BEFORE THE ILLINOIS POLLUTION CONTROL BOA

IN THE MATTER OF :

)

REVISIONS TO WATER QUALITY

)

STANDARDS FOR TOTAL DISSOLVED

)

SOLIDS IN THE LOWER DES PLAINES RIVER FOR)

EXXONMOBIL OIL CORPORATION

:

)

PROPOSED 35 Ill. Adm. Code 303.445

)

Dorothy M. Gunn, Clerk

Anand Rao

Illinois Pollution Control Board

James R . Thompson Center

100 West Randolph Street, Suite . 11-500

Chicago, Illinois 60601

William Richardson, Chief Legal Counsel

Illinois Department of Natural Resources

One Natural Resource Way

Springfield, IL 62702

Matthew J. Dunn

Division Chief, Environmental Enforcement

Illinois Attorney General

100 W. Randolph Street, 12t

h

Floor

Chicago, IL 60601

NOTICE OF FILING

Jeffrey C. Fort

Letissa Carver Reid

Elizabeth A . Leifel

Sonnenschein, Nath and Rosenthal LLP

7800 Sears Tower

233 South Wacker Drive

Chicago, Illinois 60606-6404

Susan M. Franzetti

Franzetti Law Firm, P .C .

10 S. LaSalle Street, Suite 3600

Chicago, IL 60603

Dennis L. Duffield

Director of Public Works and Utilities

City of Joliet

Department of Public Works and Utilities

921 E. Washington Street

Joliet, Illinois 60603

_P7RVED

i:0K`A

.

QF};

CE

MJNNt

0 2 20co

§TAT€

QF ILLINOIS

R06 - 24

1666h

vontrol anarc!

(Site Specific Rule

- Water)

PLEASE TAKE NOTICE

that I have today filed with the Office of the Clerk of the

Pollution Control Board the attached

PRE-FILED TESTIMONY AND RESPONSE TO

BOARD REQUEST

on behalf of the Illinois Environmental Protection Agency, a copy of which is

herewith served upon you .

ENVIRONMENTAL PROTECTION AGENCY

OF THE STATE OF ILLINOIS

By: ~ yl

Thomas Andryk

Assistant Counsel

Division of Legal Counsel

DATED: May 31, 2006

Illinois Environmental Protection Agency

1021 North Grand Avenue East

Post Office Box 19276

Springfield, Illinois 62794-9276

(217) 782-5544

THIS FILING IS SUBMITTED ON RECYCLED PAPER

---

SOLIDS IN THE LOWER DES PLAINES RIVER

)

FOR EXXONMOBIL OIL CORPORATION: )

PROPOSED 35 Ill. Adm. Code 303.445

)

(Site Specific Rule - Water)

Pre-Filed Testimony and Response to Board Request

The undersigned, as one of its attorneys, hereby provides the following information as to Illinois

EPA's Pre-Filed Testimony and its Response to the Board's Request that it address the

applicability or inapplicability of 35 Ill. Adm. Code Section 102 .210(c) prior to hearing, stating

on behalf of Respondent, Illinois Environmental Protection Agency, as follows

:

I .

TESTIMONY OF SCOTT TWAIT

My name is Scott Twait and I have been employed by Illinois EPA for over 9 years . I have been

assigned to the Water Quality Standards Unit for all of those years and have participated in

adjusted standards, site-specific water quality standards rulemakings, and variances. I hold a B.S

degree in Civil Engineering from the University of Illinois where I specialized in Environmental

Engineering .

My testimony today will be in support of the ExxonMobil Oil Corporation site-specific relief

from the total dissolved solids (TDS) Secondary Contact and Indigenous Aquatic Life standard

(35 IAC 302.407) and the TDS General Use standard (35 IAC 302 .208(g)) in the Des Plaines

River .

The petitioner is adding a Catalytic SO2 Additive Technology (DESOX) system followed by a

wet gas scrubber (WGS) and a Selective Catalytic Reduction (SCR) system to remove SO2 and

NO, from air emissions as part of a consent decree with USEPA and Illinois EPA . The addition

of the DESOX will allow the removal of SO 2 from the emissions by transferring sulfur, in stable

form, from the regenerator to the reactor, where it is released as hydrogen sulfide for

downstream recovery as elemental sulfur, thereby reducing sulfate in the plant wastewater and

minimizing dissolved solids discharged to the Des Plaines River . The DESOX, WGS, and SCR

will remove 95% of SO2 and 50% NO, at 130,000 and 9,800 pounds per day respectively. As

indicated in our November 15, 2005 meeting, ExxonMobil is adding a third tank onto the

activated sludge WWTP and will configure the process to provide an anoxic zone to denitrify,

THIS FILING IS SUBMITTED ON RECYCLED PAPER

ILLINOIS POLLUTION CONTROL BOARD

Rte-

CLE

RK',

C-

pVe

Pp~tioE

N

n

C

oneal

IN THE MATTER OF

:

)

REVISIONS TO WATER QUALITY

)

STANDARDS FOR TOTAL DISSOLVED

)

R06 - 24

---

therefore, total nitrogen loading to the stream will be reduced rather than increased as a result of

the air scrubbing. Loading of sulfates and TDS will be increased to the receiving stream ;

however, sulfates will meet water quality standards after mixing . TDS will not always meet the

water quality standards, due to seasonal loading of chlorides found in road salt from the Chicago

metropolitan area that has affected concentrations upstream of ExxonMobil

.

The subject facility discharges to the Des Plaines River at a point where 1503 .0 cfs of flow exists

upstream of the outfall during critical 7Q10 low-flow conditions. The Des Plaines River is

classified as a Secondary Contact and Indigenous Aquatic Life Use Water at the point of

discharge and is a General Use Water downstream of the 1-55 bridge. The Des Plaines River is

rated a "C" stream under the Agency's Biological Stream Characterization (BSC) program. The

Des Plaines River, Waterbody Segment, G-24, is found on the 2004 Illinois 303(d) List . The

uses impaired for this segment was aquatic life and fish consumption . The potential causes of

impairment given for the segment at that time were copper, sedimentation/siltation, other flow

regime alterations, total suspended solids (TSS), DDT (statistical guideline), PCBs, (statistical

guideline), mercury (statistical guideline), and total phosphorus (statistical guideline) . The

potential sources associated with the impairment are industrial point sources, municipal point

sources, urban runoff/storm sewers, hydrologic/habitat modification, flow

regulation/modification, contaminated sediments, and source unknown . The additional

constituents to be discharged by ExxonModbil, sulfate and TDS, therefore have no bearing on

the 303(d) status of the waterbody

.

The Illinois Department of Natural Resources was contacted on November 17, 2005 with regard

to the presence of any threatened or endangered species that may be impacted by this standard

change. IDNR terminated the consultation process on December 19, 2005 with a finding of no

threatened and endangered species or natural areas affected

.

The Agency cannot grant mixing for a discharge if the receiving stream is not meeting the water

quality standard. Since the necessary NPDES permit would require the recognition of mixing in

the Des Plaines River and the Des Plaines River has occasionally violated water quality

standards for TDS, the Agency cannot issue an NPDES permit that will accommodate this new

ExxonMobil discharge. Mixing for sulfate is allowable, however, and will extend into the

General Use portion of the river

.

The petitioners have demonstrated that TDS is not toxic to aquatic life at the concentrations that

will be found in the river provided that sulfate is the predominant anion . Toxicity test results on

TDS with the chloride to sulfate ratio that will result from the proposed discharge indicate that

even the most sensitive species tested can easily tolerate the levels likely to be found in the

receiving waters

.

In the petition for the site-specific rulemaking, the petitioners discussed compliance alternatives

that were all rejected due to cost and/or technical infeasibility. We believe that the petitioners

have shown that there are no cost-effective compliance alternatives

.

THIS FILING IS SUBMITTED ON RECYCLED PAPER

---

The Agency is in the process of proposing to change the General Use water quality standard for

sulfates and eliminate the General Use standard for TDS but has not yet filed its petition before

the Board. New aquatic life toxicity data indicates the level of sulfate that sensitive species

tolerate. This information was not available when the original water quality standards were

adopted for sulfate and TDS . Our new understanding of sulfate toxicity can be coupled with the

existing chloride standard to predict a protective level of TDS . Given the hardness of 205 mg/L

as CaCO3 and the maximum chloride concentration of 450 mg/L known for the Des Plaines

River, the proposed water quality standard based on the aquatic life toxicity of sulfate is 1,138

mg/L. If we add up the major anions, we get 450 + 1,138 = 1,588 mg/L TDS . Adding in the

major cations, a TDS concentration of about 3,000 mg/L is protective . Therefore, it has been

demonstrated that the 1,686 mg/L TDS requested as relief by ExxonMobil is well within the

TDS toxicity threshold . The 1,686 mg/L TDS in the stream in this case consists of chloride and

sulfate, plus adding in the sodium, magnesium, calcium, and all the minor ions . This site-

specific rulemaking will not result in aquatic life toxicity . For the above conclusions we relied

on the attached studies

.

This site-specific rulemaking, consisting of a new calculation of the protective level of TDS, is

consistent with 40 CFR 131 .11(b)(1)(ii). Specifically, a federal site-specific water quality

criterion would be allowed in this case because sensitive species of aquatic life have been

demonstrated to be protected by the new standard through laboratory toxicity tests . USEPA

Region 5 has given preliminary approval of the ExxonMobil site-specific standard under its

obligation to review state water quality standards under the Clean Water Act

.

