EPA RESPONSE
TO EXTERNAL PEER REVIEW COMMENTS
on the
DRAFT AQUATIC LIFE AMBIENT WATER QUALITY
CRITERIA FOR CADMIUM - 2015
November 19,2015
Office of Water
U.S. Environmental Protection Agency
Washington, DC
-------
Table of Contents
1 INTRODUCTION 1
1.1 Background 1
1.2 Peer Reviewers 1
1.3 Review Materials Provided 1
1.4 Charge Questions 1
2 EXTERNAL PEER REVIEWER COMMENTS AND EPA RESPONSES, ORGANIZED BY
CHARGE QUESTION 2
2.1 Charge Question 1 3
2.2 Charge Question 2 10
2.3 Charge Question 3 27
2.4 Charge Question 4 37
2.5 Other Comments Provided 46
3 REFERENCES CITED BY REVIEWERS 48
li
-------
1 Introduction
EPA submitted its Draft Aquatic Life Ambient Water Quality Criteria (AWQC) for Cadmium - 2015 for
contractor-led independent, external peer review from August 25 to September 14, 2015. The external peer
reviewers provided their independent responses to EPA's charge questions. This report documents the EPA's
response to the comments provided to EPA.
This report presents the four peer review charge questions and individual reviewer comments (verbatim) in
Sections 2.1 through 2.4. Section 2.5 presents additional minor comments provided by one reviewer. New
information (e.g., references) provided by reviewers is presented in Section 3. EPA separated each reviewer's
comments by charge question into distinct topics and responded to each topic individually, and also indicated
how the draft cadmium criteria document was revised in response to peer reviewer comments.
1.1 Background
EPA's Office of Water is charged with protecting ecological integrity and human health from adverse
anthropogenic, water-mediated effects, under the purview of the Clean Water Act (CWA) Section 304(a)(1).
The Agency has been working to update water quality criteria to protect aquatic life and aquatic-dependent
wildlife from the presence of cadmium in freshwater and estuarine/marine environments in order to reflect the
latest scientific knowledge.
EPA's AWQC for cadmium presents draft acute and chronic criteria expressed as concentrations of cadmium
in fresh and estuarine/marine waters (dissolved). The 2015 draft cadmium criteria document is an update to
the 2001 cadmium criteria. The 2015 draft incorporates additional toxicological data for cadmium, while
using the same criteria derivation process that was used in 2001.
1.2 Peer Reviewers
An EPA contractor identified and selected five reviewers who met the technical selection criteria provided by
EPA and who had no conflict of interest in performing this review.
The EPA contractor provided reviewers with instructions, the review document (including appendices), the
charge to reviewers) prepared by EPA, and supporting reference materials as described in the charge.
Reviewers worked individually to develop written comments in response to the charge questions.
1.3 Review Materials Provided
• Internal Draft Cadmium AWQC_ 042115 (081315).pdf
• Internal Draft Cadmium AWQC_Appendicies_7 1 15 (081315).pdf
• Appendix K Issue Summary Regarding Test Conditions and Methods...H. Azteca.pdf
• Internal Draft Cadmium AWQC References l 14 14 (081315).pdf
Background/Supplemental Material (not for review, reference only)
• Cadmium Risks to Freshwater (Mebane 2010).pdf
1.4 Charge Questions
1. Please comment on the overall clarity of the document and construction as it relates to the derivation
of each criterion.
2. Please comment on the technical approach used to derive the draft cadmium criteria; is it logical, does
the science support the conclusion, and is it consistent with the protection of freshwater and
1
-------
estuarine/marine aquatic life from acute, chronic, and bioaccumulative effects? Are the methods
described in the document scientifically sound?
3. Please comment on the data used to derive the revised criteria, including data
adequacy/comprehensiveness, and the appropriateness of the data selected and/or excluded from the
derivation of the draft criteria. Is the data used correctly for the intended purpose? Are there other
relevant data that you are aware of that should be included? If so, please provide the data along with
supporting information.
4. Are the derived criteria appropriately protective of listed species and commercially and recreationally
important species, particularly as the criteria relates to salmonids?
2 External Peer Reviewer Comments and EPA Responses, Organized by
Charge Question
The following tables list the charge questions submitted to the external peer reviewers, the external peer
reviewers' comments regarding those questions (broken into distinct topics), and EPA's responses to the peer
reviewers' comments. EPA revised the 2015 draft considering the external peer review comments, and noted
in the table where the document was edited.
2
-------
2.1 Charge Question 1
1. Please comment on the overall clarity of the document and construction as it relates to the derivation of each criterion.
Re\ iewer
K\(ern;il Peer Re\ iewer Comments Re<>;irdiii<> Ch«ir»e
Question 1
IT A Response
Re\ision Location
in 2015 Drill!
C iulmium C i ilci iit
Document
Reviewer 1
This report makes for very dull reading, but it is well-written
and it is usually clear what the author is trying to say. There
are no insightful comments or new ideas presented in this
report, but the report is laid out in a clear, logical fashion.
Thank you for your comment.
No edit needed.
Reviewer 2
Overall the document is relatively clear with formatting in a
risk assessment format which allows the reader to evaluate
each criteria. Of minor concern was the lack of inclusion of
emerging materials as sources of cadmium such as quantum
dots which do make up photovoltaic substances (mentioned).
However, the increased use of these materials as "inorganic"
Cd sources and the uncertainties surrounding the potential
absorption and effects of these materials to aquatic organisms
needs some discussion.
Information regarding quantum dots has been added to the
document.
Section 2.1
Reviewer 2
In addition, some inconsistencies were noted with regard to
sub-lethal effects mentioned in the Estuarine/Marine Acute
section. While present in this section, discussions of sublethal
effects were largely omitted in the Freshwater sections and
chronic sections of both water types.
The Estuarine/Marine Acute section was revised to remove
inconsistencies. Additionally, information about sublethal
effects in other media was added to the appropriate sections
of the document.
Section 5.1
Section 5.2
Section 5.4
Section 5.5
3
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 1
IT A Response
Rc\ision Location
in 2015 1)r:il'(
(:ul in in in C'rilcriii
Doc ii in cut
Reviewer 2
There was also inconsistencies with regard to the use of flow-
through vs. static exposures and whether more or less
uncertainty is involved in utilization of the values. For
example, flow-through methods were stated for Salmo trutta,
but methods for Morone were static or static-renewal. One
would clearly suggest the flow through values should be
given greater weight with regard to uncertainty assessments.
As it reads right now, it appears there are no differences
between using static or flow-through exposures.
Data selected to calculate the SMAV for each species
follows 1985 Guidelines recommendations. Specifically,
flow-through measured exposures are preferred and selected
for use over static and static-renewal exposure studies. If
only static or static renewal exposure studies are available,
EPA considers the study data and determines whether the
study is acceptable for inclusion considering factors, such as
known compound stability and other relevant information
presented by the study author. EPA's goal is to consider and
include as much high quality, scientifically defensible data
in its assessments as possible in order to characterize
potential response in a broad array of aquatic organisms. For
example, if a species of concern had only static renewal data
for acute studies, and EPA knew the compound was stable
in water during the test duration, the data would be
considered for inclusion if it met with the other data quality
screens EPA.
No edits needed.
4
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 1
IT A Response
Rc\ision Location
in 2015 1)r:il'(
(:ul in in in C'rilcriii
Doc ii in cut
Reviewer 2
The inability to determine salinity relationships to toxicity is
also a concern but it is likely due to varied salinity regimes
confounded with temperature and solute constituents in
experimental designs (see comments below). It is noteworthy
that a lppt value is considered "estuarine" for the Morone
value, when there are "freshwater" systems that likely have
higher conductance than this value. There should also be
some statement or better clarity documenting the lack of a
standard salinity value being utilized to compare toxicity
values. It appears that the most sensitive toxicity value is
being used regardless of the salinity.
The current statement in the Executive Summary about the
salinity relationship addresses this comment:
"Available data suggest the acute toxicity of cadmium may
be influenced by salinity, with a trend of decreasing
sensitivity to cadmium with increasing salinity. However,
this trend could not be definitively characterized and a
mathematical relationship could not be described to define
the dependency (See Section 5.4.1)." Text has been added to
elaborate on why a salinity normalization approach is not
being used in criteria development.
The estuarine/marine value is intended to be applicable to
the broad range of salinities present in non-freshwater
systems. EPA will accordingly continue to use 1 ppt as the
lowest salinity level for a salt water test. This salinity is
consistent with Mitsch and Gosselink (1986) who classify a
waterbody with a salinity of 0.5-5.0 ppt is oligohaline.
Mitsch, W.J. and J.G. Gosselink. 1986. Wetlands. Van
Nostrand Reinhold, New York. 539 pp.
Section 2.3.1
Section 5.4.1
Reviewer 2
Overall, the uncertainty analysis section should be extended
to include aspects of uncertainty with the data used for the
derivation of the criteria. As it stands presently, the emphasis
seems to be more on justification of data not utilized for the
derivations.
Data selection is consistent with the procedures presented in
the 1985 Guidelines. The Effects Characterization section
was revised to include further discussion about uncertainty
in the criteria calculations.
Section 5.1
Section 5.2
Section 5.4
Section 5.5
Reviewer 3
In general, the document language is reasonably clear.
However, throughout the document, there are several
instances where certain decisions are made that appear to be
rather arbitrary without sufficient justification as to how or
why these decisions were made (see details below).
Thank you for your comment. Please see responses to
specific comments.
No edits needed.
5
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 1
IT A Response
Rc\ision Location
in 2015 1)r:il'(
(:ul in in in C'rilcriii
Doc ii in cut
Reviewer 3
Minor comments:
p. 8 and elsewhere: use mass units rather than ppm, ppb etc.
These values are mineral deposit concentrations in units
reported by the author(s), but were changed to mg/kg based
on the comment. There are no other uses of ppm or ppb in
the document.
Section 2.1
Reviewer 3
p. 9: quantify concentrations found in impaired water
("several micrograms per liter" is vague)
The text has been revised to give a definitive value.
Section 2.1
Reviewer 3
p. 10: is the suggestion that precipitated/particulate forms of
Cd that ultimately end up in sediments are not bioavailable?
Text was added to clarify that particulate forms of cadmium
are potentially available to benthic feeders and sediment
dwellers.
Section 2.2
Reviewer 3
p. 19: do data exist for any other salts of Cd that has been
excluded?
The 1985 Guidelines note specific salts to test for metals;
onlv these salts were used. According to the Manual of
Instruction for Preparing Aquatic Life Water Oualitv
Criteria Document. Stephan 1985. Section III. Defining the
Pollutant, "for metals such as cadmium, chromium (III), and
zinc, only data from tests on chloride, nitrate, and sulfate
salts (either anhydrous or hydrated) should be used",
therefore, other data for other cadmium salts were not
included in the evaluation. Thus, studies conducted with
cadmium acetate and cadmium borate salts were not used,
nor were tests with nanoparticles and quantum dots.
Stephan, C.E. 1985. Manual of instructions for preparing
aquatic life water quality criteria documents. Draft report
dated 12-12-85. U.S. EPA. Environmental Research
laboratory, Duluth, MN. 49 pp.
No edits needed.
6
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 1
IT A Response
Rc\ision Location
in 2015 1)r:il'(
(:ul in in in C'rilcriii
Doc ii in cut
Reviewer 3
P.63: Please be explicit about how the constants in the
equations are derived for both the CMC and CCC.
The reviewed draft contained explicit information about
how the constants in the equations for the CMC and CCC
were derived:
TllC CMC x ln(hardness)-4.247)
Where, 1.103 is the acute pooled slope and;
-4.247 is calculated as
=ln(CMC at 100 hardness) - (Pooled Acute Slope x
(ln(100)))
=ln(2.3)-(l.103 x 4.605)
Similarly, the CCC=e(a8161 xln(hardness>3 663)
Where, 0.8161 is the chronic pooled slope and;
-3.663 is calculated as
=ln(CCC at 100 hardness) - (Pooled Chronic Slope x
(ln(100)))
=ln(l.l)-(0.8161 x 4.605)
No edits needed.
Reviewer 3
P. 67: Define the values listed under the two tables: (S2, L, A)
Footnotes were added to the document to define the terms S,
L and A. These terms are refer to the following: S = slope; L
= intercept; A = InFAV. FAV = Final Acute Value.
Section 4.3.1
Section 4.3.2
Section 4.4.1.
Reviewer 3
Major comments:
p. 12: "Mebane (2014) conclude that, although there were not
adequate data to establish acceptable tissue effects
concentrations for aquatic life, cadmium is unlikelv to
accumulate in tissue to levels that would result in adverse
effects to aauatic invertebrates or fish. The evaluation of
direct exposure effects is therefore considered to be more
applicable to the development of criteria for aquatic life."
