UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON D.C., 20460

			OFFICE OF PREVENTION, 

  SEQ CHAPTER \h \r 1 MEMORANDUM

             PC Code:    098301

						                         DP Barcode: 322095

SUBJECT:		Environmental Fate and Effects Division

			Response to Comment for the Aldicarb Ecological Risk Assessment 

			

FROM:		Michelle Rau Embry, Ph.D.

 			Jonathan Angier, Ph.D.

			Diana Eignor

			Jeannette Martinez

			Environmental Risk Branch II

	Environmental Fate and Effects Division (7507C)	

TO:			Bob McNally, Branch Chief

Eric Olson, Team Leader

Special Review Branch

			Reregistration Division (7508C)

THROUGH:		Dana Spatz, RAPL

			Tom Bailey, Ph.D., Branch Chief

			Environmental Risk Branch II

			Environmental Fate and Effects Division (7507C)

DATE:			November 30, 2005	

The EFED response to comments on the aldicarb ecological risk assessment
are addressed below.  A revised draft of the Ecological Risk Assessment
will be forthcoming.  

Response to Comments – Aldicarb

(Docket Control Number OPP-2005-0163)

North Carolina Peanut Growers:

Comment:  “With the use of positive displacement applicators, Temik is
applied in furrow and no material is left exposed, resulting in no
threat to wildlife.”

EPA response:  Although application and incorporation are simultaneous,
and the use of positive displacement applicators greatly reduces the
percentage of granules available on the surface, there will never be a
full 100% incorporation; some granules (~1%) will be present at the
surface.  Bioturbation and other processes following application are
also likely to cause some granules to resurface. In addition, risks to
wildlife may occur through other exposure routes, such as inhalation,
dermal, and drinking water exposure, as well as consumption of plant and
soil invertebrate food items, and incidental soil ingestion.  The
results of the Agency’s risk assessment indicate potential risk to
terrestrial and aquatic organisms resulting from the application of
aldicarb to agricultural fields.

Oklahoma State University Pesticide Safety Education Program:

Comment:  “On page 39 which lists incident reports only 7% of the
misuses were from documented registered uses of aldicarb.  We feel that
EPA should make it more clear in measuring incidents that occur on
registered uses versus non-registered uses.”

EPA response:  All of the aldicarb related incidents are discussed in
detail in the text of the Ecological Risk Assessment, as well as in
Appendix F of the document.

Comment:  “EPA states that aldicarb is toxic to bees.  We do not
dispute this but feel it is a stretch to use this in the risk assessment
since the bees would have to have direct contact with the aldicarb and
aldicarb must be incorporated into the furrow.  EPA reports that there
is a 99.96 – 100% incorporation rate for in furrow applications.”

EPA response:  The document has been revised to discuss potential
honeybee exposure due to the systemic nature of aldicarb.  Because of
its granular formulation, it is unlikely that there is a direct contact
exposure scenario for honeybees.  However, these and other insect
species could be exposed through contact with plants and soil. 
Additionally, the EPA does not report a 99.96 – 100% incorporation
rate for in furrow applications.  According to the Agency’s
“Comparative Analysis of Acute Avian Risk from Granular Pesticides”
(EPA, 1992), the incorporation efficiency for in furrow application is
99%.  This value is reflected in the Agency’s Ecological Risk
Assessment for aldicarb.

National Cotton Council of America:

Comment:  “While the preliminary risk assessment predicts nearly 100%
mortality of birds and mammals for nearly all application scenarios,
there have been few documented incidences of field kills of birds and
mammals in the U.S. associated with label uses of Temik.  This enormous
database based upon 35 years of actual use experience with Temik
obviously has not been seriously taken into consideration.”

EPA response:    SEQ CHAPTER \h \r 1 Despite the lack of incident
reports for aldicarb, avian field studies using this pesticide
definitively demonstrate potential ecological risk following labeled
application (See Section IV.A.2 of the ecological risk assessment). 
Based on the results of these field studies, it cannot be concluded that
low numbers of ecological incident reports indicate low ecological risk.
 It should be noted that there are no mammalian field studies for
aldicarb.  Additionally, there are numerous factors that may contribute
to the lack of reported incidents in the Agency’s database.  These are
discussed in detail in the risk assessment (See Section III.B.5), but
include the lack of a systemic or reliable reporting mechanism (incident
reporting is not required) and problems with carcass detection. 
Additionally, there are documented post-aldicarb application incidents
in Germany and the UK (See Section III.B.5 of the ecological risk
assessment).  

Washington State Potato Commission (WPSC):

Comment:  “Aldicarb is applied at planting directly into the open
furrow with immediate incorporation followed by potato seed piece
placement.  This application is performed using certified and approved
positive displacement equipment supplied by “lock and load”
pesticide containers.  Use of this equipment results in 100%
incorporated pesticide granules…because there is no surface applied
material, it is essentially impossible for the pesticide to be directly
applied to water, presenting virtually to hazard to aquatic species.”

EPA response:  Although application and incorporation are simultaneous,
and the use of positive displacement applicators greatly reduces the
percentage of granules available on the surface, there will never be a
full 100% incorporation; some granules (~1%) will be present at the
surface.  Bioturbation and other processes following application are
also likely to cause some granules to resurface.  Aldicarb is not
directly applied to the aquatic environment.  Calculations of aquatic
exposure estimate aquatic concentrations of aldicarb that result from
runoff and erosion from agricultural fields, which account for pesticide
inputs to aquatic systems from granules left on the surface as well as
those that are soil incorporated. 

In addition, risks to wildlife may occur through other exposure routes,
such as inhalation, dermal, and drinking water exposure, as well as
consumption of plant and soil invertebrate food items, and incidental
soil ingestion.  The results of the Agency’s risk assessment indicate
potential risk to terrestrial and aquatic organisms resulting from the
application of aldicarb to agricultural fields.

National Potato Council:

Comment:  “Therefore, without intentional misuse, there are NO
granules left on the surface.”

EPA response:  Although application and incorporation are simultaneous,
and the use of positive displacement applicators greatly reduces the
percentage of granules available on the surface, there will never be a
full 100% incorporation; some granules (~1%) will be present at the
surface for banded, in-furrow applications.  For application of aldicarb
to potatoes, the risk assessment was corrected to assume 99%
incorporation of granules for banded, in-furrow applications where the
granules are covered with a specified amount of soil.  It should be
noted, however, that assuming even 99% incorporation efficiency of the
granules, the terrestrial wildlife risk quotients exceed the Agency’s
level of concern for all crops and application rates.  

In addition, risks to wildlife may occur through other exposure routes,
such as inhalation, dermal, and drinking water exposure, as well as
consumption of plant and soil invertebrate food items, and incidental
soil ingestion.  The results of the Agency’s risk assessment indicate
potential risk to terrestrial and aquatic organisms resulting from the
application of aldicarb to agricultural fields.

American Bird Conservancy:

Comment:  “The registrant for Aldicarb, Bayer Crop Sciences, has never
submitted an avian reproductive toxicity study in support of
registration.  A complete set of avian reproductive studies, conducted
with two species as is required for all food use pesticides must be the
minimum requirement for eligiblility for re-registration.  These studies
have not been conducted, and were not metioned in the comments provided
by Bayer for this risk assessment.”

EPA response:  The aforementioned studies have been requested by the
Agency.

Comment:  “The RQs calculated for wild birds should be adjusted to
reflect the data of Mineau (1996) which indicates that the scaling
factor (1.15) used by EPA for small birds underestimates the toxicity of
carbamates, and Aldicarb in particular.  A scaling factor of 1.4 rather
than 1.15 should be used in calculation of the RQs for Aldicarb.”

EPA response:  As part of the external peer review process for
conducting refined risk assessments for pesticide impacts to non-target
wildlife, the Agency submitted to the Scientific Advisory Panel (March,
2001) a refined risk assessment model that included a process to account
for the potential for body weight scaling as it related to species
sensitivity to pesticides.  The SAP indicated that the Agency is
"…correct pay attention to scaling of the LD50s to account for
differences in bird size following the work of Mineau et al (1996).   In
general, for the majority of pesticides, not scaling for size may
seriously under-protect small species… When there is obvious scatter
in the size-sensitivity relationship, especially where there are
numerous data points, the Agency should explore the various options that
are open in order to best characterize interspecific sensitivity
differences.”  The risk quotients for the aldicarb ecological risk
assessment  were calculated using the aldicarb-specific scaling factor
of 1.4, as shown in the Mineau et al (1996) paper.  The Agency does
recognize potential limitations of the dataset, and discusses potential
uncertainties based on the use of particular scaling factors in the risk
characterization section of the document.  However, it is important to
note that the Agency level of concern for avian species are still
exceeded for all applications and application rates using no scaling
factor (1.0) as well as the default scaling factor of 1.15 to account
for difference in toxicity between different sizes of birds. 

L.J. Olsen, Inc.:

Comment:  “….The amount of aldicarb on top of the ground for birds
or animals in non-existent.”

EPA response:  Although application and incorporation are simultaneous,
and the use of positive displacement applicators greatly reduces the
percentage of granules available on the surface, there will never be a
full 100% incorporation; some granules (~1%) will be present at the
surface.  Bioturbation and other processes following application are
also likely to cause some granules to resurface. In addition, risks to
wildlife may occur through other exposure routes, such as inhalation,
dermal, and drinking water exposure, as well as consumption of plant and
soil invertebrate food items, and incidental soil ingestion.  The
results of the Agency’s risk assessment indicate potential risk to
terrestrial and aquatic organisms resulting from the application of
aldicarb to agricultural fields.

Florida Department of Agriculture and Consumer Services:

Comment:  “…the lack of reported wildlife mortality incidents from
the legitimate use of aldicarb in Florida indicates that the actual risk
to wildlife is likely to be substantially less than what is suggested in
the screening level risk assessment…”

EPA response:  Despite the lack of incident reports for aldicarb, avian
field studies using this pesticide definitively demonstrate potential
ecological risk following labeled application (See Section IV.A.2 of the
ecological risk assessment).  Based on the results of these field
studies, it cannot be concluded that low numbers of ecological incident
reports indicate low ecological risk.  It should be noted that there are
no mammalian field studies for aldicarb.  Additionally, there are
numerous factors that may contribute to the lack of reported incidents
in the Agency’s database.  These are discussed in detail in the risk
assessment (See Section III.B.5), but include the lack of a systemic or
reliable reporting mechanism (incident reporting is not required) and
problems with carcass detection.  Additionally, there are documented
post-aldicarb application incidents in Germany and the UK (See Section
III.B.5 of the ecological risk assessment).  

Comment:  “The combined assumption concerning granules exposed on the
surface and birds feeding exclusively on granules atop pesticide treated
bands, while ignoring other factors that influence the likelihood that
birds will ingest the granules, grossly overestimates risk to birds and
very likely goes beyond what one would consider a worst case scenario
under conditions allowed on the label.”  AND incorporation rate
comments.

EPA response:  Although application and incorporation are simultaneous,
and the use of positive displacement applicators greatly reduces the
percentage of granules available on the surface, there will never be a
full 100% incorporation; some granules (~1%) will be present at the
surface for banded, in-furrow applications.  For application of aldicarb
to certain crops, the risk assessment was corrected to assume 99%
incorporation of granules for banded, in-furrow applications where the
granules are covered with a specified amount of soil.  The 85%
incorporation efficiency was assumed only for broadcast applications
[Agency’s “Comparative Analysis of Acute Avian Risk from Granular
Pesticides” (EPA, 1992)].  It should be noted, however, that assuming
even 99% incorporation efficiency of the granules, the terrestrial
wildlife risk quotients exceed the Agency’s level of concern for all
crops and application rates.  

In addition, risks to wildlife may occur through other exposure routes,
such as inhalation, dermal, and drinking water exposure, as well as
consumption of plant and soil invertebrate food items, and incidental
soil ingestion.  The results of the Agency’s risk assessment indicate
potential risk to terrestrial and aquatic organisms resulting from the
application of aldicarb to agricultural fields.

Comment:  Avian toxicity scaling factor

EPA response:  As part of the external peer review process for
conducting refined risk assessments for pesticide impacts to non-target
wildlife, the Agency submitted to the Scientific Advisory Panel (March,
2001) a refined risk assessment model that included a process to account
for the potential for body weight scaling as it related to species
sensitivity to pesticides.  The SAP indicated that the Agency is
"…correct pay attention to scaling of the LD50s to account for
differences in bird size following the work of Mineau et al (1996).   In
general, for the majority of pesticides, not scaling for size may
seriously under-protect small species… When there is obvious scatter
in the size-sensitivity relationship, especially where there are
numerous data points, the Agency should explore the various options that
are open in order to best characterize interspecific sensitivity
differences.”  The risk quotients for the aldicarb ecological risk
assessment  were calculated using the aldicarb-specific scaling factor
of 1.4, as shown in the Mineau et al (1996) paper.  The Agency does
recognize potential limitations of the dataset, and discusses potential
uncertainties based on the use of particular scaling factors in the risk
characterization section of the document.  However, it is important to
note that the Agency level of concern for avian species are still
exceeded for all applications and application rates using no scaling
factor (1.0) as well as the default scaling factor of 1.15 to account
for difference in toxicity between different sizes of birds. 

Comment:  LD50/ft2 methodology

EPA response:  The Agency agrees that the LD50 per square method
utilized by the Agency to calculate risk quotients does not assume that
direct granular ingestion is the only potential exposure mechanism, nor
necessarily the most significant route of exposure.  LD50 per square
foot calculations were performed using the current parameters specified
by aldicarb application methods, including high soil incorporation
percentage (See section IV.A.3), and the Agency LOC was exceeded for
birds and mammals for all application rates and soil incorporation rates
examined.  

Florida Fruit and Vegetable Grower Association:

Comment:  “The percent acres treated amounts for citrus crops in the
state (FL) published in the revised EFED risk assessment for the
Aldicarb Reregistration Eligibility Document are exceedingly
incorrect.”

EPA response:  The values will be corrected in a forthcoming version of
the Ecological Risk Assessment.  

Comment:  “…If this were true, then citrus groves and potato fields
would be completely barren of any birds, mammals, or any other wildlife
for that matter.  Yet in reality citrus groves and potato fields are a
haven for common and endangered species alike, even in the presence of
aldicarb applications taking place.”

EPA response:  The Agency would welcome information regarding the type
and locations of species that inhabit agricultural fields where aldicarb
is used.  

Comment:  Application methodology and lack of granules on the soil
surface following proper application.

EPA response:  The Agency agrees that the LD50 per square method
utilized by the Agency to calculate risk quotients does not assume that
direct granular ingestion is the only potential exposure mechanism, nor
necessarily the most significant route of exposure.  LD50 per square
foot calculations were performed using the current parameters specified
by aldicarb application methods, including high soil incorporation
percentage (See section IV.A.3), and the Agency LOC was exceeded for
birds and mammals for all application rates and soil incorporation rates
examined.  

In addition, risks to wildlife may occur through other exposure routes,
such as inhalation, dermal, and drinking water exposure, as well as
consumption of plant and soil invertebrate food items, and incidental
soil ingestion.  The results of the Agency’s risk assessment indicate
potential risk to terrestrial and aquatic organisms resulting from the
application of aldicarb to agricultural fields.

Comment:  Lack of incident data

EPA response:  Despite the lack of incident reports for aldicarb, avian
field studies using this pesticide definitively demonstrate potential
ecological risk following labeled application (See Section IV.A.2 of the
ecological risk assessment).  Based on the results of these field
studies, it cannot be concluded that low numbers of ecological incident
reports indicate low ecological risk.  It should be noted that there are
no mammalian field studies for aldicarb.  Additionally, there are
numerous factors that may contribute to the lack of reported incidents
in the Agency’s database.  These are discussed in detail in the risk
assessment (See Section III.B.5), but include the lack of a systemic or
reliable reporting mechanism (incident reporting is not required) and
problems with carcass detection.  Additionally, there are documented
post-aldicarb application incidents in Germany and the UK (See Section
III.B.5 of the ecological risk assessment).  

Comment:  Citrus and potato application methods

EPA response:  The Agency recognizes that “broadcast” is not a
current method used to apply aldicarb to citrus or potatoes.  This
revision will be made in a forthcoming draft of the Ecological Risk
Assessment.   The risk assessment will be amended to provide values for
both the banded, soil incorporated application method (99% incorporation
efficiency) as well as the side-dress application method (85%
incorporation efficiency), where appropriate for specific crop types
[“Comparative Analysis of Acute Avian Risk from Granular Pesticides”
(EPA, 1992)].   It is important to note that calculated acute risk
quotients for both birds and mammals exceed the Agency LOC for all
crops, regardless of application method and incorporation efficiency.  

Comment:  Use of scaling factors for birds

EPA response:  As part of the external peer review process for
conducting refined risk assessments for pesticide impacts to non-target
wildlife, the Agency submitted to the Scientific Advisory Panel (March,
2001) a refined risk assessment model that included a process to account
for the potential for body weight scaling as it related to species
sensitivity to pesticides.  The SAP indicated that the Agency is
"…correct pay attention to scaling of the LD50s to account for
differences in bird size following the work of Mineau et al (1996).   In
general, for the majority of pesticides, not scaling for size may
seriously under-protect small species… When there is obvious scatter
in the size-sensitivity relationship, especially where there are
numerous data points, the Agency should explore the various options that
are open in order to best characterize interspecific sensitivity
differences.”  The risk quotients for the aldicarb ecological risk
assessment  were calculated using the aldicarb-specific scaling factor
of 1.4, as shown in the Mineau et al (1996) paper.  The Agency does
recognize potential limitations of the dataset, and discusses potential
uncertainties based on the use of particular scaling factors in the risk
characterization section of the document.  However, it is important to
note that the Agency level of concern for avian species are still
exceeded for all applications and application rates using no scaling
factor (1.0) as well as the default scaling factor of 1.15 to account
for difference in toxicity between different sizes of birds. 

Comment:  Lack of aldicarb detections in Florida ground and surface
water monitoring programs.

EPA response:  It is not clear if the lack of detections in Florida
ground and surface water monitoring programs is due to well setbacks or
carbon filters that were placed on drinking water wells that had
previously shown detection of aldicarb residues.  Therefore, risk cannot
be precluded.   

Bayer CropScience:

GENERAL COMMENTS

Terrestrial Risk Assessment

	Incorporation Efficiency (p.7 of 70)

In the March 1992 report, "Comparative Analysis of Acute Avian Risk from
Granular Pesticides", EPA used a 99% incorporation efficiency for the
applications: "Banded, cover with specified amount of soil" and
"In-furrow, drill, or shanked-in". These conditions cover virtually all
of the applications of TEMIK® that are in use today. Bayer believes
that 99 to 100% should be used for in-furrow calculations and most
banded application scenarios since the TEMIK® label requires covering
the product with soil.

EPA Response:  All model runs were conducted with an assigned
incorporation efficiency of 99%.  100% efficiency is unlikely.

Use of Beaver Creek Watershed (p.8 of 70)

The Agency has chosen to disproportionately use data from one site to
base many of its characterizations of the risk of aldicarb in aquatic
systems. Beaver Creek is not representative of “real world”
environmental conditions, particularly for estimating residue
concentrations downstream. The water samples of concern to the Agency
were taken from the Beaver Creek sampling station at the field edge. At
the point of sampling, Beaver Creek is a ditch running throughout a
cotton field. It is considered intermittent and is listed as such on
maps. Beaver Creek supports neither a resident fish population nor
significant invertebrate aquatic life throughout the year. Following any
runoff event, downstream dilution greatly reduces the residue
concentrations. Furthermore, there is significant breakdown of residues
before they reach a significant or permanent aquatic environment. The
Agency should correct its analyses and base its assessments upon
monitoring data from significant water systems (e.g., NAWQA monitoring
data). The weight of evidence clearly demonstrates that surface water
contamination from the use of TEMIK® is not a concern. If such losses
occurred uniformly in a moderate size watershed, or even somewhat
frequently over a larger area, certainly more instances of aldicarb
carbamate residues would have been present in the NAWQA, California, and
other monitoring programs.

EPA response:  A stream, ditch, or any watercourse draining an
agricultural field certainly qualifies as “real world” and cannot be
ignored as a legitimate source of potential surface water contamination,
as this water must drain into associated stream networks.  A stream or
waterway listed as ‘intermittent’ or ‘ephemeral’ may nonetheless
be important in terms of total contaminant load carried by the
higher-order stream into which it drains; indeed, many first-order
streams are intermittent, yet serve as essential components of watershed
drainage systems.  Such ephemeral water bodies may also serve as
important habitats for aquatic organisms while they are flowing – they
needn’t support populations year-round to be significant.  In fact,
data obtained from intermittent waterways draining agricultural areas
are likely the best example available for estimating potential acute
exposures.  Stream pesticide loads will be highest following
runoff-inducing rain events that occur shortly after application.  It is
these same runoff events that produce flow in the intermittent
waterways, so monitoring stream water at these times is essential to
evaluating (invariably short-lived and difficult to capture) contaminant
pulses.  Downstream dilution will certainly affect stream water
concentrations, but larger streams may also be receiving
high-concentration contaminant pulses from additional contributing
lower-order streams at the same time (especially in
agriculture-intensive regions).  The likelihood of capturing any of
these short-term high-concentration pulses under typical surface water
monitoring regimes is minimal.  In addition, under high-flow conditions,
in-stream residence time is greatly reduced, allowing for less
degradation before the compound reaches a “significant or permanent
aquatic environment.”