The Agency is currently reviewing the Secondary Contact and Indigenous Aquatic Life water

quality standards for the Lower Des Plaines River through the Use Attainability Analysis (UAA)

process. This site-specific rulemaking should remain in effect if the water quality standard for

TDS is not revised to at least 1,686 mg/L for the Lower Des Plaines River under the UAA

.

There are no other existing dischargers, in this stretch of the river, which have an elevated

discharge of TDS. The Channahon W WTF, BASF, ExxonMobil tank farm, Loder Cronklaan,

and Dow Chemical polystyrene plant are the Des Plaines River dischargers downstream of the

subject facility. Channahon is the only municipal discharge and the TDS expected in an STP

discharge would be expected to be 500 - 600 mg/L, so in effect, they are a diluter . The BASF

plant, visible from the 1-55 bridge, discharges process water and storm water, which are not

expected to have elevated TDS . There is also another ExxonMobil facility that is a tank farm

and/or a pipeline terminus . They have a boiler blowdown, but this would be minor in size and

not likely to have extremely high TDS wastewater . Loder Cronklaan makes vegetable oil

products and has no likelihood for high TDS wastewater . Finally, a Dow Chemical polystyrene

plant, which has cooling and sanitary wastewater has no potential for high TDS wastewaters of

significant size. None of these industries is categorized by the IEPA as a major discharger

.

None of these dischargers exhibited a need for water quality based effluent limits, past or

present. Regardless of the dischargers to this section of the Des Plaines River, the water quality

standard that is proposed is more stringent than what the Agency believes is protective of aquatic

life .

THIS FILING IS SUBMITTED ON RECYCLED PAPER

---

This site-specific rulemaking will not result in aquatic toxicity, there are no economically or

technically feasible alternatives, and is approvable by USEPA . I recommend that the IPCB

support the petitioners request for site-specific rulemaking for relief from the water quality

standards for TDS at 35 IAC 302.208(g) and 302.407 as written in the petition

.

II .

ILLINOIS EPA RESPONSE TO BOARD REQUEST

The Agency respectfully submits the following as its response to the Board's inquiry regarding

the applicability of 35 111 . Adm. Code 102.210. Section 102.210 (c) of the Board's regulations

requires the proponent of the rulemaking to provide a descriptive title or other description of any

published study or research report used in developing the rule, the identity of the person who

preformed the study and a description of where the public may obtain a copy of any such study

or research report. If the relied upon report or study was conducted by the agency, the agency

also has an obligation to make the underlying data available upon request

.

In the instant case, the Proponent of the rulemaking had the obligation to identify studies and

reports that it relied upon . The Agency notes that the proponent referenced on-going

investigation by the Agency and conclusions reached by the Agency from that investigation . To

its Petition for a Site-Specific Rule, Exxon Mobil attached a Jan . 9, 2004 report by Dr. Soucek

and citations to other reported toxicity research but alluded to Agency investigations . The

Agency is not the proponent of the rulemaking but in this instance, given the long history of

Agency efforts to investigate the TDS water quality standard and the need to expedite this

rulemaking, the Agency offers the following information to the Board

.

Pursuant to Section 303 (c), of the Clean Water Act, 33 U.S.C. § 1313 (c), the Agency has a

continuing duty to investigate and propose updates to its Water Quality Standards as science

becomes more refined or conditions change . The Agency has been investigating the impact of

TDS on aquatic life for several years. It convened a work group, and submitted a preliminary

draft justification documents for regulatory changes and proposed language for comment to the

work group. Although the Agency is not submitting the draft proposal for comment and

justification for regulatory change, entitled Draft Justification for Changing Water Quality

Standards for Su fate, Total Dissolved Solids and Mixing Zones (January 21, 2004)

as presented

to the work group, with this proposal for a site-specific rule, the Agency has made the draft

available to the public and believes that the proponent may have relied on the conclusions in the

justification document. This regulatory proposal is still in development; the Agency notes that

the preliminary draft justification is out-dated in some respects but the conclusions are valid

.

The Agency plans to submit a rulemaking proposal to the Board at some time in the future to the

Board after the conclusion of the regulatory development process

.

A further study by Dr. Soucek, finalized after the instant proposal, supported the Jan. 9, 2004

Soucek report. Entitled Effects of Water Quality on Acute and Chronic Toxicity of Sulfate to

Freshwater Bivalves, Ceriodaphnia dubia, and Hyalella azteca, it is available to the public at

the Agency's offices and is being submitted, with several quarterly reports, as an attachment to

Scott Twait's prefiled testimony . The Agency notes that this study was not relied upon by the

THIS FILING IS SUBMITTED ON RECYCLED PAPER

---

proponent as it was not completed at the time of the proposal . It, however, was relied upon and

considered by the Agency in reaching its conclusion to support the proposal .

The Agency supports the request for relief requested by Exxon-Mobil in its Petition for a Site-

Specific water quality standard and respectfully submits this response and additional information

in an effort to expedite this time sensitive rulemaking .

DATED: May 31, 2006

Illinois Environmental Protection Agency

1021 North Grand Avenue East

Post Office Box 19276

Springfield, Illinois 62794-9276

(217)782-5544

ILLINOIS ENVIRONMENTAL PROTECTION AGENCY

By :

~,

d'

G'

m~

Thomas Andryk

Assistant Counsel

Division of Legal Counsel

THIS FILING IS SUBMITTED ON RECYCLED PAPER

---

STATE OF ILLINOIS

)

5S

COUNTY OF SANGAMON

)

PROOF OF SERVICE

I, the undersigned, on oath state that I have served the attached PRE-FILED TESTIMONY AND

RESPONSE TO BOARD REQUEST and NOTICE OF FILING upon the person to whom it is directed, by

placing a copy in an envelope, with proper first class postage pre-paid, addressed to

:

Dorothy M. Gunn, Clerk

Jeffrey C . Fort

Anand Rao

Letissa Carver Reid

Illinois Pollution Control Board

Elizabeth A. Leifel

James R. Thompson Center

Sonnenschein, Nath and Rosenthal LLP

100 West Randolph Street, Suite 11-500

7800 Sears Tower

Chicago, Illinois 60601

233 South Wacker Drive

Chicago, Illinois 60606-6404

William Richardson, Chief Legal Counsel

Illinois Department of Natural Resources

Dennis L. Duffield

One Natural Resource Way

Director of Public Works and Utilities

Springfield, IL 62702

Department

ptJoliet

of Public Works and Utilities

921 E. Washington Street

Matthew J. Dunn

Joliet, Illinois 60603

Division Chief, Environmental Enforcement

Illinois Attorney General

100 W . Randolph Street, 12'" Floor

Susan M. Franzetti

Chicago, IL 60601

Franzetti Law Firm, P .C .

10 S. LaSalle Street, Suite 3600

Chicago, IL 60603

and mailing it from Springfield, Illinois on May 31, 2006 with sufficient postage affixed as indicated

above .

SUBSCRIBED AND SWORN TO BEFORE ME

thisc?/ day of

, 2006

71" 0) j

THIS FILING IS SUBMITTED ON RECYCLED PAPER

OFFICIAL SEAL

c

CYNTHIA L. WOLFE

NOTARY PUBLIC, STATE OF IWNOIS

MY COMMISSION EXPIRES 3.20

.2007

;[

S

S(hAA

•

~..y¢7,..}fjQ4S'OSSVSSS,\I,

.:. . : ...5{.7J

---

Effects of Water Quality on Acute and Chronic Toxicity of Sulfate

to Freshwater Bivalves,

Ceriodaphnia dubia,

and

Hyalella azteca .

First Quarterly Report

Submitted to :

Edward Hammer and Dertera Collins

United States Environmental Protection Agency

Region 5, Water Division, 77 West Jackson Boulevard

Chicago, Illinois 60604

December 21, 2004

Illinois Natural History Survey, Champaign, IL

U.S. EPA Region 5

INNS

Sept 1, 2004

Page 1 of 11

fulnoisEPAExhibit

No .

--

---

Background

While there are no Federal water quality criteria (WQC) for the protection of freshwater

life for total dissolved solids (TDS), sulfate, or sodium (U .S . EPA 1999), several states,

including Minnesota, Indiana, and Illinois, are at various stages in the process of

developing standards for sulfate . Water quality standards are developed to protect

designated uses, aquatic life uses in this case, but the economic impacts of these standards

are important considerations as well . For example, after existing sulfate standards were

enacted in Illinois, the Illinois Pollution Control Board adopted exceptions to the

standards to provide relief for a number of coalmines that were enduring severe economic

hardship. In fact, -60% of coalmines have expired permits in Illinois because of

violations of the sulfate standard, and -50% of those have been expired for more than

three years (T. Frevert, pers. comm.). The current "General Use" standard of 500 mg/L in

Illinois is based on the value thought to be protective of livestock . Consultation with

appropriate authorities revealed that livestock were capable of tolerating much higher

levels of sulfate. In light of these factors, the Illinois Environmental Protection Agency

(IL EPA) is actively pursuing an update of the sulfate standards based on scientific

research, and is close to proposing an updated standard (R . Mosher, IL EPA, pers .

comm.) .