This line of reasoning is questionable on many levels.
Establishing critical tissue effects thresholds that work across
species is problematic, especially in invertebrates, because
species vary in their abilities to store/sequester Cd in
EPA concurs with the reviewer about the difficulty in
characterizing dietary exposure and establishing critical
tissue effects thresholds for bioaccumulated metals.
Text has been added to discuss these points and incorporate
the work of other researchers, such as Mebane (2006), who
discuss cadmium bioaccumulation and dietary exposure in
detail and the uncertainties currently associated with its
evaluation. In particular, text was added to detail how
cadmium can bioaccumulate in aquatic organisms through
multiple exposure routes including ingestion and direct
exposure, and how total uptake depends on the cadmium
Section 5.6.1
7
-------
Rc\ icwcr
K\(crn;il I'ccr Rc\ icwcr Comments Rc«i:irdin«i C'h«ir»c
Question 1
IT A Response
Revision Location
in 2015 1)r:il'(
(:ul in in in C'rilcriii
Doc ii in cut
physiologically inert forms. However, this does not mean that
bioaccumulated metals are non-toxic as is implied by the
language in this document. I think Mebane is being grossly
misquoted here (aside from the fact that there is no 2014
reference). Here are some quotes from his 2010 document that
directly refute the underlined text above:
"Thus the consequences of elevated tissue residues or effects
of dietary exposures may be important when estimating
protective thresholds for cadmium and other pollutants
(McCarty and Mackay, 1993; Meyer and others, 2005)." P. 32
"A diet of cadmium-contaminated green algae Chlorella sp
caused reduced growth in the amphipod Hyalella azteca in a
recent study (Ball and others, 2006)." P. 38
"Dietary cadmium exposures appear to be an important risk
for at least some invertebrates. The data reviewed on dietary
effects of cadmium to invertebrates indicated that adverse
effects could occur at concentrations realistic in cadmium-
polluted waters". P. 38
"Toxicity to mayflies from feeding on cadmium-contaminated
algal mats at environmentally realistic concentrations was
observed (Irving and others, 2003). P. 38
I understand that dealing with dietary exposures is incredibly
inconvenient in the context of the 1985 Guidelines, but
pretending that they are not important in 2015 is irresponsible
because we know better. The Irving et al., 2003 study
referenced above provides direct evidence that diet derived
Cd can be problematic in this aquatic insect example.
concentration, exposure route and the duration of exposure.
Text was also added to clarify that there does not appear to
be a consistent relationship between body burden and
toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
8
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 1
IT A Response
Rc\ision Location
in 2015 1)r:il'(
(:ul in in in C'rilcriii
Doc ii in cut
Reviewer 4
I found the overall clarity of the document to be quite good. I
especially appreciated the document being generally
organized in a risk assessment format. I think this is very
useful, particularly the Problem Formulation section that
outlines various sources, potential exposure pathways and
receptors. I hope EPA will use this overall structure for future
criteria documents as well. I also like all of the comparisons
to previous Cd criteria documents. This makes key changes to
the criteria very transparent.
Thank you for your comment.
No edits
Reviewer 4
My only significant criticism of the overall format is that
there are a number of redundancies where information is
presented multiple times, often the exact same wording (for
example, Section 5.4.1 is redundant of earlier text in the
document). I encourage EPA to consider consolidating and
reducing these redundancies.
The document was reviewed and revised to minimize
redundant text.
Various locations
Reviewer 4
An additional minor point is that it is unclear how the data
tables in the appendices are organized. They don't seem to be
listed alphabetically by either common or scientific name. It
would be useful if they were.
Data tables are organized as recommended by the 1985
Guidelines (phylogenetically) and text was added to each
table in the Appendices to clarify this.
Appendices
Reviewer 5
Generally sufficient. Problem formulation section seemed a
bit of a forced fit, as if added to satisfy a new stylist protocol.
Thank you for your comment.
No edits needed.
9
-------
2.2 Charge Question 2
2. Please comment on the technical approach used to derive the draft cadmium criteria; is it logical, does the science support the conclusion,
and is it consistent with the protection of freshwater and estuarine/marine aquatic life from acute, chronic, and bioaccumulative effects? Are the
methods described in the document scientifically sound?
Re\ iewer
K\(ern;il Peer Re\ iewer Comments Re<>;irdin<> C'hjir»e
Question 2
KIW Response
Re\ ision
Locution in 2015
Drsil'l C iulniiiini
C'rileriii
Document
Reviewer 1
This report is rather antiquated in its thinking. It basically
assumes that Cd is accumulated only from the aqueous phase
rather than from both the aqueous phase and ingested food.
Over the past 10-15 years, it has been shown that many
toxicants, including Cd and other metals, can be
bioaccumulated from food as well as from the aqueous phase.
Indeed, a number of laboratory, field, and modeling studies
have shown that diet can be the dominant source of metals for
marine invertebrates and fish. The relative importance of diet
has been shown to vary with species, but it is rarely a minor
source and sometimes (for some fish species, for example) the
predominant source. Moreover, once accumulated from diet,
Cd can reach sensitive organs within animals that are not
reached by Cd taken up from the aqueous phase. Therefore,
the toxic response of an animal to either ambient Cd or body
burden Cd can vary considerably, depending on whether the
source is ingested food or solute in ambient water. Thus,
dissolved metal may be sorbed onto exoskeletons in
crustacean zooplankton (often the most sensitive species, as
the author points out) but this does not directly affect the
animal because the metal (Cd in this case) bound to chitosan
on the exoskeleton does not interact with metabolic processes,
whereas metal assimilated from ingested food can enter into
internal tissues where it may interfere with a variety of
metabolic and reproductive processes. I saw no
acknowledgement of the possible significance of dietary Cd
on aquatic (freshwater or marine) animals in this report, and
EPA concurs with the reviewer about the multiple potential
exposure routes and the complexity of characterizing these
routes and establishing critical tissue effects thresholds for
bioaccumulated metals.
Text has been added to discuss these points and incorporate
the work of other researchers, such as Mebane (2006), who
discuss cadmium bioaccumulation and dietary exposure in
detail and the uncertainties currently associated with its
evaluation. In particular, text was added to detail how
cadmium can bioaccumulate in aquatic organisms through
multiple exposure routes including ingestion and direct
exposure, and how total uptake depends on the cadmium
concentration, exposure route and the duration of exposure.
Text was also added to clarify that there does not appear to be
a consistent relationship between body burden and
toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
Section 5.6.1
10
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
yet numerous papers describing such effects appeared in the
reference section. In looking over appendices, many of these
reports were not used, often for what appear to be spurious
reasons or misinterpretations of studies. In some cases, dietary
metals could be 1 -2 orders of magnitude more toxic than
dissolved metals to freshwater cladocerans and marine
copepods, for example. In the case of Cd, an EC50 value of 5
nM (-0.5 |_ig/L) was observed in copepods in a study by Hook
& Fisher (cited in this report) if the animal had been fed food
exposed to that Cd concentration, whereas the measured LC50
value based on a dissolved Cd source was 200 times greater.
Also, measuring growth or mortality, as is often the case in
simple toxicity tests, would have missed the effect—rather the
reproductive capability of the copepods was affected by the
dietary Cd, but no mortality was observed at environmentally
realistic concentrations. Because dissolved Cd concentrations
are typically at very low concentrations in natural waters (at
least 10-fold lower in surface seawater, for example), the lower
EC50 value derived from dietary rather than dissolved sources
still indicates that Cd is unlikely to cause toxic effects in most
natural waters.
Reviewer 2
With a few notable exceptions, the technical approach for the
freshwater acute and chronic derivations appear valid.
Incorporation of hardness normalization is warranted given the
likelihood that Cd and Ca compete for similar biological and
abiotic sites. In addition, the increased number of species
extending the SSDs is also an excellent step forward in
confirming proposed criteria.
Thank you for your comment.
No edits needed.
Reviewer 2
Of concern is the approach utilized for the chronic
estuarine/marine values. Utilization of ACRs with freshwater
fish or other organism to derive estuarine/ marine values is not
appropriate, especially when the criteria concentrations are
increased. It is also unclear why freshwater salmonid values
The use of a freshwater ACR to derive estuarine/ marine
values is described as an acceptable approach in the 1985
Guidelines, and was used in the draft criterion document
reviewed by the external peer reviewers.
Section 2.7.3
Section 4.4.2
Section 5.5.1
11
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
were not utilized for the ACRs, as many reside in
estuarine/marine environments (see salmonid comments
below).
Based on the peer reviewer comment, the estuarine/marine
ACR approach was re-examined and revised for the 2015 draft
proposal for public comment. The revised FACR incorporates
data for seven genus-level ACRs and was derived using data
for marine species and a diversity of freshwater species, many
of which have taxonomically-related marine species. ACRs
used to derive the FACR incorporate data for five freshwater
fish species, three freshwater invertebrate species, and two
acutely sensitive estuarine/marine mysids.
Reviewer 3
Bioaccumulative effects of Cd are largely ignored in this
document.
Text has been added to discuss these points and incorporate
the work of other researchers, such as Mebane (2006), who
discuss cadmium bioaccumulation and dietary exposure in
detail and the uncertainties currently associated with its
evaluation. In particular, text was added to detail how
cadmium can bioaccumulate in aquatic organisms through
multiple exposure routes including ingestion and direct
exposure, and how total uptake depends on the cadmium
concentration, exposure route and the duration of exposure.
Text was also added to clarify that there does not appear to be
a consistent relationship between body burden and
toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
Section 5.6.1
12
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 3
My comments for this section are divided into 2 parts: 1. The
technical approach according to the 1985 Guidelines, and 2.
The technical approach in light of our current understanding of
cadmium bioaccumulation, effects, and deficiencies in the
traditional testing approaches.
The technical approach according to the 1985 Guidelines
A. What is the rationale for use of EC20 values for the chronic
toxicity assessment? I understand that a MATC approach
(based on NOEC and LOECs) has its issues, and I'm generally
in favor of more statistically robust approaches such as the use
of an EC level based on entire datasets. But why is a 20%
effect level chosen here? This value seems rather high. There
should be some rationale for choosing this value, and this
rationale should be clearly articulated in the text. How do we
know that a 20% effect level has no impacts at the population
level?
The endpoint for chronic exposure is the EC2o, which
represents a 20 percent effect/inhibition concentration. This is
in contrast to a concentration that causes a low level of
reduction in response, such as an EC5 or ECio, which is rarely
statistically significantly different from the control treatment.
U.S. EPA selected an EC2o to estimate a low level of effect
that would be statistically different from control effects, but
not severe enough to cause chronic effects at the population
level (see U.S. EPA 1999c). Reported NOECs (No Observed
Effect Concentrations) and LOECs (Lowest Observed Effect
Concentrations) were only used for the derivation of chronic
criterion when an EC2o could not be calculated for the genus.
A NOEC is the highest test concentration at which none of the
observed effects are statistically different from the control. A
LOEC is the lowest test concentration at which the observed
effects are statistically different from the control. When
LOECs and NOECs are used, a Maximum Acceptable
Toxicant Concentration (MATC) is calculated, which is the
geometric mean of the NOEC and LOEC.
Regression analysis was used to characterize a concentration-
effect relationship and to estimate concentrations at which
chronic effects are expected to occur. For the calculation of
chronic criterion, point estimates were selected for use as the
measure of effect over a MATC, as MATCs are highly
dependent on the concentrations tested. Point estimates also
provide additional information that is difficult to determine
with an MATC, such as a measure of effect level across a
range of tested concentrations.
U.S. EPA. 1999c. 1999 Update of ambient water quality
criteria for ammonia. EPA-822-R-99-014. National Technical
Information Service, Springfield, VA.
No edits needed.
13
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 3
Only 3 species (all fish) were used to generate the hardness
correction for the freshwater chronic toxicity data set. D.
magna and P. promelas data were not used because only
MATCs were available and not EC20s. Is it not possible to
estimate EC20's from these datasets? The use of only 3
species to make this very important hardness adjustment
would seem to add a significant level of uncertainty to the
final analysis, especially since 2 of species used have
divergent slopes. ANCOVA (p=0.08) based on data from 3
species was used to say that the slopes 0.32, 1.46 and 1.08 are
not different and can be pooled. Is this defensible? Shouldn't a
conservative slope estimate be chosen here.... especially in
light of the fact that a 20% effect level is much higher than an
MATC or EC05 would be?
EC20S were not estimated for species other than the three fish
species because the data necessary to calculate EC2o point
estimates were not provided by the authors. EC2o point
estimates were preferentially selected for use over aNOEC or
LOEC as the measure of effect, as NOECs and LOECs, which
are the basis of the MATCs, are highly dependent on the test
concentrations selected. Furthermore, point estimates provide
additional information that is difficult to determine using
NOEC and LOEC effect measures, such as a measure of effect
level across the range of tested concentrations, and the
confidence intervals around those measures of effect.