Field Dissipation (p.8 of 70)

Many of the more recent studies on the EPA Guideline Status report
(listed under 164-1 and 164-5) were apparently not considered. The
degradation of aldicarb carbamate residues in the United States
(summarized in Jones and Estes, 1995) is well understood, with half
lives ranging from 0.5 to 3.5 months.

EPA response:  A Field Dissipation half-life range of 15-105 days has
been assigned

I.  EXECUTIVE SUMMARY

Page: 4     Paragraph: 1     Lines: 10-13 

EPA comment:  “Risks to aquatic organisms are currently assessed based
on modeled estimated environmental concentrations (EECs) for the cotton,
potato, citrus, pecan, and soybean uses of the chemical.  These five
aquatic scenarios were chosen because they are the major aldicarb use
crops and are representative of specific geographic use areas as well as
application rates, and are representative of the other uses.”

Bayer response:  The five major use crops for TEMIK® are cotton,
citrus, potatoes, peanuts and sugarbeets.  The application and use of
TEMIK® on pecans is unique and is not representative of any other
application use or practice.  See the report and DVD, “TEMIK® 15G
Brand Aldicarb Product Usage and Methods of Application” submitted
with this response (Hall, 2005).

EPA response:  The aquatic modeling scenarios chosen for this risk
assessment were based on currently available Agency models.  All
potential modeling scenarios will be forthcoming in a future draft of
the risk assessment.   PRZM-EXAMS model scenarios are intended to be
representative of crop regions, rather than a particular state (e.g.,
the Mississippi cotton scenario includes hydrological data from the
southeastern US, not solely the state of Mississippi).  Therefore, the
MS cotton scenario is representative of cotton grown in the southeastern
states.

Despite these modeling limitations, the scenarios chosen to model
potential aldicarb exposure in the aquatic environment are considered
adequately representative of the potential uses.  The PRZM-EXAMS
scenarios that will be included in the revised risk assessment  include:
 citrus (CA, FL), cotton (CA, MS, NC, TX), sugar beet (CA, MN),
sugarcane (LA), pecans (GA), potatoes (ID), sorghum (KS), soybean (MS),
peanut (NC), sweet potato (NC), and tobacco (NC).  It should be noted
that PRZM-EXAMS scenarios are not available for all crops and crop
regions.

Page: 4     Paragraph: 8     Lines: 1-2 

EPA comment:  “The chronic level of concern is also exceeded for
freshwater fish . . ."

Bayer response:  Chronic levels of concern for fish are NOT exceeded if
standard EPA risk assessment methods are used.  The Agency’s standard
method is to use the lowest chronic NOAEC value observed in the chronic
RQ calculation.  The lowest fish chronic NOAEC is 78 ppb.  If this value
is used, the resulting RQs range from 0.02 to 0.36 and these are all
below the Agency’s LOC.

Rather than use the experimentally determined NOAEC of 78 ppb which was
determined in a test with the fathead minnow, the agency instead has
calculated an expected no-effect concentration (ENEC) for the bluegill
sunfish of 0.46 ppb and used this in the RQ calculation.  Bayer believes
the ENEC was calculated inappropriately for reasons presented in detail
later (see Bayer’s comments on EFED’s Ecological Effects
Characterization on pages 27-29). Such calculations carry a high degree
of uncertainty and it is preferable to use experimentally measured
values when an appropriate test has been performed.  In this case, an
experimentally measured chronic NOAEC is available for a fish species
that has been demonstrated to be approximately equal in sensitivity to
aldicarb as the bluegill sunfish. This species is the sheepshead minnow.
 The reported LC50 in this species ranges from 41 to 170 ppb with a
geometric mean of 83 ppb.  The reported LC50 values for the bluegill
sunfish range from 52 to 115 ppb with a geometric mean of 72 ppb.  Based
on this comparison, tests with the sheepshead minnow should be a
reasonably good surrogate for tests with the bluegill sunfish.  The
chronic NOAEL value experimentally determined for the sheepshead minnow
is 50 ppb.  If this value is used in the fish chronic risk assessment,
resulting RQ values range from 0.03 to 0.56, and these are all below the
Agency’s LOC.  Based on this analysis, it can be concluded that
aldicarb poses a minimal chronic risk to both freshwater and saltwater
fish.

EPA response:  This comment is addressed in detail in the EPA Response
to Bayer’s Comment for page 30, Paragraphs 1-2.

Page: 5     Paragraph: 1     Lines: 1-2 

EPA comment: “In addition to risk based exposure estimates from
modeling, there were also exceedances of the Agency levels of concern
based on monitoring data.”

Bayer response:  This is true for one field monitoring data point
(Beaver Creek), but is a misleading summary of the totality of field
monitoring data which generally shows that for more than 99% of the
time, the Agency’s levels of concern are not exceeded.  The
overwhelming weight of the evidence from monitoring studies points to
the conclusion that aldicarb concentrations very rarely exceed the
Agency’s levels of concern for aquatic organisms.

EPA response:  Conditions that area likely to be of greatest concern for
aquatic organisms (at least in terms of acute effects) are unlikely to
be sampled under even the most rigorous sampling regimens; that is,
elevated stream flow enhanced by surface runoff (as a result of
rainfall) that occurs shortly after application of the chemical to the
field.  Thus, the Beaver Creek datum is probably most representative of
this type of event, since flow is most likely to occur within this
‘intermittent’ channel as a result of the aforementioned conditions.
 If approximately 1% of other samples showed levels of concern (and
likely none of these were obtained during storm events that followed
application) then there is probably much more of concern that simply
goes unrecorded.

Page: 5     Paragraph: 2     Lines: 1-4 

EPA comment:  “Using Multiple Lines of Evidence (such as use
scenarios, average or “typical” application rates, registrant
submitted toxicity studies, open literature data, and field monitoring
data), lead to the same conclusion: Aldicarb poses acute risks
(mortality) to birds, mammals, and aquatic organisms.  In addition,
there are chronic reproductive effects in fish and aquatic
invertebrates.”

Bayer response:  The open literature and field incident reports do not
support a conclusion of high risk to birds, fish or aquatic
invertebrates.  Surface water monitoring data indicate that aldicarb
concentrations in real world rivers and streams are almost always far
below toxic thresholds to aquatic species.  There are no documented
cases of chronic reproductive effects in fish and aquatic invertebrates,
and none would be expected to occur based on the surface water
monitoring data.  There have been very few documented incidents of field
kills of wild birds or mammals associated with the 35-year use of the
end use product TEMIK®.  Recorded incidents have almost always been due
to misuse, improper application, or in association with other
insecticides.  Thus, the data from incident monitoring programs stands
in stark contrast to the Agency’s assessment which predicts 100%
mortality of birds and mammals for nearly all application scenarios. 
Such a risk prediction is obviously erroneous when one considers the
multiple lines of evidence.

The Agency has not used all available information.  Conspicuously absent
in the Agency’s assessment is any mention of the numerous published
studies by Dr. Louis Best and colleagues at Iowa State University on the
factors that influence whether birds are likely to ingest pesticide
granules.  These studies indicate that availability of granules per unit
area on the soil surface is a relatively unimportant factor in
comparison to features such as size, shape, composition and color of
granules.  These studies demonstrate that birds rarely ingest TEMIK®
granules even when they are readily available to them.  More details
including relevant citations are provided in the Line-By- Line comments
which follow.

An objective use of multiple lines of evidence would consider whether
real world observations confirm whether the magnitude of risk is as high
as suggested by the Agency’s calculated risk quotients.  Bayer
believes multiple lines of evidence clearly point to a conclusion that
risks in the real world posed by aldicarb to birds, mammals and aquatic
organisms are not as high as indicated by the screening assessment.  The
Agency’s risk assessment should point this out.

EPA response:   As discussed in the chapter, there several surface water
sampling studies have detected aldicarb at concentrations that are above
the toxic thresholds to aquatic species (See Section III.A. 2 of the
ecological risk assessment).

  SEQ CHAPTER \h \r 1 Despite the lack of incident reports for aldicarb,
avian field studies using this pesticide definitively demonstrate
potential ecological risk following labeled application (See Section
IV.A.2 of the ecological risk assessment).  Based on the results of
these field studies, it cannot be concluded that low numbers of
ecological incident reports indicate low ecological risk.  It should be
noted that there are no mammalian field studies for aldicarb. 
Additionally, there are numerous factors that may contribute to the lack
of reported incidents in the Agency’s database.  These are discussed
in detail in the risk assessment (See Section III.B.5), but include the
lack of a systemic or reliable reporting mechanism (incident reporting
is not required) and problems with carcass detection.  Additionally,
there are documented post-aldicarb application incidents in Germany and
the UK (See Section III.B.5 of the ecological risk assessment).  

Toxicity tests performed by the registrant indicate the potential for
chronic reproductive effects in fish and aquatic invertebrates, and the
chronic EEC values as estimated using PRZM-EXAMS for different crops and
application rates suggest potential chronic risk to these organisms (See
section IV.A.1 of the ecological risk assessment for details).  Chronic
risk quotients exceed the Agency LOC of 1.0 for all crops and
application rates modeled for freshwater fish, freshwater invertebrates,
and estuarine/marine invertebrates.  Documented cases of chronic
reproductive effects in fish and invertebrates are not available because
there is currently no system for monitoring or assessing such effects in
the field.  If the registrant is aware of such monitoring efforts that
have shown no effects, we urge the registrant to submit that information
to the Agency so that we may consider it on the merits of the study
quality.

Although the focus of this comment is on direct granule ingestion, it
should be noted that the LD50 per square foot index utilized by the
Agency to calculate risk quotients does not assume that this route of
exposure is the only potential exposure mechanism.  The index is
intended to provide insight into the possible consequences if material
in a given surface area is bioavailable to a non-target organism,
regardless of exposure route. Numerous exposure routes may be possible,
including plant ingestion, soil ingestion, inhalation, drinking water,
soil invertebrate ingestion, AND direct granule ingestion.  However the
exposure term in the model is not quantative for any specific route of
exposure for all potentially applicable routes.  the effects term in the
LD50 per square foot index is based on the oral LD50 toxicity endpoint, 
 This use of the LD50 endpoint does not suggest that all exposure must
occur through the oral ingestion of granules, ratehr it is being
utilized as a surrogate for a total body burden toxicity endpoint.  It
is recognized that a pesticide may exert different acute toxic potency
when administered by varying routes. If  that difference in toxic
potency, supported by credible toxicity data, is great enough to
significantly alter risk conclusions that uncertainty would be
acknowledged in the risk characterizatipon.  There is another known
limitation in the calculation of the LD50 per square foot RQ.  This is
the assumption of bioavailability  based solely on the amount of
pesticide present on the soil surface. This assumption has contributed
to a general misunderstanding that the model is only accounting for
potential direct granule ingestion.  If pesticide is bioavailable from
soil strata below the surface, it is possible that the LD50 per square
foot calculation used in this screening-level risk assessment may
underestimate exposure, due to the lack of consideration of these other
exposure routes in the actual calculation.  As shown from the earthworm
fugacity model calculation (See section IV.A.3), ingestion of soil
invertebrates by terrestrial birds and mammals is another significant
route of exposure, as is incidental soil ingestion (See section  IV.
E.3.e). Additionally, the systemic nature of aldicarb makes ingestion of
plant food items a significant route of exposure that is not
quantitatively considered in this risk assessment.  Additional exposure
routes, discussed above, include inhalation, dermal, and drinking water
exposure.

The Agency has reviewed the numerous papers referred to by the
registrant that pertain to granular pesticide ingestion in bird species.
 Parameters such as granule size, color, shape, and matrix do affect the
direct consumption of aldicarb granules by birds.  However, several
facts should be noted:  1) The papers cited by the registrant pertain to
ingestion of granular pesticides only by avian species (there is also
potential mammalian exposure, which is not addressed in these articles),
and 2) There are definitive field studies which demonstrate the
potential for avian species to consume aldicarb granules.

II.  PROBLEM FORMULATION

A.  Introduction

Page: 6     Paragraph: 1     Lines: 15-18 

EPA comment:  “Risks to aquatic organisms are assessed based on
modeled EECs for the cotton, potato, citrus, pecan, and soybean uses of
the chemical.  These five aquatic scenarios were chosen because they are
the major aldicarb use crops and are representative of specific
geographic use areas as well as application rates.”

Bayer response:  The five major use crops for TEMIK® are cotton,
citrus, potatoes, peanuts and sugarbeets.  The application and use of
TEMIK® on pecans is unique and is not representative of any other
application use or practice.  See the report, “TEMIK® 15G Brand
Aldicarb Product Usage and Methods of Application” submitted with this
response.

EPA response:  See response for “Page: 4     Paragraph: 1     Lines:
10-13” above.

B. Stressor Source and Distribution

Page: 8     Paragraph: 11     Line: 1 

EPA comment:  “Approximately 4.8 millions pounds of aldicarb active
ingredient (ai) are used per year on 4.9 million acres (BEAD)
Quantitative Usage Analysis, August 9, 2004).”

Bayer response:  Doane and Bayer’s internal tracking information
indicate that during 2004 actual applications were 4.53 million pounds
of aldicarb active ingredient on 5.45 million acres.

EPA response:  The correct usage information as specified by Doane and
Bayer (2004) have been changed in the document. 

Page: 9     Table: 1     Line: 4 

EPA comment:  "264-426, Temik® Brand 15G Aldicarb Pesticide for Sale
and Use in CA only.”

Bayer response:  The correct name is, "264-426, TEMIK® Brand 15G
Aldicarb Pesticide for Sale and Use in California Only.”

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: 9     Table: 1     Line 5 

EPA comment: "264-417, Temik® Brand 15G Aldicarb Pesticide for Use on
Citrus only.”

Bayer response:  The correct name is, "264-417, TEMIK® Brand 15G CP
Aldicarb Pesticide.”

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: 9     Table: 1     Line: 6 

EPA comment:  "264-523, Temik® Brand 15G NW Aldicarb Pesticide for Use
on Potatoes.”

Bayer response:  The correct name is, "264-523, TEMIK® Brand 15G NW
Aldicarb Pesticide.”

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: 9     Paragraph: 2     Lines: 9-10 

EPA comment:  “Nation-wide single application rates for aldicarb range
from 0.4 lbs active ingredient per acre (lbs ai/A) for sorghum to a
maximum of 10.05 lbs ai/A for pecans.”

Bayer response:  The lowest labeled use rate for TEMIK® Brand 15G
Aldicarb Pesticide is 2.0 lbs product per acre for cotton (in Texas,
Oklahoma and New Mexico only).  This rate equals 0.3 lbs ai/A.  The
TEMIK® 15G label allows the further reduction of the cotton rate of 0.3
lbs ai/A by ½ (to 0.15 lbs ai/A) if seeds and TEMIK® are applied by
the hill-drop method.

EPA response:  The sentence was changed in the risk assessment to read
“Nation-wide single application rates for aldicarb range from 0.15 lbs
active ingredient per acre (lbs ai/A) for certain cotton applications to
a maximum of 10.05 lbs ai/A for pecans.”

Page: 10     Table: 2     Cotton 

EPA comment:  Column 4.  # of applications (interval).  “Cotton.  2
(30 days)”

Bayer response:  The TEMIK® label does not specify an application
interval for cotton.  Delete the reference “(30 days).”

EPA response:  Due to the lack of label language, the reference to a 30
day application interval was removed.  Therefore, the assumed
application interval for cotton in the forthcoming revised risk
assessment will be 7 days, and this value will be applied to the cotton
modeling scenarios to calculate EEC values.  EFED originally used the
30-day application interval because the label states that the first
cotton application should be made at plant, and the second at first
squaring.  According to the University of Georgia College of
Agricultural and Environmental Sciences, Cooperative Extension Service,
the time from plant to first square for cotton is approximately 38 days.
 However, due to the registrant’s objection, this value will be
changed to 7 days.  Changes will be reflected in a future draft of the
document.

C.  Conceptual Model

Page: 15     Paragraph: 2     Lines: 1-4 

EPA comment:  "Immediately following application of aldicarb and prior
to soil incorporation, granules are expected to be available at the soil
surface on agricultural sites.  Wildlife exposure could result . . .”

Bayer response:  Granule application and soil incorporation are
essentially simultaneous events.  Application equipment used in modern
agriculture provides both deposition of the granules and soil
incorporation without the need for separate operations. Therefore, under
normal use granules would not be expected to be available at the soil
surface.  See the report and DVD, “TEMIK® 15G Brand Aldicarb Product
Usage and Methods of Application” submitted with this response (Hall,
2005).

EPA response:  Although application and incorporation are simultaneous,
there will never be a full 100% incorporation; some granules (~1%) will
be present at the surface.  Bioturbation and other processes following
application are also likely to cause some granules to resurface. 

Page: 15    Paragraph: 2    Lines: 4-5 

EPA comment:  "Later soil incorporation of the granules is expected to
result in the movement of aldicarb down into the soil column.”

Bayer response:  This sentence is incorrect and confusing.  How does
EFED define “later”?  TEMIK® granules are incorporated into the
soil at the same time they are applied.  There is no time lag between
application and incorporation during which a bird could be exposed. 
This sentence should be deleted.  See the report and DVD, “TEMIK® 15G
Brand Aldicarb Product Usage and Methods of Application” submitted
with this response (Hall, 2005).

EPA response:  The sentence was changed to read, “Soil incorporation
of the granules is expected to result in the movement of aldicarb down
into the soil column.”  

Page: 15    Paragraph: 3	Lines: 1-2

EPA comment:  "Terrestrial organisms will be exposed mostly to granules,
except under conditions where there is surface saturation and dissolved
aldicarb remains in pooled water at the surface.”

Bayer response:  The sentence should be changed to read, “Terrestrial
organisms will be potentially exposed mostly to granules, except . .
.” The word “potentially” should be added because the vast
majority of terrestrial organisms will not be exposed at all.  As
already explained in our response to Page 15, Paragraph 2, Line 1,
TEMIK® is applied to crops and covered or incorporated in a single and
continuous process.  There is no time between the granules hitting the
soil and the granules either being completely covered by soil or
incorporated into the soil, thus, there is no exposure.  See the report
and DVD, “TEMIK® 15G Brand Aldicarb Product Usage and Methods of
Application” submitted with this response (Hall, 2005).  Furthermore,
TEMIK® is not applied to saturated soil and if there is sufficient
rainfall to saturate the soil after application all of the TEMIK®
granules will have disintegrated and the aldicarb will have moved down
into the soil.  As EPA is aware, TEMIK® granules are composed of gypsum
and disintegrate when they become wet.  As mentioned elsewhere in the
EFED chapter, aldicarb is soluble and moves down in soil and out of the
area where birds and mammals would be exposed.

EPA response:  This sentence was removed from the document and replaced
by, “Terrestrial organisms will be potentially exposed to aldicarb and
aldicarb residues through multiple pathways, including direct contact
with granules (ingestion, dermal, inhalation), ingestion of plant food
items, soil, and earthworms.  Additionally, terrestrial organisms may be
exposed to aldicarb in drinking water pooled on the field surface if
application is followed by irrigation.”  Although TEMIK® is not
applied to saturated soil, if (after application) rainfall intensity
exceeds infiltration capacity of the soil, there will be temporary
pooling of water at the surface.  Moreover, there is the influence of
microtopography on spatially heterogeneous infiltration patterns. 
Episaturation is most likely to occur in small swales and depressions,
where fine-grain material also tends to accumulate (slowing infiltration
rates); these pockets can contain standing water long after most of the
water at the surface has drained or infiltrated.  These features may
present transient pockets of dissolved chemical.

Page: 15     Paragraph: 4     Lines: 1-5 

EPA comment:  "Wildlife exposure could also result from a number of
other exposure pathways and wildlife actions and behaviors including
inhalation of dust particulates: dermal intake . . .”

Bayer response:  This sentence is complete speculation without any
evidence to support these claims.

EPA response:  This sentence is part of the conceptual model, which is a
group of working hypotheses about how aldicarb is likely to reach (i.e.,
exposure pathways) and affect ecological entities of concern on and
adjacent to a treated agricultural field.  The conceptual model examines
all potential routes of exposure.   The Agency is currently working on
models to assess potential inhalation, dermal, and drinking water
exposure to wildlife.  Until these models are complete and approved for
use in the screening-level ecological risk assessment or data is
provided to the Agency, these routes of exposure will be discussed as
“other potential routes of exposure.”

D.  Key Uncertainties and Information Gaps

Page: 18     1.  Ecotoxicity Information Gaps 

EPA comment: ”The following studies are requested. 71-4(a) Avian
Reproduction – Quail and 71-4(b) Avian Reproduction – Duck”

Bayer response:  Bayer believes avian reproduction studies would provide
no useful information and therefore should not be required because (1)
the Agency has no methods to use the toxicity data produced by these
studies in a risk assessment, (2) TEMIK® granules do not persist in the
field (they rapidly disintegrate with the first rainfall, or upon
contact with moist soil), and so the important route of exposure for
birds (ingesting granules) will not occur over a long period of time,
and (3) studies in other animals (mammals, daphnids, etc.) all point to
the conclusion that aldicarb is an acute toxicant only (due to rapid
metabolism and reversibility of the cholinesterase inhibition, animals
can withstand repeated exposure up to levels that are nearly lethal). 
Aldicarb is only formulated as a granular product (TEMIK® Brand
Aldicarb Pesticide).  The Agency’s policy for assessing avian risks of
granular products is to calculate LD50s per square foot.  The Agency has
no methodology for calculating an EEC for avian diets for granular
products.  Thus, if avian reproduction NOAECs were available, they could
not be used in a risk assessment. As confirmation of this, one only has
to look at the way chronic mammalian test results were used in this risk
assessment.  They weren’t!  The Agency has high quality chronic
toxicity studies of aldicarb with mammals yet the entry in Table 20
(page 38) for the mammalian chronic toxicity endpoint used in the risk
assessment is “N/A” presumably for “not applicable”.  There were
chronic NOAELs from tests with mammalian species but they weren’t used
in any RQ calculations for the reasons indicated above.  What is the
point of requiring two very expensive (>$100,000 each) studies when the
data generated will not affect the outcome of, or even be used in, a
risk assessment?  The requirement of these studies should be waived.