Sodium is one of the most common major cations in high TDS effluents, but calcium and

chloride are usually present in mine-impacted waters as well

. While major ion or TDS

toxicity is caused by osmoregulatory stress from the combination of all cations and

anions, chloride standards currently exist, and Illinois plans to additionally regulate

for

sulfate in order to address the major non-chloride component of TDS in these waters

.

Therefore, studies (funded by IL EPA, and IL Coal Association) were conducted by

Soucek (2004; In Press Environmental Toxicology and Chemistry) to (1) generate LC50s

(lethal concentration to 50% of a sample population) and LCIOs (lethal concentration to

10% of a sample population) for sulfate with selected freshwater invertebrates

(Ceriodaphnia dubia, Chironomus tentans, Hyalella azteca,

and Sphaerium simile) in

U.S. Environmental Protection Agency's (US EPA, 1993) moderately hard reconstituted

water (MHRW) and (2) determine the effects of laboratory water composition, water

hardness, and test organism acclimation on the acute toxicity of sulfate to

Ceriodaphnia

dubia and Hyalella azteca (Soucek, 2004). In these previous studies (Soucek, 2004), the

mean LC50s, expressed as mg S04 27L, in moderately hard, reconstituted water (MHRW ;

U.S. EPA, 1993) ranged from: 512 to 1,4134 mg/L. The LC50 generated for the

amphipod,

Hyalella

(512 mg/L) was surprisingly low, given that it is known as a

euryhaline organism (Ingersoll et al ., 1992), but as will be discussed below, water quality

data, including other cations and anions present, are critical for predicting the responses

of freshwater organisms (especially

Hyalella)

to elevated sulfate concentrations .

INNS

September I, 2004

Page 2 of 12

---

INHS

September 1, 2004

Page 3 of 12

The composition of the dilution water used during testing in the Soucek (2004) study had

a dramatic effect on the toxicity of sulfate to Hyalella . Whereas the 96-hour LC50 in

MHRW was 512 mg/L, the LC50 increased to 2,855 mg/L when using a "Reformulated

Moderately Hard Reconstituted Water" (RMHRW, Smith et al ., 1997). The LC50 for C .

dubia also increased from 2,050 in MHRW to 2,526 mg/L in RMHRW. Both dilution

waters were similar in terms of hardness (-90-106 mg/L as CaCO3), alkalinity, and pH,

but RMHRW had a higher chloride concentration and different calcium to magnesium

ratio than that in MHRW. An additional experiment conducted at the Illinois Natural

History Survey (INHS) laboratory, but not included in the Soucek (2004) report, indicated

that when sulfate (-2,800 mg/L) and hardness (106 mg/L) were held constant, percent

survival ofH. azteca was positively correlated with chloride concentration (up to 67 mg

Cl -/L). These experiments illustrate the need to further characterize the interacting effects

of chloride and sulfate on aquatic organisms .

Another factor that appears to have a strong effect on the toxicity of sulfate is the

presence of other major cations, in this case, calcium and magnesium, measured as

hardness. In the previous study (Soucek 2004), increased hardness reduced the toxicity of

sulfate to Hyalella and had a dramatic effect on the 48-hour LC50 for C. dubia,

increasing from 2,050 at a hardness of 90 to > 2,900 mg/L at hardness values higher than

194 mg/L as CaCO 3 .

Others have observed reduced toxicity of saline solutions due to

increased hardness as well (e.g ., Dwyer et al. 1992; Mount et al. 1997) .

While a great deal of progress was made in the understanding of sulfate toxicity under

varying water quality conditions, several important data gaps remain. In the previous

studies (Soucek 2004), the fingernail clam, Sphaerium simile, had a lower LC 10 than that

of C. dubia, but because of the temporal nature of its availability, this bivalve was only

tested in MHRW . It remains unclear whether or not a mollusk will have the same

physiological response as two crustaceans to increased chloride or hardness in these

experiments with sulfate. The principal inorganic anion of crustacean blood, or

hemolymph, is chloride, and it has been suggested that low chloride concentrations may

limit the distribution of at least one euryhaline amphipod (Corophium curvispinum) in

freshwaters (Bayliss and Harris, 1988) . However, in the unionid mussel, Toxolasma

texasiensis, chloride and bicarbonate are equally important anions in the hemolymph (see

McMahon and Bogan 2001). Because bicarbonate is readily available via respiration and

metabolism, this mussel may not depend on external chloride concentrations to the extent

that some crustaceans do. If this is the case, the protective effect of chloride observed for

Hyalella and Ceriodaphnia might not be manifest in some unionoidean bivalves . The

hardness effect observed in the Soucek (2004) study may be more widespread among

aquatic phyla, because calcium simply reduces gill permeability (Lucu and Flik, 1999

; Pic

and Maetz, 1981). However, McMahon and Bogan (2001) state that unionoideans

"generally lose capacity for osmotic and volume regulation above 3-4 ppt" (salinity)

.

TDS is a rough measure of salinity, and the TDS of a sample of RMHRW with 2000

---

mg/L sulfate is 2.9 g/L or ppt (Soucek unpublished data) . Further experiments with

freshwater bivalves are required to determine if there is an absolute TDS level that is

tolerable, or if the limit depends upon water quality characteristics such as chloride

concentration and hardness .

An additional data gap is the fact that all of the tests conducted in the Soucek (2004)

study were acute exposures of 48 to 96 hours. Sublethal effects of sulfate in longer-term

exposures are unknown to this date. Given that, under the new limit proposed by IL EPA,

a continuous, long-term release at a given concentration will be allowed chronic testing

should be conducted to determine potential sublethal impacts

.

The purpose of the present study is to further provide data to support an appropriate

sulfate criterion for the protection of aquatic life in Illinois. Therefore, the objectives of

the current study are to build on previous studies conducted to support development of a

sulfate criterion for protection of aquatic life by (1) determining the effects of hardness on

toxicity of sulfate to bivalves, (2) determining the toxicity of sulfate to juvenile unionid

mussels, (3) determining the short-term (7 days) chronic toxicity of sulfate to

Ceriodaphnia dubia, (4)

determining the effects of chloride on acute toxicity of sulfate to

Hyalella

and

Ceriodaphnia,

and (5) determining the effects of hardness on toxicity of

sulfate to

Hyalella

at a critical chloride level, i.e ., the chloride concentration at which

sulfate is significantly less toxic to

Hyalella

as determined in

#4

above .

Project Objectives

Based on existing data gaps described above, and because of the desire to expedite the

process of updating sulfate standards in several states with limits based on scientific

research, the following tasks will be conducted

:

Examine hardness and chloride effects on acute sulfate toxicity to bivalves

.

The first experiments conducted in this task will include testing the acute toxicity of

sulfate to

Sphaerium simile

in RMHRW (Smith et al ., 1997). The other two organisms

tested in this water had markedly different responses compared to those observed in

MHRW; whereas the 96-hour LC50 in MHRW for

Hyalella

was 512 mg/L, the LC50

increased to 2,855 mg/L in RMHRW. The LC50 for C.

dubia

also increased from 2,050

in MHRW to 2,526 mg/L in RMHRW. However, both of these organisms are

crustaceans. Sulfate LC50s will be generated for S.

simile

at two chloride concentrations

(5 and 30 mg/L) and three hardness levels (100, 200, 300 mg/L)

.

Because they will be

conducted with organisms collected from the field, some of the tests may be eliminated

depending on availability of test organisms at the collection site

.

In addition, we will

generate sulfate LC50s using freshwater unionid mussel juveniles as test organisms, to

INHS

September 1, 2004

Page 4 of 12

---

INHS

September 1, 2004

Page 5 of 12

determine if this family is more sensitive to sulfate than the family Sphaeriidae (fingernail

clams) .

Conduct 7-day chronic sulfate toxicity tests withCeriodaphnia dubia

.

To test the hypothesis that the acute safe level is similar to the chronic safe level, we will

conduct 7-day chronic, survival/reproduction tests with Ceriodaphnia dubia in both

MHRW and RMHRW. Tests will be conducted according to ASTM methods (2002b),

and endpoints generated will include the number of young (both live and dead recorded

separately) produced by each first generation C.

dubia,

and survival of first generation C .

dubia .

EC20s and EC50s for survival and reproduction will be generated as will No

Observable Adverse Effects Concentrations (NOAEC) and Least Observable Adverse

Effects Concentrations (LOAEC) for both endpoints .

Determine the effects of chloride on acute toxicity of sulfate to

Hyalella

and

Ceriodaphnia,

Chloride standards currently exist in Illinois (500 mg/L, R. Mosher, pers. comm.), and the

State plans to additionally regulate for sulfate in order to address the major non-chloride

component of TDS in these waters. Because IL EPA proposes to eliminate the TDS

standard, opting instead to regulate sulfate and chloride, the interacting toxic effects of

these two anions must be characterized . Soucek (manuscript accepted with minor

revisions) has already shown that incrementally increasing chloride from 5 to 67 mg/L

reduces sulfate toxicity to

Hyalella .

We will further characterize this interaction by

determining sulfate LC50s over a wide range of chloride concentrations (10, 15, 20, 25,

100, 300, 500 mg/L) with hardness held constant. These tests will be conducted with

bothH. azteca and C. dubia .

Test effects of hardness on toxicity of sulfate to

Hyalella

at critical chloride level .