Correspondence has been sent to the authors who did not
provide raw data for their studies, so EC20S can be calculated if
the data are available. Additional EC2oS were calculated based
on their responses.
An additional analysis was conducted to determine if the
inclusion of 3 MATCs from the Chapman Manuscript for D.
magna could be included in the hardness relationship along
with the new EC20s. This additional data supported the same
conclusion that a pooled slope could be generated with a
slightly different slope of 0.7977. Values were edited to reflect
this new pooled slope.
Section 3.1.2
Appendix C
Reviewer 3
The most acutely sensitive marine genus, Tigriopus was not
used in the analysis. The rationale was that it falls below the
5th percentile of the distribution. Isn't the whole point of the
SSD to determine what is protective of 95% of the species?
(Not 95% of the remaining taxa after sensitive taxa are
arbitrarily removed from the dataset). Shouldn't all of the data
be used here?
The 1985 Guideline recommendations were followed in that
the four GMAVs closest to the 5th percentile are used to
estimate the FAY.
Section 2.5
Section 2.7.2
14
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 3
The use of 2 ACRs from freshwater species in the
development of a marine chronic criterion is dubious on many
fronts. The justification for doing this needs to be articulated.
If justifiable, the authors should then justify their choices as to
why these 2 species were chosen. The reason given in the text
is that the freshwater species were chosen on the basis of
being acutely sensitive. However the purpose of ACRs is to
evaluate the potential for the chemical to cause chronic
toxicity. Use of an acutely sensitive species for ACR choice
should theoretically result in species with low ACRs, and in
this case, this is borne out. The freshwater invertebrate L.
silquoidea has a reported ACR of 2.727, suggesting that is
chronically not very toxic. However, the ACRs for most
species are considerably higher: (see below)
Mebane (2010) list ACRs for freshwater invertebrates:
Ephemerella: 158.67
Physa: 47.6
Aplexa: 28.5 and 47.87
Ceriodaphnia: 12.41 and 31.5
Daphnia: 65, 155, 112, 13
Hyalella: 17.5
This document lists the following freshwater invertebrate
ACRs:
Aplexa: 49.7
Lymnea: 12.81
Ceriodaphnia: 19.82
Daphnia: 57.3
With all of these values to choose from, 2.727 is clearly not a
representative ACR for freshwater invertebrates. Since the use
of a "mean ACR" is being applied across taxa, shouldn't the
The use of a freshwater ACR to derive estuarine/ marine
values is consistent with the 1985 Guidelines. However, based
on the peer reviewer comment, the estuarine/marine ACR
approach was re-examined and revised in the 2015 draft
proposal for public comment. The revised FACR incorporates
data for seven genus-level ACRs and was derived using data
for both marine species and a diversity of freshwater species,
many of which have taxonomically-related marine species.
The revised FACR of 8.291 was derived from a geometric
mean of genus-level ACRs for the following:
• Estuarine/marine mysids, Americamysis bahia and A.
bigelowi
• Cladocerans, Ceriodaphnia dubia and Daphnia (D.
magna and D. pulex)
• Mottled sculpin, Cottus bairdii
• Salmonids, Oncorhynchus (0. mykiss, 0.
tshawytscha) and Salmo (S. trutta)
• Fathead minnow, Pimephales promelas
The seven ACRs differ by a factor of < 11.95, which
approximates the factor of 10 or less recommended by the
1985 Guidelines. The ACRs for salmonids were less than
2.0 and were therefore raised to 2.0 to be consistent with
the 1985 Guidelines. The ACRs for the other freshwater
species were not used for the revised FACR because they
have no taxonomically-related marine species (e.g.,
pulmonate snails) and/or the ACRs appear to be outliers.
The description of and rational for the new estuarine/marine
ACR approach is provided in the post-peer review 2015 draft
document.
Section 2.7.3
Section 4.4.2
Section 5.5.1
15
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
values be representative? Would it make sense to have higher
ACRs apply to invertebrates and lower ACRs apply to fish
since fish generally have low ACRs and inverts generally have
high ACRs?
Reviewer 3
Technical approach based on what we understand about the
world post 1985:
Cadmium has been demonstrated to be toxic to practically
every in vitro system it has been tested in. We strive to limit
human dietary exposures in part because it is a known
carcinogen and is nephrotoxic after dietary exposure. Effects
of Cd on antioxidant physiology are well described in several
species including aauatic insects. What evidence can we point
to suggest that bioaccumulated Cd is not toxic to aquatic
organisms? This is a fundamental flaw in this document.
Text has been added to discuss these points and incorporate
the work of other researchers, such as Mebane (2006), who
discuss cadmium bioaccumulation and dietary exposure in
detail and the uncertainties currently associated with its
evaluation. In particular, text was added to detail how
cadmium can bioaccumulate in aquatic organisms through
multiple exposure routes including ingestion and direct
exposure, and how total uptake depends on the cadmium
concentration, exposure route and the duration of exposure.
Text was also added to clarify that there does not appear to be
a consistent relationship between body burden and
toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
Section 5.6.1
Reviewer 3
We have a major and important disconnection between what
traditional laboratory tests (using only direct aqueous
exposures) and what field ecologists tell us about metal effects
in aquatic insects. Because insects are such important players
in freshwater ecosystems, and are the focus of CWA-driven
biomonitoring programs, we have numerous examples of
stream community structure being impaired by metal
Criteria were derived considering lab water-based exposures
using procedures that are consistent with the 1985 Guidelines.
Additional discussion has been added to address the
uncertainty of using lab-based tests to determine protective
field concentrations and the importance of dietary exposures
to this faunal group.
Section 2.5
Section 5.1.3
Section 5.6.1
16
-------
Rc\ icwcr
K\(crn;il I'ccr Rc\ icwcr Comments Rc«i:irdin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
exposures. Yet lab (aqueous) tests generally suggest that
insects are insensitive to Cd. Work in our laboratory has used
Cd uptake and depuration kinetics to clearly demonstrate that
96 hour exposures are insufficient to elicit toxicity in aquatic
insects are ecologically relevant concentrations (Buchwalter et
al. 2007, Buchwalter et al. 2008, Poteat and Buchwalter 2014,
Poteat and Buchwalter 2014). We have also shown that
periphyton is a major sink for Cd, and is readily
bioaccumulated in insects (Xie et al. 2010). We have also
showed that Cd exposure does not negatively affect Ca
transport in insects (Poteat and Buchwalter 2014) (as it is
known to do in acutely sensitive taxa), and Ca provides little
protective effects on Cd uptake (Poteat et al. 2012). Finally,
we show that diet derived (but not water derived) Cd affects
antioxidant physiology suggesting that dietary exposures may
be more challenging to aquatic insects that aqueous exposures
(Xie and Buchwalter 2011). These findings mirror those of
Irving et al., 2003. All of these findings point towards short-
term, water-only exposures are insufficient for evaluating
metal toxicity in this important faunal group (see (Poteat and
Buchwalter 2014) for discussion of these findings).
In addition, generally good agreement has been reported for
microcosm studies/whole effluent toxicity test results with
corresponding field observed effects (Clements and Kiffney
1996; Clements et al. 2002; Norberg-King 1986). Mebane
(2006) compared chronic criterion values and apparent effects
values from ecosystem studies and field surveys and
concluded that the data showed mostly good agreement
between the laboratory-based predictions and effects observed
in the field surveys or ecosystem experiments.
EPA concurs with the reviewer about the importance of
considering the dietary exposure route. Text has been added to
discuss this and incorporate the work of other researchers,
such as Mebane (2006), who discuss cadmium
bioaccumulation and dietary exposure in detail and the
uncertainties currently associated with its evaluation. In
particular, text was added to detail how cadmium can
bioaccumulate in aquatic organisms through multiple exposure
routes including ingestion and direct exposure, and how total
uptake depends on the cadmium concentration, exposure route
and the duration of exposure. Text was also added to clarify
that there does not appear to be a consistent relationship
between body burden and toxicological effect, and an
acceptable tissue effect concentration cannot be defined for
aquatic life at this time. Bioaccumulation and effect level data
that are available indicate that cadmium is unlikely to
accumulate in tissue to levels that would result in adverse
effects to aquatic invertebrates or fish at the calculated chronic
criterion concentrations. For this reason, the evaluation of
direct exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
17
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
Overall yes, I think the technical approach is scientifically
sound and consistent with the protection of aquatic life. I do,
however, have some specific significant comments for EPA to
consider which I list below.
Thank you for your comment.
No edits needed
Reviewer 4
Page 15: EPA concludes that most changes in Cd toxicitv can
be explained by changes in hardness and therefore
incorporation of the BLM into this revision is not necessary. I
strongly disagree with this statement. Every study I'm aware
of in which a range of DOC and pH have been measured has
shown that these parameters strongly influence Cd toxicity.
Just because the majority of laboratory studies are conducted
in laboratory waters with low DOC and do not measure
dissolved organic carbon (DOC), does not provide a valid
rationale for not using the BLM (biotic ligand model).
Obviously, in the natural environment, DOC varies widely. I
would think the objective of the criteria is to ensure that they
are protective/predictive of toxicity in the natural
environment, not in artificial laboratory waters.
EPA revised the text to indicate that hardness is a critical
factor in determining toxicity, and additional water quality
parameters such as DOC, alkalinity, and pH may also
influence cadmium toxicity. As the external peer reviewer
noted, the objective of the criteria is to ensure that they are
protective and predictive of toxicity in the natural
environment. The addition of consideration of DOC would
generally yield higher criteria values. Thus the focus on
hardness only in this draft is expected to be protective. The
EPA may consider the applicability of the BLM including
parameters such as DOC in future revisions of the cadmium
criteria.
Criteria were derived considering lab water-based exposures
using procedures that are consistent with the 1985 Guidelines.
Additional discussion has been added to address the
uncertainty of using lab-based tests to determine safe field
concentrations and the importance of dietary exposures to this
faunal group.
In addition, generally good agreement has been reported for
microcosm studies/whole effluent toxicity test results with
corresponding field observed effects (Clements and Kiffney
1996; Clements et al. 2002; Norberg-King 1986). Mebane
(2006) compared chronic criterion values and apparent effects
values from ecosystem studies and field surveys and
concluded that the data showed mostly good agreement
between the laboratory-based predictions and effects observed
in the field surveys or ecosystem experiments.
Section 2.3.1
Section 2.5
Section 5.1.3
18
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
Page 34: Following up on the previous comment regarding not
using the BLM, why did EPA only consider a multiple linear
regression with alkalinity? Why not pH and/or DOC? It is
quite possible that pH autocorrelates with hardness as well
given this is the case for most artificial laboratory waters
(though not as consistent for natural waters), but there will not
be an autocorrelation with DOC. This is a really important
water quality parameter that EPA is ignoring.
Text relating directly to alkalinity was removed from the
document and replaced with the text discussed in the previous
response to comment. EPA notes that integrating DOC into
the analysis would be expected in most cases to make the
criteria less stringent. Thus, while recognizing the need to
consider applicability of the BLM in future cadmium criteria
updates, it is notable that the inclusion of DOC in the BLM
approach will likely not make the criteria more stringent or
conservative.
Section 2.3.1
Section 2.5
Section 5.1.3
Reviewer 4
Page 50-51: Is the studv bv Vover et al. (1974). the onlv studv
where the effects of salinity on Cd toxicity was not consistent
or are there multiple studies with this problem? If it's only this
one study, it's not clear why the general trend would be
ignored. I don't think EPA would ignore the hardness
relationship in freshwater if only a single study was
inconsistent with the general trend. It is a concern that there is
an obvious and significant salinity effect for the Neomysis
integer data (p. 51), which is one of the four taxa used for the
criteria derivation, and yet this obvious effect is ignored and
the geometric mean is used to develop the species mean acute
values (SMAV). Does EPA consider a test performed at a
salinity of 1 ppt to be a marine test?
Based on the peer reviewer's comments, additional analysis of
the relationship between salinity and cadmium toxicity was
conducted. As discussed in a previous comment, text has been
added detailing why a salinity normalization approach was not
used in the criteria development. A salinity-toxicity trend
could not be definitively characterized and a mathematical
relationship could not be described to define the dependency.
EPA will continue to use the 1 ppt as the lowest salinity level
for a saltwater test, consistent with the development of
estuarine/marine criteria, which is applicable to the broad
range of salinities characterized by these habitats. This salinity
is also consistent with Mitsch and Gosselink (1986) who
classify a waterbody with a salinity of 0.5-5.0 ppt as
oligohaline.