EPA response:  Although the current LD50 per square foot method does not
quantitatively calculate an avian chronic RQ value, this does not
preclude the need for an avian reproduction study.  Due to the systemic
nature of aldicarb and the persistence of the equally toxic aldicarb
sulfoxide in plant tissue, ingestion of plant food items, soil, and soil
invertebrates in fields treated with aldicarb are major chronic exposure
pathways.  The use of the earthworm fugacity model (see Ecological risk
assessment, Section IV.A.3), indicates potentially high levels of
aldicarb residues in this important food item, and these residue values,
coupled incidental soil ingestion rates, may be used to quantitatively
calculate an RQ value for birds if the toxicity data is provided.

However, avian studies are still required by the Agency, whether a
quantitative RQ is calculated or not.  Knowledge of the potential for
aldicarb to cause chronic reproductive effects in birds is necessary in
order to perform a complete ecological risk assessment.  Although the
chronic mammalian data was not used to calculate an RQ value, the
results from these studies suggest the potential for deleterious chronic
effects, and this fact was discussed in the risk characterization.  The
purpose of the RQ calculation is to allow for a numerical estimation of
potential risk; inability to perform an RQ calculation does not preclude
risk, and therefore does not provide adequate grounds for waiving data
requirements.  

Page: 18     Paragraph: 5     Lines: 1-2 

EPA comment:  data gap for “161-2 Photodegradation in Water
(sulfoxide, sulfone)"

Bayer response:  Field monitoring and laboratory studies under
controlled conditions demonstrated that it is unlikely for any
significant aldicarb sulfoxide and aldicarb sulfone residues to occur in
open bodies of water, or to exist for sufficient duration to allow for
photodegradation to play an important role in the environmental impact
of aldicarb.  Non guideline studies have shown that photodegradation in
water is limited for both aldicarb sulfoxide and aldicarb sulfone. 
Therefore, further investigation of the photodegradation of the two
carbamates is not warranted. Refer to the discussion section for more
details.

EPA response:  Although it is true that aldicarb sulfoxide and aldicarb
sulfone are more likely to be detected in ground water than surface
water (and thus generally not subject to photodegradation), most local
groundwater eventually discharges into surface water bodies under
baseflow conditions.  Nevertheless, this is probably not a major issue,
and it is agreed that this additional information is probably
unnecessary.

Page: 18     Paragraph: 6     Lines: 1-3 

EPA Comment:  “164-1 Terrestrial Field Dissipation (parent aldicarb,
sulfoxide, sulfone)”

Bayer response:  The registrant has identified numerous field
dissipation studies conducted for aldicarb and previously submitted to
the Agency that have not been included in the preliminary EFED risk
assessment.  The degradation of aldicarb carbamate residues in the
United States (summarized in Jones and Estes, 1995) is well understood,
with half lives ranging from 0.5 to 3.5 months.  The discussion section
provides a complete response to this comment.

EPA response:  The Agency agrees with this statement.  Further studies
should be unnecessary.  A field dissipation half-life range of 15-105
days has been assigned.

E. Analysis Plan

Page: 18     Paragraph: 9     Lines: 5-8 

EPA comment:  “Risks to aquatic organisms are assessed based on
modeled EECs for the cotton, potato, citrus, pecan, and soybean uses of
the chemical.  These five aquatic scenarios were chosen because they are
the major aldicarb use crops and are representative of specific
geographic use areas as well as application rates.”

Bayer response:  The five major use crops for TEMIK® are cotton,
citrus, potatoes, peanuts and sugarbeets.  The application and use of
TEMIK® on pecans is unique and is not representative of any other
application use or practice.  See the report and DVD, “TEMIK® 15G
Brand Aldicarb Product Usage and Methods of Application” submitted
with this response (Hall, 2005).

EPA response:  See response for “Page: 4     Paragraph: 1     Lines:
10-13” above.

Page: 21     Table: 5     Line: 4 

EPA comment:  “Broadcast” is listed a method of application.

Bayer response:  TEMIK® is not applied to agricultural crops using
broadcast methods of application.

EPA response:  The following label language is present on EPA Reg.No.
264-322,  SLN No. FL-870002, and SLN No. NC 780021, respectively: (1)
“Apply at planting in furrow, as a band over the row, or
broadcast….. If banded over row or broadcasted, …..”,  (2)
“Broadcast granules to the soil surface on both sides of the ……”
, and (3) “Apply granules as overall broadcast…..”. The Agency
will keep “Broadcast” as a listed method of application. 

III.  ANALYSIS

A.  Exposure Characterization

Page: 24     Paragraph: 2 

EPA Comment:  “At this time EPA lacks valid guideline data on the
aquatic metabolism of aldicarb . . .”

Bayer response:  EFED should consider rewriting the entire paragraph, to
better focus on degradation from the guideline studies, and include the
data from MRID Nos. 45592107, 45592108, and 45592109.  Other specific
issues in this paragraph are listed below.  The study of Vink (1997) is
not an appropriate study since sediment is not included and the work
described by Lightfoot (1986) shows that the presence of sediment
enhances degradation under aerobic and anaerobic conditions.

EPA response:  MRID #44592107 has been used to determine that the
aerobic aquatic half-life of parent + degradates is 12 days (prior
parameter value had been 120 days).  Vink et al. (1997) is no longer
used.

Page: 24     Paragraph: 2     Lines: 1-9 

EPA Comment:  “At this time EPA lacks valid guideline data on the
aquatic metabolism of aldicarb. . . . In a Guideline laboratory
anaerobic aquatic metabolism study (MRID 43805701) aldicarb degraded
into acid, nitrile and alcohol forms, with a half life of 3 hours.  No
sulfone or sulfoxide were formed in this study, suggesting that
anaerobic degradation can detoxify aldicarb residues very rapidly.  The
value of this study is in doubt, though; redox potential was
inconsistent and variable throughout the study period (127 days),
contrary to guideline requirements. In addition, no discernable pattern
of formation and decline of degradates was observed (data were highly
variable and inconsistent).”

Bayer response:  The anaerobic aquatic study meets the requirements of
guideline 162-4, and therefore the reported aldicarb half-life of 2
hours is scientifically valid, and should be used as the degradation
rate under anaerobic conditions throughout the EFED risk assessment. 
Throughout the study, the test systems were maintained under a reducing
environment as required by the current guidelines, with redox potentials
(Eh) of -113 to -181 mV. According to the classification of Wolfe, et.
al., a reducing environment exists at Eh of -50 to -200 mV.  The
reducing conditions are also consistent with the lack of the oxidative
sulfoxide and sulfone degradates observed in aerobic studies.  Also, the
oxygen content showed anaerobic conditions (0.08-0.27 ppm) throughout
the study.

With regard to formation and decline of degradates, a clear pattern is
observed, with aldicarb aldehyde forming and completely degrading within
one day, and aldicarb nitrile, aldicarb alcohol, aldicarb acid reaching
maximum concentrations in the first week, and then declining only
slightly during the remainder of the study.  The “variable and
inconsistent” data noted by the reviewer, is recognized, but the
overall conclusions of the study remain acceptable.  In consideration of
the reviewer’s noted deficiencies, the determination of the aldicarb
half-life must be recognized as scientifically valid, due to its
complete degradation to non-carbamate moieties, within 1 day of
application.  The short half-life also agreed with work conducted by EPA
with aldicarb sulfoxide and sulfone by Wolfe (1985).

A detailed response to the EFED data evaluation record for this study is
provided in Holmsen (2000), and a full kinetic evaluation of the
formation and decline of degradates is provided in Ramanarayanan (2000).

EPA response:  The Agency disagrees with the assertion that this study
is valid.  The oxygen contents cited above (0.08-0.27 ppm) reflect
sub-oxic rather than anoxic conditions.  Redox values varied too greatly
too quickly for this study to be declared acceptable.  Patterns of
degradate formation and decline were inconsistent as well.  An anaerobic
aquatic input parameter of 24 days (2X aerobic aquatic half-life) has
been assigned (was formerly 120 days).

Page: 24     Paragraph: 2     Line: 9 

EPA Comment:  “However, the fate of the parent aldicarb under
anaerobic aquatic conditions (particularly ground water) is of less
concern that that of the degradates (aldicarb sulfoxide and aldicarb
sulfone) which have been

detected in groundwater long after application of the parent chemical
had ceased (e.g., degradates detected in Long Island, NY groundwater
decades after usage was stopped).”

Bayer response:  The EPA is incorrect in citing the example of
persistence in Long Island ground water as an example of slow anaerobic
aquatic degradation rates.  The main reason for the persistence of
aldicarb residues in Long Island ground water is that the upper part of
the aquifer is highly aerobic, not anaerobic.  As EPA points out the
degradation rate under anaerobic conditions is faster than under aerobic
conditions.  The decline of aldicarb sulfoxide and aldicarb sulfone
residues in Long Island ground water has been consistent with a
half-life of about five years.  Wells currently with residues in excess
of 7 ppb (NY guideline) comprise about 1.5 percent of the wells that
have had been identified with residues greater than 7 ppb during the
past 25 years.  Degradation in surface water under aerobic and anaerobic
conditions is more rapid than in ground water.  One explanation is that
microbial degradation appears to be a contributing process in (aerobic
and anaerobic) surface water, while degradation in ground water is
primarily chemical (hydrolysis catalyzed by the presence of ferrous ion
on solid surfaces under anaerobic conditions).

EPA response: The Agency concurs that the situation in Long Island, NY
is largely due to the relatively aerobic nature of the shallow
groundwater in that region, although the acidic characteristics of this
groundwater may also be a factor; aldicarb residues appear to be more
persistent in low-pH environments where hydrolysis is the principle mode
of degradation.

Page: 24     Paragraph: 2     Line: 15 

EPA Comment:  “. . . potentially equally toxic metabolite of aldicarb
sulfone, hydroxymethyl aldicarb sulfone . . .”

Bayer response:  The speculation by EPA that hydroxymethyl sulfone could
be as toxic as aldicarb sulfone is incorrect.  The toxicity of
hydroxymethyl aldicarb sulfone is about two orders of magnitude less
than aldicarb sulfone.  A LD50 value for rat of 2,460 mg/kg bw was
reported by Weil, and Carpenter (1972), MRID No. 00054419.

EPA response:  The sentence has been changed.  It is recognized that
aldicarb sulfone is significantly less toxic than parent aldicarb. 
Parent aldicarb and aldicarb sulfoxide are considered equally toxic.

Page: 24     Paragraph: 2     Line: 29 

EPA Comment:  “The metabolite hydroxymethyl aldicarb sulfone, which
forms from aldicarb sulfone under both aerobic and anaerobic conditions,
can persist for long periods in oxic, suboxic, and anoxic groundwater
within aquifers, which may account for their detection long after above
ground application of the parent aldicarb has been terminated.”

Bayer response:  This statement is purely speculative in stating that
the detection of aldicarb sulfoxide and aldicarb sulfone in ground water
is due to formation from the hydroxymethyl aldicarb sulfone.  The
formation of aldicarb sulfoxide and aldicarb sulfone from the
hydroxymethyl aldicarb sulfone will not occur under aerobic conditions
(see discussion section for more details) and is unlikely to occur under
anaerobic conditions (although the formation of aldicarb sulfone and
aldicarb sulfoxide under anaerobic conditions is not of concern since
they are not persistent under anaerobic conditions.)  Therefore, the
degradation rates obtained for aldicarb sulfoxide and aldicarb sulfone
in the aquatic aerobic and anaerobic metabolism studies are suitable for
use as input parameters in the surface water modeling simulations.  The
suitability of these degradation rates is further confirmed by the
similarity of results of repeat studies described at the end of this
response.  As discussed in the next response, the repeat studies did not
show the formation of hydroxymethyl aldicarb sulfone.

In addition, field data shows that the persistence of aldicarb sulfone
(and aldicarb sulfoxide) can be explained by the aerobic conditions
within the aquifer.  For example, data from the Netherlands show that
aerobic conditions can exist for 20-30 m and while aldicarb sulfoxide
and aldicarb sulfone are traveling downward in this portion of the
aquifer, residues degrade rather slowly.  However, rapid degradation
occurred as soon as anaerobic conditions were encountered in the
aquifer.  There is no reason to assume, as EPA suggests, that aerobic
degradation in ground and surface water should be similar.  In surface
water, microbial degradation is more likely to be important.

Upon the receipt of the EPA comments on hydroxymethyl aldicarb sulfone,
the aerobic and anaerobic studies for aldicarb sulfone were reviewed and
the formation of hydroxymethyl aldicarb sulfone was thought to be
unlikely. Therefore, both anaerobic and aerobic metabolism studies for
aldicarb sulfone were repeated using water and sediment collected from
the same location.  The preliminary results are presented in the
discussion section. Although aldicarb sulfone has disappeared from the
system, the studies are ongoing to characterize the decline of the
metabolites.  The registrant intends to submit final reports for these
studies in January.  The main difference between the earlier studies and
the latest results is the lack of hydroxymethyl aldicarb sulfone in both
studies.  The cause of

the difference is likely the result of misidentification in the earlier
study since hydroxymethyl aldicarb sulfone is rapidly converted to
aldicarb sulfone nitrile.  However, the formation or lack of formation
of hydroxymethyl aldicarb sulfone is not a major factor in risk
assessment since even if it is formed, its toxicity is quite low and
will not revert back to aldicarb sulfone under aerobic conditions. 
Degradation of aldicarb sulfone was slightly faster in the latest
studies (aerobic aquatic degradation rates of 5.0 and 3.0 days in the
earlier and latest studies; anaerobic aquatic degradation rates of 3.5
and 2.4 days in the earlier and latest studies.  The material balances
in the latest studies were also excellent.

EPA response:  The Agency agrees that the persistence of aldicarb
sulfoxide and sulfone in ground waters may not be the result of
conversion from the hydroxymethyl aldicarb sulfone form.  However,
registrant states above that sulfoxide and sulfone residues move
downward within an oxic aquifer until they reach a deeper anaerobic
zone, whereupon they commence to degrade rapidly.  This is inconsistent
with common patterns of groundwater movement within aquifers.  The
dominant direction for groundwater is usually lateral, not downward. 
The primary characteristic of groundwater movement through aquifers is
stratified horizontal laminar flow; downward movement within an aquifer
is typically facilitated by further rainfall.  Thus, aerobic groundwater
near the upper boundary of the water table can stay oxic over relatively
long (primarily horizontal) distances and remain oxic until discharged
into a surface water body.

Page: 24     Paragraph: 3     Line: 7 

EPA Comment:  “However, the monitoring data in areas with historical
contamination confirm the high persistence of total aldicarb residues in
ground water.”

Bayer response:  The monitoring data confirm that aldicarb carbamate
residues are persistent under aerobic conditions in ground water.  The
persistence is related to temperature and the time required for movement
to deeper anaerobic ground water.  Changes to TEMIK® labels made in the
late 1980’s have essentially eliminated residues in drinking water
wells resulting from applications made after this date.  This has been
accomplished by eliminating uses in highly vulnerable areas such as Long
Island, the northeastern United States, Wisconsin, and northern coastal
California and instituting well set-backs in some of the use areas.

EPA response:  While temperature is likely an important factor, laminar
stratification within aquifers can inhibit movement into deeper
anaerobic layers, thus increasing persistence.  The implementation of
setbacks in some potentially sensitive regions is as yet inconclusive;
many of the wells sampled before and after setback requirements that
showed marked diminution of aldicarb residues also were fitted with
carbon filters in the interim.

Page: 25     Paragraph: 1     Line: 7 

EPA Comment:  “The quality of the lab studies was deemed questionable
because the half-life values from the field dissipation studies were
consistently (and substantially) longer than those from the lab studies
. . .”

Bayer response:  This statement is not correct and is probably due to
confusion between total carbamate residues and parent residues.  The
half-lives of total carbamate residues in the laboratory studies
conducted during the past 20 years are generally longer than measured in
the field studies although degradation rates for parent in laboratory
studies are faster than for total carbamate residues observed in field
studies.  Some of the degradation rates for total carbamate residues
observed in the earlier laboratory studies conducted under more moist
conditions than according to current guidelines show rates comparable to
field studies.

EPA response:  The Agency agrees with the above response.

Page: 25     Paragraph: 1     Line: 13 

EPA Comment:  “Therefore an intermediate value of 60 days, derived
from the field studies, was used.”

Bayer response:  As shown by the results in Jones and Estes (1995), this
is a reasonable value to use in the northern United States.  For
applications in the southern half of the United States, a half-life of
15-30 days would be more appropriate.

EPA response:  An aerobic soil half-life of 55 days has been determined
from re-evaluation of results from19 soils (Upper 90th pct bound on mean
for combined parent+sulfoxide+sulfone half-life from 19 soils). 
Although surface soils in the southern U.S. would presumably be warmer
than in the north (thus speeding degradation rates), temperature
differentials decrease dramatically with depth; subsurface temperatures
become more consistent with increasing depth throughout the world.  In
addition, during the growing season in northern climates (when
application occurs), there is less difference between northern and
southern region soil temperatures than would appear based on average
annual soil temperatures.  The input value of 55 days should be used in
all regions, as it is the most protective.

Page: 26     Table: 8     Line: 1 

EPA Comment:  “Input value of 60 days for aerobic soil”

Bayer response:  As shown by the results in Jones and Estes (1995), this
is a reasonable value to use in the northern United States.  For
applications in Mississippi, Georgia, and Florida, a half-life of 15-20
days would be more appropriate.

EPA response:  An aerobic soil half-life of 55 days has been determined
from re-evaluation of results from19 soils (Upper 90th pct bound on mean
for combined parent+sulfoxide+sulfone half-life from 19 soils). 
Although surface soils in the southern U.S. would presumably be warmer
than in the north (thus speeding degradation rates), temperature
differentials decrease dramatically with depth; subsurface temperatures
become more consistent with increasing depth throughout the world.  In
addition, during the growing season in northern climates (when
application occurs), there is less difference between northern and
southern region soil temperatures than would appear based on average
annual soil temperatures.  The input value of 55 days should be used in
all regions, as this is the most protective.

Page: 26     Table: 8     Line: 2 

EPA Comment:  “Input value of 120 days for aerobic aquatic”

Bayer response:  An input value of 5 days for the aerobic aquatic
half-life is appropriate based on the guideline aerobic aquatic
water/sediment studies for aldicarb, aldicarb sulfoxide and aldicarb
sulfone (MRID Nos. 45592107, 45592107, 45592107, respectively).  The
aldicarb aerobic aquatic half-life was 5.5 days – although the
half-life is based on the total system, it represents the water
half-life since aldicarb was only detected in the sediment in only

one interval, at low concentrations.  The aldicarb sulfoxide aerobic
aquatic half-life was 5 days for both the total system and the water
phase (from EFED Data Evaluation Record).  The aldicarb sulfone aerobic
aquatic half-life was 3.5 days for the whole system and water phase
(from EFED Data Evaluation Record).  Although the data evaluation
records note some minor deficiencies in the sulfoxide and sulfone
studies, these deficiencies do not affect the scientifically valid
calculation of the half-lives, and thus, an aerobic aquatic half-life of
5 days is appropriate.  As mentioned previously and described in more
detail in the discussion section, repeat studies with excellent material
balances showed half-life values of 3.0 and 2.4 days for aerobic and
anaerobic aquatic degradation of aldicarb sulfone.  The rapid aquatic
degradation of aldicarb under aerobic conditions is also supported by
the guideline study of Skinner (1995; MRID No. 43805702), which showed a
half-life of 9 hours.  Persistence in ground water is not an indication
of persistence in surface water due to microbial activity in surface
water.

EPA response:  MRID #44592107 has been used to determine that the
aerobic aquatic half-life of parent + degradates is 12 days (3X the
value of 4 days determined from this single acceptable study).  MRID
#43805702 is not a valid study.

Page: 26     Table: 9     Line: 1 

EPA comment:  Application Scenario:  “MS Cotton, (4.05 lbs ai/A) (1
app) = max”

Bayer response:  This table appears to indicate that 4.05 lbs ai/A of
aldicarb could be made in a single application, however, for use in
cotton in the state of Mississippi the TEMIK® 15G label restricts a
single applications to at- plant 1.5 lbs ai/A (10 lbs product/A) or a
post-emergent side-dress application of 3.0 lbs ai/A (20 lbs/A).

EPA response:  As stated above (see comment regarding Page 4, Paragraph
1, Lines 10-13), the MS cotton modeling scenario is representative of
cotton-growing regions of the southeast United States, and is not solely
representative of cotton grown in the state of Mississippi.  The Agency
agrees with Bayer that the ‘one time maximum application rate’
should be will be made in the revised risk assessment.  However, it
should be noted that the this revision of the one time maximum
application rate will not change the overall conclusions of the risk
assessment; there are still risks to terrestrial and aquatic species
resulting from the use of aldicarb on cotton at all application rates.