Having determined the critical chloride concentration required for

Hyalella

to show some

resistance to sulfate (task 3), we will determine how hardness affects the toxicity of

sulfate to

Hyalella

at this chloride concentration by generating LC50s at hardness values

of 100, 200, 300, 400, and 500 mg/L (as CaCO 3) and a constant chloride concentration

.

---

INHS

September 1, 2004

Page 6 of 12

Methods

Invertebrates selected for testing include

Ceriodaphnia dubia, Hyalella azteca, Sphaerium

simile

(Pelecypoda, Sphaeriidae), and a juvenile freshwater unionid mussel (species to be

determined). The cladoceran,

Ceriodaphnia dubia,

was cultured in-house (Soucek

laboratory, INHS) according to U.S. EPA (1993) methods. Prior to testing,

C.

dubia

were

fed a diet of

Pseudokirchneriella subcapitata

(also known as

Raphidocelis subcapitata

or

Selenastrum capricornutum)

and a Yeast-Cereal Leaves-Trout Chow (YCT) mixture at a

rate of 0.18 ml each per 30-m1 water, daily. Cultures were maintained at 25 °C, and a 16 :8

(L:D)photoperiod in Moderately Hard Reconstituted Water (MHRW; U.S. EPA, 1993) .

Amphipods,

Hyalella azteca,

were cultured in-house (Soucek laboratory, INHS) according

to U.S. EPA (1994a) methods in RMHRW at 22 °C and a 16 :8 (L:D) photoperiod for at

least 7 d prior to testing

.

Hyalella

were fed a diet of

Pseudokirchneriella subcapitata

and

a Yeast-Cereal Leaves-Trout Chow (YCT) mixture as well. Sphaeriid clams were and will

be collected from Spring Creek, near Loda, Illinois, (Iroquois County) and acclimated to

MHRW at 22 °C and a 16:8 (L:D) photoperiod for 5-7 d prior to testing . Clams collected

from this site were previously identified to species by Dr. Gerald Mackie, of the

University of Guelph, Department of Zoology, Guelph, Ontario, Canada

.

For toxicity testing, a pure (99%) grade of anhydrous Na2SO4 served as the source of

sulfate. A concentrated solution of this salt as well as a sample of laboratory deionized

water, will be acidified to pH <2.0 and analyzed for priority metal concentrations at the

Illinois State Water Survey (Champaign, IL) using inductively coupled plasma-atomic

emission spectrometry according to U.S. EPA (1994b) methods to determine if samples

are contaminated with trace metals .

For static, non-renewal acute toxicity tests, conducted according to ASTM E729-96

methods (2002a), treatments are comprised of a 75% dilution series (i .e., the 100%

concentration is serially diluted by 25%), rather than the standard 50%, because major ion

toxicity tests often cause 100% mortality in one concentration and 0% mortality in the

next

highest

concentration if the spread is too great. For the C.

dubia

and

H. azteca

tests,

five to six concentrations were tested, with four replicates tested per concentration, five

organisms per replicate . Tests with C.

dubia

were conducted for 48 h with a 16:8 (L:D)

photoperiod at 25 °C

.

H. azteca

and S.

simile

were exposed for 96 h at 22 °C and a 16 :8

(L:D) photoperiod.

C.

dubia

and

H. azteca

were exposed in 50-m1 glass beakers with 5

organisms per beaker, and for

H. azteca, 1

g of quartz sand was added to each beaker to

serve as substrate. Clam tests were and will be conducted in 150-m1 glass beakers (no

substrate). All clams used are juveniles .

C.

dubia

used in tests were less than 24-h old,

and

H. azteca

were -third instar (7 - 14 d old). Percent survival in each replicate was

recorded every 24 h and at the end of the exposure period . A dissecting microscope was

used to assess survival of

Hyalella

and

Simile .

For results to be acceptable, controls must

have had at least a 90% survival rate .

---

tNHS

September 1, 2004

Page 7 of 12

Chronic testing will be conducted according to guidelines described in ASTM E 1295-01

(2002). Ten replicates will be used per sulfate concentration with one organism per

replicate. Endpoints will include the number of young (both live and dead recorded

separately) produced by each first generation C.

dubia,

and survival of first generation C.

dubia .

Standard water chemistry parameters were measured at both the beginning and the end of

each exposure period, including temperature, pH, conductivity, dissolved oxygen,

alkalinity and hardness. The pH measurements were made using an Accumet (Fisher

Scientific, Pittsburgh, PA, USA) model AB15 pH meter equipped with an Accumet ® gel-

filled combination electrode (accuracy < ± 0 .05 pH at 25 °C). Dissolved oxygen was

measured using an air-calibrated Yellow Springs Instruments (RDP, Dayton, OH, USA)

model 58 meter with a self-stirring BOD probe . Conductivity measurements were made

using a Mettler Toledo® (Fisher Scientific, Pittsburgh, PA, USA) model MC226

conductivity/TDS meter. Alkalinity, and hardness were measured (beginning of tests

only) by titration as described in APHA et al . (1998). Samples from each treatment will

be analyzed to confirm sulfate concentrations by ion chromatography at the INHS Aquatic

Chemistry Laboratory, Champaign, IL

.

LC50 values were calculated using the Spearman-Karber method . To increase confidence

in LC50 values, three assays were/will be conducted for each objective

(i.e., either

organism or chloride X hardness combination) . This will provide a stronger estimate of

the mean LCSO value for each species .

In experiments testing effects of hardness of sulfate toxicity, hardness will be increased by

adding enough CaSO4 and MgSO4 to achieve the nominal hardness values. Then Na2SO4

will be added as was done with the standard MHRW. Whole carboys will be made at each

elevated hardness level and this water will used as both diluent and control ; therefore,

each concentration within a given test will have the same hardness

(i.e., [Ca21J and [Mg2.1

will not change with dilution). In experiments testing the influence of chloride on the

toxicity of sulfate toH.

azreca,

chloride, as NaCl, was added at appropriate concentrations

to solutions with a hardness of -100 mg/L.

---

INHS

September l, 2004

Page 8 of 12

Progress to date:

Hardness and chloride effects on acute sulfate toxicity to bivalves

.

In September, enough organisms were collected to conduct only two tests . One was conducted in

USEPA MHRW and the other in RMHRW . LC50s based on nominal concentrations were 1643

mg S041L for MHRW and 1864 mg S04/L. The LC50 for MHRW is on the low end of the range

previously generated by Soucek (2004), but these values are not yet based on measured

concentrations .

Because of rainfall and excessively high river levels in October and November, further sphaeriid

collections were impossible and remaining tests will be conducted in the Spring as soon as clams

are again accessible. In addition, tests with unionid glochidia and juveniles will be conducted by

USGS in the Spring of 2005 when test organisms are available from gravid females .

Effects of chloride on acute toxicity of sulfate to Hyalella and Ceriodaphnia .

Nearly all of the tests with Hyalella for this objective have been completed, and a trend in the

results has developed based on LC50s calculated from nominal sulfate concentrations (Table I

and Fig. 1). It appears that increasing chloride concentration from 10 to 33 mg/L results in a

sharp decrease in sulfate toxicity. In the Soucek (2004) study, RMHRW had a chloride

concentration of 33, and this water resulted in a mean LC50 of 2,855 mg S04/L . The 33 mg CI/L

treatment in this study resulted in a mean LC50 of 1,825 mg SO41L, but it is important to

remember that the solutions in these tests have a lower calcium concentration than in RMHRW

(23 mgt compared to 32 mgt in RMHRW) despite having the same hardness. In the current

study, a Ca:Mg ratio of 2.33:1 (compared to 5.40:1 in RMHRW) was selected to reflect the

median ratio for streams in Illinois . Calcium is probably more important in mediating sulfate

toxicity than magnesium, thus accounting for the relatively large difference in LC50 values

despite nearly identical hardness values. It is my hope that further experiments will support this

contention .

Above 33 mg Cl/L, LC50 values rose again slightly in the 100 mg CIt treatment but then

dropped in the 300 and 500 mg CUL tests (Table 1 and Fig . 1). If these data are viewed in a

slightly different way, another interesting trend appears (Figs . 2 and 3). I compared

conductivities and total dissolved solids (TDS, calculated) at LC50 concentrations for the

different chloride treatments, and found that for the lower concentrations, both conductivity and

TDS of the LC50 concentration increased with increasing chloride, as was the case for sulfate

concentration, until a threshold was reached . Then, at the higher chloride concentrations (100 to

500 mgt), no further benefit was provided by chloride above 100 mgt . In fact, it appears that

at 100 mg Cl/L, a threshold of 4,300 pS/cm or 3 .1 gt (conductivity and TDS, respectively) is

reached. At higher chloride concentrations, less sulfate is required to reach the critical

conductivity and TDS values, so LC50s in terms of sulfate decrease at 300 and 500 mg CIIL . In

other words, it appears that for Hyalella, in the range of 100 to 500 mg CUL, toxicity is reached

at a fixed conductivity or TDS and if there is more chloride, less sulfate is required to reach this

threshold and vice versa

.

---

Table 1. Influence of chloride concentration on toxicity of sulfate to Hyalella azteca. All tests were conducted at 22

°C for 96 hours. Chloride and sulfate values shown are nominal concentrations, and all treatments within a given test

had the same chloride concentration and hardness. Mean hardness for all tests was -100 mg/L CaCO 3 , and the

Ca:Mg used was 2.33:1 (mg/L:mg:L) to reflect the median Ca:Mg ratio in Illinois (Clark Olson, IEPA,pers. com.)