Note that for the salinity in the lppt Morone exposure, the
conductance in this experiment was 1,600 uS/cm at 25 C. This
was approximately three times the conductance compared to
fresh hard water.
Mitsch, W.J. and J.G. Gosselink. 1986. Wetlands. Van
Nostrand Reinhold, New York. 539 pp.
Section 2.3.1
Section 5.4.1
19
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
Unfortunately some aspects of the document lead to answering
both parts of the charge question 2 with answers of "no." I am
only commenting on aspects which to me did not follow the
available science, deviate from the principles of the 1985
"Guidelines" or otherwise have logical problems. While
Stephan et al. (1985) Guidelines for derivation of aquatic life
criteria are 30 years old and aspects of the science have
progressed such that some details may not fit, they include
solid principals that should continue to guide the approach.
Key among Stephan et al. guiding concepts is from their p. 3:
"The guidelines were intended to provide the same level of
protection as would an (infeasible) approach of conducting
field tests on a wide variety of unpolluted bodies of water,
adding various amounts of the material to each body of water
in order to determine the highest concentration that would not
cause any unacceptable long-term or short-term effects on the
aquatic organisms or their uses" Further (p. 10), These
National Guidelines have been developed on the theory that
effects which occur on a species in appropriate laboratory
tests will generally occur on the same species in comparable
field situations. All North American bodies of water and
resident aquatic species and their uses are meant to be taken
into account." Not bodies of water for which conditions are
optimal - all bodies of water.
Thus, a key concept behind the logic of criteria derivation is
that criteria be suitable for diverse, natural water bodies, and
laboratory data should attempt to encompass comparable field
situations. The draft document instead moves towards a very
different concept of only using data from an idealized
aquaculture setting, without regard to whether the species
occurs in the wild in waters with "suboptimal" conditions.
We concur that an objective of the 1985 Guidelines is to
provide for the development of criteria that are applicable to a
variety of field conditions. The testing procedures must,
however, be conducted with organisms that are determined to
be fundamentally healthy and with tests that meet with a
consistent set of standards in order to evaluate test
acceptability and develop criteria that are not impacted by
testing artifacts and that are applicable on a national basis.
This approach is consistent with internationally-recognized
and broadly applied approaches for developing effects
analyses for toxicants, relying on such reproducible laboratory
data because they are designed to be as free from confounding
influences as possible, in order to permit for robust,
unconfounded consideration of risk for a given chemical, and
relative risk across chemicals. States, tribes, and other end
users can then consider site-specific conditions and variables
in the development of standards that are applicable to their
specific end use, such as the application to a particular water
body or region. A further discussion of uncertainty regarding
differences between laboratory and field conditions and
implications for criteria has been added to the document.
Section 2.5
Section 5.1.3
20
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
Drilling down on Hyalella
Most fundamentally, by throwing out all long-term test
endpoints for the most sensitive genus (Hyalella) this
document strays from a guiding principle of the Guidelines
that criteria are to protect diverse natural waters. Criteria are
indeed developed using laboratory data, but they are not
intended to apply to laboratory waters; they are intended to
apply to natural waters. This disconnect between laboratory-
based derivation of numeric water quality criteria and
application to natural waters has repeatedly debated in the
literature, with me chiming in specifically with cadmium
(Mebane 2010).
In essence, optimal aquaculture conditions are defined for
culturing Hyalella azteca, and chronic tests in which less than
15 mg/L chloride was present in dilution waters, or control
growth, survival, and reproduction did not meet expectations.
These were control growth (>0.35 mg at 28 days and >0.5 mg
at 42 days), survival (80% at 42d) and reproduction (>6 per
young). No explanation was found in the document why
researchers were tasked to drill down on Hyalella, as any
commonly used test organism could have been similarly
scrutinized. Absent explanation, the inference is that Hyalella
must have been chosen because it was the most sensitive
organism, and there was a desire to exclude data if this
heightened sensitivity could be shown to be an artifact of
stressful laboratory culture conditions. In essence this logic
requires the following implicit assumptions. Since only
Hyalella data obtained from laboratory test waters >15 mg/L
are to be used for criteria development, it follows that:
In ambient waters, Hyalella (and presumably other freshwater
amphipods) are only expected to occur in waters with >15
mg/L chloride; Alternatively if Hyalella do in fact occur in
In addition to the response to the previous comment,
additional text has been added to further detail the decision
process that was used, based on the recently-completed
evaluation, to determine which Hyalella tests were included in
the evaluation. The basic premise behind the selection of
specific Hyalella tests, based on the consideration of test
conditions, is that in order to develop robust comparative
toxicity tests, animal husbandry conditions should be optimal
and provide for low control mortality and optimal control
growth to decrease control noise, increase the ability to
capture low level effects, and thus understand the implications
of introducing a toxicant into the system even at low levels.
EPA developed this test condition/husbandry analysis for
Hyalella after repeatedly observing extremely high variability
in Hyalella test results for the same compound under different
lab/husbandry conditions (e.g., chloride concentrations in test
water) and considering less than optimal control survival,
growth and reproduction. These analyses led to the
determination of conditions under which repeatable results
could be obtained by minimizing interfering confounders,
such as water chemistry and diet.
These analyses were not developed to exclude sensitive tests.
EPA's analyses are developed with the goal of generating high
quality and scientifically defensible predictions of
concentrations, that if not exceeded beyond the specified
frequency and duration, will be protective of aquatic life.
Section 2.5
Section 5.1.3
Section 5.2.1
21
-------
Rc\ icwcr
K\(crn;il I'ccr Rc\ icwcr Comments Rc«i:irdin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
waters with lower chloride concentrations, the criteria are
only intended to apply to waters with >15 mg/L.
Chloride is an important factor affecting the toxicity of
cadmium to Hyalella (and presumably other related but less
well studied amphipods or freshwater crustaceans). If so, then
it follows that:
Chloride should be included in the criteria derivation and
factored into the criteria. Per the Guidelines (p32), "when
enough data are available to show that the chronic toxicity is
similarly related to a water quality characteristic, the
relationship should be taken into account .... If two or more
factors affect toxicity, multiple regression analysis should be
used."
Alternatively, while not specifically mentioned in the
guidance, if data were insufficient for the covariance or
multiple regression analyses endorsed, it would seem
reasonable to establish different criteria in brackets, such as
waters <15 mg/L chloride or >15 mg/L chloride.
Alternatively, if chloride is not an important factor affecting,
then there is no reason to factor it into the criteria
development.
However, Appendix K does not address the question of
whether chloride is a factor affecting cadmium toxicity, all
that has been established is that Hyalella growth and
reproductive output is greatest in waters with chloride >15
mg/L. This is not unexpected. Freshwater environments
usually have an osmolarity far less than blood plasma, and
energy requirements to maintain hydromineral balance
increase in more dilute waters (e.g., Wendelaar Bonga and
Lock 2008). Fish in dilute waters don't grow well either. For
instance, about 80% of the restaurant/retail rainbow trout sold
in the United States come from a 30 mile stretch known as the
22
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Thousand Springs area of southern Idaho. There the constant
chloride of about 20 mg/L, hardness of about 180 mg/L and
temperature of 15°C provide optimal energy conversions and
growth per unit feed. It would follow just as logically that only
rainbow trout data that were generated from waters with
chloride >15 mg/L or so should be used, because that
optimizes growth? Why would it not follow that only acute
data in which organisms were fed should be used, because
starvation stresses organisms? This seems to be internally
inconsistent logic.
Reviewer 5
The reason why Appendix K was requested was never stated.
It should be. I assume the reason must be a presumption that if
organisms do not grow and reproduce at high rates, then they
will "too sensitive" or not represent responses expected in
natural conditions. It is not obvious that this is the case.
McNulty et al. (1999) showed that starved amphipods exposed
to low levels of cadmium survived better than controls.
However, even if optimal diets do produce higher (less
sensitive) growth and reproduction effects with Cd and
Hyalella, the universal use of optimal diets could lead to
underestimation of the toxicity risks experienced by wild
populations, which may experience limited food availability.
In the wild, organisms don't live in optimal conditions. Even
in the center of their ranges, conditions are seldom optimal all
of the time. Organisms also live in marginal conditions, for
they tend to expand their ranges to the limits of their
physiological tolerances. See for example France's (1996)
description of Hyalella living on the margins of lakes with
tolerable mineral content (France 1996). Similarly, Gibbons
and Mackie (1991) showed that increasing reproductive output
of H. azteca was associated with increasing sulfate, calcium
hardness, sediment particle size, conductivity, alkalinity,
seston, and the organic matter of the fine sediment. This
The basic premise behind the selection of specific Hyalella
tests, based on the consideration of test conditions, is that in
order to develop robust comparative sensitivity analyses,
animal husbandry conditions should be optimal and provide
for low control mortality and optimal control growth to
decrease control noise, increase the ability to capture low level
effects, and thus understand the implications of introducing a
toxicant into the system even at low levels.
EPA developed this test condition/husbandry analysis for
Hyalella after repeatedly observing extremely high variability
in Hyalella test results for the same compound under different
lab/husbandry conditions (e.g., chloride concentrations in test
water) and considering less than optimal control survival,
growth and reproduction. These analyses led to the
determination of conditions under which repeatable results
could be obtained by minimizing interfering confounders,
such as water chemistry and diet.
These analyses were not developed to exclude sensitive tests.
EPA's analyses are developed with the goal of generating high
quality and scientifically defensible predictions of
concentrations that if not exceeded beyond the specified
Section 2.5
Section 5.1.3
Section 5.2.1
Section 5.6.1
23
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
consistent with Appendix K, but begs the question, what are
effects of Cd in these suboptimal waters? Why assume that if
Cd criteria are needed, they should only be developed from
exposures in high hardness, but then blindly extrapolate results
to low chloride, low hardness conditions using tests with other
organisms? This is further logical problem with Appendix K's
rationale - as noted in appendix K, waters with hardness less
than 80 mg/L tend to have chloride less than 10 mg/L. Does
the hardness-toxicity relation predict safe conditions for
Hyalella at low hardness? No way to know.
frequency and duration will be protective of aquatic life.
We concur that an objective of the 1985 Guidelines is to
provide for the development of criteria that are applicable to a
variety of field conditions. The testing procedures must,
however, be conducted with organisms that are determined to
be fundamentally healthy and with tests that meet with a
consistent set of standards in order to evaluate test
acceptability and develop criteria that are not impacted by
testing artifacts and that are applicable on a national basis.
Reviewer 5
I've poked around a bit the literature on Hyalella life histories
under different environmental stresses in an effort to include
extrapolate organism-level effects of Cd to potential
population-level effects (Mebane 2010). While by no means
exhaustive, and by now a bit dated, this leads to some other
thoughts on the expected control survival, growth, and
reproduction in long term tests in Appendix K. With control
survival, in at least some wild populations, I estimated half-
month survival rates for juveniles of about 0.9, or close to a
5% decline per week (Mebane 2010, Table II). This is higher
than the 2-3% noted in Appendix K, and suggests that in the
wild, survival to 42-days would likely be less than 80%. With
regards to growth, while some wild populations grew as much
as those in the laboratory settings discussed in Appendix (>0.5
mg at sexual maturity), this cannot be assumed in all natural
waters. Cooper (1965) reported average dry weights of adults
Hyalella were 0.2 mg in a population in a warm, shallow lake
in Michigan. Gibbons and Mackie (1991) reported mean
weights of Hyalella at maturity were only 0.1 mg, and weights
of all Hyalella were only 0.3 mg. Thus the 0.35 at day 28 and
0.5 mg at day 42 may be higher than that expected in some
natural settings. Gibbons and Mackie (1991) reported ranges
The basic premise behind the selection of specific Hyalella
tests, based on the consideration of test conditions, is that in
order to develop robust comparative sensitivity analyses,
animal husbandry conditions should be optimal and provide
for low control mortality and optimal control growth to
decrease control noise, increase the ability to capture low level
effects, and thus understand the implications of introducing a
toxicant into the system even at low levels.
EPA developed this test condition/husbandry analysis for
Hyalella after repeatedly observing extremely high variability
in Hyalella test results for the same compound under different
lab/husbandry conditions (e.g., chloride concentrations in test
water) and considering less than optimal control survival,
growth and reproduction. These analyses led to the
determination of conditions under which repeatable results
could be obtained by minimizing interfering confounders,
such as water chemistry and diet.
These analyses were not developed to exclude sensitive tests.
EPA's analyses are developed with the goal of generating high
quality and scientifically defensible predictions of
Section 2.5
Section 5.1.3
Section 5.2.1
Section 5.6.1
24
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
of brood per female ranged from 6-15, which is consistent
with appendix K. However, Strong (1972), his fig 4, showed
sometimes natural brook sizes may be as low as 3 per female.