Page: 26     Table 9     Line: 2 

EPA comment:  Application Scenario:  “MS Cotton, (4.95 lbs ai/A) total
(2 apps) = max”

Bayer response:  This table appears to indicate that 4.95 lbs ai/A of
aldicarb could be made in a total of two applications, however, for use
in cotton in the state of Mississippi the TEMIK® 15G label restricts
usage to a maximum single at-plant application of 1.5 lbs ai/A (10 lbs
product/A) and a post-emergent side-dress application of 3.0 lbs ai/A
(20 lbs product/A) for a seasonal maximum of 4.50 lbs ai/A (30 lbs
product/A).

EPA response:  As stated above (see comment regarding Page 4, Paragraph
1, Lines 10-13), the MS cotton modeling scenario is representative of
cotton-growing regions of the southeast United States, and is not solely
representative of cotton grown in the state of Mississippi. The Agency
agrees with Bayer that the ‘two times application scenario’ should
be changed from 4.95 lbs ai/A (2 applications) to 2.48 lbs ai/A (2
applications) Corrections will be make in the revised risk assessment. 
.  However, it should be noted that the this revision of the one time
maximum application rate will not change the overall conclusions of the
risk assessment; there are still risks to terrestrial and aquatic
species resulting from the use of aldicarb on cotton at all application
rates.

Page:  26    Table 9	Footnote

EPA comment:  Footnote at the bottom of the table reads
“*Discrepancies between EECs in this document and the 2002 document
are due to changes in the meteorological data set (particularly
rainfall) used for the PRZM model runs in the interim.”

Bayer response:  In earlier comments to the Agency, Bayer pointed out
that this statement is false.  Bayer suggested that the main reason for
the discrepancies between EECs in this document and the 2002 document is
the assumption of how much pesticide is available on the surface of the
soil and susceptible to run-off.  In this document, 15% is used.  In the
2002 document, the chemical was assumed to be evenly mixed in the top
few inches of the soil, and so less was assumed available for runoff. 
Bayer further pointed out that the assumption of 15% that is used in
this modeling is a gross overestimate for real-world use of TEMIK®,
most of which is applied in-furrow or as a band below the surface of the
soil.  The Agency’s default assumption of 1% availability is more
appropriate for the vast majority of TEMIK® applications.  Under this
assumption, the EECs obtained would be 15-fold lower than the values
reported in Table 9.  While it is acceptable for the Agency to calculate
EECs for worst-case scenarios, it is important that EECs for typical
scenarios also be presented.  In this case, the Agency’s typical
values are not very typical at all because they do not reflect the
application method (in-furrow) that is predominantly used in many of the
crops considered.  In these instances, the real “typical” EECs
should be at least 15-fold lower.

In the Agency’s memorandum, “Environmental Fate and Effects Division
Error-Only Corrections and Revised Aldicarb Science Chapter”, dated
May 31, 2005, the Agency stated (page 8, paragraph 1) that “. . . it
appears that there are multiple causes for the discrepancies [in the
aquatic EECs] between the older 2002 document and the most recent
document, with changes in meteorological data sets being a relatively
minor contributor.”  Thus, the Agency has confirmed that this footnote
is in error.  The footnote should either be deleted or it should contain
an accurate explanation for why the modeling results changed so much
between the 2002 and current documents.

EPA response:  While it is true that changes in the meteorological data
appear to have a relatively minor effect on changes in EECs between this
document and previous iterations, the 15% vs. 1% availability cited
above is irrelevant.  The model runs used to obtain EECs do not
differentiate in this fashion (these availability percentages are
applied at a later point).  The PRZM/EXAMS model (operated using the
‘CAM 1 – soil incorporated’ setting) assumes 99% incorporation
within the top 5 cm (1-4 cm below surface), decreasing linearly with
depth.  Decreasing amount with depth is more consistent with real
agricultural practices than uniform incorporation with depth.  The
differences between the 2002 and 2005 documents are largely due to
changes in fate parameter inputs used in the model.

Page: 27     Paragraph: 1     Lines: 4-6 

EPA Comment:  “In NAWQ monitoring sites, aldicarb and its sulfone and
sulfoxide transformation products were detected infrequently (about 0.1
percent of all samples) at low concentrations (total residues <1.6
ppb).”

Bayer response:  The registrant agrees with the characterization of the
NAWQA data that residues were detected infrequently.  However, an
examination of the detections indicates that many of these infrequent
detections are suspect.  The discussion section provides a complete
response to this comment including discussion of data from other
monitoring programs.

EPA response:  Capturing high-end concentrations in streams is hindered
by the inability to obtain many samples at times corresponding to
post-application rainfall events.  Results obtained from typical
monitoring programs should be expected to reflect the lower, not upper,
range of concentrations.

Page: 27     Paragraph: 1     Lines: 6-8 

EPA Comment:  “However, targeted monitoring in smaller streams suggest
that aldicarb may occasionally pose a contamination hazard.  For
example, Williams and Harris (1996) found substantially higher aldicarb
concentrations in small southeastern streams after a rainfall . . .”

Bayer response:  The concentrations in this intermittent stream are
sufficient to result in residues in downstream rivers and reservoirs,
which is inconsistent with the findings of surface water monitoring
programs.  Therefore, the residues observed at this site must be a rare
occurrence.  Intermittent streams are not an appropriate location to
apply risk quotients.

EPA response:  Intermittent streams are likely the MOST appropriate
locations to apply risk quotients, as they are more apt to convey water
(and contaminants) under conditions where there is significant runoff
from associated agricultural fields.  Thus samples can generally be
obtained only when conditions resemble those where high-end
concentrations might be expected.

Page: 27     Paragraph: 2     Lines: 3-7 

EPA Comment:  “In the case of aldicarb, which is applied only in a
granular form, the method used in calculating terrestrial EECs took into
account this granular formulation and its soil incorporation.  The model
assumes that only 1% (in-furrow application) or 15% (banded application)
of the applied granules remain on the surface and have the potential for
terrestrial animal exposure.”

and 

Page: 27     Table: 10

EPA comment:  “Table 10.  Terrestrial EECs for aldicarb, calculated
based on application method, application rate (maximum and average), and
% unincorporation of granules.”

Bayer response:   While it is appropriate to consider the formulation
and soil incorporation during an aldicarb application when conducting
exposure modeling, it is also appropriate to use values that reflect how
the product is actually being applied in the field.  A survey of TEMIK®
usage and methods of application conducted in 2001 across the U.S.
(Hall, 2003) showed for, instance, that greater than 99% of cotton
applications are applied in-furrow or shanked.  Also see the report and
DVD, “TEMIK® 15G Brand Aldicarb Product Usage and Methods of
Application” submitted with this response (Hall, 2005).  By following
US EPA (1992), the assumed incorporation efficiency for in-furrow or
shanked applications is 99%.  The assumed incorporation for banded and
covered with a specific amount of soil is also 99%.  The assumed
incorporation for banded and lightly incorporated (i.e. with just a
press wheel) is 85%. Values listed below should be incorporated into
Table 10 (page 27) to reflect a more realistic assessment of application
types by crop.

EPA response:  Revisions are forthcoming in the future draft of the
document in light of the information provided by the registrant. 
Although the information provided by the registrant indicates that most
of the aldicarb applied to cotton is applied using in-furrow or shanking
methods, the pesticide label allows for applications using side-dress,
which may leave a higher percentage of granules available on the soil
surface.  For the purposes of a screening-level ecological risk
assessment, the Agency must use the application parameters on the
currently approved label that provide the most conservative estimation
of risk.  Therefore, unless it is removed from the label, it is assumed
that the side-dress application method can be used for aldicarb
application to cotton.  The risk assessment was amended to provide
values for both the banded, soil incorporated application method (99%
incorporation efficiency) as well as the side-dress application method
(85% incorporation efficiency)[“Comparative Analysis of Acute Avian
Risk from Granular Pesticides” (EPA, 1992)].  

Page: 27     Paragraph: 2     Line: 11 

EPA comment:  “EEC = [Application rate x 453,950 mg/lb] / [# of rows
per acre x bandwidth x row length]” and

Page: 27     Table: 10 

EPA comment:  “Table 10.  Terrestrial EECs for aldicarb, calculated
based on application method, application rate (maximum and average), and
% unincorporation of granules.”

Bayer response:  The Agency's method of calculating the amount of
TEMIK® granules per ft2 only within the application band and ignoring
the fact that only a small fraction of the crop field is actually
treated does not make practical sense and is inappropriate for a
granular product that is unattractive to birds, is applied to non-food
source crops, and is efficiently incorporated.  Birds do not forage only
in pesticide-treated bands.  The Agency’s methodology produces
artificially inflated hazard indices for band applications in comparison
to a broadcast application with the same soil incorporation efficiency. 
Intuitively, treatment of 100% of the field area represents a greater
hazard than treatment of 3 to 10% of the field area, but the Agency’s
methodology calculates 10-fold greater exposure for the band application
that for the broadcast application.  There is no scientific
justification for performing the calculation this way.

Using cotton, as an example from Table 10, page 27, of the preliminary
risk assessment, the EEC calculation is (4.05 lb ai/Acre X 15%
unincorporated granules X 453,950 mg/lb) / (13 rows/acre X 0.33 ft
bandwidth X 1005 ft) = 64 mg ai/row ft2.   Note that this assessment is
based on exposure to the product only within the 4-inch band.  The
balance of the area, 90% of the total area, which does not have any
TEMIK® applied is ignored in these calculations.  Correcting for the
percentage of applied area per acre, 4,311.5 row ft2 / 43,560 ft2/acre,
reduces this value to 6.33 mg ai/ Acre.

EPA response:  As noted above, the LD50 per square foot method is an
index of the total aldicarb available throughout the area.  Although
band treatments result in less toxicant having to be applied (on a per
acre basis), they also result in higher concentrations being applied
over the banded area.  Studies have been conducted which show that birds
confined to band-treated areas suffer greater mortality than birds
confined to broadcast treated areas (at the same pound/acre basis). 
This clearly suggests that band treatment applications actually increase
the probability of an animal consuming a lethal dose (Felthousen, 1977).
 The treated row calculation does not factor in the area between rows
that is not treated, since birds and other wildlife focus their foraging
activity on the disturbed and/or treated area (USEPA, 1992).  The
Agency’s method of calculating the amount of toxicant per square foot
is has been approved by the SAP and is the currently accepted method as
discussed in the “Overview of the Ecological Risk Assessment
Process” document (2004). 

B.  Ecological Effects Characterization

Page: 29     Paragraph: 1     Lines: 7-9 

EPA comment:  “A supplemental study from the open literature also
concluded that aldicarb was moderately toxic to the fathead minnow
Pimephales promelas with a reported 48-hour EC50 was 8860 ppb (Moore et
al. 1998).”

Bayer response:  In addition to the study cited, Pickering and Gilliam
(1982) reported the LC50 of the fathead minnow to be 1370 ppb.  This
study has been accepted as a core study by the Agency and was used in
EFED’s 2001 risk assessment.

EPA response:  The Pickering and Gilliam (1982) study with a fathead
minnow LC50 of 1370 ppb will be added to the text and Table 11 as an
additional registrant submitted study in a forthcoming risk assessment. 
However, this value will not be used in risk quotient calculations
because it is not the most sensitive endpoint.

Page: 30     Paragraphs: 1-2 

EPA comment:  “A freshwater fish early life-stage test using aldicarb
(99% ai) has been conducted with the fathead minnow.  The NOAEC was 78
ppb, the LOAEC was 156 ppb and the MATC (median acute toxicity
concentration = geometric mean of the NOAEC and LOAEC) was 100 ppb. . .
. However, according to the acute freshwater fish data the bluegill
sunfish was the most sensitive freshwater fish tested with an LC50  of
52 ppb.  This LC50 of 52 ppb is lower than the NOAEC of 78 ppb for the
fathead minnow, indicating that the bluegill is much more sensitive to
aldicarb than the fathead minnow.  Therefore, chronic risk to fish was
also assessed using a predicted chronic ENEC for the bluegill sunfish
estimated using the acute-to-chronic ratio for the fathead minnow.”

Bayer response:  The extrapolation procedure the Agency has used to
calculate a predicted chronic ENEC for the bluegill has many sources of
uncertainty and should be avoided if an acceptable straightforward
procedure for deriving an appropriate chronic fish endpoint is
available.  For example, the acute-to-chronic ratio (ACR) calculation is
highly susceptible to variability found in acute and chronic toxicity
values.  There are several fathead minnow acute and chronic values the
Agency could have chosen to use in the ACR calculation.  The combination
of values that were used are the ones that result in the largest
possible ACR (114), which in turn yields the lowest possible estimate of
the ENEC for the bluegill.  In the 2001 draft chapter, EFED calculated
an ACR for the fathead minnow that was an order of magnitude lower. 
Based on the toxic mechanism of action of aldicarb, a high acute to
chronic ratio is quite unexpected.  It is well known that animals that
do not die from exposure to carbamate pesticides usually recover fully
and effects on reproduction and growth are generally not seen at
concentrations far below those that are lethal.  For example, the rat
LD50 value is 0.9 mg/kg and the rat chronic NOAEL is 0.4 mg/kg/day. 
Thus, the ACR for the rat is 2.3.

A more straightforward approach to the problem is use the sheepshead
minnow chronic NOAEC as a surrogate for the bluegill.  The sheepshead
minnow has been demonstrated to be approximately equal in sensitivity to
aldicarb as the bluegill sunfish.  The reported LC50 for the sheepshead
minnow ranges from 41 to 170 ppb with a geometric mean of 83 ppb.  The
reported LC50 values for the bluegill sunfish range from 52 to 115 ppb
with a geometric mean of 72 ppb.  Based on this comparison, tests with
the sheepshead minnow should be a reasonably good surrogate for tests
with the bluegill sunfish.  The chronic NOAEC value experimentally
determined for the sheepshead minnow is 50 ppb.  It is more reasonable
to use this value to estimate the NOAEC for the bluegill than to
extrapolate an ACR from the fathead minnow.   Bayer believes this value
should be used in the risk assessment.  It also represents the lowest
fish chronic NOAEC value available that has been experimentally measured
for aldicarb. Therefore, use of this value in the risk assessment is
consistent with the Agency’s standard practice.

EPA response:  According to the Agency’s “Overview of the Ecological
Risk Assessment Process in the Office of Pesticide Programs, USEPA”
(2004), the acute and chronic endpoints can be selected from the most
sensitive species tested within that organism group (page 35, Section
V.A.5).  Therefore, the acute to chronic ratio calculation to determine
the ENEC for the bluegill sunfish is estimating the chronic endpoint for
the most acutely sensitive freshwater fish species.  This is procedure
is according to the Overview document specifications.

The use of the sheepshead minnow NOAEC (an estuarine/marine fish
species) as a surrogate for the bluegill sunfish will be considered and
evaluated in the forthcoming ecological risk assessment in addition to
the estimated bluegill ENEC.  A discussion of the range of RQs will be
included.

Page: 34     Paragraphs: 2-3 

EPA comment:  “The chronic risk to freshwater invertebrates is based
on the estuarine/marine chronic study using Mysidopsis bahia . . .￿

“A supplemental chronic toxicity study was submitted that evaluated
the effect of aldicarb (99.9% ai) on  Daphnia magna.  An EC50 of 90 ppb,
NOAEC of 20 ppb, and LOAEC of 60 ppb were reported for mortality and
immobilization.  An EC50 with a range of 190 to 570 ppb, NOAEC of 190
ppb and LOAEC greater than 190 ppb were reported for reproductive
effects.  The most sensitive endpoint was reproductive effects [MRID No.
45592112].  This study is classified as supplemental and is not
upgradeable to core due to the study’s deviations. Therefore, the
aquatic risk assessment will utilize the mysid shrimp data (NOAEC = 1.0
ppb).”

Bayer response: The risk assessment for freshwater invertebrates should
be carried out with Daphnia magna chronic data available from the
supplemental study.  Studies rated as supplemental by the Agency are
considered to be scientifically sound and suitable for use in risk
assessments.  In this case, since the most sensitive endpoint was
mortality/immobility, the MATC (geometric mean of NOAEC and LOAEC, = 35
ppb) should be used as the chronic toxicity endpoint.  The Agency’s
statement, “The most sensitive endpoint was reproductive effects” is
incorrect. Since a scientifically sound Daphnia magna chronic study is
now available, it is no longer appropriate to use the Mysid chronic
toxicity value in the freshwater invertebrate risk assessment.

EPA response:    SEQ CHAPTER \h \r 1 According to the memorandum dated
December 14, 1993, from A. Maciorowski, EFED to P. Poli, SRRD (D196663):
“EEB has previously received and reviewed a valid Mysid shrimp chronic
toxicity study that resulted in a calculated MATC of 1-1.5 ppb. Since
the acute data indicates that the Mysid shrimp is more sensitive to
aldicarb than Daphnia magna (Mysid shrimp LC50 = 16 ppb, Daphnia magna
LC50 = 410.7 ppb) and EEB does have valid Mysid shrimp chronic toxicity
data, EEB is willing to waive the requirement for the Guideline 72-4(b)
Freshwater Invertebrate Life Cycle Study. However, as a result all
aquatic risk assessments utilizing invertebrate chronic toxicity data
will be based on the Mysid shrimp data.”  Therefore, the requirement
for the Guideline Study 72-4(b) was waived at the registrant’s request
with the stipulation the mysid shrimp NOAEC of 1.0 ppb would be used in
the risk calculations.

In addition, studies rated as supplemental by the Agency are considered
to be scientifically sound and suitable for use in risk assessments at
the discretion of the EPA scientist.  According to the the Agency’s
“Overview of the Ecological Risk Assessment Process in the Office of
Pesticide Programs, USEPA” (2004), professional judgement is used to
determine whether or how toxicological endpoints are included in the
risk assessment (p. 46, Section V.B.2).  The supplemental chronic
Daphnia magna study is supplemental and not upgradeable to a core study
due to significant study deviations.

EPA agrees the sentence “  SEQ CHAPTER \h \r 1 The most sensitive
endpoint was reproductive effects [MRID 45592112].” is incorrect and
will be corrected to “  SEQ CHAPTER \h \r 1 The most sensitive
endpoint was   SEQ CHAPTER \h \r 1 mortality and immobilization [MRID
45592112].” in the forthcoming risk assessment.

Page: 34     Paragraph: 4 

EPA comment:  “Based on results from the preferred test species,
sheepshead minnow, the LC50 falls in the range of 41 to 170 ppb.” . .
. . “The most sensitive estuarine/marine fish species tested was the
sheepshead minnow with an LC50 of 41 ppb . . .”

Bayer response:  When there are multiple scientifically sound studies on
the same test species, the geometric mean of the individual study
measurements should be used as the reference value for the species.  In
this case, there are two scientifically sound acute toxicity studies
with the sheepshead minnow.  One produced an LC50 estimate of 41 ppb,
the other produced an LC50 estimate of 170 ppb.  The geometric mean of
these two values, 83 ppb, should be used as the toxicity reference value
for the sheepshead minnow.  It is poor scientific practice to ignore
studies that produced higher toxicity estimates and use only the study
that produced the lowest value.

EPA response:  As stated in the Agency’s “Overview of the Ecological
Risk Assessment Process in the Office of Pesticide Programs, USEPA”
(2004), the lowest LC50 value can be used as an input to the risk
quotient methods of expressing risk (Section V.B.2).  Therefore, the use
of the sheepshead minnow LC50 value of 41 ppb in the risk assessment is
correct.

Page: 36     Paragraph: 5     Lines: 1-2 

EPA comment:  “The LD50 is 1.0 mg/kg for aldicarb, and the LD50 for
aldicarb sulfone is 33.5 mg/kg.  The most sensitive species tested for
both aldicarb and aldicarb sulfone is the mallard duck.”

Bayer response:  There are multiple scientifically sound studies
available for the mallard.  Three studies have been conducted with adult
birds per EPA guidelines.  These produced LD50 values of 1.0 mg/kg
(Beavers and Fink, 1979), 3.4 mg/kg (Hudson, Tucker and Haegele, 1979)
and 4.44 mg/kg (Hudson, Tucker and Haegele, 1972).  When there are
multiple scientifically sound studies on the same test species, the
geometric mean of the individual test measurements should be used as the
toxicity reference value for the species.  In this case, the geometric
mean is 2.5 mg/kg.  Based on this value, the mallard may not be the most
sensitive species to aldicarb.

The entry for mallard LD50 in Table 20 (page 38) should be changed
accordingly to 2.5 mg/kg.

EPA response:  As stated in the Agency’s “Overview of the Ecological
Risk Assessment Process in the Office of Pesticide Programs, USEPA”
(2004), the lowest LD50 value is used as an input to the risk quotient
methods of expressing risk (Section V.B.2).  Therefore, the mallard duck
LD50 value of 1.0 mg/kg used in the risk assessment is correct.  

Pages: 37     Paragraph: 2 

EPA comment:  “No avian reproduction studies were submitted by the
registrant.  Avian reproduction studies on the mallard duck and bobwhite
quail using the TGAI are required for aldicarb because the following
conditions are met: (1) birds may be subject to repeated or continuous
exposure to the pesticide, especially preceding or during the breeding
season, and (2) the pesticide is stable in the environment to the extent
that potentially toxic amounts may persist in animal feed.”