LC50s were generated using the Spearman-Karber method .

500

0

8

t

0

100

200

300

400

chloride (mg/L)

500

600

INHS

September 1, 2004

Page 9 of 12

Figure 1 . Influence of chloride concentration on toxicity of sulfate to Hyalella azteca.

See table one for details on

test conditions .

Chloride (mg/L)

number of tests

mg/L

conducted

Mean LC50

mg sulfate/L

10

3

1,387

15

2

1,632

20

3

1,562

25

3

1,854

33

2

1,825

100

3

1,938

300

3

1,691

500

2

1,469

2500

2000

-

---

5000

0

UU

A

4500 -

9

0

4000

-

I

;,

3500

-

3000

-

O

U

2500

0

100

200

300

400

chloride (mg/L)

500

600

INNS

September 1, 2004

Page 10 of 12

Figure 2. Relationship between chloride concentration and solution conductivity at the LC50 concentration in

toxicity tests sulfate with

Hyalella

azteca. Conductivity values were calculated from equations generated with linear

regression of measured conductivities attest sulfate concentrations

.

3 .50

3.30

-

3.10

-

0

290-

U

-t 2.70 -

A

i 2

.50 -

2.30

-

2.10

-

1.90

-

1.70

-

1 .50

0 .00

100.00

200.00

300.00

400.00

chloride (mg/L)

500.00

600.00

Figure 3 . Relationship between chloride concentration and total dissolved solids at the LC50 concentration in

toxicity tests sulfate to

Hyalella

azteca. Total dissolved solids values were calculated using nominal concentrations

of all ions present in solution (excluding H+ and OH-) at LC50 concentrations

.

I

---

Tests have been conducted with

Ceriodaphnia

at three chloride concentrations, and thus far, it

appears the sulfate toxicity increases when chloride increases from 100 to 300 to 500 mg/L

(Table 2). This objective has not progressed as far as would have been expected because a

problem occurred in that the C .

dubia

cultures were either not producing healthy neonates or

were not reproducing at all . A number of weeks were spent isolating the problem and it was

discovered through the process of elimination that the

Pseudokirchneriella

culture being used as

food for the organisms was causing toxicity. This problem has been rectified and the

contaminated food has been replaced with

Pseudokirchneriella

from Aquatic Research

Organisms (Hampton, New Hampshire). As a result, C.

dubia

cultures are again robust and

producing ample healthy neonates for testing . The tests shown in table 2 were recently

conducted using these organisms .

Table 2. Influence of chloride concentration on toxicity of sulfate to

Hyalella azteca. All

tests were conducted at 22

°C for 96

hours. Chloride and sulfate values shown are nominal concentrations, and all treatments within a given test

had the same chloride concentration and hardness . Mean hardness for all tests was -100 mg/L CaCO3, and the

Ca:Mg used was 2.33:1 (mg/L:mg:L) to reflect the median Ca:Mg ratio in Illinois (Clark Olson,

IEPA,

pers .

com .)

LC50s were generated using the Spearman-Karber method

.

Chloride (mg/L)

Mean LC50

mx/L mg sulfate/L

100

2,357

300

1,895

500

1,400

Effects of hardness on toxicity of sulfate to

Hyalella

at critical chloride level

and

7-day chronic sulfate toxicity tests with

Ceriodaphnia dubia,

These objectives will be completed during the first quarter of 2005

.

Projected accomplishments for first quarter, 2005

After finishing the remaining tests with C.

dubia

and

H. azteca

examining effects of chloride on

sulfate toxicity, the remaining two objectives: examining the effects of hardness on toxicity of

sulfate to

Hyalella

at critical chloride level, and 7-day chronic sulfate toxicity tests with

Ceriodaphnia dubia, will

commence. It is my expectation that the chronic testing and most, if

not all of the tests examining hardness effects will be completed in the first quarter of 2005

.

Then, in the Spring, when bivalves are available, the last remaining objective will be completed

.

The remaining time in the project period will then be used for data analysis and report

preparation .

INNS

September

1, 2004

Page

I I of 12

---

Literature cited

American Public Health Association (APHA), American Water Works Association, Water

Environment Federation. 1998. Standard methods for the examination of water and

wastewater, 20th ed. American Public Health Association, Washington, DC

.

ASTM (American Society for Testing and Materials). 2002a. Standard guide for conducting

acute toxicity tests on test materials with fishes, macroinvertebrates, and amphibians

.

E729-96. American Society for Testing and Materials, Philadelphia, PA, USA.

ASTM (American Society for Testing and Materials) . 2002b. Standard guide for conducting

three-brood, renewal toxicity tests with

Ceriodaphnia dubia

.

E1295-01. American

Society for Testing and Materials, Philadelphia, PA, USA

.

Bayliss D, Harris RR . 1988. J Comp Physiol 158:81-90

Dwyer FJ, Burch SA, Ingersoll CG, and Hunn JB

.

1992. Environ Toxicol Chem 11 :513-520 .

Ingersoll CG, Dwyer FJ, Burch SA, Nelson MK, Buckler DR, and Hunn JB

. 1992. Environ

Toxicol Chem 11 :503-511

.

Lucu C, and Flik G. 1999. Am J Physiol 276:R490-R499

.

McMahon, RF, and Bogan, AE. 2001. Mollusca: Bivalvia, in JH Thorp and AP Covich (eds .)

Ecology and Classification of North American Freshwater Invertebrates, 2nd Edition.

Academic Press, San Diego, CA. pp 352-353

.

Mount DR, Gulley DD, Hockett JR, Garrison TD, and Evans JM. 1997. Environ Toxicol Chem

16:2009-2019

.

Pic P, and Maetz J. 1981. J Comp Physiol B 141 :511-521

.

Soucek, DJ. 2004. Effects of hardness, chloride, and acclimation on the acute toxicity of sulfate

to freshwater invertebrates . Final Report to: Illinois Environmental Protection Agency

and Illinois Coal Association

.

Smith ME, Lazorchak JM, Herrin LE, Brewer-Swartz S, and Thoeny WT

. 1997. Environ

Toxicol Chem 16:1229-1233 .

U.S. EPA. 1985. Guidelines for deriving numerical water quality criteria for the protection

of

aquatic organisms and their uses. PB85-227049. Washington, DC .

U .S. EPA. 1993. Methods for measuring the acute toxicity of effluents and receiving waters to

freshwater and marine organisms, 4 d' ed. EPA/600/4-901027F, Cincinnati, OH

.

U.S. EPA. 1994a. Methods for measuring the toxicity and bioaccumulation of sediment

associated contaminants with freshwater invertebrates . EPA/600/R-94/024. U.S

.

Environmental Protection Agency, Washington, DC

.

U.S. EPA. 1994b. Methods for the determination of metals in environmental samples. EPA/600/R-

4/111. U.S. Environmental Protection Agency, Cincinnati, OH

.

U .S. EPA. 1999. National Recommended Water Quality Criteria - Correction . EPA 822-Z-99-

001, Washington, DC .

Zipper CE. 2000. Coal mine reclamation, acid mine drainage and the Clean Water Act, in R .

Bamhisel, W. Daniels, and R. Darmody (eds) Reclamation of Drastically Disturbed

Lands. American Society of Agronomy, Madison, WI

.

INHS

September 1, 2004

Page 12 of 12

---

Effects of Water Quality on Acute and Chronic Toxicity of Sulfate

to Freshwater Bivalves,

Ceriodaphnia dubia,

and

Hyalella azteca .

Second Quarterly Report

Submitted to :

Edward Hammer and Dertera Collins

United States Environmental Protection Agency

Region 5, Water Division, 77 West Jackson Boulevard

Chicago, Illinois 60604

April 10, 2005

Illinois Natural History Survey, Champaign, IL

U.S. EPA Region 5

INHS

April 10, 2005

Page I of 16

Illinois EPA Exhibit No

.

3

---

Background

While there are no Federal water quality criteria (WQC) for the protection of freshwater

life for total dissolved solids (TDS), sulfate, or sodium (U .S . EPA 1999), several states,

including Minnesota, Indiana, and Illinois, are at various stages in the process of

developing standards for sulfate . Water quality standards are developed to protect

designated uses, aquatic life uses in this case, but the economic impacts of these standards

are important considerations as well . For example, after existing sulfate standards were

enacted in Illinois, the Illinois Pollution Control Board adopted exceptions to the

standards to provide relief for a number of coalmines that were enduring severe economic

hardship. In fact, -60% of coalmines have expired permits in Illinois because of

violations of the sulfate standard, and -50% of those have been expired for more than

three years (T. Frevert, pers. comm.). The current "General Use" standard of 500 mg/L in

Illinois is based on the value thought to be protective of livestock .

Consultation with

appropriate authorities revealed that livestock were capable of tolerating much higher

levels of sulfate. In light of these factors, the Illinois Environmental Protection Agency

(IL EPA) is actively pursuing an update of the sulfate standards based on scientific

research, and is close to proposing an updated standard (R . Mosher, IL EPA, pers

.

comm.) .

Sodium is one of the most common major cations in high TDS effluents, but calcium and

chloride are usually present in mine-impacted waters as well

. While major ion or TDS

toxicity is caused by osmoregulatory stress from the combination of all cations and

anions, chloride standards currently exist, and Illinois plans to additionally regulate

for

sulfate in order to address the major non-chloride component of TDS in these waters

.