In sum, the logical problems of how Appendix K's analyses
are used in the document are analogous to the metaphor of not
seeing the forest because of all the trees. Some trees were
examined in great detail (lab performance of Hyalella) but it
misses the point that the comparisons of acceptable conditions
should be again performance in the wild.
concentrations that if not exceeded beyond the specified
frequency and duration will be protective of aquatic life.
We concur that an objective of the 1985 Guidelines is to
provide for the development of criteria that are applicable to a
variety of field conditions. The testing procedures must,
however, be conducted with organisms that are determined to
be fundamentally healthy and with tests that meet with a
consistent set of standards in order to evaluate test
acceptability and develop criteria that are not impacted by
testing artifacts and that are applicable on a national basis.
Reviewer 5
Other items:
Problem formulation: It is germane to note that in natural
waters, Cd is always in association with Zn, usually at about
mass ratios of 1:200 (Wanty et al. 2009).
This information was added to Section 2.1 of the document.
Section 2.1
Reviewer 5
p. 12,1 was not quoted quite accurately. "Mebane (2014 2006)
concluded that, although there were not adequate data to
establish acceptable tissue effect concentrations for aquatic
life, cadmium is unlikely to accumulate in tissue to levels that
would result in adverse effects to aquatic invertebrates or fish,
at calculated chronic criterion concentrations, which were
lower than that chronic criterion concentration derived here. "
This report is variously cited as Mebane (2006), Mebane
(2010), or Mebane (2014). The suggested citation is,
"Mebane, C.A. 2006. Cadmium risks to freshwater life:
derivation and validation of low-effect criteria values using
laboratory and field studies. U.S. Geological Survey Scientific
Investigation Report 2006-5245 (2010 rev.).
http://pubs. usgs.gov/sir/2006/5245/."
The 2010 revision only corrected minor mistakes, and did not
include any updated literature reviews.
Text was added as suggested and the citation was fixed.
References and
various locations
in document
Section 2.3
Section 5.6.1
25
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 2
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
p. 28, the approach of requiring data used in the hardness-
toxicity regressions to have a 3X spread and 100 mg/L
absolute difference between the highest and lowest value was
indeed used in the 2001 version, but was not really presented
as policy. In contrast, my colleagues and I found that
hardness-toxicity relations were more reliable from test series
that concurrently tested the same cohort of organisms in
waters with different hardness, than were ad hoc collections of
found data tested under different conditions at different
hardness levels (Mebane et al. 2012).
Where available, giving concurrent test series data obtained at
different hardnesses precedence over general hardness-toxicity
compilations would be warranted.
The approach of requiring data used in the hardness-toxicity
regressions to have a 3X spread and 100 mg/L absolute
difference between the highest and lowest value was
established when updating the cadmium document in 2001.
This practice has been followed for all subsequent criteria
document updates because it was found that the variability
associated with different test conditions that are associated
with multiple studies can sometimes be so great that it masks
the hardness/toxicity relationship.
Data for each species are first reviewed to determine if they
are potentially suitable for use in the hardness-toxicity
evaluation. The data are initially considered regardless of
source/test condition (laboratory, dilution water, temperature,
etc.). However, if the hardness/toxicity data are widely
scattered, we then attempt to decrease uncertainty introduced
by the differing test conditions by focusing on those studies
specifically evaluating the toxicity relationship. In addition,
studies are excluded when only a single acute toxicity value
was available and where multiple tests were conducted at the
same hardness. When different life stages were used at test
initiation, only data for the same life stage is evaluated. The
end result is that the most defensible data are used to develop
the hardness-toxicity slope.
No edits
26
-------
2.3 Charge Question 3
3. Please comment on the data used to derive the revised criteria, including data adequacy/comprehensiveness, and the appropriateness of the
data selected and/or excluded from the derivation of the draft criteria. Is the data used correctly for the intended purpose? Are there other relevant
data that you are aware of that should be included? If so, please provide the data along with supporting information.
Re\ iewer
K\(ern;il Peer Re\ iewer Comments Re<>;irdin<> C'hjir»e
Question 3
KIW Response
Re\ ision
Locution in 2015
Drsil'l C iulniiiini
C'rileriii
Document
Reviewer 1
As noted above, the author chose to ignore many relevant
studies that did not conform with standard EPA toxicity
protocols. But the problem is that these protocols basically
ignore the fact that animals eat, hardly a realistic scenario and
are too simplistic in looking only at growth and mortality.
Typically, the test organisms are exposed to dissolved Cd at
varying concentrations, but in the absence of food.
Occasionally, some artificial food (fish flakes or the like) is
presented once every several days (sometimes never!) to keep
the animals alive. But these studies are hardly representative
of what happens in natural waters.
Studies that were determined not to be acceptable were
presented in Appendix J, along with a rationale for their
exclusion. However, the exclusion of food in acute tests is
standard practice in EPA, ASTM and internationally-
harmonized toxicity test protocols. This is based on the
potential for food to alter the exposure concentration and/or
bioavailability of the chemical. This approach is consistent
with procedures stated on page 14 of the 1985 Guidelines:
"Except for test with saltwater annelids and mysids, results of
acute tests during which the test organisms were fed should
not be used, unless data indicate that the food did not affect
the toxicity of the test material".
No edits needed.
Reviewer 2
The use of additional species for SSD reduced uncertainty and
greatly improved criteria assessments for freshwater. The QA
evaluations of data usefulness was adequate and the data
selected for the acute responses was correctly used for the
intended purpose. The mechanistic assumption that adverse
effects are primarily related to calcium uptake at the gill, is
accurate for acute effects. Consequently, the data used for
derivation of the criteria for acute effects is valid.
Thank you for your comment.
No edits needed.
27
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 2
However, with regard to chronic effects, there are other targets
once absorption of cadmium occurs, particularly the kidney,
brain and gonad. In addition to specific interactions with
signaling proteins, Cd clearly binds sulfhydral groups of
proteins within targets disrupting cellular maintenance. The
latter two tissue targets above are likely involved in the
reproductive effects observed with chronic exposures. Cd
clearly disrupts the Hypothalmic Pituitary Gonadal axis and
gonadal function in fish (Vetillard, and Bailhache 2005). It
reduces vitellogenin in females and accumulates in kidney
upon chronic exposures either via diet or water (Szczerbik et
al. 2006; Thomann et al. 1997).
It is understood that tissue data from these organs are limited,
but studies that have these data, or the fact that these data are
limited should be discussion points of the uncertainty analysis.
Clearly, discussions of uncertainty regarding accumulation are
needed, particularly in light of limited data for chronic effects
in estuarine/marine organism. The statement "Aquatic
organisms are considered to be more susceptible to cadmium
from direct aqueous exposure than through bioaccumulation
and the development of criteria protective of direct exposure
effects are considered more applicable to the development of
criteria for aquatic life" is clearly biased toward acute toxicity
and should be re-visited with particular emphasis on
reproductive effects of cadmium which likely result from
accumulation and not direct exposure.
EPA recognizes the difficulty in characterizing dietary
exposure and establishing critical tissue effects thresholds
associated with bioaccumulated metals.
The information provided regarding tissue targets has been
added to the Uncertainty section along with additional
discussion about limitations in the organ data.
Text has been added to discuss these points and incorporate
the work of other researchers, such as Mebane (2006), who
discuss cadmium bioaccumulation and dietary exposure in
detail and the uncertainties currently associated with its
evaluation. In particular, text was added to detail how
cadmium can bioaccumulate in aquatic organisms through
multiple exposure routes including ingestion and direct
exposure, and how total uptake depends on the cadmium
concentration, exposure route and the duration of exposure.
Text was also added to clarify that there does not appear to be
a consistent relationship between body burden and
toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
Section 5.6.1
28
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 2
With regard to reproduction, it is unclear what endpoint data is
being used to determine the effect values in the Appendices.
Tests are provided in terms of exposure duration, but it is
unclear whether growth, survival or reproduction is being
utilized as the endpoint. Again, given the potential for
reproductive effects upon chronic exposure, reproduction
would be expected to be the most sensitive endpoint. If other
endpoints were used then the uncertainties inherent to these
endpoints should be discussed. Clearly growth and survival
effects have likely difference mechanisms and targets than that
of reproduction.
The most sensitive acceptable endpoint is used for each study.
The endpoint for each exposure concentration was added for
each study in the table.
Appendix C
Reviewer 2
It is also significantly disappointing that data from the same 2
species in 1980s are still the only two species being used to
derive the 2015 values. In addition, it is puzzling how criteria
values can be raised for estuarine/marine organisms when the
same degree of uncertainty exists (only 2 species) in each year
criteria were assessed. To add in data from freshwater
organisms for ACR estimates increases uncertainty and does
not reduce it. Therefore, the 2001 value should stay as is, or be
reduced because of the uncertainty associated with its
derivation.
Additional acceptable estuarine/marine chronic data were not
found based on an extensive literature search that was
conducted in 2014. The CCC was calculated using the
saltwater FAV and FACR as recommended in the 1985
Guidelines.
The use of a freshwater ACR to derive estuarine/ marine
values is described as an acceptable approach in the 1985
Guidelines, and was used in the draft criterion document
reviewed by the external peer reviewers.
Based on the peer reviewer comment, the estuarine/marine
ACR approach was re-examined and revised for the 2015 draft
proposal for public comment. The revised FACR incorporates
data for seven genus-level ACRs and was derived using data
for marine species and a diversity of freshwater species, many
of which have taxonomically-related marine species. ACRs
used to derive the FACR incorporate data for five freshwater
fish species, three freshwater invertebrate species, and two
acutely sensitive estuarine/marine mysids.
Section 2.7.3
Section 4.4.2
Section 5.5.1
29
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 3
Practically all relevant work related to bioaccumulated Cd and
the importance of dietary exposures is ignored, (see (Barata et
al. 2002, Barata et al. 2002, Buchwalter et al. 2008, Cain et al.
2004, Croteau et al. 2003, Hare et al. 2001, Hare et al. 2003,
Irving et al. 2003, Klaassen et al. 1999, Luoma and Rainbow
2005, Luoma et al. 2009, Luoma and Carter 1991, Martin et
al. 2007, Timmermans et al. 1992, Wallace et al. 2003, Xie et
al. 2010, Xie and Buchwalter 2011, Xie et al. 2008) for some
examples)
The referenced materials were evaluated and added, as
applicable, to the bioaccumulation uncertainty section.
Section 5.6.1
Reviewer 3
I suspect that there are other reviewers who can comment
more directly on the issues with Hyalella data, so I will refrain
from doing so here.
Thank you for your comment.
No edits needed.
Reviewer 4
Overall, I found the data used by EPA to derive the criteria to
be comprehensive and generally sound. There are a few
specific data where I have concerns that EPA should consider
as described below.
Thank you for your comment.
No edits needed.
Reviewer 4
Page 51: I'm verv concerned that EPA is still allow ing studies
in which test concentrations were unmeasured as being
acceptable for WQC derivation. This is particularly
concerning when they are for one of the four taxa used to
calculate the criteria. In my opinion, these studies should not
be included.
The use of unmeasured acute test study results is an acceptable
approach in the 1985 Guidelines under specific conditions.
The lack of measured exposure concentrations in an acute
toxicity test does not invalidate the results if there is a
demonstration that the material tested is stable during the
testing period. Only if there were observed solubility problems
(e.g., precipitant present) would the data be suspect and
therefore potentially not acceptable.
No edits needed.
30
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
Page 68:1 agree with EPA's use of freshwater ACRs to
supplement the limited marine ACRs for the purpose of
deriving a final marine ACR. However, I question whether use
of the ACR for Lampsilis siliquoidea is appropriate. There are
obviously a number of factors that influence the ACR, but a
major factor is the life history of the organism and the life
stage selected for the acute toxicity test used to derive the
ACR. It seems to me that freshwater mussels have a unique
life history with no real analog in marine systems (marine
bivalves have a different life history). Consequently, use of
this of the ACR for this species to derive a marine ACR seems
inappropriate. I think use of an ACR for daphnids would be
more appropriate and representative of the life history of the
most acutely sensitive taxa in marine systems, the copepod
Tirgriopus.
Based on peer reviewer comments, calculation of the FACR
was revised and does not include Lampsilis. The revised
FACR incorporated data for seven genus-level ACRs and was
derived from data for marine species and a diversity of
freshwater species, many of which also have taxonomically-
related marine species. ACRs used to derive the FACR
incorporate data for five freshwater fish species, three
freshwater invertebrate species (including applicable data for
daphnids), and two acutely sensitive estuarine/marine mysids.