Bayer response:  Bayer believes avian reproduction studies would provide
no useful information and therefore should not be required because (1)
the Agency has no methods to use the toxicity data produced by these
studies in a risk assessment, (2) TEMIK® granules do not persist in the
field (they rapidly disintegrate with the first rainfall, or upon
contact with moist soil), and so the important route of exposure for
birds (ingesting granules) will not occur over a long period of time,
and (3) studies in other animals (mammals, daphnids, etc.) all point to
the conclusion that aldicarb is an acute toxicant only (due to rapid
metabolism and reversibility of the cholinesterase inhibition, animals
can withstand repeated exposure up to levels that are nearly lethal). 
Aldicarb is only formulated as a granular product (TEMIK® Brand
Aldicarb Pesticide).  The Agency’s policy for assessing avian risks of
granular products is to calculate LD50s per square foot.  The Agency has
no methodology for calculating an EEC for avian diets for granular
products.  Thus, if avian reproduction NOAECs were available, they could
not be used in a risk assessment. As confirmation of this, one only has
to look at the way chronic mammalian test results were used in this risk
assessment.  They weren’t!  The Agency has high quality chronic
toxicity studies of aldicarb with mammals yet the entry in Table 20
(page 38) for the mammalian chronic toxicity endpoint used in the risk
assessment is “N/A”.  There were chronic NOAELs from tests with
mammalian species but they weren’t used in any RQ calculations for the
reasons indicated above.  What is the point of requiring two very
expensive (>$100,000 each) studies when the data generated will not
affect the outcome of, or even be used in, a risk assessment?  The
requirement of these studies should be waived.

EPA response:  See response to Page: 18 1. Ecotoxicity Information Gaps

Page: 39     Paragraph: 1     Lines: 2-3 

EPA comment:  “A field study by Hawkes et al. (1996) demonstrated
reduced cover-seeking in mourning doves and bobwhite quail following
exposure to a lethal aldicarb dose.”

Bayer response:  The statement is technically correct, but we suggest
the following rewording to make the main finding clearer and more
accurate.  We suggest the statement be changed to read, “A field study
by Hawkes et al. (1996) demonstrated that mourning doves and bobwhite
quail administered a lethal dose became immobilized before they were
able to seek cover.”  We suggest the rewording to make it clear that
this study demonstrated that the knock-down effect of the compound is so
rapid it is unlikely that birds leave aldicarb treated fields and die
off-site where they are less likely to be found or noticed.

EPA response:  The wording was changed in the aforementioned paragraph.

Page: 39     5.  Incident Data Review 

EPA Comment:  “There have been 29 incidents related to aldicarb
reported in the Environmental Incident Information system (EIIS)
database (reported to the Agency from 1988 to 2005).  Of these 29
incidents . . .

Bayer response:   The fact that the EIIS reports 29 incidents is not
applicable to EFED’s argument that aldicarb kills fish, birds and
other wildlife.  As stated by EFED, only 2 of the 29 incidents in this
18-year period are associated with a registered use of aldicarb.  Of
these two incidents, one (I000165-052, 6/12/92) involves a fish kill in
North Carolina following application to corn (a misuse) as well as
tobacco (a registered use).  The other incident cited by EFED cannot be
identified from the discussion on pages 39-40 or from Appendix F.

In this document EFED has corrected the record somewhat by pointing out
that the vast majority of incidents in their data base are clearly
attributable to misuse, and hardly any are from registered use. 
However, they do not clearly state what is the obvious and most
important finding from reviewing the incident record: that in 35 years
of widespread use in the United States, including use in California
where a concerted effort is made by state personnel to investigate
pesticide poisoning incidents, there is not a single document field bird
field kill incident associated with registered use of aldicarb in
agricultural crops.  An objective review of the available evidence
should point this out.

EFED correctly states that “. . . before a pesticide incident can be
reported or investigated, the dead animals must be found.”  EFED’s
discussion as to why dead animals may not be found following application
is rather disingenuous and specious because there are numerous reports
of both bird and fish kills associated with other pesticides.  And there
are numerous incidents associated with illegal misuse of aldicarb which
demonstrates that wildlife killed by aldicarb does not always go
unnoticed and unreported.  Rather than developing true and factual
assessments of risk, it appears that EFED is simply putting forth
theories to justify hypothetical risk assessment calculations.

A comparison between the incident record associated with use of granular
carbofuran and granular aldicarb is both appropriate and instructive. 
Both aldicarb and carbofuran are highly toxic carbamate insecticides
formulated as granules.  Numerous bird kill incidents have been reported
for granular carbofuran, but not for aldicarb.  The reason for this
difference has very little to do with the inherent toxicity of these two
active ingredients to birds (they are both highly toxic) or the number
of LD50s/ft2 that result when these products are applied.  Instead, the
reasons granular aldicarb pose a low risk is because of well defined
application practices that essentially eliminate the possibility of
exposure to birds, continuous and aggressive product stewardship, the
fact that TEMIK® granules are inherently unattractive to birds as
either food or grit, and because TEMIK® granules do not remain intact
long in the environment since they rapidly disintegrate upon contact
with moist soil.  Granular carbofuran, on the other hand, poses a high
risk because it is formulated on a sand carrier that is very attractive
to birds as grit, and does not physically break down in the environment.
 So long as rainwater doesn’t wash the active ingredient off of the
granule, carbofuran granules can remain intact for months and bird kills
have occurred a month or more post application.  The Agency’s 1992
Comparative Analysis of Acute Avian Risk from Granular Pesticides
presented data on acres treated by crop (see EPA 1992, Appendix 3). 
Granular aldicarb was estimated to be used annually on 5.035 million
acres, mainly row cropland where it is applied to bare soil at planting
time.  Granular carbofuran was estimated to be used annually on 5.025
million acres, mainly row cropland where it was applied to bare soil at
planting time. Although the specific crops differ somewhat, both are
applied to large bare ground fields and the species of wildlife that use
such fields do not differ substantially.  For example, there is no
reason to think that more large, conspicuous species of wildlife used
cropland treated with carbofuran than used cropland treated with
aldicarb.  The two compounds have similar toxicokinetic profiles in
vivo.  Both are fast acting, reversible cholinesterase inhibitors.
Balcomb et al. (1984) reported average time to death for birds dosed
near the LD50 level was 14 minutes with carbofuran, and 18 minutes for
aldicarb.  Kendall (1990) and (1992) reported that mourning doves and
bobwhite quail dosed with aldicarb at levels near the LD50 died as
quickly as within 5 to 15 minutes.  Clearly there is little difference
between these compounds in how rapidly birds become incapacitated if
they ingest a lethal dose.  Based on the time to death data, effects
occur only slightly faster, if at all, in carbofuran poisoned birds in
comparison to aldicarb.  The difference is certainly not enough to
explain why more than 400 incidents have been reported with carbofuran
and only 2 for aldicarb

EPA response:  It is the Agency’s policy to report all of the
incidents reported to the EIIS database.  All of the incidents reported
to the database are discussed in detail in the accompanying appendix
(See Appendix F), as well as in the text.  The second incident that was
reported to the database resulting from registered use is Incident
I012089-009, which resulted in damage to 1050 acres of potato crops. 
This incident is referenced in Appendix F.  The comparision of aldicarb
to carbofuran has been removed from the risk assessment.   

  SEQ CHAPTER \h \r 1 Aldicarb and carbofuran belong to a group of
pesticides known as N-methyl carbamates.  As such, it is reasonable to
expect that aldicarb and carbofuran could have similar toxicity
profiles, similar environmental fate properties, similar target pests,
and similar ecological risk issues.  Both pesticides are fast acting,
“quick knock-down” agents that are very efficacious against a wide
variety of agricultural pest species including various sucking insects,
mites and nematodes.  Both of these chemicals act as actylcholinesterase
inhibitors and have been classified as being very highly toxic to birds
and mammals on an acute basis.  In addition, they are very highly toxic
to freshwater fish, freshwater invertebrates, estuarine/marine fish, and
moderately to very highly toxic to estuarine/marine invertebrates on an
acute basis.  Both pesticides are formulated into granular products that
are typically soil incorporated when applied.

From an environmental fate perspective, both pesticides are highly water
soluble and relatively non-persistent in the environment, while their
toxic degradates can persist for longer periods of time.

With these similar properties, it could be presumed that ecological risk
to non-target organisms would be comparable.  In fact, acute avian and
mammalian risk quotients for aldicarb and carbofuran greatly exceed
Agency levels of concern.  However, while an analysis of documented
incident reports and field studies for carbofuran appears to support the
predicted ecological risk, this does not hold true for aldicarb.  In
contrast to aldicarb, where there have been 17 incidents involving bird
kills of generally less than 10 animals, there have been 131 reported
carbofuran bird kill incidents, many of which involved tens or hundreds
of animals.  One-hundred of the carbofuran incidents were attributed to
granular use, and 31 were a result of flowable application.  It should
be noted that the incidents attributed to carbofuran use have
significantly declined following the 1991 phase-out (only 8 incidents
involving granular carbofuran have been reported post-1991).  In
addition, there have been numerous incidents involving mammal deaths
following carbofuran application, whereas there are relatively few
involving aldicarb.

The following discussion provides background on this issue and proposes
possible explanations.  Further data are needed before any conclusions
can be drawn, and it should be noted that this is not an in-depth
analysis of all potential justifications for incident report
dissimilarities.

Use Comparison

Aldicarb has been used since 1970 on a variety of crops in the United
States, including dry beans, cotton, citrus, peanuts, pecans, potatoes,
sorghum, soybeans, sugarcane, sugarbeets, sweet potatoes, and
ornamentals, as well as on alfalfa, beans, coffee, tobacco, and yams in
specific states and territories.  It is applied only in granular form,
and approximately 4.5 million pounds of aldicarb active ingredient (ai)
are used per year on 5.45 million acres (Bayer, 2004).  Total pounds of
aldicarb applied in the US have increased in recent years, rising from
1.4 million pounds ai applied in 1991 (NASS) to the current 2004 rate of
4.5 million pounds.  

Carbofuran was first registered in 1969 and is available in flowable,
granular, and wettable powder formulations.  Registered uses include
alfalfa, artichokes, bananas, clover, coffee, corn, cotton, cucumbers,
flax, garlic, grapes, melons, peanuts, peppers, plantains, potatoes,
pumpkins, small grains (wheat, oats, barley), sorghum, soybeans, squash,
strawberries, sugar beets, sugarcane, sunflowers, tobacco, Bermuda
grass, forest trees, non-bearing fruit trees, and ornamentals.  The
granular form was used until 1991, when it was voluntarily phased-out to
reduce risk to avian species.  Currently, only 2,500 pounds ai/year of
granular carbofuran is approved for use on minor crops, including
bananas, curcubits, pine seedlings, and spinach for seed.  Flowable
forms are applied via ground or aerial spray.  Currently, 1.5 million
pounds ai of flowable carbofuran is applied per year in the United
States.  This represents a significant reduction in use: approximately 3
million pounds were used in 1990, prior to the granular phase-out.

As stated above, the majority of the reported carbofuran incidents
occurred pre-1991 and were related to granular carbofuran use. 
According to 1992 agriculture census data (NCFAP, 1992), carbofuran was
applied to approximately 260 million acres, whereas aldicarb was applied
to only 30 million acres.  This difference is largely due to
carbofuran’s numerous large crop uses (e.g., corn), whereas aldicarb
use is restricted to fewer crops that are less widespread.  This large
difference in total acreage to which the two chemicals were applied is a
possible explanation for the disparity in ecological incident reports
seen between aldicarb and carbofuran during the time that granular
carbofuran was widely used.

Toxicity

Both carbofuran and aldicarb are classified as very highly toxic to
birds (LD50 values of 0.24 ppm and 1 ppm, respectively) and mammals
(LD50 values of 6 ppm and 0.9 ppm, respectively) on an acute basis.  In
addition, they are very highly toxic to freshwater fish, freshwater 
invertebrates, estuarine/marine fish, and moderately to very highly
toxic to estuarine/marine invertebrates on an acute basis.  

Aldicarb and Carbofuran Toxicity Comparison

	Carbofuran	Aldicarb

Avian Acute Oral	LD50  Range of 0.24 ppm to 7.9 ppm 

(highly to very highly toxic)	LD50 of 1.0 ppm 

(highly to very highly toxic)

Avian Subacute Dietary	LC50  Range of 79 to 714 ppm

(very highly toxic to moderately toxic) 	LC50  71 ppm 

(very highly toxic)

Mammalian Acute oral	LD50 of 6.0 ppm	LD50 of 0.9 ppm



A granular pesticide feeding study performed on adult house sparrows and
red-winged blackbirds demonstrated similar time to first death between
aldicarb and carbofuran (Balcomb et al., 1984).  House sparrows died
within 14 minutes following acute exposure to carbofuran, and 18 minutes
with aldicarb.  Red-winged blackbirds demonstrated similar values of 12
minutes for carbofuran and 15 minutes for aldicarb.  

Aldicarb is an order or magnitude more toxic to mammals than carbofuran.
 The  in vitro inhibition of cholinesterase activity by aldicarb is
spontaneously reversible in mammals, and symptoms (salivation, cramps,
trembling, and sedation) usually subside within six hours in mammals
(Risher et al., 1987).   Carbofuran is also highly toxic after acute
oral exposure in mammals.  Toxic symptoms indicative of cholinesterase
inhibition are observed within minutes of exposure and can continue for
up to three days (WHO 2004).

A mammal exposed to carbofuran (unless death is the immediate outcome)
will have difficulty leaving the site of exposure due to cholinesterase
inhibition and may be repeatedly exposed to the carbofuran for up to
three days.  This could lead to the higher numbers of mammalian wildlife
kills that are observed in fields and adjacent properties. Aldicarb, on
the other hand, requires higher concentrations for cholinesterase
inhibition and the effects are spontaneously reversible within 6 hours. 
Mammals would experience the effects at the site of exposure but would
return to normal behaviors within six hours and would be able to move
from the area.

Field Studies

Despite the lack of incident reports for aldicarb, avian field studies
using this pesticide definitively demonstrate potential ecological risk
following labeled application (See Section IV.A.2).  Based on the
results of these field studies, it cannot be concluded that low numbers
of ecological incident reports indicate low ecological risk.  It should
be noted that there are no mammalian field studies for aldicarb.

Differences in Granular Size

Despite the fact that both aldicarb and carbofuran are granular N-methyl
carbamates, the size, shape, and percent active ingredient of the
granules could affect incident reporting.  Carbofuran granules range
from 3% to 15% ai, whereas aldicarb granules are 10% or 15% ai.  Given
that the majority of incident reports do not specify the granule type,
it is not plausible to directly compare the two chemicals (e.g., one
could expect less incidents for aldicarb if a lower percent ai granule
was applied).  Additionally, granules of carbofuran can range in size
from 0.35 to 0.63 mg, and aldicarb granules range from 0.45 to 0.50 mg
(Balcomb et al., 1984), leading to potentially large differences in the
amount of pesticide contained per granule of each chemical.  

Differences in Granular formulation

As stated above, both aldicarb and carbofuran exist in granular form,
although carbofuran granular use has been severely restricted since
1991.  Carbofuran granules are a purple silica pellet, whereas aldicarb
granules use a gypsum or corncob carrier and are brownish/black in
color.  Several studies have investigated the differences in pellet
consumption based on composition.  The high visibility of carbofuran
makes the granules more visible and attractive to birds, whereas the
more neutral-toned aldicarb pellets blend into the soil background (Best
and Fischer, 1992).  Additionally, it was demonstrated that house
sparrows showed little attraction to gypsum or corn cob-type pellets
(containing no pesticide) and high attraction to silica pellets (Best
and Gionfriddo, 1994).  Other factors, such as soil texture and color,
weather conditions, pesticide application rate, and species foraging
habits could also significantly affect granular consumption.

Although it is possible that there are significant differences in
granule consumption by bird or mammal species between carbofuran and
aldicarb,  these laboratory studies represent only a few of the numerous
bird species potentially exposed to these pesticides in the wild, and do
not examine potential mammalian granule consumption.  Additional routes
of exposure, such as dermal contact, ingestion of contaminated worms,
insects, plants, soil, and water, and inhalation must also be
considered.  Further information exploring this hypothesis is warranted
before any definitive conclusions can be drawn.

Publicity

Publicity involving carbofuran bird kill incidents dramatically
increased incident reporting in the 1980s and early 1990s.  An Avian
Risk Reduction Plan was put in place by the Virginia Pesticide Control
Board and FMC (carbofuran registrant) in 1990 to reduce avian risk while
allowing continued pesticide use in Virginia, which involved intensive
monitoring (Stinson et al., 1994).  Additionally, the American Bird
Conservancy (ABC) and other organizations brought wide-spread publicity
to bird kill incidents in New York, leading to an increase in reported
incidents.  Conversely, similar publicity resulting in increased
incident reporting did not occur for aldicarb, a possible explanation
for the low number of reported aldicarb-related incidents.  

Conclusions

In summary, there are numerous explanations for the large differences in
carbofuran and aldicarb incident reporting, including pesticide use,
toxicity, species affected, granule size and formulation, field studies,
exposure pathways, and publicity.  Although these hypotheses attempt to
explain the seemingly inconsistent magnitude of incident reporting
between aldicarb and carbofuran, they are by no means an exhaustive
list.  They are presented as a partial listing of uncertainties and
unknowns that must be considered when comparing the two chemicals.

IV.  RISK CHARACTERIZATION

A.  Risk Estimation – Integration of Exposure and Effects Data

Page: 42     Table: 21 

EPA comment:  “Table 21: Acute and chronic risk quotients for
freshwater fish”

Bayer response:  The LOC exceedences for acute and chronic risk
indicated in Table 21 come from RQ calculations that assumed 15% of
granules applied remain on the soil surface.  This assumption about the
proportion of granules left unincorporated is the Agency’s standard
assumption for band-incorporated, side-dress and broadcast-incorporated
applications (US EPA, 1992).  For in-furrow, drilled, or shanked-in
applications and banded applications covered with a specific amount of
soil, the standard assumption is 1% of granules available according to
the Agnecy’s Comparative Analysis of Acute Avian Risk from Granular
Pesticides (EPA 1992).  In a survey of TEMIK® usage and methods of
applications, Hall (2003 and 2005) showed that the vast majority of
TEMIK® applications are made via in-furrow, drilled or shanked-in
applications.  These data were discussed previously in comments on Table
10.  Thus, for the vast majority of real-world TEMIK® applications, the
RQs are 15-fold less than what is indicated in Table 21, and there would
be no acute LOC exceedences.  The chronic RQs presented in Table 21 over
predict chronic risk levels for 2 reasons (1) the EECs are too high
because (a) the assumed number of granules remaining on the surface is
overestimated (for reasons presented immediately above), and (b) the
aquatic half-life assumed in the modeling is too long (60-120 days used
vs. the 3- day half-life found in aquatic metabolism studies) and (2) an
inappropriate chronic toxicity value is used (see our comments on the
fish chronic NOAEC value, Page: 30, Paragraphs: 1-2).

EPA response:  The explanation of EEC calculations and % incorporation
is discussed in detail in the EPA response to Bayer’s comments for
page 27, Table 10 and will be addressed in the forthcoming risk
assessment.  EPA disagrees that an inappropriate chronic toxicity value
was used (see EPA Response to Bayer’s Comment, page 30, Paragraphs
1-2).

Page: 43    Table: 22 

EPA comment:  “Table 22: Acute risk quotients for freshwater
invertebrates.  Risk quotients for freshwater invertebrates based on an
adult Daphnia laevis EC50 of 20 ppb using technical grade Aldicarb
(Moore et al. 1998).”

Bayer response:  The toxicity value used in these RQ calculations comes
from a test with Chironomus tetans, not Daphnia laevis.

As discussed in comments immediately above regarding freshwater fish
risk estimation, the EECs presented are based on unrealistic assumptions
of proportion of granules left unincorporated.  Incorporating the
TEMIK® usage data from Hall (2003 and 2005), the following more
complete table of RQ values and LOC exceedences is obtained. Data in
bold type are new and data in regular type are same as presented in EFED
Table 22.

EPA response:  The explanation of EEC calculations and % incorporation
is discussed in detail in the EPA response to Bayer’s comments for
page 27, Table 10 and will be addressed in the forthcoming risk
assessment.

EPA agrees that the toxicity value used in the RQ calculations come from
a test with Chironomus tetans and Table 22 will be corrected to reflect
this correction in the forthcoming risk assessment.

Page: 44    Table: 23 

EPA comment:  “Table 23: Chronic risk quotients for freshwater
invertebrates.  Chronic RQ calculated using NOAEC for mysid shrimp
(Americamysis bahia) (1.0 ppb).”

Bayer response:  As mentioned in comments on the effects assessment, the
chronic toxicity value that should be used in this assessment is the
MATC from a chronic test with Daphnia magna.  This chronic toxicity
value is 35 ppb.  This is the lowest chronic NOAEC available from tests
with freshwater invertebrates.   Daphnia magna is a freshwater species;
Americamysis bahia is an estuarine/marine species.

As discussed in comments above regarding freshwater fish risk
estimation, the EECs presented are based on unrealistic assumptions of
proportion of granules left unincorporated.  Incorporating the TEMIK®
usage data from Hall (2003 and 2005), the following more complete table
of RQ values and LOC exceedences is obtained.  Data in bold type are
new; data in regular type are same as presented in EFED Table 23.