Therefore, studies (funded by IL EPA, and IL Coal Association) were conducted by

Soucek (2004; In Press Environmental Toxicology and Chemistry) to (1) generate LC50s

(lethal concentration to 50% of a sample population) and LC lOs (lethal concentration to

10% of a sample population) for sulfate with selected freshwater invertebrates

(Ceriodaphnia dubia, Chironomus tentans, Hyalella azteca,

and

Sphaerium simile)

in

U.S. Environmental Protection Agency's (US EPA, 1993) moderately hard reconstituted

water (MHRW) and (2) determine the effects of laboratory water composition, water

hardness, and test organism acclimation on the acute toxicity of sulfate to

Ceriodaphnia

dubia

and

Hyalella azteca

(Soucek, 2004). In these previous studies (Soucek, 2004), the

mean LC50s, expressed as mg S04 2/L, in moderately hard, reconstituted water (MHRW

;

U.S. EPA, 1993) ranged from: 512 to 1,4134 mg/L

The LC50 generated for the

amphipod,

Hyalella

(512 mg/L) was surprisingly low, given that it is known as a

euryhaline organism (Ingersoll et al ., 1992), but as will be discussed below, water quality

data, including other cations and anions present, are critical for predicting the responses

of freshwater organisms (especially

Hyalella)

to elevated sulfate concentrations .

INNS

April 10, 2005

Page 2 of 16

---

INHS

April 10, 2005

Page 3 of 16

The composition of the dilution water used during testing in the Soucek (2004) study had

a dramatic effect on the toxicity of sulfate to

Hyalella .

Whereas the 96-hour LC50 in

MHRW was 512 mg/L, the LC50 increased to 2,855 mg/L when using a "Reformulated

Moderately Hard Reconstituted Water" (RMHRW, Smith et al ., 1997). The LC50 for C .

dubia

also increased from 2,050 in MHRW to 2,526 mg/L in RMHRW. Both dilution

waters were similar in terms of hardness (-90-106 mg/L as CaCO3), alkalinity, and pH,

but RMHRW had a higher chloride concentration and different calcium to magnesium

ratio than that in MHRW. An additional experiment conducted at the Illinois Natural

History Survey (INNS) laboratory, but not included in the Soucek (2004) report, indicated

that when sulfate (-2,800 mg/L) and hardness (106 mg/L) were held constant, percent

survival of

H. azteca

was positively correlated with chloride concentration (up to 67 mg

Cl -/Q. These experiments illustrate the need to further characterize the interacting effects

of chloride and sulfate on aquatic organisms

.

Another factor that appears to have a strong effect on the toxicity of sulfate is the

presence of other major cations, in this case, calcium and magnesium, measured as

hardness. In the previous study (Soucek 2004), increased hardness reduced the toxicity of

sulfate to

Hyalella

and had a dramatic effect on the 48-hour LC50 for C .

dubia,

increasing from 2,050 at a hardness of 90 to > 2,900 mg/L at hardness values higher than

194 mg/L asCaCO3. Others have observed reduced toxicity of saline solutions due to

increased hardness as well (e.g., Dwyer et al. 1992; Mount et al. 1997) .

While a great deal of progress was made in the understanding of sulfate toxicity under

varying water quality conditions, several important data gaps remain. In the previous

studies (Soucek 2004), the fingernail clam,

Sphaerium simile,

had a lower LC 10 than that

of C. dubia,

but because of the temporal nature of its availability, this bivalve was only

tested in MHRW. It remains unclear whether or not a mollusk will have the same

physiological response as two crustaceans to increased chloride or hardness in these

experiments with sulfate. The principal inorganic anion of crustacean blood, or

hemolymph, is chloride, and it has been suggested that low chloride concentrations may

limit the distribution of at least one euryhaline amphipod

(Corophium curvispinum)

in

freshwaters (Bayliss and Harris, 1988) . However, in the unionid mussel,

Toxolasma

texasiensis,

chloride and bicarbonate are equally important anions in the hemolymph (see

McMahon and Bogan 2001). Because bicarbonate is readily available via respiration and

metabolism, this mussel may not depend on external chloride concentrations to the extent

that some crustaceans do. If this is the case, the protective effect of chloride observed for

Hyalella

and

Ceriodaphnia

might not be manifest in some unionoidean bivalves . The

hardness effect observed in the Soucek (2004) study may be more widespread among

aquatic phyla, because calcium simply reduces gill permeability (Lucu and Flik, 1999

; Pic

and Maetz, 1981). However, McMahon and Bogan (2001) state that unionoideans

"generally lose capacity for osmotic and volume regulation above 3-4 ppt" (salinity) .

TDS is a rough measure of salinity, and the TDS of a sample of RMHRW with 2000

---

mg/L sulfate is 2.9 g/L or ppt (Soucek unpublished data). Further experiments with

freshwater bivalves are required to determine if there is an absolute TDS level that is

tolerable, or if the limit depends upon water quality characteristics such as chloride

concentration and hardness .

An additional data gap is the fact that all of the tests conducted in the Soucek (2004)

study were acute exposures of 48 to 96 hours. Sublethal effects of sulfate in longer-term

exposures are unknown to this date. Given that, under the new limit proposed by IL EPA,

a continuous, long-term release at a given concentration will be allowed chronic testing

should be conducted to determine potential sublethal impacts

.

The purpose of the present study is to further provide data to support an appropriate

sulfate criterion for the protection of aquatic life in lllinois . Therefore, the objectives of

the current study are to build on previous studies conducted to support development of a

sulfate criterion for protection of aquatic life by (1) determining the effects of hardness on

toxicity of sulfate to bivalves, (2) determining the toxicity of sulfate to juvenile unionid

mussels, (3) determining the short-term (7 days) chronic toxicity of sulfate to

Ceriodaphnia dubia,

(4) determining the effects of chloride on acute toxicity of sulfate to

Hyalella

and

Ceriodaphnia,

and (5) determining the effects of hardness on toxicity of

sulfate to

Hyalella

at a critical chloride level,

i.e ., the chloride concentration at which

sulfate is significantly less toxic to

Hyalella

as determined in #4 above .

Project Objectives

Based on existing data gaps described above, and because of the desire to expedite the

process of updating sulfate standards in several states with limits based on scientific

research, the following tasks will be conducted

:

Examine hardness and chloride effects on acute sulfate toxicity to bivalves

.

The first experiments conducted in this task will include testing the acute toxicity of

sulfate to

Sphaerium simile

in RMHRW (Smith et al., 1997). The other two organisms

tested in this water had markedly different responses compared to those observed in

MHRW; whereas the 96-hour LC50 in MHRW for

Hyalella

was 512 mg/L, the LC50

increased to 2,855 mg/L in RMHRW. The LC50 for C .

dubia

also increased from 2,050

in MHRW to 2,526 mg/L in RMHRW. However, both of these organisms are

crustaceans. Sulfate LC50s will be generated for

S. simile

at two chloride concentrations

(5 and 30 mg/L) and three hardness levels (100, 200, 300 mg/L) . Because they will be

conducted with organisms collected from the field, some of the tests may be eliminated

depending on availability of test organisms at the collection site. In addition, we will

generate sulfate LC50s using freshwater unionid mussel juveniles as test organisms, to

INNS

April 10, 2005

Page 4 of 16

---

INNS

April 10, 2005

Page 5 of 16

determine if this family is more sensitive to sulfate than the family Sphaeriidae (fingernail

clams) .

Conduct 7-day chronic sulfate toxicity tests with

Ceriodaphnia dubia

.

To test the hypothesis that the acute safe level is similar to the chronic safe level, we will

conduct 7-day chronic, survival/reproduction tests with

Ceriodaphnia dubia

in both

MHRW and RMHRW. Tests will be conducted according to ASTM methods (2002b),

and endpoints generated will include the number of young (both live and dead recorded

separately) produced by each first generation C. dubia, and survival of first generation C.

dubia .

EC20s and EC50s for survival and reproduction will be generated as will No

Observable Adverse Effects Concentrations (NOAEC) and Least Observable Adverse

Effects Concentrations (LOAEC) for both endpoints

.

Determine the effects of chloride on acute toxicity of sulfate toHvalellaand

Ceriodaphnia.

Chloride standards currently exist in Illinois (500 mg/L, R. Mosher, pers. comm .), and the

State plans to additionally regulate for sulfate in order to address the major non-chloride

component of TDS in these waters . Because IL EPA proposes to eliminate the TDS

standard, opting instead to regulate sulfate and chloride, the interacting toxic effects of

these two anions must be characterized . Soucek (manuscript accepted with minor

revisions) has already shown that incrementally increasing chloride from 5 to 67 mg/L

reduces sulfate toxicity to Hyalella. We will further characterize this interaction by

determining sulfate LC50s over a wide range of chloride concentrations (10, 15, 20, 25,

100, 300, 500 mg/L) with hardness held constant . These tests will be conducted with

both H.

azteca

and C.

dubia .

Test effects of hardness on toxicity of sulfate toHvalellaat critical chloride level

.

Having determined the critical chloride concentration required for

Hyalella to show some

resistance to sulfate (task 3), we will determine how hardness affects the toxicity of

sulfate to Hyalella at this chloride concentration by generating LC50s at hardness values

of 100, 200, 300, 400, and 500 mg/L (as CaCO3) and a constant chloride concentration

.