Section 2.7.3
Section 4.4.2
Section 5.5.1
Section 5.5.2
Reviewer 4
Table 17: Whv is the pH 6.0 test for H. azteca excluded? This
is within the range of test pH values (6.0-9.0) normally
considered by EPA. Additionally, earlier in the document it
was stated that hardness was the only water quality parameter
that mattered for normalizing Cd toxicity data. I disagree with
that statement, but if EPA is going to argue other water quality
parameters are not important, then I don't see how it can then
exclude data for this reason.
In addition to the pH being below the level accepted by EPA
for tests (6.5-9.0), Br and CI" concentrations were not provided
and the dilution water was comprised of well water that was
significantly diluted from a hardness of 380 mg/L to 15.3
mg/L (Mackie 1989). The Hyalella memo found in Appendix
K of the draft criteria document states that "Natural waters
with hardness less than 80 mg/L typically have <10 mg Cl/L".
The rationale for the exclusion of this study from the criteria
derivation was clarified in the document table.
Table 17
31
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
Table 18:1 agree with EPA's re-evaluation of the Hvalella
data and their application of water quality and performance
criteria for test acceptability. However, I'm concerned about
the study EPA retained for purposes of criteria derivation for
several reasons. First, I do not believe use of a 10-d survival
endpoint constitutes a chronic study as defined in Stephan et
al. (1985). EPA has excluded a number of other studies from
use in criteria derivation for this reason (e.g., the 21-d survival
study on the sea starlet anemone, p. 81) in this document that
creates a major internal inconsistency. Having said that, it
could be argued that inclusion of this sub-chronic data is
warranted given that it is the lowest toxicity value in the data
set and exclusion of the data would be non-conservative in
terms of environmental protection (as opposed to including
sub-chronic data for insensitive species). However, using this
logic why would the 7-d survival/growth data with the
fountain darter then be excluded?
In response to peer reviewers' comments, a further
examination of this issue was conducted. Thus, after further
evaluation, the full-life cycle study by Ingersoll and Kemble
(2001) was found to satisfy the acceptability criteria for H.
azteca and was used to replace the lOd study used in the
previous draft of the document. This change is based on
consultation with the study author, where it was determined
that techniques used to measure length data are likely to more
accurately reflect growth than the originally-reported direct
weight measurements. Since the original study was conducted,
this laboratory has developed a robust empirical relationship
between amphipod length and weight. Applying the formula,
the 28-d average control length translates into a weight that is
above the minimum control performance values listed in
Appendix K of the draft criteria document. The average
control reproduction for this study also met minimum
performance values. Although the feeding rate used in this test
was below that recommended for H. azteca, the finding that
control organisms met the performance criteria of tests using a
higher feeding rate supports retaining these data for use in
deriving the AWQC.
Section 3.1.2
Section 5.2.1
32
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
My second concern is whether the sensitivity of H. azteca is
real? Given that these 10-d data come from a 42-d study that
fails to meet control performance criteria, how does EPA
know that these animals weren't already stressed at 10 d and
inappropriately sensitive? Given both the duration and
performance issues associated with these data, in my opinion
they should not be used for WQC derivation. However, I
strongly encourage EPA to conduct a 28- or 42-d Hyalella
study that meets the necessary performance criteria. Finally,
after Table 18, EPA has descriptions of each of the chronic H.
azteca studies and rationale for their rejection but did not
include a description of the Ingersoll and Kemble study that
was accepted and the rationale for use of the 10-d survival
endpoint. This should be added to the document.
Please see response to previous comment. As indicated in the
response, the full-life cycle study by Ingersoll and Kemble
(2001) was found to satisfy the acceptability criteria for H.
azteca and was used to replace the lOd study used in the
previous draft of the document.
The Agency is interested in obtaining information regarding
new toxicity tests on H. azteca as noted in the Federal Register
Notice to be issued announcing the availability of the 2015
draft cadmium criteria document for public comment.
Section 3.1.2
Section 5.2.1
33
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
As noted in the response above, the exclusion of most Hyalella
data is doubtfully justifiable, because the criteria for doing so
are questionable. However, even with these Appendix K
criteria as they are, the Ingersoll and Kemble data
reproductive data should not have been excluded. The 42d
reproductive endpoint from that test met the Appendix K
criteria for control survival and brood size (6.35 per female).
The 28 day endpoint was presumably excluded because of low
growth as weight. However, the organisms were not weighed,
but rather lengths were measured and weights were inferred
from lengths. Regardless, by the stated logic, it would follow
to exclude the 28-day endpoint with low (estimated) weight.
But to then pick an acute survival endpoint (10-day) instead of
the 42-day reproductive endpoint is inexplicable.
The entry for this test in Table 2 is misleading. Saying the test
was a life cycle test, but then using an acute endpoint, is
misleading. I estimated the EC20 for reduced reproduction to
be about 1.2 (ig/L using logistic regression, or the MATC
(geomean of LOEC and NOEC) would be 0.98 (ig/L.
In response to peer reviewers' comments, a further
examination of this issue was conducted. Thus, after further
evaluation, the full-life cycle study by Ingersoll and Kemble
(2001) was found to satisfy the acceptability criteria for H.
azteca and was used to replace the lOd study used in the
previous draft of the document. This change is based on
consultation with the study author, where it was determined
that techniques used to measure length data are likely to more
accurately reflect growth than the originally-reported direct
weight measurements. Since the original study was conducted,
this laboratory has developed a robust empirical relationship
between amphipod length and weight. Applying the formula,
the 28-d average control length translates into a weight that is
above the minimum control performance values listed in
Appendix K of the draft criteria document. The average
control reproduction for this study also met minimum
performance values. Although the feeding rate used in this test
was below that recommended for H. azteca, the finding that
control organisms met the performance criteria of tests using a
higher feeding rate supports retaining these data for use in
deriving the AWQC.
Section 3.1.2
Section 5.2.1
34
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
Other specific points on data used or not used.
Durations of tests
If 30-day tests with salmonids that started with fry consistently
yield more sensitive results than 60-day tests that started with
eggs or embryos, why ignore all the shorter, more sensitive
tests. The Guidance counsels to beware of tests in which
acclimation probably occurred during resistant states.
Chapman (1985) recently described this problem. It would
make more sense to exclude the less sensitive data, rather
exclude the more sensitive data.
Use of the life cycle (LC) tests over the early life stage (ELS)
tests in the draft reviewed by the external peer reviewers was
consistent with the 1985 Guidelines. It was noted that there
was no consistent pattern of early life stage tests being more
sensitive than life cycle tests for salmonids
Subsequently, based upon peer reviewer comments, use of
sensitive salmonid tests was reconsidered and changes in the
approach were made for the 2015 draft criteria. Specifically,
ELS tests were used to calculate the revised SMCV in
instances where they were more sensitive than the LC tests
(e.g., Salmo trutta).
Section 3.1.2
Section 5.2.2
Appendix C
Reviewer 5
Likewise with Mottled Sculpin, there's doubtfully anything
special about 28-day exposures over 21-day exposures. Besser
et al. (2007) ran two tests, one 28-day and one 21-day test.
The 28-day was less sensitive, and it was used with the other
ignored. There is no established ASTM protocol for Mottled
Sculpin, and the ASTM (1998) mention of "28 to 120-day
(depending on species) continuous exposure" tests for early-
life stage tests refers back to their species-specific appendices.
The SMCV/GMCV for the most sensitive fish species, Cottus
bairdii, is from the results of one 28-d ELS test (Besser et al.
2007). The other study reported in the same paper, a 21-d ELS
study, was not used quantitatively for the criteria derivation
because there is a lack of guidance on the most appropriate
duration for ELS tests with this species. U.S. EPA and ASTM
guidance implies that ELS tests should last at least 28 days, so
these data were not added to Appendix C. However, it is
noteworthy that incorporating these data would only change
the CCC slightly; the SMCV would change from 1.721 to
1.470 (ig/L, and the criteria would only change by one-
hundredth of a microgram, from 0.80 to 0.79 (ig/L total
cadmium.
Section 5.2.2
Reviewer 5
Other data
(Calfee et al. 2014) and (Wang et al. 2014) report acute and
chronic data with White Sturgeon and Rainbow Trout. The
same data are reported in Environmental Toxicology and
Chemistry, but Wang is paywalled, so I would use the open
access USGS report version.
The cited papers were reviewed and the additional acceptable
data were added to the appropriate tables and appendices.
Appendix A
Appendix C
Table 7
Table 20
Sections 5.1
Section 5.8.1
35
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 3
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
An acute test with Mottled Sculpin, (Cottus bairdi) and Cd
was attributed to Mebane et al. (2012). We tested Shorthead
Sculpin, Cottus confusus.
This was an error. The species name will be corrected, as
appropriate.
Various locations
36
-------
2.4 Charge Question 4
4. Are the derived criteria appropriately protective of listed species and commercially and recreationally important species, particularly as the
criteria relates to salmonids?
Rc\ iewcr
Kxlcrnnl I'ccr Re\ iewer Comments Rc<>;ir(lin<> Chsirgc
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drill! C iulniiiini
Crilcriii
Document
Reviewer 1
I agree with the author that marine animals are less at risk than
freshwater animals, and this is primarily due to the strong
chloro-complexation of Cd in seawater, thereby reducing the
bioavailability of Cd. Consequently, marine bioconcentration
factors are often 1 -2 orders of magnitude higher in freshwater.
Thank you for your comment.
No edits needed.
Reviewer 1
I also agree that plants (e.g., phytoplankton) are less sensitive
to Cd than animals, and thus it is appropriate to focus on the
animals.
Thank you for your comment.
No edits needed.
37
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 1
I think that the criteria that the author generated for dissolved
Cd have taken into consideration many of the key issues
influencing this (e.g., water hardness) are probably ok, but by
missing the effects of dietary Cd, the report is missing a large
part of the overall story. This is not to suggest that ambient Cd
concentrations are unsafe for animals, but the derived criteria
are probably over-estimates of the safe levels of Cd.
Text has been added to discuss dietary exposure and
incorporate the work of other researchers, such as Mebane
(2006), who discuss cadmium bioaccumulation and dietary
exposure in detail and the uncertainties currently associated
with its evaluation. In particular, text was added to detail how
cadmium can bioaccumulate in aquatic organisms through
multiple exposure routes including ingestion and direct
exposure, and how total uptake depends on the cadmium
concentration, exposure route and the duration of exposure.
Text was also added to clarify that there does not appear to be
a consistent relationship between body burden and
toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
Section 5.6.1
Reviewer 1
Another complicating issue is the influence of dissolved
organic carbon and its effect on Cd bioavailability. Thus,
expressing Cd toxicity as a function of body burden is
appropriate; the caveats associated with this approach have
been appropriately discussed in the report.
EPA revised the text to indicate that in addition to hardness,
which is a critical factor in determining toxicity, other water
quality parameters such as DOC, alkalinity, and pH may also
influence cadmium toxicity. EPA notes that integrating DOC
into the analysis would be expected in most cases to make the
criteria less stringent. Thus, while recognizing the need to
consider the applicability of the BLM in future cadmium
criteria updates, which would incorporate DOC, the inclusion
of DOC in a BLM would be unlikely to make the criteria more
stringent or conservative.
Section 2.3.1
Section 2.5
Section 5.1.3
38
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 2
Salmonids are clearly one of the more sensitive species with
regard to Cd toxicity. Not only are they very sensitive, they
are commercially important, and possess several species that
are listed as endangered and threatened in the US. The
proposed criteria are appropriate for freshwater conditions
since many of the studies used to derive the criteria focused on
freshwater treatments to rainbow trout. However, only one
study evaluated Cd toxicity in coho salmon smolts in saltwater
conditions, and this was at nearly full seawater strength (28
ppth). Of concern is the fact that many salmonids including
strains of 0. mykiss (steelhead) are anadromous and often
come in contact with Cd at lower salinities (5-15 ppth). While
the agency should be applauded for normalizing toxicity to
hardness to improve freshwater criteria, there is a critical need
to understand the impacts of salinity on Cd toxicity
particularly in anadromous salmonid species. Of additional
concern is the lack of discussion of sublethal impacts of Cd
particularly to olfaction (Williams and Gallagher 2013) which
significantly alters return rates of salmon (Baldwin et al.
2009). Return metrics are population level endpoints that
should supersede standard repro/survival/growth. These
should also be topics of discussion with regard to uncertainty.
Additional acceptable estuarine/marine toxicity data were not
available for salmonids. Additional text (and references) was
added to the appropriate uncertainty sections to emphasize the
absence of these data.
