birds (5 males and 5 females),
were individually caged and offered feed in two feed pans per cage; one
pan contained feed mixed with carbofuran and the other pan was feed only
(no carbofuran).  

A control group with 10 birds (5 males and 5 females) was also
maintained throughout the study; no carbofuran was offered to these
birds at any time. 

The levels of carbofuran in feed for the six treatments were 1, 3, 5,
10, 30, and 135 mg ai/kg feed which corresponded to mean measured
concentrations in feed of 0.67, 2.2, 3.9, 9.6, 30 and 145.4 mg ai/kg
feed.  

Carbofuran was mixed in feed from a stock solution in distilled water. 
At each treatment level, ducks received treated feed in one feed tray
and untreated feed in a second feed tray to provide a choice of diet;
the location (left or right side) were alternated daily.  

Control birds received a single feed choice, provided in two separate
feeders to mimic conditions in the treatment cages.  

Exposure occurred for 5 days (five 24-hour intervals).  

Non-treated feed was then provided to each cage in a single feeder for
three additional days.

Observations of birds were taken during acclimation and throughout the
study, including the post-treatment recovery period.  Body weights were
measured for each animal to the nearest 0.1 g on the morning of day
–18, -12,, -11 (begin acclimation), -5, -4, -3, -2, -1, time 0
(morning of study initiation, prior to feeding), at the end of day +1,
+5 and +8.  Body weights on days –11, -1, 0, 1, 5, and 8 were used to
calculate proportional change in body weight over time in each group. 
Daily feed consumption was measured in all groups from day –14 to day
+8.  Feed consumption measurements for days –3 through +8 were used in
statistical analyses and calculations.    

No mortality or clinical signs of toxicity were observed in mallards
exposed for 5 days to 0.67, 2.2, 3.9, 9.6, 30 and 145.4 mg ai/kg feed. 
However, birds in the 3.9, 9.6, 30 and 145.4 mg ai/kg feed treatments
had significant reductions in total (treated and untreated) feed
consumption (Fig 3).  The 50% food avoidance concentration (FAC50),
defined as the concentration at which test animal consume equal amounts
of treated and untreated feed was 10 mg ai/kg feed in this study.  Food
intake rates (FIR) and body weight measurements were used to calculate
food repellency factors (RF) for all test groups.  However, data suggest
that from day –1 the test birds developed habits during the
acclimation period that led to favouritism toward one feeder (side) or
another (mean group RF ranging from 0.17 to 1.23).  

 

Figure 3.  Day 5 feed total feed consumption .

Based on our review we conclude that the study provides useful but
limited insight into avoidance behaviour of mallard ducks to
carbofuran-treated feed.  The results indicate that avoidance of
carbofuran-contaminated feed cannot be entirely precluded; a more
refined test is needed to better estimate avoidance behaviour.  A
refined study would include those study attributes characterized by an
OECD workgroup on avian avoidance studies.  Some important
considerations include, creating a hunger-stressed condition in birds,
using pens instead of cages (to more closely resemble reality) and
providing food in a non-concentrated source (distributing feed
throughout the pen). Because there are no guidelines for this study and
because of general issues related to study design it is classified
SUPPLEMENTAL.  

Although the study is of limited utility for avoidance, per se, it
provides some insight into the effect of carbofuran on feed consumption.
 It is apparent from Fig. 3 that there is an effect of carbofuran on
total feed consumption.  Fig. 4 shows individual data points and a
preliminary attempt at fitting a line to the data.  We recognize that a
linear regression is perhaps inappropriate for these data, given the
wide scatter; it may be difficult to find a regression curve that
explains more than 17% of the variation (note the r-squared). 
Statistical analysis of these data does indicate that the slope is
significantly different than zero; however, the fit statistics suggest
that factors other than carbofuran dose may contribute to the
relationship.

 

Figure 4.  Day 5 feed reduction vs. Estimated carbofuran dose.

Effect of Food Matrix on Carbofuran Acute Toxicity

Study 3:  Stafford, J.M.  2007.  Assessment of the Differential Toxicity
of Carbofuran to Northern Bobwhite When Dosed With a Single Aqueous
Bolus Versus the Same Dose Mixed in Feed, Springborn Smithers
Laboratories Protocol No.: 030907/comparative/tox/ bobwhite/Furadan 4F. 
Springborn Smithers Laboratories, 2900 Quakenbush Rd., Snow Camp, North
Carolina 27349. Springborn Smithers Study Number 282.4113 and FMC Study
Number A2007-6204-01. FMC Corporation, P.O. Box 8, Route 1 and
Plainsboro Rd. Princeton, NJ 08543.  May 15, 2007.  Unpublished report.
MRID 47152901

Study 4:  Stafford, J.M.  2007.  Assessment of the Differential Toxicity
of Carbofuran to Mallard Ducks When Dosed as a Single Aqueous Bolus
Versus the Same Dose Mixed in Feed, Springborn Smithers Laboratories
Protocol No.:0309207/comparative/tox/ mallard/Furadan 4F.  Springborn
Smithers Laboratories, 2900 Quakenbush Rd., Snow Camp, North Carolina
27349. Springborn Smithers Study Number 282.4112 and FMC Study Number
A2007-6204. FMC Corporation, P.O. Box 8, Route 1 and Plainsboro Rd.
Princeton, NJ 08543.  May 31, 2007.  Unpublished report.  MRID 47143706