Methods

Invertebrates selected for testing include

Ceriodaphnia dubia, Hyalella azteca, Sphaerium

simile (Pelecypoda, Sphaeriidae), and a juvenile freshwater unionid mussel (species to be

determined). The cladoceran,

Ceriodaphnia dubia, was

cultured in-house (Soucek

laboratory, INHS) according to U .S. EPA (1993) methods. Prior to testing, C . dubia were

fed a diet of

Pseudokirchneriella subcapitata

(also known as

Raphidocelis subcapitata

or

Selenastrum capricornutum) and a Yeast-Cereal Leaves-Trout Chow (YCT) mixture at a

rate of 0.18 ml each per 30-m1 water, daily. Cultures were maintained at 25 °C, and a 16:8

---

INHS

April 10, 2005

Page 6 of 16

(L:D)

photoperiod in Moderately Hard Reconstituted Water (MHRW; U.S. EPA, 1993) .

Amphipods,

Hyalella azteca,

were cultured in-house (Soucek laboratory, INHS) according

to U.S. EPA (1994a) methods in RMHRW at 22 °C and a 16 :8 (L:D) photoperiod for at

least 7 d prior to testing .

Hyalella

were fed a diet of

Pseudokirchneriella subcapitata

and

a Yeast-Cereal Leaves-Trout Chow (YCT) mixture as well. Sphaeriid clams were and will

be collected from Spring Creek, near Loda, Illinois, (Iroquois County) and acclimated to

MHRW at 22 °C and a 16 :8 (L:D) photoperiod for 5-7 d prior to testing . Clams collected

from this site were previously identified to species by Dr. Gerald Mackie, of the

University of Guelph, Department of Zoology, Guelph, Ontario, Canada .

For toxicity testing, a pure (99%) grade of anhydrous Na2SO4 served as the source of

sulfate. A concentrated solution of this salt as well as a sample of laboratory deionized

water, will be acidified to pH <2 .0 and analyzed for priority metal concentrations at the

Illinois State Water Survey (Champaign, IL) using inductively coupled plasma-atomic

emission spectrometry according to U.S. EPA (1994b) methods to determine if samples

are contaminated with trace metals

.

For static, non-renewal acute toxicity tests, conducted according to ASTM E729-96

methods (2002a), treatments are comprised of a 75% dilution series (i.e., the 100%

concentration is serially diluted by 25%), rather than the standard 50%, because major ion

toxicity tests often cause 100% mortality in one concentration and 0% mortality in the

next highest concentration if the spread is too great. For the C.

dubia

and

H. azteca

tests,

five to six concentrations were tested, with four replicates tested per concentration, five

organisms per replicate. Tests with C.

dubia

were conducted for 48 h with a 16:8 (L:D)

photoperiod at 25 °C .

H. azteca

and S.

simile

were exposed for 96 h at 22 °C and a 16:8

(L:D) photoperiod .

C.

dubia

and

H. azteca

were exposed in 50-m1 glass beakers with 5

organisms per beaker, and for

H. azteca, 1

g of quartz sand was added to each beaker to

serve as substrate. Clam tests were and will be conducted in 150-m1 glass beakers (no

substrate). All clams used are juveniles .

C.

dubia

used in tests were less than 24-h old,

and

H. azteca

were -third instar (7 - 14 d old). Percent survival in each replicate was

recorded every 24 h and at the end of the exposure period. A dissecting microscope was

used to assess survival of

Hyalella

and

Simile.

For results to be acceptable, controls must

have had at least a 90% survival rate

.

Chronic testing will be conducted according to guidelines described in ASTM E 1295-01

(2002). Ten replicates will be used per sulfate concentration with one organism per

replicate. Endpoints will include the number of young (both live and dead recorded

separately) produced by each first generation C.

dubia,

and survival of first generation C.

dubia.

Standard water chemistry parameters were measured at both the beginning and the end of

each exposure period, including temperature, pH, conductivity, dissolved oxygen,

---

INHS

April 10, 2005

Page 7 of 16

alkalinity and hardness. The pH measurements were made using an Accumet ®(Fisher

Scientific, Pittsburgh, PA, USA) model AB15 pH meter equipped with an Accumet ®gel-

filled combination electrode (accuracy < ± 0 .05 pH at 25 °C). Dissolved oxygen was

measured using an air-calibrated Yellow Springs Instruments (RDP, Dayton, OH, USA)

model 58 meter with a self-stirring BOD probe . Conductivity measurements were made

using a Mettler Toledo (Fisher Scientific, Pittsburgh, PA, USA) model MC226

conductivity/TDS meter. Alkalinity, and hardness were measured (beginning of tests

only) by titration as described in APHA et al. (1998). Samples from each treatment will

be analyzed to confirm sulfate concentrations by ion chromatography at the INHS Aquatic

Chemistry Laboratory, Champaign, IL.

LC50 values were calculated using the Spearman-Karber method. To increase confidence

in LC50 values, three assays were/will be conducted for each objective

(i.e., either

organism or chloride X hardness combination) . This will provide a stronger estimate of

the mean LC50 value for each species .

In experiments testing effects of hardness of sulfate toxicity, hardness will be increased by

adding enough CaSO4 and MgSO4 to achieve the nominal hardness values. Then Na2SO4

will be added as was done with the standard MHRW . Whole carboys will be made at each

elevated hardness level and this water

will

used as both diluent and control ; therefore,

each concentration within a given test will have the same hardness

(i.e.,

[C2*]

and [Mg

2l

j

will not change with dilution) . In experiments testing the influence of chloride on the

toxicity of sulfate to

H. azteca,

chloride, as NaCl, was added at appropriate concentrations

to solutions with a hardness of -100 mg/L.

Progress to date:

Hardness and chloride effects on acute sulfate toxicity to bivalves

.

In September, enough organisms were collected to conduct only two tests. One was conducted in

USEPA MHRW and the other in RMHRW. LC50s based on nominal concentrations were 1643

mg SO4/L for MHRW and 1864 mg SO 4/L. The LC50 for MHRW is on the low end of the range

previously generated by Soucek (2004), but these values are not yet based on measured

concentrations .

Currently, water levels have dropped sufficiently for collection of sphaeriids, and enough

individuals for two tests were collected on April 9, 2005 . Tests will be initiated with these

organisms on April 11, 2005, and more clams will be collected in subsequent weeks for further

tests .

---

INNS

April 10, 2005

Page 8 of 16

Effects of chloride on acute toxicity of sulfate to

Hvalella

and

Ceriodaphnia .

All of the tests with

Hyalella

for this objective have been completed, and a trend in the results

has developed based on LC50s calculated from nominal sulfate concentrations (Table 1 and Fig

.

1). It appears that increasing chloride concentration from 10 to 33 mg/L results in a sharp

decrease in sulfate toxicity. In the Soucek (2004) study, RMHRW had a chloride concentration

of 33, and this water resulted in a mean LC50 of 2,855 mg S0 4/L. The 33 mg CUL treatment in

this study resulted in a mean LC50 of 1,825 mg S04/L, but it is important to remember that the

solutions in these tests have a lower calcium concentration than in RMHRW (23 mg/L compared

to 32 mg/L in RMHRW) despite having the same hardness . In the current study, a Ca:Mg ratio

of 2.33:1 (compared to 5 .40:1 in RMHRW) was selected to reflect the median ratio for streams in

Illinois. Calcium is probably more important in mediating sulfate toxicity than magnesium, thus

accounting for the relatively large difference in LC50 values despite nearly identical hardness

values .

Above 33 mg Cl/L, LC50 values rose again slightly in the 100 mg Cl/L treatment but then

dropped in the 300 and 500 mg Cl/L tests (Table I and Fig. 1). If these data are viewed in a

slightly different way, another interesting trend appears (Fig . 2). I compared total dissolved

solids (TDS, calculated) at LC50 concentrations for the different chloride treatments, and found

that for the lower concentrations, TDS of the LC50 concentration increased with increasing

chloride, as was the case for sulfate concentration, until a threshold was reached . Then, at the

higher chloride concentrations (100 to 500 mg/L), no further benefit was provided by chloride

above 100 mg/L. In fact, it appears that at 100 mg Cl/L, a threshold of -3 .1 g/L TDS is reached

.

At higher chloride concentrations, less sulfate is required to reach the critical TDS value, so

LC50s in terms of sulfate decrease at 300 and 500 mg CI/L . In other words, it appears that for

Hyalella,

in the range of 100 to 500 mg Cl/L, toxicity is reached at a fixed TDS and if there is

more chloride, less sulfate is required to reach this threshold and vice versa

.

Table 1 . Influence of chloride concentration on toxicity of sulfate to

Hyalella azteca

.

All tests were conducted at 22

°C for 96 hours. Chloride and sulfate values shown are nominal concentrations, and all treatments within a given test

had the same chloride concentration and hardness . Mean hardness for all tests was -100 mg/L CaCO3, and the

Ca:Mg used was 2.33:1 (mg/L:mg:L) to reflect the median Ca:Mg ratio in Illinois (Clark Olson, TEPA, pers. com.)

LC50s were generated using the Spearman-Karber method

.