EPA response:  The explanation of EEC calculations and % incorporation
is discussed in detail in the EPA response to Bayer’s comments for
page 27, Table 10 and will be addressed in the forthcoming risk
assessment.

The use of the mysid shrimp NOAEC of 1 ppb is appropriate (See EPA
Response to Bayer’s comment page 34, Paragraphs 3 to 4).

Page:  45     Table: 24 

EPA comment:  “Table 24: Acute and chronic risk quotients for
estuarine/ marine fish.  Risk quotients for estuarine/marine fish based
on a sheepshead minnow (MRID No. 40228401) LC50 of 41 ppb and sheepshead
minnow (MRID No. 00066341) NOAEC of 50 ppb using technical grade
aldicarb.”

Bayer response:  As discussed in comments above regarding freshwater
fish risk estimation, the EECs presented are based on unrealistic
assumptions of proportion of granules left unincorporated. 
Incorporating the TEMIK® usage data from Hall (2003 and 2005), the
following more complete table of RQ values and LOC exceedences is
obtained. Thus, for the vast majority of real-world TEMIK®
applications, the RQs are 15-fold less than what is indicated in Table
24, and there would be no acute LOC exceedences.

No chronic RQs were presented in Table 24, but since a valid chronic
study with the sheepshead minnow is available, chronic RQs may be
calculated.  These are all below the Agency’s LOC, so it can be
concluded that chronic risks to estuarine and marine fish are minimal.

The following table recalculates acute and chronic RQs and LOC
exceedences for estuarine/ marine fish by considering the TEMIK® usage
information provided by Hall (2003 and 2005) and using the appropriate
fish chronic toxicity value.  The table does not correct for the
overestimated half-life values used in the modeling, so the RQ values
presented are still somewhat biased on the high side.  Data in bold type
are new and data in regular type are exactly as presented in EFED Table
24.

EPA response:  The explanation of EEC calculations and % incorporation
is discussed in detail in the EPA response to Bayer’s comments for
page 27, Table 10 and will be addressed in the forthcoming risk
assessment.

  SEQ CHAPTER \h \r 1 Chronic risk quotients could not be calculated
since available data indicate the NOAEC is greater than the acute LC50. 
A review of the studies suggests that adult sheepshead minnows are  more
sensitive than the juveniles.  Additional data are needed to resolve
this inconsistency (e.g., a sheepshead minnow full life-cycle test).  A
more detailed discussion of RQs calculated with the sheepshead minnow of
50 ppb will be included in the forthcoming risk assessment.

Page:  45-46     Table: 25 

EPA comment:  “Table 25: Acute and chronic risk quotients for
estuarine/ marine invertebrates.  Risk quotients for estuarine/marine
invertebrates based on a Pink Shrimp (Penaeus duorarum MRID No.
40228401) LC50 of 12 ppb and Mysid Shrimp (Americamysis bahia MRID No.
00066341) NOAEC of 1 ppb using technical grade aldicarb.”

Bayer response:  As discussed in comments above regarding freshwater
fish risk estimation, the EECs presented are based on unrealistic
assumptions of proportion of granules left unincorporated. 
Incorporating the TEMIK® usage data from Hall (2003 and 2005), the
following more complete table of RQ values and LOC exceedences is
obtained. Thus, for the vast majority of real-world TEMIK®
applications, the RQs are 15-fold less than what is indicated in Table
25.

The following table recalculates RQs and LOC exceedences by considering
the TEMIK® usage information provided by Hall (2003 and 2005).  The
table does not correct for the overestimated half-life values used in
the modeling, so the RQ values presented are still somewhat biased on
the high side.  Data in bold type are new; data in regular type are
exactly as presented in EFED Table 25.

EPA response:  The explanation of EEC calculations and % incorporation
is discussed in detail in the EPA response to Bayer’s comments for
page 27, Table 10 and will be addressed in the forthcoming risk
assessment.

Page: 47     Paragraph: 2. 

EPA Comment:  “The LD50 values entered into T-REX are adjusted for
animal class (20, 100 and 1000g birds and 15, 35 and 1000g mammals)
using the following equations: . . . “ and

Page: 48     Paragraph: 1     Lines: 3-7 

EPA Comment:  “However, research by Mineau et al. (1996) suggests that
the scaling factor for aldicarb should be 1.4, rather than 1.15. 
Substituting the aldicarb-specific scaling factor of 1.4 into the above
avian equation results in RQ values as high as 22,000 (20g bird, maximum
cotton application rate).  It is therefore possible that the RQ
calculations present in this risk assessment are underestimated.”

Bayer response:  The body-size extrapolation method should not be used
unless it can be demonstrated that a significant statistical
relationship generally holds between body size and toxicological
sensitivity.  In the case of birds, this was NOT shown by Mineau et al.
in the cited paper (Mineau et al. 1996).  These authors reported the
slope (1.15) of their regression of LD50 and body mass was not
significantly different from 1.0, the slope under the null hypothesis
that sensitivity does not differ with body size.  Therefore, there is
not sufficient justification to reject the null hypothesis of no
difference in toxicological sensitivity for birds of different body
sizes.

In the case of aldicarb, Mineau et al. (1996) reported a slope of 1.40
and this slope was significantly greater than 1. However, in their
analysis, they used an LD50 estimate for the mallard duck of 3.4 mg/kg,
and not the 1.0 mg/kg, the estimate EPA has used in the current
assessment.  If one substitutes a value of 1.0 for the mallard and
repeats the Mineau et al. regression analysis for aldicarb, the slope is
no longer significantly different from 1.

Intuitively, it does not make sense to do this body size adjustment for
birds in the current assessment.  The purpose of this procedure is to
take into account that smaller birds are more sensitive than larger
birds.  But for aldicarb, the largest bird tested (mallard) has been
identified by the Agency as being the most sensitive species, so the
underlying assumption doesn’t hold.  It is especially inappropriate to
apply this extrapolation method, as the Agency has done, to the lowest
of several LD50 values that have been generated for the mallard.  In
this case, the Agency’s procedure of choosing the lowest available
LD50 estimate from among 10+ species that have been tested already
adequately accounts for interspecies variability in sensitivity.  The
mallard value of 1.0 mg/kg should be used in the RQ calculations without
further adjustment.

Contrary to the statement on the bottom of page 48, the Agency’s
extrapolation procedure has likely resulted in an overestimate of the RQ
values, not an underestimate.

EPA response:  As part of the external peer review process for
conducting refined risk assessments for pesticide impacts to non-target
wildlife, the Agency submitted to the Scientific Advisory Panel (March,
2001) a refined risk assessment model that included a process to account
for the potential for body weight scaling as it related to species
sensitivity to pesticides.  The SAP indicated that the Agency is
"…correct pay attention to scaling of the LD50s to account for
differences in bird size following the work of Mineau et al (1996).   In
general, for the majority of pesticides, not scaling for size may
seriously under-protect small species… When there is obvious scatter
in the size-sensitivity relationship, especially where there are
numerous data points, the Agency should explore the various options that
are open in order to best characterize interspecific sensitivity
differences.”  The risk quotients for the aldicarb ecological risk
assessment  were calculated using the aldicarb-specific scaling factor
of 1.4, as shown in the Mineau et al (1996) paper.  The Agency does
recognize potential limitations of the dataset, and discusses potential
uncertainties based on the use of particular scaling factors in the risk
characterization section of the document.  However, it is important to
note that the Agency level of concern for avian species are still
exceeded for all applications and application rates using no scaling
factor (1.0) as well as the default scaling factor of 1.15 to account
for difference in toxicity between different sizes of birds. 

Page: 47     Paragraph: 3     Lines: 2-4 

EPA Comment:  “As previously stated, the method used in calculating
terrestrial EECs took into account the granular formulation of the
product and its soil incorporation.” and

Pages: 47-50     Tables: 27 and 28 

EPA comment:  “Table 27.  Avian Acute Risk Quotients for Aldicarb
(maximum and average use rates)” and “Table 28.  Mammalian Acute
Risk Quotients for Aldicarb (maximum and average use rates).”

Bayer response:  While it is appropriate to consider the formulation and
soil incorporation during an aldicarb application when conducting
exposure modeling, it is also appropriate to use values that reflect how
the product is actually being applied in the field.  A survey of TEMIK®
usage and methods of application conducted in 2001 across the U.S. 
(Hall, 2003) showed, for instance, that greater than 99% of cotton
applications are applied in-furrow or shanked.  Also see the report and
DVD, “TEMIK® 15G Brand Aldicarb Product Usage and Methods of
Application” submitted with this response (Hall, 2005).  By following
US EPA (1992), the assumed incorporation efficiency for in-furrow or
shanked applications is 99%.  Values listed below should be incorporated
into Tables 27 and 28 (Pages: 47-50) to reflect a more realistic
assessment of application types by crop.

EPA response:  For the purposes of a screening-level ecological risk
assessment, the Agency must use the application parameters on the
currently approved label that provide the most conservative estimation
of risk.  Therefore, unless it is removed from the label, it is assumed.
 The risk assessment was amended to provide values for both the banded,
soil incorporated application method (99% incorporation efficiency) as
well as the side-dress and broadcast application methods (85%
incorporation efficiency), where appropriate.  

Pages: 47-50     Tables 27, 28 and 29 

EPA Comment:  the acute avian and mammalian risk quotients presented in
Tables 27, 28 and 29.

Bayer response:  As indicated previously, Bayer disagrees with the
calculations of both the EEC (mg of aldicarb available per square foot)
and toxicity value (the LD50 value should be for the TEMIK® granule
rather than aldicarb, and no body-size scaling adjustment should be used
in the calculation).  Bayer believes the values presented in these
tables are at least 1 to 2 orders of magnitude too high since the
Agency’s methodology produces artificially inflated hazard indices for
band applications in comparison to a broadcast application with the same
soil incorporation efficiency.

EPA response:  The issues concerning body-size scaling are addressed in
the comment for Page 46, Paragraph 3, Lines 2-4, above.  For the
purposes of a screening-level ecological risk assessment, the Agency
must use the application parameters on the currently approved label that
provide the most conservative estimation of risk.  Therefore, unless it
is removed from the label, it is assumed to be a potential application
method.  The risk assessment will be amended to provide values for both
the banded, soil incorporated application method (99% incorporation
efficiency) as well as the side-dress application method (85%
incorporation efficiency), where appropriate.   It is important to note
that calculated acute risk quotients for both birds and mammals exceed
the Agency LOC for all crops, regardless of application method and
incorporation efficiency. Discussion of the LD50 index calculation and
banded application are provided in response to the comment on Page 27,
Paragraph 2, Line 11, above.  

Page: 52     Paragraph: 4     Lines: 1-2 

EPA Comment:  “This indicates that there is potential risk to
honeybees as well as other beneficial insects.”

Bayer response:  A different conclusion is stated on Page 37,  Paragraph
5, Lines 1-2, states that, “. . . aldicarb due to its granular
formulation, its use is not expected to result in honey bee exposure.”

EPA response:  The document has been revised to discuss potential
honeybee exposure due to the systemic nature of aldicarb.  Because of
its granular formulation, it is unlikely that there is a direct contact
exposure scenario for honeybees.  However, these and other insect
species could be exposed through contact with plants and soil.

B.  Risk Description – Interpretation of Direct Effects

Pages: 53-54     Risks to Aquatic Animals: Summary of Major Conclusions
EPA Comment:  The entire Acute Risk and Chronic Risk summaries

Bayer response:  As presented in an earlier section, Bayer recalculated
RQ values and reached the following conclusions that differ from the
Agency.

EPA response:  RQs will be recalculated in the forthcoming risk
assessment.

C.  Threatened and Endangered Species Concerns

Page: 55     1. Taxonomic Groups Potentially at Risk 

EPA Comment:  Comments refer to the entire section:  “1. Taxonomic
Groups Potentially at Risk”

Bayer response:  The list of taxonomic groups potentially at risk must
be considered a very preliminary starting point since it is based on a
screening assessment in which unrealistic exposure estimates have been
used.  Even simple adjustments to the aquatic EECs, such as those
provided in our recalculated RQ estimates, show that many of the
triggered concerns are unfounded.  A more refined assessment should be
performed to determine which taxonomic groups are really at risk for the
various crops and application scenarios.

EPA response:  The list of taxonomic groups potentially at risk is
considered a preliminary screening of species that could be directly or
indirectly affected by the Federal Action in the action area.   SEQ
CHAPTER \h \r 1 For listed species assessment purposes, the action area
is considered to be the area affected directly or indirectly by the
Federal action and not merely the immediate area involved in the action.
 At the initial screening-level, the risk assessment considers broadly
described taxonomic groups and so conservatively assumes that listed
species within those broad groups are collocated with the pesticide
treatment area.  This means that terrestrial plants and wildlife are
assumed to be located on or adjacent to the treated site and aquatic
organisms are assumed to be located in a surface water body adjacent to
the treated site.  The assessment also assumes that the listed species
are located within an assumed area, which has the relatively highest
potential exposure to the pesticide, and that exposures are likely to
decrease with distance from the treatment area.  The Use
Characterization of this risk assessment presents the pesticide use
sites that are used to establish initial collocation of species with
treatment areas.    SEQ CHAPTER \h \r 1 If the assumptions associated
with the screening-level action area result in RQs that are below the
listed species LOCs, a "no effect" determination conclusion is made with
respect to listed species in that taxa, and no further refinement of the
action area is necessary.  Furthermore, RQs below the listed species
LOCs for a given taxonomic group indicate no concern for indirect
effects upon listed species that depend upon the taxonomic group covered
by the RQ as a resource.  However, in situations where the screening
assumptions lead to RQs in excess of the listed species LOCs for a given
taxonomic group, a potential for a "may affect" conclusion exists and
may be associated with direct effects on listed species belonging to
that taxonomic group or may extend to indirect effects upon listed
species that depend upon that taxonomic group as a resource.  In such
cases, additional information on the biology of listed species, the
locations of these species, and the locations of use sites could be
considered to determine the extent to which screening assumptions
regarding an action area apply to a particular listed organism. These
subsequent refinement steps could consider how this information would
impact the action area for a particular listed organism and may
potentially include areas of exposure that are downwind and downstream
of the pesticide use site.

Page: 58-59     3. Probit Slope Analysis 

EPA Comment:  Entire section on “Birds” and entire section on
“Mammals”

Bayer response:  The entire section on “Birds” and the entire
section on “Mammals” should be deleted.  The amount of toxicant per
square foot is an arbitrarily defined number—it is not a valid
estimate of pesticide intake for individual birds or mammals.  It is
scientifically indefensible to calculate percent mortality risk as is
done here.

EPA response:  According to the Overview Document, this is one of the
Agency’s established methodologies of assessing the risk to birds and
mammals. For further questions regarding Agency methodology, the
overview document may be consulted. 

Page:61   Paragraph:  7  Lines:  1-4

EPA comment:  “The high acute toxicity of alcidarb to honeybees may
lead to mortality to this and other
insect-pollinators.”…”Additionally, the potential risk to bird
species from aldicarb use could also affect bird pollinated plant
species.”

Bayer response:  These statements are speculative, lack foundation, and
conflict with the statement on Page 37, Paragraph 5, Lines 1-2, which
states, “…aldicarb due to its granular formulation, its use is not
expected to result in honey bee exposure.”  Risks to pollinating
species are minimal for insecticides applied in granular form to the
soil.  There is no evidence that systemic residues will pose a risk to
any pollinating species.

EPA response:  The list of taxonomic groups potentially at risk, i.e.
insect pollinators, is considered a conservative and preliminary
screening of species that could be directly or indirectly affected by
the Federal Action in the action area. A further refinement of this
assessment will determine which species may experience a ‘likely to
adversely’ effect by this Federal Action.

D.  Description of Assumptions, Uncertainties, Strengths, and
Limitations

Page: 62     3.a. The LD50 /sq ft Index     Lines: 1-4 

EPA comment:  “The concept was based on field observations of DeWitt
(1966) who suggested that ecological effects are expected to occur when
exposure residues that equal or exceed the LD50 value for a pesticide,
as determined in laboratory studies, are reached in the field."

Bayer response:  DeWitt (1966) proposed a possible relationship between
exposure levels in the field and the quantity that resulted in mortality
in sub-chronic feeding studies, not acute oral toxicity studies.

EPA response:  The DeWitt (1966) paper states that, “Bird LC50 tests
furnish measures of absolute and relative toxicities, but do not provide
the means for relating the LC50 values to the quantities (mg/kg) which
birds might ingest of absorb when exposed to areas which may have been
treated with the toxicants…The data do not establish a basis for
relating dietary concentrations (ppm) causing mortality to quantities
applied for control of pests.  However, they indicate a possible
relationship between the lethal quantity (mg/bird) as determined in
short term feeding tests, and the quantity of toxicant per unit area.”
 The comparison between exposure levels in the field and the LD50 is a
correct comparison.

Page: 63    b. Uncertainties Associated with LD50 /sq ft Index	Lines:
1-6

EPA comment:  “Risk quotients based on the LD50 /sq. ft. hazard index
have been criticized as being too conservative and overestimating
“real world” risk.  It has been argued that the method greatly
oversimplifies the exposure component to hazard assessment by not
specifically addressing the temporal and spatial situations that
non-target wildlife species experience under field conditions.  Although
this is somewhat correct there are still many other exposure related and
toxicological factors that are not accounted for by the index which may
actually underestimate risk from this method.”

Bayer response:  The problem of the LD50/sq. ft method is not that it
“oversimplifies the exposure component to hazard assessment by not
specifically addressing the temporal and spatial situations that
non-target wildlife species experience under field conditions”.  The
problem is that it doesn’t estimate exposure in a biologically
meaningful way.  Exposure is simply assumed to equal the amount of
toxicant in an arbitrarily defined unit area.  This is obviously not
going to be a valid assumption.  Although it might be possible to come
up with reasons for why the LD50/ft2 index may “actually underestimate
risk”, there is no empirical evidence to support such a conclusion and
a mountain of evidence to refute it.  For example, Fischer and Best
(1995) showed that under actual field conditions granule consumption was
a tiny fraction of the number of granules available per square foot even
when the most attractive granule type was applied.  Stafford and Best
(1998) showed that mortality rates were consistently low (10- 15%)
across exposure levels ranging from 50 to 2000 LD50s/ft2.  In the
preliminary risk assessment, the predicted risk level according to the
LD50/ft2 method was 100% in all groups.  It is difficult to understand
how the Agency could envision this index as underestimating risk of
granular products when for most registered products it predicts that
100% bird mortality is expected to occur most of the time!

EPA response:  Although the focus of this comment is on direct granule
ingestion, it should be noted that the LD50 per square method utilized
by the Agency to calculate risk quotients does not assume that this
route of exposure is the only potential exposure mechanism.  The index
is intended to provide an idea of numerous exposure routes, including
plant ingestion, soil ingestion, inhalation, drinking water, soil
invertebrate ingestion, AND direct granule ingestion.  However, a known
limitation in the calculation of the LD50 per square foot RQ is that the
calculation is based solely on the amount of granules present on the
soil surface, and in effect only accounts for potential direct granule
ingestion.  Therefore, it is probable that the LD50 per square foot
calculation used in this screening-level risk assessment may be an
underestimate of exposure, due to the lack of consideration of these
other exposure routes in the actual calculation.  As shown from the
earthworm fugacity model calculation (See section IV.A.3), ingestion of
soil invertebrates by terrestrial birds and mammals is another
significant route of exposure, as is incidental soil ingestion (See
section  IV. E.3.e). Additionally, the systemic nature of aldicarb makes
ingestion of plant food items a significant route of exposure that is
not quantitatively considered in this risk assessment.  Additional
exposure routes, discussed above, include inhalation, dermal, and
drinking water exposure.

Page: 65     Paragraph 2:     Line: 3 and Aldicarb Soil Concentration 

EPA comment:  “. . . the following soil concentrations can be
calculated for a depth of 1 cm.” and “aldicarb soil concentration =
112 mg/kg”

Bayer response:  The aldicarb soil concentration here does not agree
with the “aldicarb soil concentration = 7.47 mg/kg” cited on Page
53, Paragraph 2.  The assumption on page 53 is that aldicarb granules
are evenly mixed into the top 15 cm of soil.  Here the Agency assumes
the mixing depth is only 1 cm.  The Agency states that actual
incorporation depths may range from 5 to 20 cm.  If this is so, then of
what possible relevance is a calculation based on the assumption that
the mixing depth is only 1 cm.  This would be an illegal use of the
product!  The calculated value therefore has no relevance to an
assessment of risk of legally applied aldicarb.

EPA response:  The soil calculations and subsequent earthworm fugacity
model equations were changed to account for the label-specified soil
depth for most aldicarb application methods of approximately 3 inches
(7.62 cm).  

Page: 67     6. Data Gaps and Limitations of the Risk Assessment,     a:
Ecotoxicity Data Gaps 

EPA comment:  “71-4(a) Avian reproduction – Quail, 71-4(b) Avian
Reproduction. – Duck”

Bayer response:  The lack of these data have no bearing on the outcome
of the risk assessment because, as is stated by EFED on Page 47,
Paragraph 3, Lines 6-7,  “Quantitative chronic risk assessments are
not currently performed for granular pesticides on terrestrial
organisms.”  The lack of this information is therefore not a data gap.