Section 5.4.1
Section 5.5.2
Reviewer 2
Lastly, the issue of climate change is largely missing from the
document. Acidification (particularly with metal availability)
and temperature issues are also likely to impact sensitive
species (e.g. salmonids). Sea level rise will also cause
saltwater intrusion into salmonid spawning habitats and affect
"estuarine/marine" criteria. Evaluation of these stressors
should be focal points for future criteria assessment
particularly for salmonids. Overall, while the values for
freshwater are likely safe for salmonids, the values for
estuarine/marine are highly uncertain and deserve further
evaluation.
Thank you for your comment. EPA revises the criteria
documents based on the best available scientific information at
the time of development and based on current conditions in
the environment. Criteria documents are then periodically
revised to incorporate the latest scientific information based
on toxicity and consideration of applicable environmental
conditions.
No edits
39
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 3
This seems to be the case if we assume that only aqueous
exposures matter. Evidence for dietary toxicity is less
compelling than for invertebrates, so for these fish species, the
criteria are likely more protective for these species than they
are for invertebrates.
EPA concurs with the reviewer about the difficulty of
characterizing dietary exposure and establishing critical tissue
effects thresholds for bioaccumulated metals.
Text has been added to discuss dietary exposure and
incorporate the work of other researchers, such as Mebane
(2006), who discuss dietary exposure and cadmium
bioaccumulation in detail and the uncertainties currently
associated with its evaluation. In particular, text was added to
detail how cadmium can bioaccumulate in aquatic organisms
through multiple exposure routes including ingestion and
direct exposure, and how total uptake depends on the
cadmium concentration, exposure route and the duration of
exposure. Text was also added to clarify that there does not
appear to be a consistent relationship between body burden
and toxicological effect, and an acceptable tissue effect
concentration cannot be defined for aquatic life at this time.
Bioaccumulation and effect level data that are available
indicate that cadmium is unlikely to accumulate in tissue to
levels that would result in adverse effects to aquatic
invertebrates or fish at the calculated chronic criterion
concentrations. For this reason, the evaluation of direct
exposure effects to organisms via water is considered
applicable to the development of criteria that is protective of
aquatic life.
Section 5.6.1
40
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
Yes, I think the criteria as derived will be protective of
salmonids. However, I'm concerned about the exclusion of the
fountain darter data from the derivation. EPA argues that the
acute data should be excluded because the test was fed and
that the chronic data should be excluded because the study
was only 7 d in duration (i.e., not true chronic). Generally, I
agree with both of these decisions, but from my perspective,
these rules are in place to prevent the inclusion of data
indicating organisms are insensitive due to inappropriate test
conditions (i.e., food reducing metal bioavailability, short test
durations missing sensitive endpoints). However, this is not
the case with the darter data, which indicate this species is
very sensitive despite test conditions that would tend to reduce
their sensitivity. EPA also seems to infer (p. 86) that the
fountain darter data has limited applicability because this
species has a limited distribution. However, the genus
Etheostoma is widespread throughout central and eastern U.S.
with a number of listed species at both the state and federal
level. Hence these data a representative for a genus that is
under considerable threat. Given this, I think it would be
important to assess how inclusion of these data would impact
derivation of the freshwater Cd WQC.
Text has been added to clarify that data eliminated were not
used in criteria derivation because the test organisms were fed
and the duration was too long for an acute test and too short
for a true ELS test. EPA also added text indicating the genus
Etheostoma is widespread, with some of species representing
those of special concern. It is important that states evaluate the
potential occurrence of these species when establishing site-
specific standards.
Section 5.8.1
Section 5.8.2
41
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 4
Page 87:1 don't think the statement that dividing the LC50 bv
two is expected to result in a concentration with effects no
different than the control is correct. Dividing the LC50 by two
will result in an "LC-low". I agree that across a range of
species and toxicants, dividing by two equates to a values that
approximates the NOEC. However, it does not equate to an
LCO, which is inferred by this statement. Please clarify.
Dividing the FAV by a factor of two to derive a CMC is the
standard approach used by the Agency to derive its 304(a)
acute criterion recommendations, consistent with the 1985
Guidelines. The FAV is a statistical estimate of the 5th
percentile of a set of LC50s. The LC50 is defined as the
concentration that kills 50% of the exposed organisms. Thus,
by definition, the FAV, as defined in the 1985 Guidelines, is a
concentration that would be lethal to 50% of organisms with a
sensitivity greater than 95% of genera. Since the FAV is a
concentration that may affect 50 percent of the 5th percentile
or 50 percent of a sensitive species, this value cannot be
considered to be protective of that percentile or that species.
Therefore, per the 1985 Guidelines, to derive the CMC EPA
divides the FAV by a factor of 2 with the intention of defining
a concentration that will not affect the majority of organisms.
The rationale for adjusting the FAV to derive the CMC is
explained in item 6 on page 17 of the 1985 Guidelines. The
basis for this adjustment factor is an analysis of data from 219
acute toxicity tests showing that the mean concentration lethal
to 0-10% of the test population was 0.44 times the LC50 or the
LC50 divided by 2.27. The data and analysis on which the
2.27 value is based is described in the Federal Register on
May 18, 1978 (43 FR 21506-21518). Best professional
judgment was used to round the FAV "adjustment factor" of
2.27 to 2 in revisions of the Guidelines that occurred
subsequent to the 1987 Federal Register notice. The use of the
factor became final EPA guidance in the 1985 Guidelines.
Section 5.1.3
42
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
Not necessarily, although to definitively answer this would
take a considerably more thorough review to determine than
was presented in the document, or could be done
independently in the time available. I note that NMFS (2012)
in Oregon concluded the 2001 CMC of 2.0 (ig/L could
jeopardize some salmonids and that the CCC of 0.25 (ig/L
would not jeopardize listed salmonids under their prevue.
Thus the draft 2015 criterion of 2.2 (ig/L would presumably be
a concern. Conversely, NMFS (2011) concurred with EPA
that Idaho acute and chronic criteria of 1.34 and 0.55 (ig/L
respectively would not jeopardize listed anadromous
salmonids. I did not attempt to reconcile the three documents.
However, I think part of the discrepancies may be in the
manner of analyses. In the draft document, data from long-
term exposures to salmonids that began with sensitive fry life
stage are excluded in favor of data from tests that began with
eggs or alevins. While all fish have some life stage-sensitivity
interaction, with at least salmonids sensitivity increases with
size up to at least 0.4g ww, and maybe up to lg or more
(Hansen et al. 2002; Mebane et al. 2012). With other fish, the
newly hatched stage may be more sensitive, or life events such
as the onset of exogenous feeding may be related to a stressful
and sensitive stage (Wang et al. 2014).
Use of the life cycle (LC) tests over the early life stage (ELS)
tests is consistent with the 1985 Guidelines. Furthermore,
there is no consistent pattern of early life stage tests being
more sensitive than life cycle tests for salmonids However, to
account for this discrepancy, ELS tests were used to calculate
revised SMCV in instances where they were more sensitive
than the LC tests (e.g., Salmo trutta).
Section 3.1.2
Section 5.2.2
43
-------
Rc\ icwcr
K\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
There are some instances of inappropriate averaging using
resistant life stages. Bull trout at the most sensitive (~lg) were
averaged with results of test with yearling brook trout to
produce a nonsense genus mean acute value of 126 (ig/L.
Stephan et al. advise against pooling species mean values
when they differ by more than a factor of 10; these differed by
a factor of 1000X.
This issue was re-examined ion depth based on the peer
reviewer's comments. The SMAVs for bull trout and brook
trout do differ by more than a factor of 10 (factor of 708X),
most likely due the different sensitivities of the fish used to
initiate the tests. The freshwater and estuarine/marine acute
databases also include several genera where two or more
widely different SMAVs (>10x factor) are available for
estimating the GMAV. In this case the 1985 Guidelines
recommend that some or all of the values probably should not
be used in calculations. To resolve this issue, only the more
sensitive SMAV (primarily due to a more sensitive life stage
tested) was used to calculate the GMAV, thereby ensuring
protection of the genus. It is important to note that the FAV
can be lowered to protect the most sensitive SMAV for a
commercially or recreationally important species to be
conservative. This was the case for the acute freshwater value
for both the 2001 AWQC and the current 2015 draft criteria
update.
Table 7
Table 10
Section 5.1.3
Section 5.4.1
44
-------
Rc\ icwcr
L\(crn;il Peer Rc\ icwcr Comments kc«i:irclin«i C'h«ir»c
Question 4
KIW Response
Rc\ ision
Locution in 2015
Drsil'l C iulmium
Crilcrisi
Document
Reviewer 5
The draft document evaluates protection of listed species by
rolling up species data to a hardness-normalized species mean
acute value (SMAV) and comparing that with the criteria.
Because the accuracy of hardness-normalization is uncertain,
but the criteria values can be calculated with certainty for any
hardness, a more informative way to evaluate the data with
listed species is to compare the criteria values for the
conditions of each test of interest with listed species to the
effects magnitude of effects to listed species at a given criteria.
If the test concentrations causing an adverse effect are close to
criteria concentrations, such as if the EC50s were within a
factor of 2 (or maybe 2.5 to 3 to be on the safe side), then
evaluate the actual adverse effects observed at the criteria
concentrations. The SMAV approach involves a lot of data
manipulation and may lose sensitive life stages or strains.
The criteria document provides the available toxicity data for
listed species. A separate document is in development that
addresses the detailed analyses of protection of Listed
salmonid species. Additionally, states and tribes have the
opportunity to use the toxicity data provided in this document,
as appropriate, to their address their specific situation.
Section 5.8
45
-------
2.5 Other Comments Provided
Re\ iewer
Com mciils
KIW Response
Re\ ision
Locution
Reviewer 4
Additional Minor Comments
Page 11: Note that Cd does not form complexes with Ca as
stated, but rather competes with Ca for uptake and Ca
channels. Please correct.
Text was revised as suggested.
Section 2.3
Reviewer 4
Page 11: While Atli and Canli did observe a reduction in NKA
activity in their study, it's a significant overstatement to say
disruption of Na homeostasis is a mechanism of action for Cd.
To the best of my knowledge, it hasn't been observed in any
other study that has investigated this potential mechanism.
Text was revised as suggested.
Section 2.3
Reviewer 4
Page 11: If Cd inhibits catalase, glutathione reductase, SOD,
etc., it seems to me this is direct inhibition of anti-oxidant
processes, not indirect as stated.
Text was revised as suggested.
Section 2.3
Reviewer 4
Page 12: Regarding the relationship between Cd tissue
burdens and toxicity, see also the analysis by Adams et al.
(2011).
EPA recognizes the difficulty in characterizing dietary
exposure and establishing critical tissue effects thresholds
associated with bioaccumulated metals (the identified paper
has been reviewed and text has been added to the document).
Section 5.6.1
Reviewer 4
Page 50: Tigriopus is a copepod, not a mysid, as indicated in
the second paragraph.
Text was revised as suggested.
Section 3.2.1
Reviewer 4
Page 58: Please specific at the top of p. 58 which two
freshwater ACRs were used in the calculation of the marine
ACR.
Text was revised to be clearer in the selection of ACRs used to
calculate the FACR.
Section 3.2.2
Section 5.5.1
Reviewer 4
Table 18: Change the test duration for the Borgmann studies
to 42 d rather than 6 w to make the units consistent with the
rest of the table.
Text was revised as suggested.
Table 18
Reviewer 4
Page 83: It should be mentioned that both BCFs and BAFs are
inversely related to exposure concentration which explains
much of the variation in BCFs/BAFs (McGeer et al. 2003,
DeForest et al. 2007).
Text was revised as suggested.
Section 5.6
46
-------
Re\ iewer
Com mciils
KIW Response
Re\ ision
Locution
Reviewer 4
Table 21: Taking a final look through Table 211 note that
EPA has included several species that are not resident to N.
America (Oreochromis spp., Danio rerio, Xenopus laevis).
Unless this requirement has changed, they should be removed
from the data set.
Naturally/wild reproducing North American species
populations are considered for inclusion in the document.
Each has been verified as such. Please see the following links
for the species mentioned:
http://nas.er.usgs.gov/queries/factsheet.aspx?SpeciesID=67
http://nas.er.usgs.gov/queries/FactSheet.aspx?speciesID=505
http://nas.er.usgs.gov/queries/FactSheet.aspx?speciesID=468
http://nas.er.uses.eov/aueries/FactSheet.aspx?speciesID=466
No edits
Reviewer 5
Unfortunately, the compressed time period for this review (2
weeks, which works out to several hours on evenings and
weekends for volunteer reviewers), makes a comprehensive
review of a document of this length and complexity infeasible.
Thank you for your comment. EPA appreciates the comments
that were provided during the time available for review.
No edits
47
-------
3 References Cited by Reviewers
Adams, W.J., R. Bluet, K. V. Brix, D. K. DeForest, A. S. Green, J. S. Meyer, J. C. McGeer, P. R. Paquin, P.