Two studies were submitted that evaluated the differential toxicity of
carbofuran to birds (Bobwhite quail and Mallard ducks) when dosed as a
single aqueous bolus dose versus the same dose mixed in feed.  The
objective of these tests was to compare the toxicity of various dose
levels of carbofuran to Northern Bobwhite (Colinus virginianus) and
Mallard ducks (Anas platyrhynchos) under two exposure scenarios: 1) when
dosed with an aqueous bolus after 15 hours of fasting and 2) when
similar doses are mixed with feed and intubated into the crop of test
birds fasted for at least 15 hours.  The studies quantify the difference
in mortality resulting from these two exposure scenarios.  

All birds that received the aqueous bolus doses were given the test
substance in a solution of distilled water introduced directly into the
crop via syringe and dosing needle at a volume of < 5 mL solution / kg
body weight. In the case of birds receiving a food slurry dose, the
appropriate volume of test substance solution (< 5 ml solution / kg body
weight) was added to the food portion of the slurry dose and mixed
thoroughly in a 50-mL dosing syringe cylinder. Food slurry dose
preparation proceeded by placing the pre-weighed feed into the dosing
syringe cylinder, thoroughly stirring the test substance solution into
the feed, then adding the pre-weighed aliquot of distilled water, again
thoroughly stirring the mixture, and finally fitting the plunger and
feeding needle to the syringe and expelling any excess air prior to
delivery to the test animal.

Behavioral symptoms of intoxication and mortality were the endpoints
measured during this study. Test animals were monitored immediately
after dosing for evidence of regurgitation. Additionally, birds were
observed at 10 minutes, 30 minutes, one hour, and two hours post-dosing,
and once again late in the afternoon. Thereafter, the treated animals
were observed at least once daily for five days. Behavioral symptoms of
toxicity and time of death were recorded, in addition to body weights
and dose calculations. All surviving birds were euthanized at the end of
the study.

The relative toxicity of carbofuran resulting from the food slurry bolus
doses versus that of the aqueous bolus doses can be approximated by
comparing the dose that caused similar number of mortalities and by
comparing the estimated LD50s. For Bobwhite quail, the dose that caused
10% mortality was 1.75 mg/kg for the aqueous bolus dose and 7.08 mg/kg
for the food slurry bolus dose; this is a 4-fold reduction in toxicity
when the carbofuran is in a food-slurry bolus.  The aqueous bolus dose
that caused 80-90% mortality was 3.53 mg/kg while the dose for the food
slurry was 12.5 mg/kg; this is a 3.5-fold reduction in toxicity with the
food slurry bolus dose.  Although more dose groups would be beneficial
in estimating an LD50, the available dataset when used to calculate and
LD50 indicate that there is a 3.9-fold difference between the LD50 from
the food slurry bolus dose and the aqueous bolus dose.  For Mallard
ducks, the results were less robust since many of the birds regurgitated
the dose, suggesting that effects likely occurred at a lower functional
dose.  At the aqueous dose of 0.81 mg/kg, there was 100% mortality while
for birds dosed with the food slurry bolus at the same dose there was
20% mortality.  The data were inadequate for robust LD50 estimation,
however, estimates using the moving average method indicated that the
LD50 for the aqueous dose was 0.496 (95% CL: 0-0.81) and for the food
bolus dose was 1.0 (95% CL: 0.46-1.61).  There is a 2-fold difference
between the LD50s although the confidence intervals overlap suggesting
there is no difference.  Hence for Mallard ducks it is difficult, based
on the available data to estimate what, if any, effect the food bolus
has on acute toxicity estimates.  

These two studies provide limited insight into potential food matrix
effects on acute toxicity of carbofuran to birds.  At face value, it
seems that the effect of a food bolus could reduce toxicity up to
4-fold, although it may have no effect.  Moreover, the studies offer no
insight into how toxicity or exposure might be altered for birds
consuming wild forage items and how this may impact energy acquisition. 


Based on our review we conclude that these two studies provide useful
insight into the effect of food matrix on carbofuran acute toxicity, but
are most appropriate as screening studies for potential avoidance
behaviours.   Because there are no guidelines for this study type these
studies are classified SUPPLEMENTAL.  

Conclusions

Overall, the four studies submitted by FMC provided some insight into
the potential risks of carbofuran to avian species.  However, there
still remains considerable uncertainty regarding some elements of study
design and how results from these laboratory toxicity studies apply to
birds in field conditions.  Birds in the field are coping with a number
of environmental conditions and variables that will impact the potential
for adverse effects associated with pesticide use.  Moreover, there can
be considerable inter-species variability in sensitivity to toxicants
and in behavioral patterns further complicating study interpretation and
risk projections. Specifically, EPA concludes that the four studies
recently submitted by FMC do not thoroughly address uncertainties the
studies were intended to address.  Of the four studies, the ChE
inhibition and recovery study contributed to our understanding of the
mode of action and recovery of carbofuran toxicity to birds.  The feed
avoidance and food matrix studies provided some insight into exposure to
birds but considerable uncertainty still remains.    

 PAGE   

 PAGE   3 

Study Terminated at  6 h