Bayer believes avian reproduction studies would provide no useful
information and therefore should not be required because (1) the Agency
has no methods to use the toxicity data produced by these studies in a
risk assessment, (2) TEMIK® granules do not persist in the field (they
rapidly disintegrate with the first rainfall, or upon contact with moist
soil), and so the important route of exposure for birds (ingesting
granules) will not occur over a long period of time, and (3) studies in
other animals (mammals, daphnids, etc.) all point to the conclusion that
aldicarb is an acute toxicant only (due to rapid metabolism and
reversibility of the cholinesterase inhibition, animals can withstand
repeated exposure up to levels that are nearly lethal).  Aldicarb is
only formulated as a granular product (TEMIK® Brand Aldicarb
Pesticide).  The Agency’s policy for assessing avian risks of granular
products is to calculate LD50s per square foot.  The Agency has no
methodology for calculating an EEC for avian diets for granular
products. Thus, if avian reproduction NOAECs were available, they could
not be used in a risk assessment.  As confirmation of this, one only has
to look at the way chronic mammalian test results were used in this risk
assessment.  They weren’t!  The Agency has high quality chronic
toxicity studies of aldicarb with mammals yet the entry in Table 20
(page 38) for the mammalian chronic toxicity endpoint used in the risk
assessment is “N/A” presumably for “not applicable”.  There were
chronic NOAELs from tests with mammalian species but they weren’t used
in any RQ calculations for the reasons indicated above.  What is the
point of requiring two very expensive (>$100,000 each) studies when the
data generated will not affect the outcome of, or even be used in, a
risk assessment?  The requirement of these studies should be waived.

EPA response:  See response for comment, Page 18, 1. Ecotoxicity
Information Gaps, above.

Page: 67     Paragraph: 7:     Line: 1 

EPA comment:  data gap for “161-2 Photodegradation in Water
(sulfoxide, sulfone)"

Bayer response:  Field monitoring and laboratory studies under
controlled conditions demonstrate that it is unlikely for any
significant aldicarb sulfoxide and aldicarb sulfone residues to occur in
open bodies of water, or to exist for sufficient duration to allow for
photodegradation to play an important role in the environmental impact
of aldicarb. Non guideline studies has shown that photodegradation in
water is limited for both aldicarb sulfoxide and aldicarb sulfone. 
Therefore, further investigation of the photodegradation of the two
carbamates is not warranted.  Refer to discussion section for more
details.

EPA response:  The Agency concurs.

Page: 67     Paragraph: 7     Line: 2 

EPA Comment:  “164-1 Terrestrial Field Dissipation (parent aldicarb,
sulfoxide, sulfone)”

Bayer response:  The registrant has identified numerous field
dissipation studies conducted for aldicarb and previously submitted to
the Agency that have not been included in the preliminary EFED risk
assessment.  The degradation of aldicarb carbamate residues in the
United States (summarized in Jones and Estes, 1995) is well understood,
with half lives ranging from 0.5 to 3.5 months.  The discussion section
provides a complete response to this comment.

EPA response:  A Field Dissipation half-life range of 15-105 days has
been assigned.

Page: A-1     Table: A-1     Line 4 

EPA comment: "264-417, Temik® Brand 10% Granular Aldicarb Pesticide for
Use on Citrus only.”

Bayer response:  The correct name is, "264-417, TEMIK® Brand 15G CP
Aldicarb Pesticide.”

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: A-1     Table: A-1     Line: 5 

EPA Comment:  264-426, Temik® Brand 15G Aldicarb Pesticide for Sale and
Use in CA only.”

Bayer response:  The correct name is, "264-426, TEMIK® Brand 15G
Aldicarb Pesticide for Sale and Use in California Only.”

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: A-1     Table: A-1     Line: 6 

EPA comment:  "264-523, Temik® Brand 15G NW Aldicarb Pesticide for Use
on Potatoes.”

Bayer response:  The correct name is, "264-523, TEMIK® Brand 15G NW
Aldicarb Pesticide.”

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: A-2     Table: A-1     Line: 6 

EPA comment:  “FL800003”

Bayer response:  The correct registration number is FL880003.

EPA response:  The agency agrees with Bayer’s comment. Corrections
were made.

Page: A-3     Table: A-1     Line: 7 

EPA comment:  “PR960001”

Bayer response:  This registration is not on our maintenance list.

EPA response:  This SLN (PR 960001) was issued by the Agency for
distribution and use of TEMIK 15G on yams in Puerto Rico on September
4th, 1996.

Pages: A-2 - A-3     Table: A-1 EPA comment:

Bayer response:  These SLN registrations for TEMIK® 15G Aldicarb
Pesticide are not listed in table A-1: GA790016 soybeans; MT000003
potatoes; NC770012 peanuts; OK780009 peanuts; TX780013 peanuts; VA770003
peanuts.

EPA response:  The agency agrees with Bayer’s comment. All these SLNs
are cancelled and will be included in the table.

EPA comment:  “PR96000100, PR, yams”

Bayer response:  This registration is not on our maintenance list.

EPA response:  The SLN number should read ‘PR 960001’. It was issued
by the Agency for distribution and use of TEMIK 15G on yams in Puerto
Rico.

Pages: A-4 - A-5     Table: A-2 EPA comment:

Bayer response:  These SLN registrations are not listed in table A-2: 
GA790016 soybeans; MT000003 potatoes; NC770012 peanuts; OK780009
peanuts;  TX780013 peanuts;  VA770003 peanuts.

EPA response:  The agency agrees with Bayer’s comment. All these SLNs
are cancelled and will not be included in the table.

DISCUSSION

I. DATA REQUIREMENTS – ENVIRONMENTAL FATE 

161-2. Photodegradation in Water: The agency has requested additional
data for aldicarb sulfoxide and aldicarb sulfone. However, aerobic and
anaerobic aquatic metabolism studies of aldicarb have shown that
oxidation of aldicarb to aldicarb sulfoxide and aldicarb sulfone did not
occur in biologically active pond water and sediment systems (studies
listed below). Aldicarb sulfoxide and aldicarb sulfone are the major
degradation products of aldicarb in soil. Since aldicarb is incorporated
in the soil, the majority of the two carbamate residues are located 

below the soil and thus not exposed to sunlight. The same conclusion
applies to any residues reaching ground water. Surface run-off and
drainage may result in some carbamate residues reaching open bodies of
water. In addition to the dilution, extensive and rapid aerobic and
anaerobic degradation takes place, which minimize the exposure to
sunlight. 

Non-guideline studies (Andrawes and Holsing, 1976a and 1976b) showed
that S-methyl 14C-aldicarb sulfoxide and S-methyl 14C-aldicarb sulfone
were very stable to photolysis under UV irradiation (>290nm). Only 2% of
aldicarb sulfoxide was degraded in 14 days of continuous illumination
while aldicarb sulfone was degraded with a half-life of 37.8 and 35.5
days in the nonsensitized and sensitized reactions respectively. The
major degradation product of aldicarb sulfone was aldicarb sulfone
nitrile. Minor amounts of aldicarb sulfone oxime, aldicarb sulfone
aldehyde, aldicarb sulfone amide, and water-soluble components were
reported.

1. Andrawes, N. R. and Holsing, C. G. (1976a) Photostability of Aldicarb
Sulfoxide. Union Carbide Project 

Report, File No. 22936, December 13, 1976. (MRID No. 00068246). 

2. Andrawes, N. R. and Holsing, C. G. (1976b) Photochemical
transformation of UC 21865 Sulfocarb 

Pesticide. Union Carbide Project Report, File No. 22936, December 13,
1976. (MRID Nos. 00053376, 

45592105). 

3. Skinner, W. (1995) Aerobic Aquatic Metabolism of(Carbon 14)-Aldicarb:
Lab Project Number: 467W-1: 

467W:EC-94-274. Unpublished study prepared by PTRL West, Inc. (MRID No.
43805702). 

4. Skinner, W. and Jao, N. (1995) Anaerobic Aquatic Metabolism of
(Carbon 14)-Aldicarb: Lab Project 

Number: 468W-1: 468W:EC-94-275. Unpublished study prepared by PTRL West,
Inc. MRID No. 

43805701 

5. Jesudason, P.A. (2001a) Aerobic Aquatic metabolism of [S- Methyl-14C]
Aldicarb Under Laboratory 

Conditions at 25ºC (OPP. §162-4). Aventis Crop Science, Research
Triangle Park, NC, Study Number 

602YT, December 20, 2001, 127 pages. (MRID No. 45592107) 

6. Lui, A.C. (2001) Aerobic Aquatic metabolism of [S- Methyl-14C]
Aldicarb Sulfoxide Under Laboratory 

Conditions at 25ºC (OPP. §162-4). Aventis Crop Science, Research
Triangle Park, NC, Study Number 

604YT, December 13, 2001, 135 pages. (MRID No. 45592108).

7. Jesudason, P.A. (2001b) Aerobic Aquatic metabolism of [S- Methyl-14C]
Aldicarb Sulfone Under 

Laboratory Conditions at 25ºC (OPP. §162-4). Aventis Crop Science,
Research Triangle Park, NC, Study 

Number 603YT, December 20, 2001, 149 pages. (MRID No. 45592109). 

8. Rupprecht, J.K. (2001) Anaerobic Aquatic metabolism of [S-
Methyl-14C] Aldicarb Sulfoxide Under 

Laboratory Conditions at 25ºC (OPP. §162-4). Aventis Crop Science,
Research Triangle Park, NC, Study 

Number 606YT, December 13, 2001, 107 pages. (MRID No. 45592110). 

9. Huang, M.N. (2001) Anaerobic Aquatic metabolism of [S- Methyl-14C]
Aldicarb Sulfone Under 

Laboratory Conditions at 25ºC (OPP. §162-4). Aventis Crop Science,
Research Triangle Park, NC, Study 

Number 605YT, December 13, 2001, 109 pages. (MRID No. 45592111). 

EPA response:  Agency concurs; there is presently no need to provide
additional studies on the rates of photodegradation of aldicarb
sulfoxide and aldicarb sulfone in water.

164-1. Terrestrial Field Dissipation: EPA has requested data for parent
aldicarb, aldicarb sulfoxide, and aldicarb sulfone. 

On pages 17 and 69 of the preliminary risk assessment EPA states that
the data requirement for 164-1, Terrestrial Field Dissipation, is not
satisfied. Appendix B of the preliminary risk assessment for aldicarb
lists the environmental fate data requirements for aldicarb for each of
the various guidelines. Guideline 164-1, Terrestrial Field Dissipation
(Page B-2), lists thirteen MRID numbers for studies that the Agency
reviewed in coming to the above conclusion that this guideline is not
satisfied. Five of the documents reviewed were considered supplemental
while eight were considered invalid. The 13 studies included in the
Agency review were the following: 

1. Gunther, F.A.; Carman, G.E.; Baines, R.C.; et al. (1975) Aldicarb
(Temik) ) Residues in Oranges, Orange 

Leaves, and Soil after Soil Application in an Orange Grove. Unpublished
study received Oct 10, 1975 

under 6G1689; prepared in cooperation with Univ. of
California--Riverside, Citrus Research Center and 

Agricultural Experiment Station, Dept. of Entomology and others,
submitted by Union Carbide Corp., 

Washington, D.C.; CDL: 096440-H. (MRID No. 00036313). 

2. Romine, R.R.; Meeker, R.L. (1972) Temik Aldicarb Pesticide: Leaching
of Aldicarb into Sandy Soil with 

Irrigation of a Temik Treated Sugar Beet Field: UCC Project Report No.
17079. Unpublished study 

received Aug 4, 1977 under 1016-79; submitted by Union Carbide Corp.,
Arlington, Va.; CDL:231503-U. 

(MRID No. 00068252). 

3. Union Carbide Corporation (1965) Residue Data: UC 21149 .
Compilation; unpublished study received 

Jan 25, 1966 under 6G0473; CDL:090525-O. (MRID No. 00080815). 

4. Union Carbide Corp. (1973) Residues: Temik 10G and Temik 15G .
Unpublished study received Jun 21, 

1974 under 1016-78; CDL: 026641-A. (MRID No. 00101910). 

5. Union Carbide Corp. (1971) Magnitude of the Residues: Aldicarb .
Compilation; unpublished study 

received May 3, 1972 under 2F1188; CDL:091000-I. (MRID No. 00101923). 

6. Kearby, W.; Ercegovich, C.; Bliss, M. (1970) Residue studies on
aldicarb in soil and Scotch pine. Journal 

of Economic Entomology 63(4):1317-1318. Also In unpublished submission
received Dec 6, 1977 under 

1016-69; submitted by Union Carbide Corp., Arlington, VA; CDL:096671-X.
(MRID No. 00102078). 

7. Bull, D.L. (1968) Metabolism of UC-21149
(2-Methyl-2-(methythio)propionaldehyde o - 

(methylcarbamoyl)oxime) in cotton plants and soil in the field. Journal
of Economic Entomology 

61(6):1598-1602. Also In unpublished submission received Jan 18, 1977
under 1016-Ex-37; submitted by 

Union Carbide Corp., Arlington, Va.; CDL:228979-D. (MRID No. 00053364). 

8. Andrawes, N.; Bagley, W.; Herrett, R. (1968a) Temik Metabolism: Fate
of C14-Temik in Cultivated Soil: 

File No. 9218. Unpublished study received Apr 18, 1969 under 9F0798;
submitted by Union Carbide 

Corp., New York, NY; CDL:091372-K. (MRID No. 00101935).

9. Union Carbide Corp. (1969) Soil: Decline of Aldicarb . (Unpublished
study received Aug 20, 1970 under 

0F1008; CDL:091748-M). MRID No. 00101968. 

10. Woodham, D.; Edwards, R.; Reeves, R.; et al. (1973) Total toxic
aldicarb residues in soil, cottonseed, and 

cotton lint following a soil treatment with the insecticide on the Texas
high plains. J. Agr. Food Chem. 

21(2):303-307. Also In unpublished submission received Dec 9, 1977 under
1016-69; submitted by Union 

Carbide Corp., Arlington, VA; CDL:096670-V. (MRID No. 00102061). 

11. Andrawes, N.; Bagley, W.; Herrett, R. (1971) Fate and carryover
properties of Temik aldicarb pesticide 2- 

methyl-2-(methylthio)-propionaldehyde O-(methylcarbamoyl)oxime in soil.
Agricultural and Food 

Chemistry 19(4):727-730. Also In unpublished submission received Dec 6,
1977 under 1016-69; submitted 

by Union Carbide Corp., Arlington, VA; CDL:096671-C. (MRID No.
00102064). 

12. Clarkson, V.; Weiden, M. (1968) Temik Insecticide: The Persistence
of Temik in an Agricultural Soil As 

Indicated by Field and Laboratory Bioassay: File No. 10490. Unpublished
study received Apr 18, 1969 

under 9F0798; Union Carbide Corp., New York, NY; CDL:091372-L. (MRID No.
00101936). 

13. Andrawes, N.; Bagley, W.; Herrett, R. (1968b) Temik Metabolism:
Degradation and Carry-over Properties 

of ... (Temik) in Soil: File No. 10494. Unpublished study received Apr
18, 1969 under 9F0798; submitted 

by Union Carbide Corp., New York, NY; CDL: 091372-M. (MRID No.
00101937). 

The Pesticide Assessment Guidelines: Subdivision N, states that the
purpose of the field soil dissipation study is “. . . to determine the
extent of pesticide residue dissipation under actual field use
conditions. . . ” by evaluating the “. . . mobility, degradation,
and dissipation of the residues.” The above reports and publications
that are dated between 1965 and 1975 and other studies conducted in the
1980’s and 1990’s do address the field dissipation of aldicarb and
the sulfoxide and sulfone metabolites. The field studies referenced and
summarized below (see Table 2, Jones and Estes, 1995) have evaluated the
degradation and movement of aldicarb residues in 40 plots in 14 states,
involved eight different crops, and have collected and analyzed
approximately 19,000 soil samples. The following studies were not
reviewed in the preliminary EFED risk assessment: 

14. Hornsby, A. G., Rao, P. S. C., and Jones, R. L. (1990) Fate of
Aldicarb in the Unsaturated Zone Beneath a 

Citrus Grove, Water Resour. Res., 26, 2287-2302. 

15. Jones, R. L.; Kirkland, S. D.; Chancey, E. L.; Proter, K. S.;
Walker, M.; Ferro, D. N. (1991) Measurement 

of Aldicarb Degradation and Movement in Upstate New York and
Massachusetts Potato Fields: Rhône- 

Poulenc Ag Company, NY State Water Resources Institute and Department of
Entomology, University of 

Massachusetts, August 19, 1991. J. Contam. Hydro., 10: 251-271 (1992).
(MRID Nos. 40493302, 

40493303). 

This study, conducted on plots in potato fields in Massachusetts and New
York, was a cooperative effort of 

state agencies, universities and the manufacturer. In Massachusetts the
soil was a very fine sandy loam 

while in New York the soil was a loamy fine sand. Aldicarb was applied
to the plots at a rate of 2.24 kg/ha 

and 3.45 kg/ha in the Massachusetts and New York plots, respectively.
Soil samples were taken at each 

plot just prior to treatment, as well as one, two, four, and six months
post-treatment. Post-treatment 

samples were collected from four subplots within the treated area. At
each post-treatment sampling 

interval, 16 cores were collected down to a depth of 1.8 to 4.8 meters
in Massachusetts and to a depth of 1.2 

meters to 1.8 meters in New York. Samples were analyzed for residues of
aldicarb, aldicarb sulfoxide and 

aldicarb sulfone as individual analytes. Results from both the
Massachusetts and New York potato fields 

showed that total carbamate residues (aldicarb + aldicarb sulfoxide +
aldicarb sulfone) degraded at a rate 

corresponding to a half-life of about 1.1 months. 

16. Hegg, R. O.; Shelley, W. H.; Jones, R. J.; and Romine, R. R. (1988)
Movement and Degradation of 

Aldicarb Residues n South Carolina Loamy Sand Soil, Agriculture,
Ecosystems and Environment, 20 

(1988) 303-315. (MRID No. 40493318).

The movement and degradation of aldicarb residues was measured in a
loamy sand soil in Barnwell 

County, South Carolina. Aldicarb was applied at a rate of 3.4 kg/ha
(maximum label rate for soybeans) 

incorporated as a 0.15-meter band over the center of the seed row. Soil
samples were collected prior to 

application as well as 35, 70, and 131 days after treatment. Samples
were taken to a depth of 1.8 meters. 

Analysis of soil samples for residues of aldicarb, aldicarb sulfoxide
and aldicarb sulfone as individual 

analytes indicated a half-life of total carbamate residues (aldicarb +
aldicarb sulfoxide + aldicarb sulfone) 

of approximately 9 days. 

17. Jones, R. L. (1987) Central California Studies on the Degradation
and Movement of Aldicarb Residues., J. 

Contam. Hydrol., 1: 287-298.). (MRID No. 40493308). 

Studies were conducted to measure the degradation and movement of
aldicarb residues (aldicarb, aldicarb 

sulfoxide and aldicarb sulfone) in soil at three central California
locations. The test sites, which consisted 

of a tomato field near Manteca and vineyards near Livingston and Fresno
are representative of conditions 

in central California under which aldicarb residues would be most likely
to reach drinking water. Results 

indicate that total carbamate residues (aldicarb + aldicarb sulfoxide +
aldicarb sulfone) degrade in soil with 

a half-life of 1.5-2.0 months. 

18. Jones, R. L. (1991) Measurement of Aldicarb Degradation and Movement
for Winter Applications to 

Grapes: Rhône-Poulenc Ag Company, May 8, 1991. (MRID No. 40493304). 

This study, which measured the degradation and movement of aldicarb
residues following winter 

application to grapes, was conducted in flood-irrigated vineyards at two
sites in California. One site 

contained fine sandy loam while the other site contained a loamy sand.
Aldicarb was applied to each site at 

a rate of 4.48 kg/ha. Soil samples were taken at one, two, four and six
months following application. At 

one site additional samples were taken at eight months following
application. Soil samples were analyzed 

for individual carbamate residues (aldicarb, aldicarb sulfoxide and
aldicarb sulfone). The results indicated 

the degradation rate of total carbamate residues (aldicarb + aldicarb
sulfoxide + aldicarb sulfone) resulting 

from winter applications (3.5 months) was significantly longer than
previously measured with spring 

applications (1.5-2.0 months). It should be noted that this is not a
typical application scenario since 

aldicarb is not registered for winter application. 

19. Norris, F. A. (1991) A Terrestrial Soil Dissipation Study with
Aldicarb the Active Ingredient of Temik® 

Brand Insecticide, Applied to Cotton at Two Field Sites in California:
Rhône-Poulenc Ag Company, July 

24, 1991. 

The degradation and movement of aldicarb residues was measured following
application of aldicarb to 

cotton in California. The soils in the two sites were a low organic loam
soil and a sandy clay loam soil. 

The aldicarb was applied at a rate of 1.18 kg ai/ha in furrow at
planting followed two months later by a 

sidedress application at a rate of 2.35 kg ai/ha. Analysis of soil
samples for aldicarb, aldicarb sulfoxide 

and aldicarb sulfone (individually) showed that total carbamate residues
had a half-life of 1.5 to 2.0 

months. 

20. Jones, R. L.; Hornsby, A. G.; and Rao, P. S. C.; Degradation and
Movement of Aldicarb in Florida Citrus 

Soils, Pestic. Sci., 1988. 23, 307-325. (MRID No. 40493322). 