S. Rainbow and C. M. Wood (2011). "Utility of tissue residues for predicting effects of metals on aquatic
organisms." Integr. Environ. Assess. Manag. 7(1): 75-98.
ASTM. 1998. Standard guide for conducting early life-stage toxicity tests with fishes. Method E1241-98.
Pages 29 in Annual Book of ASTM Standards, volume 11.04. American Society for Testing and Materials,
West Conshohocken, PA.
Baldwin, DH, Spromberg, JA, Collier, TK and NL Scholz. 2009. A fish of many scales: extrapolating
sublethal pesticide exposures to the productivity of wild salmon populations. Ecological Applications
19:2004-2015.
Barata, C., D. J. Baird, S. E. Mitchell, and A. M. V. M. Soares. 2002. Among- and within-population
variability in tolerance to cadmium stress in natural populations of Daphnia magna: implications for
ecological risk assessment. Environ. Toxicol. Chem. 21:1058-1064.
Barata, C., S. J. Markich, D. J. Baird, and A. M. V. M. Soares. 2002. The relative importance of water and
food as cadmium sources to Daphnia magna Straus. Aquat. Toxicol. 61:143-154.
Besser, JM, Mebane, CA, Mount, DR, Ivey, CD, Kunz, JL, Greer, EI, May, TW, and Ingersoll, CG. 2007.
Relative sensitivity of mottled sculpins (Cottus bairdi) and rainbow trout (Oncorhynchus mykiss) to toxicity
of metals associated with mining activities. Environ Toxicol Chem 26:1657-1665.
Buchwalter, D. B., D. J. Cain, C. A. Martin, L. Xie, S. N. Luoma, and T. Garland, Jr. 2008. Aquatic insect
ecophysiological traits reveal phylogenetically based differences in dissolved cadmium susceptibility. Proc.
Nat. Acad. Sci. 105:8321-8326.
Buchwalter, D. B., D. J. Cain, W. H. Clements, and S. N. Luoma. 2007. Using biodynamic models to
reconcile differences between laboratory toxicity tests and field biomonitoring with aquatic insects. Environ.
Sci. Technol. 41:4821-4828.
Cain, D. J., S. N. Luoma, and W. G. Wallace. 2004. Linking metal bioaccumulation of aquatic insects to their
distribution patterns in a mining-impacted river. Environ. Toxicol. Chem. 23:1463-1473.
Calfee, RD, Little, EE, Puglis, HJ, Beahan, E, Brumbaugh, WG, and Mebane, CA. 2014. Acute Sensitivity of
White Sturgeon (Acipenser transmontanus) and Rainbow Trout (Oncorhynchus mykiss) to Cadmium, Copper,
or Zinc in Laboratory Water-Only Exposure. Pages 5-34 in Ingersoll, CG, and Mebane, CA, editors. Acute
and chronic sensitivity of white sturgeon (Acipenser transmontanus) and rainbow trout (Oncorhynchus
mykiss) to cadmium, copper, lead, or zinc in laboratory water-only exposures. U.S. Geological Survey
Scientific Investigations Report 2013-5204.
Chapman, GA. 1985. Acclimation as a factor influencing metal criteria. Pages 119-136 in Bahner, RC, and
Hansen, DJ, editors. Aquatic Toxicology and Hazard Assessment: Eighth Symposium (STP 891-EB), volume
STP 891. American Society for Testing and Materials (ASTM), Philadelphia.
Cooper, WE. 1965. Dynamics and production of a natural population of a fresh-water amphipod, Hyalella
azteca. Ecol Monogr 35:377-394.
48
-------
Croteau, M.-N., L. Hare, and A. Tessier. 2003. Difficulties in relating Cd concentrations in the predatory
insect Chaoborus to those of its prey in nature. Can. J. Fish. Aquat. Sci. 60:800-808.
DeForest, D. K., K. V. Brix and W. J. Adams. 2007. "Assessing metal bioaccumulation in aquatic
environments: the inverse relationship between bioaccumulation factors, trophic transfer factors and exposure
concentration." Aquat. Toxicol. 84: 236-246.
France, RL. 1996. Biomass and production of amphipods in low alkalinity lakes affected by acid
precipitation. Environ Pollut 94:189-193.
Gibbons, WN, and Mackie, GL. 1991. The relationship between environmental variables and demographic
patterns of Hyalella azteca (Crustacea: Amphipoda). J N Am Benthol Soc 10:444-454.
Hansen, JA, Welsh, PG, Lipton, J, Cacela, D, and Dailey, AD. 2002. Relative sensitivity of bull trout
(Salvelinus confluentus) and rainbow trout (Oncorhynchus mykiss) to acute exposures of cadmium and zinc.
Environ Toxicol Chem 21:67-75.
Hare, L., A. Terrier, and L. Warren. 2001. Cadmium accumulation by invertebrates living at the sediment-
water interface. Environ. Toxicol. Chem. 20:880-889.
Hare, L., A. Tessier, and U. Borgmann. 2003. Metal sources for freshwater invertebrates: pertinence for risk
assessment. Human Ecol. Risk Assessment 9:779-793.
Irving, E. C., D. J. Baird, and J. M. Culp. 2003. Ecotoxicological responses of the mayfly Baetis tricaudatus
to dietary and waterborne cadmium: implications for toxicity testing. Environ. Toxicol. Chem. 22:1058-1064.
Klaassen, C. D., J. Liu, and S. Choudhuri. 1999. Metallothionein: An intracellular protein to protect against
cadmium toxicity. Annu. Rev. Pharmacol. Toxicol. 39:267-294.
Luoma, S. N. and J. L. Carter. 1991. Effects of trace metals on aquatic benthos, pp. 261-300 In M. C.
Newman and A. W. Mcintosh (eds.), Metal Ecotoxicology Concepts & Applications. Lewis Publishers, Inc.,
Chelsea, Michigan.
Luoma, S. N. and P. S. Rainbow. 2005. Why is metal bioaccumulation so variable? Biodynamics as a
unifying concept. Environ. Sci. Technol. 39:1921-1931.
Luoma, S. N., D. J. Cain, and P. S. Rainbow. 2009. Calibrating Biomonitors to Ecological Disturbance: a
New Technique for Explaining Metal Effects in Natural Waters. Integrated Environmental Assessment and
Management 6:199-209.
Martin, C. A., D. J. Cain, S. N. Luoma, and D. B. Buchwalter. 2007. Cadmium ecophysiology in seven
stonefly (Plecoptera) species: Delineating sources and susceptibility. Environ. Sci. Technol. 41:7171-7177.
McGeer, J.C., K.V. Brix, J.M. Skeaff, D.K. DeForest, S.I. Brigham, W.J. Adams and A.S. Green. 2003. "The
inverse relationship between bioconcentration factor and exposure concentration for metals: implications for
hazard assessment of metals in the aquatic environment." Environ. Toxicol. Chem. 22(5): 1017-1037.
McNulty, EW, Dwyer, FJ, Ellersieck, MR, Greer, EI, Ingersoll, CG, and Rabeni, CF. 1999. Evaluation of
ability of reference toxicity tests to identify stress in laboratory populations of the amphipod Hyalella azteca.
Environ Toxicol Chem 18:544-548.
49
-------
Mebane, CA, Dillon, FS, and Hennessy, DP. 2012. Acute toxicity of cadmium, lead, zinc, and their mixtures
to stream-resident fish and invertebrates. Environ Toxicol Chem 31:1334-1348.
Mebane, CA. 2010. Relevance of risk predictions derived from a chronic species-sensitivity distribution with
cadmium to aquatic populations and ecosystems. Risk Anal 30:203-223.
Poteat, M. D. and D. B. Buchwalter. 2014. Calcium uptake in aquatic insects: influences of phylogeny and
metals (Cd and Zn). J. Exp. Biol. 217:1180-1186.
Poteat, M. D. and D. B. Buchwalter. 2014. Four reasons why traditional metal toxicity testing with aquatic
insects is irrelevant. Environ. Sci. Technol. 48:887-888.
Poteat, M. D. and D. B. Buchwalter. 2014. Phylogeny and Size Differentially Influence Dissolved Cd and Zn
Bioaccumulation Parameters among Closely Related Aquatic Insects. Environ. Sci. Technol. 48:5274-5281.
Poteat, M. D., M. Diaz-Jaramillo, and D. B. Buchwalter. 2012. Divalent metal (Ca, Cd, Mn, Zn) uptake and
interactions in the aquatic insect Hydropsyche sparna. J. Exp. Biol. 215:1575-1583.
Stephan, C. E., D. I. Mount, D. J. Hansen, J. H. Gentile, G. A. Chapman and W. A. Brungs. 1985. Guidelines
for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses.
Duluth, U.S. Environmental Protection Agency, Environmental Research Laboratory: 98 pp.
Stephan, CE, Mount, DI, Hansen, DJ, Gentile, JH, Chapman, GA, and Brungs, WA. 1985. Guidelines for
deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. U.S.
Environmental Protection Agency, EPA 822-R-85-100, NTIS PB85 227049, Duluth, Narragansett, and
Corvallis.
Strong, DR, Jr. 1972. Life history variation among populations of an amphipod (Hyalella azteca). Ecology
53:1103-1111.
Szczerbik, P, Mikolajczyk, T, Sokolowska-Mikolajczyk, M, Socha, M, Chyb, J and P. Epler. 2006. Influence
of long-term exposure to dietary cadmium on growth, maturation and reproduction of goldfish (subspecies:
Prussian carp Carassius auratus gibelio B.) Aquatic Toxicology 77:126-135.
Thomann, RV, Shkreli, F. and S. Harrison. 1997. A pharmacokinetic model of cadmium in rainbow trout.
Environmental Toxicology and Chemistry 16:2268-2274.
Timmermans, K. R., E. Spijkerman, and M. Tonkes. 1992. Cadmium and zinc uptake by two species of
aquatic invertebrate predator from dietary and aqueous sources. Can. J. Fish. Aquat. Sci. 49:655-662.
Vetillard, A. and T. Bailhache. 2005. Cadmium: An endocrine disrupter that affects gene expression in the
liver and brain of juvenile rainbow trout. Biology of Reproduction 72:119-126.
Wallace, W. G., B.-G. Lee, and S. N. Luoma. 2003. The subcellular compartmentalization of Cd and Zn in
two bivalves. I. The significance of metal-sensitive fractions (MSF) and biologically-detoxified metal (BDM).
Mar. Ecol. Prog. Ser. 249:183-197.
50
-------
Wang, N, Ingersoll, CG, Dorman, RA, Kunz, JL, Hardesty, DK, Brumbaugh, WG, and Mebane, CA. 2014.
Chronic Sensitivity of White Sturgeon (Acipenser trans montanus) and Rainbow Trout (Oncorhynchus
mykiss) to Cadmium, Copper, or Zinc in Laboratory Water-Only Exposure. Pages 35-76 in Ingersoll, CG, and
Mebane, CA, editors. Acute and chronic sensitivity of white sturgeon (Acipenser transmontanus) and rainbow
trout (Oncorhynchus mykiss) to cadmium, copper, lead, or zinc in laboratory water-only exposures. U.S.
Geological Survey Scientific Investigations Report 2013-5204.
Wanty, RB, Verplanck, PL, San Juan, CA, Church, SE, Schmidt, TS, Fey, DL, DeWitt, EH, and Klein, TL.
2009. Geochemistry of surface water in alpine catchments in central Colorado, USA: Resolving host-rock
effects at different spatial scales. Appl Geochem 24: 600-610.
Wendelaar Bonga, SE, and Lock, RAC. 2008. The osmoregulatory system. Pages 401-416 in Di Giulio, RT,
and Hinton, DE, editors. Toxicology of Fishes. CRC Press, Boca Raton, Florida.
Williams CR and EP Gallagher. 2013. Effects of cadmium on olfactory mediated behaviors and molecular
biomarkers in coho salmon (Oncorhynchus kisutch) Aquatic Toxicology 140-141:295-302.
Xie, L. and D. B. Buchwalter. 2011. Cadmium exposure route affects antioxidant responses in the mayfly
Centroptilum triangulifer. Aquat. Toxicol. 105:199-205.
Xie, L. T., D. Lambert, C. Martin, D. J. Cain, S. N. Luoma, and D. Buchwalter. 2008. Cadmium biodynamics
in the oligochaete Lumbriculus variegatus and its implications for trophic transfer. Aquat. Toxicol. 86:265-
271.
Xie, L., D. H. Funk, and D. B. Buchwalter. 2010. Trophic transfer of Cd from natural periphyton biofilms to
the grazing mayfly Centroptilum triangulifer in a life cycle test. Environmental Pollution. 158:272-277.
51
------- |