The movement and degradation of aldicarb residues was monitored in two
Florida citrus groves. Soil core 

samples were collected at 0.3 or 0.6 meter depth increments six times
over an eight month period and were 

analyzed for aldicarb, aldicarb sulfoxide and aldicarb sulfone as
individual analytes. These analyses 

showed a relatively rapid degradation of aldicarb followed by an
increase in aldicarb sulfoxide and aldicarb 

sulfone. The levels of the carbamate metabolites peaked after about 50
days and began to decline. The 

dissipation rate of aldicarb carbamate residues in surface soils
corresponded to a half-life of 10 to 20 days 

at both field sites. The half-life in subsoils appeared to be greater
than 60 days.

21. Wyman, J. A.; Jones, R. L.; Medina, J.; Curwen, D.; and Hansen, J.
L. Environmental Fate Studies of 

Aldicarb and Aldoxycarb Applications to Wisconsin Potatoes, Journal of
Contaminant Hydrology, 2 (1987) 

61-72. (MRID No. 40493314). 

Aldicarb and aldoxycarb (aldicarb sulfone) were applied to potato fields
in Central Wisconsin to study the 

degradation and movement of their carbamate residues within the soil
profiles. Aldicarb was applied at 

planting at a rate of 3.36 kg/ha and at emergence at a rate of 2.24
kg/ha. Aldoxycarb (aldicarb sulfone) was 

applied at planting at a rate of 3.36 kg/ha. Soil samples were analyzed
for total carbamate residues 

(aldicarb, aldicarb sulfoxide and aldicarb sulfone). Total carbamate
residues following application of 

aldicarb and aldoxycarb residues degraded at similar rates with
half-lives ranging from 0.9 to 1.3 months. 

22. Wyman, J. A.; Jensen, J. O.; Curwen, D.; Jones, R. L. and Marquardt,
T. E. (1985) Effects of Application 

Procedures and Irrigation on Degradation and Movement of Aldicarb
Residues in Soil. Environ. Toxicol. 

Chem., 4: 641-651. 

This was an earlier study to Wyman, et.al., above where aldicarb was
applied to potato fields in Wisconsin 

at the same rate as above. Sample analysis gave a half-life for total
carbamate residues of 1.2 to 2.0 

months, similar to the above results. 

23. Jones, R. L., Hansen, J. L., Romine, R. R., and Marquardt, T. E.
(1986) Unsaturated Zone Studies on 

Degradation and Movement of Aldicarb and Aldoxycarb Residues. Environ.
Toxicol. Chem., 5: 361-372 

Studies were reported in which aldicarb was applied to cotton in Arizona
(at planting and emergence), corn 

in Michigan and Indiana, tobacco in North Carolina and Virginia, and
potatoes in Washington. Results of 

aldoxycarb (aldicarb sulfone) applications to tomatoes in Florida,
cotton in Arizona and tobacco in North 

Carolina and Virginia were also reported. These studies showed the
half-life of total carbamate residues 

(aldicarb, aldicarb sulfoxide and aldicarb sulfone) following
application of aldicarb to be from 0.7 to 2.1 

months. The half-life of total carbamate residues following application
of aldoxycarb (aldicarb sulfone) 

was found to be from 0.6 to 1.3 months. 

24. Jones, R. L., Rourke, R. V. and Hansen, J. L. (1986) Effects of
Application Methods on Movement and 

Degradation of Aldicarb Residues in Maine Potato Fields. Environ.
Toxicol. Chem., 5: 167-173 

The effects of application methods on the movement and degradation of
aldicarb was studied in a Maine 

potato field by comparing movement and degradation of aldicarb applied
in furrow at planting with 

aldicarb applied as a top dress at emergence. Movement was greater from
the at-planting application with 

residues found in the top 0.6 meters of soil while the top-dress
application at emergence showed residues 

down to 0.3 meters. The half-life of total carbamate residues in these
cooler Maine soils was 

approximately 3 to 3.5 months. 

As indicated in Jones and Estes (1995), the movement and degradation of
aldicarb residues in soil is a complex process affected by soil
properties, hydrogeological properties, climate conditions and
agricultural practices. The environmental fate database for aldicarb is
substantial, including field research, monitoring studies, laboratory
experiments, potable well analyses and modeling simulations. The field
dissipation studies, which were conducted in numerous locations
(Arizona, California, Florida, Maine, Massachusetts, New York, North
Carolina, South Carolina, Virginia, Washington, Wisconsin, etc.)
representing varying climates, crops and soil conditions, show that
total aldicarb carbamate residues (aldicarb, aldicarb sulfoxide and
aldicarb sulfone) degrade in soil with a half-life of approximately one
month (range of 0.5 to 3.5 months). This degradation rate is sufficient
to prevent aldicarb residue movement to the water table under most use
conditions. In the relatively few areas where aldicarb residues reach
ground water, the primarily lateral movement of the ground water and
continuing degradation usually limit the presence of aldicarb residues
to shallow ground water near treated fields. 

The above referenced studies conducted from 1983 to 1990 do, in fact,
“. . . determine the extent of pesticide residue dissipation under
actual field use conditions . . .” by evaluating the “. . .
mobility, degradation, and dissipation of the residues . . .” and
therefore, satisfy the requirements for the field soil dissipation
study. This position is supported by the OPP Chemical Review Management
System Guideline Status Report for Aldicarb (Run Date: 7/20/01), which
lists the Guideline Status for Guideline 164-1, Terrestrial Field Soil
Dissipation, as “Acceptable/Satisfied. Risk assessments for aldicarb
residues are based upon dissipation of total carbamate residues
(aldicarb, aldicarb sulfoxide and aldicarb sulfone) and, therefore, do
not actually require individual analysis for the three carbamates.
However, as indicated above, many of the studies do analyze separately
for aldicarb, aldicarb sulfoxide and aldicarb sulfone. Therefore, these
studies evaluate the dissipation of not only aldicarb, but the sulfoxide
and sulfone metabolites as well. In addition, studies were conducted
where aldoxycarb (aldicarb sulfone) was applied in the field and the
degradation and movement of aldoxycarb measured with time. 

These soil dissipation studies are supported by numerous other soil
dissipation studies as well as ground water monitoring studies and
modeling simulations. Data from all of these studies allow a good
assessment of the mobility, degradation, and dissipation of aldicarb
residues in the environment. Therefore, there is no need to conduct any
additional aldicarb field soil dissipation studies. 

EPA response:  Agency concurs; there is currently sufficient data to
determine that the field dissipation half-life for total aldicarb
carbamate residues (parent aldicarb, plus sulfoxide and sulfone forms)
ranges from 15-105 days (0.5-3.5 months).

II. SURFACE WATER CONCENTRATIONS FROM FIELD MONITORING 

On Page 26, Paragraph 1, the preliminary EFED risk assessment states
that aldicarb may occasionally pose a contamination hazard in low-order
streams in high use areas. The monitoring general experience of the
registrant indicates that the only agricultural setting where residues
can be present in significant concentrations outside treated fields in
more than an infrequent occurrence is in bedded citrus grown on
flatwoods soils in southern Florida, where irrigation/drainage ditches
run between the rows of trees. In addition to samples collected from
surface water bodies, samples of irrigation runoff water have also been
sampled. For examples in California, state officials (D. A. Gonzalez and
D. J. Weaver, Monitoring Concentrations of Aldicarb and its Breakdown
Products in Irrigation Water Runoff and Soil from Agricultural Fields in
Kern County, 1985, Environmental Hazards Assessment Program, State of
California Department of Food and Agriculture, Environmental Monitoring
and Pest Management, March 1986) collected 11 water samples from eight
fields including water standing in furrows between the crop rows, water
flowing in a drainage ditch, and water from a sump pit. Aldicarb
carbamate residues were not found in any sample. This is similar to the
experience of the registrant in a 1984 study near Livingston, California
(Merced County), in which the irrigation water was sampled as it moved
past an aldicarb treated section in a vineyard (data on file). No
aldicarb residues were detected. Summaries of the aldicarb analyses from
NAWQA program and the California Surface Water Monitoring Database are
also included later in this discussion section. 

All of the factors (other than the rainfall events on the day of
application) contributing to the transient aldicarb carbamate residues
of several hundred ppb in the ditch with intermittent flow upstream of
the Beaver Creek sampling station 07030249 in 1991 and 1993 are not
known to the registrant. Much less was found at another nearby sampling
station 07030241 in which samples were collected at the same time during
1991 (analyses of samples from this site were not reported during the
relevant time in 1993) but which drained a smaller watershed also
composed principally of silt loam soils. An explanation of why the
differences between the two sites would be quite helpful. Possible
explanations include differences in incorporation, site characteristics,
or even cleanup practices. It is known that the ditch is recorded on
maps as "intermittent" and physical examination revealed that cotton is
planted (and may have been treated) to the edge of the ditch, making it
a part of the treated field and therefore, any residues should be
considered in-field in nature. The ditch does not contain sufficient
water to support a resident fish population.

The lack of significant findings of aldicarb carbamate residues in
surface water outside of Florida bedded citrus helps demonstrate the
losses in the ditch upstream of the Beaver Creek sampling station
07030249 are rare. As EPA points out, losses of this magnitude might
result in concentrations of around 10 ppb in small reservoirs. If this
were so, even higher concentrations would be found in small creeks and
streams downstream of surface water bodies the size of intermittent
creek that was sampled in the Beaver Creek watershed. If such losses
occurred uniformly in a moderate size watershed, or even somewhat
frequently over a larger area, certainly more instances of aldicarb
carbamate residues would have been present in the NAWQA, California, and
other smaller monitoring programs. The infrequent detection of residues
in these programs and the total absence of residues at the levels
expected based on the Beaver Creek data indicate that these specific
data taken from an intermittent ditch located within (adjacent to) a
cotton field should not be the sole basis of a higher tier aquatic
exposure assessment. The weight of evidence indicates that exposure of
aldicarb to aquatic organisms in surface water is extremely low and is
achieved by the successful mitigation measures utilized during the
application of the product. 

EPA response:   Although the relatively infrequent instances of aldicarb
residues detected in streams (as cited above) may qualify as “rare”,
they are nonetheless real and must be addressed.  Indeed, the conditions
that appear to have contributed to the fairly high detections in Beaver
Creek (most notably, the occurrence of rainfall shortly after
application) may be somewhat unusual, yet they should be expected to
happen occasionally.  The detection in a separate sample obtained at the
same time from a nearby station supports the contention that this is not
an example of a false positive detection; although the second detection
exhibited lower concentrations, this could easily be explained by the
smaller drainage area and other characteristics of the site (e.g.,
topography, timing and extent of rainfall, soil physical properties,
sampling location, etc.).  Contaminant ‘spikes’ in stream water are
typically very short-lived and transient – peak concentrations will
usually occur somewhere near the early portion of the hydrograph peak
(representing the beginning of a large-scale influx of surface runoff
water into the channel).  It should be expected that in any standard
stream monitoring regimen that there would seldom be samples taken at
precisely those times when peak concentrations would be expected to
occur (that is, early in a high-intensity rainfall event that follows
soon after chemical application).  Thus, infrequent samples with
relatively high concentrations should more likely be considered
representative of these atypical but not unprecedented circumstances
rather than indicative of false data.  These occurrences must be
considered as part of any high-end estimate of potential risks.

Summary of Surface Water Data from the NAWQA Program 

On Page 26, Paragraph 1, the preliminary EFED risk assessment summarized
available surface water monitoring data from the NAWQA program as having
only a few detections (0.1 percent of all sites) with a maximum
concentration of 1.6 ppb. Table 1, Summary of Surface Water Data from
the NAWQA Program, summarizes the aldicarb analyses available from this
program. The source is the USGS data base, which was downloaded by state
from their web site during the second quarter of 2005. The registrant
previously did an analysis of the data through the fourth quarter of
2000. A comparison of the two sets of data show that as expected more
samples were added to the data base; however, a number of samples
present in the data base in 2000 were not present in the data base in
2005. 

Of the 6262 samples analyzed, aldicarb residues were detected in 21
samples (0.34 %). However, the analyses of 18 of these samples were
suspect for several reasons (only one carbamate compound present or
little or no aldicarb use in the county), leaving only 3 samples with
credible detections (0.05 %). The concentration of total aldicarb
carbamate residues in these three samples were 1.68, 0.085, and 0.014
ppb. Table 2 provides information on surface water samples with
detectable residues: 

The small number of detections of aldicarb residues in the surface water
samples from the NAWQA program supports EPA’s conclusion that aldicarb
residues have not been widely detected in surface waters. 

Summary of Surface Water Data from the California Surface Water
Monitoring Database 

On Page 26, Paragraph 1, the preliminary EFED risk assessment summarized
surface water monitoring data from various sources. One source not
included in this discussion is the California Surface Water Monitoring
Database. The number of analyses and the detections of aldicarb
carbamate residues are summarized in Table 3. 

The California Surface Water Monitoring Database shows three samples
with detectable aldicarb residues, all in Stanislaus County in 1991 to
1992. A sample taken on July 2, 1991 at the San Joaquin River Hills
Ferry sampling station in Stanislaus County showed detectable
concentrations of aldicarb (0.12 ppb) and aldicarb sulfoxide (0.28 ppb).
Follow-up sampling at the same site was conducted with a high frequency
of sampling (4 more samplings in July and 8 samplings in August). All
the samples showed no detectable concentrations of aldicarb or its
metabolites except for the sample collected on August 27, 1991 that
showed 0.05 ppb of aldicarb sulfone, but no detectable aldicarb or
aldicarb sulfoxide. Sampling on February 18, 1992 at the Turlock
Irrigation District Drain #5 sampling site in Stanislaus county showed
detectable concentrations of aldicarb sulfone (0.26 ppb), but aldicarb
and aldicarb sulfoxide were not detectable in that sample. The absence
of aldicarb sulfoxide in this sample makes this analysis 

suspect.

EPA response:    The Agency disagrees with the criteria used to
disqualify detections in surface water samples.  There is no
justification for disregarding cases where only one or two of the
aldicarb residues are detected.  All three (parent aldicarb, aldicarb
sulfoxide, aldicarb sulfone) need not co-occur.  In addition, subsequent
sampling that yields no detectable residues where earlier samples had
been positive does not invalidate the previous sample detections; it is
unreasonable to expect that residue concentrations should remain stable
over time, particularly in streams.  On the contrary, it should be
expected that levels will be highly temporally variable within the same
stream.  When detections occur in areas where aldicarb is not currently
in use, a historical background for the region is necessary; if the
chemical had never been used in the larger watershed it can be presumed
that the detection is invalid.  However, if the compound had been used
in the past, its detection in surface water may reflect (baseflow)
groundwater contributions to the stream – residues in groundwater,
which can persist for long periods under certain conditions, may
ultimately be discharged to the surface.  Thus some of these seemingly
suspect samples may reflect past rather than present usage.

III. FORMATION AND TRANSFORMATION OF HYDROXYMETHYL ALDICARB 

SULFONE AND THE RESULTS OF ONGOING AEROBIC AND ANAEROBIC 

METABOLISM STUDIES FOR ALDICARB SULFONE 

Potential for Hydroxymethyl Aldicarb Sulfone to Revert to Aldicarb
Sulfone 

As discussed earlier, hydroxymethyl aldicarb sulfone was reported as a
metabolite of aldicarb sulfone in aerobic and anaerobic aquatic
metabolism studies (MRID Nos. 45592109 and 45592111). Repeat studies
with the same system (results presented later in this section)have shown
that hydroxymethyl aldicarb sulfone is likely not a metabolite of
aldicarb sulfoxide and that the misidentification in these studies
occurred as a result of a rapidly degrading standard. 

However, in the unlikely event that hydroxymethyl aldicarb sulfone is
formed, it will not revert to the parent sulfone under aerobic
conditions. The registrant agrees with the reviewer’s comment on DER
Report (Aerobic Biotransformation of Aldicarb Sulfone in Water Sediment
System, EPA MRID No. 45592109) pages 3 and 17, that hydroxymethyl
sulfone will undergo oxidation to corresponding carboxylic acid. These
types of amino carboxylic acids are extremely unstable and will
eliminate carbon dioxide instantaneously and upon hydrolysis yield
sulfone 

oxime. In addition, as stated in the DER report hydroxymethyl aldicarb
sulfone will undergo hydrolysis to form sulfone oxime, or elimination to
form aldicarb sulfone nitrile. The nitrile upon hydrolysis will yield
sulfone amide and further to aldicarb sulfone acid. Similarly aldicarb
sulfone oxime will undergo hydrolytic/oxidative pathway to yield
aldicarb sulfone acid. Aldicarb sulfone acid and aldicarb sulfone amide
upon hydrolysis yield methane sulfonic acid which will be the ultimate
transformation product. 

Therefore under aerobic conditions the hydroxymethyl aldicarb sulfone
will transform into a methane sulfonic acid and it will not revert to
the parent aldicarb sulfone (although transformation back to aldicarb
sulfone might be possible under anaerobic conditions, aldicarb sulfone
degrades rapidly under such conditions). 

Results of Ongoing Aerobic Aquatic Metabolism Study for [S-Methyl
14C]Aldicarb Sulfone 

Aldicarb sulfone is a major metabolite of parent compound aldicarb. The
aerobic biotransformation of radiolabeled S-methyl aldicarb sulfone,
(2-Methyl-2-(methylsulfonyl)propanal-O-methylcarbamoyloxime, was studied
in a pond 

water/sediment system (water: pH 7.1, dissolved organic carbon = 4.8
ppm, sediment: texture = loamy sand, pH 6.2, organic carbon = 2.5%) from
Clayton, North Carolina United States for 30 days in the dark at 25 ± 1
ºC. Aldicarb sulfone was applied at a rate of 0.28mg a.i./L, which
closely approximates the single maximum field use rate of 5.6 kg/ha to a
depth of 200 cm. The sediment/water ratio used was 1:4. The average
material balance total material balance in the water sediment system was
96% of applied activity. (Table 4) 

The concentration of [S-methyl14C] in water decreased from 98.6% at day
zero to 1% at day 14. The concentration of [S-methyl14C] in sediment
decreased from 3.4% at day zero to 1% at day 7. Aldicarb sulfone
degraded very fast with a half-life of 3.00 days (Figure 1). 

The major transformation products were identified as aldicarb sulfone
alcohol (26%), aldicarb sulfone acid (35%), aldicarb sulfone nitrile
(25%), aldicarb sulfone amide (26%), and aldicarb sulfone oxime(9%).
There was no major metabolite detected with an intact carbamate moiety.
There were two other minor polar transient metabolites range 3-5% which
appeared day three and five only and then degraded (Table 2).

Non extractable residues were very low and accounted for 9% of applied
activity at day 30. Excellent mineralization to CO2was observed and
accounted for 39% of the applied activity at day 30. 

Metabolic Pathway. Aldicarb sulfone degraded very rapidly in pond water
in the pond water and sediment system with almost complete degradation
in two weeks after application. A proposed metabolic pathway is shown in
Figure 2. The pathway involves hydrolysis and elimination of methyl
carbamoyl group to form aldicarb sulfone oxime, and aldicarb sulfone
nitrile respectively. Aldicarb sulfone nitrile undergoes further
hydrolysis and oxidation to aldicarb sulfone amide and aldicarb alcohol,
which undergoes further oxidization to form aldicarb sulfone acid.

EPA response:  The Agency generally concurs with the statement that
reversion of hydroxymethyl aldicarb sulfone to aldicarb sulfone is
likely not a major concern, and that if such does occur (under anaerobic
conditions) the aldicarb sulfone should degrade fairly rapidly. 
However, groundwater discharge zones – usually low-lying areas which
are commonly reducing environments (e.g., riparian zones, wetlands,
etc.) – typically have rapidly changing redox conditions, both
spatially and temporally.  How this might affect these chemical
transformations remains largely unknown.

 

Results of Ongoing Anaerobic Aquatic Metabolism Study for [S-Methyl
14C]Aldicarb Sulfone 

Aldicarb sulfone is a major metabolite of parent compound aldicarb. The
anaerobic biotransformation of radiolabeled S-methyl aldicarb sulfone,
(2-Methyl-2 (methylsulfonyl)propanal-O-methylcarbamoyloxime, was studied
in a pond water/sediment system (water: pH 7.1, dissolved organic carbon
= 4.8 ppm, sediment: texture = loamy sand, pH 6.2, organic carbon =
2.5%) from Clayton, North Carolina United States for 30 days* in the
dark at 25 ± 1 ºC. Aldicarb sulfone was applied at a rate of 0.28mg
a.i./L, which closely approximates the single maximum field use rate of
5.6 kg/ha to a depth of 200 cm. The sediment/water ratio used was 1:3.
The average total material balance in the water sediment system was 100%
of applied activity (Table 6). 

The concentration of [S-methyl14C] in water decreased from 98.6% at day
zero to 1% at day 14. The concentration of [S-methyl14C] in sediment
decreased from 3.4% at day zero to 1% at day 7. Aldicarb sulfone
degraded very fast with a half-life of 2.4 days (Figure 3). 

The major transformation products were identified as aldicarb sulfone
alcohol accounting for 50.8%, aldicarb sulfone acid for 16.5%, and
aldicarb sulfone nitrile accounting for 32.3% of the applied activity in
total system. There was no major metabolite detected with intact
carbamate moiety. There were two other minor transient metabolites
ranging from 1-3% of applied, which appeared on day one and three only
and then degraded (Table 7).

 

Non extractable residues were very low and accounted for less than 2% of
applied activity at day 30. 

EPA response:  The Agency is very grateful for the continued research
into the behavior, transformations, and characteristics of the various
degradates of aldicarb.

 PAGE   

 PAGE   14 

	