Chloride (mg/L)

mgfL

number of tests

conducted

Mean LC50

mg sulfate/L

10

3

1,387

15

3

1,563

20

3

1,562

25

3

1,854

33

3

1,799

100

3

1,938

300

3

1,691

500

3

1,470

---

2500

2000

7

1500

N

S

M

0

100

200

300

400

500

600

chloride (mg/L)

O

F

2.30

2.10

1 .90

1.70

1.50

I

0

100

200

300

400

500

600

chloride

INNS

April 10, 2005

Page 9 of 16

Figure 1 . Influence of chloride concentration on toxicity of sulfate to Hyalella azteca.

See table one for details on

test conditions

.

Figure 2. Relationship between chloride concentration and total dissolved solids at the LC50 concentration in

toxicity tests sulfate to Hyalella azteca.

Total dissolved solids values were calculated using nominal concentrations

of all ions present in solution (excluding H+ and OH-) at LC50 concentrations,

3.50

3.30

-

3.10 -

2.90 -

y =0.2175Ln(x)+ 1 .8951

U 2.70

RZ = 0.7317

N 2.50

y

---

I14IS

April 10, 2005

Page 10 of 16

All of the tests with

Ceriodaphnia

for this objective also have been completed. The trend of

decreased toxicity with increasing chloride at the lower end of the range as seen for

Hyalella is

not as clear with

Ceriodaphnia

(Table 2 and Fig. 3). It appears that at chloride concentrations

between 10 and 100 mg/L sulfate toxicity is fairly constant (with some noise in the data), with

LC50 values between 2,200 and 2,500 mg SO4/I. (nominal concentrations). In the Soucek (2004)

study,

Ceriodaphnia

did not respond as strikingly as

Hyalella

when tested in RMHRW compared

to MHRW so these results may not be entirely surprising .

As was the case with

Hyalella,

LC50 values in terms of sulfate were lower at the 300 and 500 mg

Cl/L concentrations compared to the lower chloride concentrations (Table 2 and Fig . 3).

Comparing TDS (calculated) at LC50 concentrations for the different chloride treatments at the

lower range of chloride concentrations, TDS values at LC50 concentrations were higher for

Ceriodaphnia

(3 .5 - 3.8 g/L) than they were for

Hyalella

(2.4 - 3 .1 g/L). The overall trend was

different for

Ceriodaphnia

as well with an overall linear trend of decreasing TDS at LC50 with

increasing chloride concentration (Fig . 4)

.

Table 2. Influence of chloride concentration on toxicity of sulfate to

Ceriodaphnia dubia. All

tests were conducted

at 25 °C for 48 hours. Chloride and sulfate values shown are nominal concentrations, and all treatments within a

given test had the same chloride concentration and hardness. Mean hardness for all tests was -100 mg/L CaCO

3 , and

the Ca:Mg used was 2.33:1 (mg/L:mg:L) to reflect the median Ca:Mg ratio in Illinois (Clark Olson,

TEPA,

pers .

com.) LC50s were generated using the Spearman-Karber method .

Chloride (mg/L)

mg/L

number of tests

conducted

Mean LC50

mg sulfate/L

10

3

2,469

15

3

2,289

20

3

2,419

25

3

2,272

100

3

2,417

300

3

1,914

500

3

1,496

---

3000

2500

y°

2000

h

1500

°

1000

U

..1

N

500 -

0

0

100

200

300

400

500

600

chloride (mg/L .)

Figure 3. Influence of chloride concentration on toxicity of sulfate to Ceriodaphnia dubia

.

See table 2 for details on test conditions .

chloride

INHS

April 10, 2005

Page I I of 16

Figure 4. Relationship between chloride concentration and total dissolved solids at the LC50 concentration in

toxicity tests sulfate to Ceriodaphnia dubia. Total dissolved solids values were calculated using nominal

concentrations of all ions present in solution (excluding H+ and OH-) at LC50 concentrations

.

3.9

3.7

3.5

0

h

J

3.3

-

O

3.1

-

F

2.9

-

2.7

-

y =-0.001x+ 3.6964

R2 = 0.628

2.5

1

0

100

200

300

400

500

600

---

INNS

April 10, 2005

Page 12 of 16

Effects of hardness on toxicity of sulfate to

Hvalella

at critical chloride level

A number of tests have been completed for this objective, but not all of them. Thus far, as was

seen with

Ceriodaphnia dubia

in the Soucek (2004) study, a general trend of decreased toxicity

with increased hardness is developing with the data (Fig . 5). Only one test has been conducted at

hardness = 500 mg/L, and that value is rather low, but it is expected that the overall linear trend

will smooth out once all of the tests are completed

3500

3000

2500

2000

1500

1000

500

0

y= 1.585x+ 1947.3

R2 = 0.3189

0

100

200

300

400

500

600

700

hardness (mg/L)

Figure 5. Influence of hardness on toxicity of sulfate to

Hyalella

azteca at chloride = 25 mg/L .

See table one for details on test conditions

.

7-day chronic sulfate toxicity tests with

Ceriodaphnia dubia

.

All of the tests for this objective have been completed (three chronic tests in each dilution water),

and results were quite consistent from test to test . As depicted in figures 6-9, chronic toxicity in

terms of survival and reproduction was less in RMHRW compared to MHRW. Table 3

summarizes the Least Observable Adverse Effects Concentrations (LOAEC) and No Observable

Adverse Effects Concentrations (NOAEC) for each dilution water. While LOAECs for

reproduction are quite low, it should be noted that I have had a continuous, self sustaining,

reserve culture

of

Ceriodaphnia dubia

in MHRW spiked with 1,000 mg SO4/L since at least

August of 2004. Organisms used in chronic testing for this study were cultured in MHRW or

RMHRW as appropriate

.

---

INHS

April 10, 2005

Page 13 of 16

Table 3. Summary of mean NOAEC and LOAEC concentrations for survival and reproduction of

Ceriodaphnia

dubia

in 7-day static-renewal three brood toxicity tests with sulfate

.

Values are in terms of mg sulfate/L. All tests

were conducted at 25 °C. Chloride and sulfate values shown are nominal concentrations. MHRW = Moderately

Hard Reconstituted Water (USEPA 1993)

. RMHRW = Reformulated Moderately Hard Reconstituted Water (Smith

et al., 1997)

Dilution water

number of tests Survival Reproduction

conducted NOAEC

LOAEC

NOAEC

LOAEC

MHRW

3

1,727

2,273

780

934

RMHRW 3 2,264 3,000 906 1,195

Projected accomplishments for second quarter, 2005

After finishing the remaining tests with

H. azteca

examining effects of hardness on toxicity of

sulfate to

Hyalella

at critical chloride level, one objective remains

: Hardness and chloride effects

on acute sulfate toxicity to bivalves

. As stated above, a first batch of fingernail clams has been

collected and testing will begin immediately, to be finished as soon as possible given water levels

remain appropriate for further collections . Juvenile unionid mussels will be available

approximately May 1, 2005 and testing at the USGS, Columbia Environmental Research Center,

Columbia, MO, will take approximately one week

.

In addition, all values reported herein are based on nominal sulfate and chloride concentrations .

Sample analysis is on-going and thus far, all measured values have been within approximately

5% of nominal values, so results reported here are not expected to change substantially when

chemical analysis is completed . The remaining time in the project period will then be used for

data analysis and report preparation .

---

INNS

April 10, 2005

Page 14 of 16

Figure 6. Mean percent survival of

Ceriodaphnia

dubia in 3-brood, static-renewal, chronic toxicity tests with sulfate

in Moderately Hard Reconstituted Water (MHRW) . Bars and error bars indicate means and standard deviations for

three separate tests,

Figure 7. Mean reproduction of Ceriodaphnia

dubia

in 3-brood, static-renewal, chronic toxicity tests with sulfate in

Moderately Hard Reconstituted Water (MHRW). Bars and error bars indicate means and standard deviations for

three separate tests.

---

Figure 8. Mean percent survival of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests with sulfate

in Reformulated Moderately Hard Reconstituted Water (RMHRW) . Bars and error bars indicate means and standard

deviations for three separate tests .

INNS

April 10, 2005

Page 15 of 16

Figure 9. Mean reproduction of Ceriodaphnia dubia in 3-brood, static-renewal, chronic toxicity tests with sulfate in

Reformulated Moderately Hard Reconstituted Water (RMHRW) . Bars and error bars indicate means and standard

deviations for three separate tests.

---

INHS

April 10, 2005

Page 16 of 16

Literature cited

American Public Health Association (APHA), American Water Works Association, Water

Environment Federation. 1998. Standard methods for the examination of water and

wastewater, 20th ed. American Public Health Association, Washington, DC

.

ASTM (American Society for Testing and Materials). 2002a. Standard guide for conducting

acute toxicity tests on test materials with fishes, macroinvertebrates, and amphibians

.

E729-96. American Society for Testing and Materials, Philadelphia, PA, USA

.

ASTM (American Society for Testing and Materials) . 2002b. Standard guide for conducting

three-brood, renewal toxicity tests with

Ceriodaphnia dubia .

E1295-01 . American

Society for Testing and Materials, Philadelphia, PA, USA

.

Bayliss D, Harris RR. 1988. J Comp Physiol 158 :81-90

Dwyer FJ, Burch SA, Ingersoll CG, and Hunn JB . 1992. Environ Toxicol Chem 11 :513-520 .

Ingersoll CG, Dwyer FJ, Burch SA, Nelson MK, Buckler DR, and Hunn JB . 1992. Environ

Toxicol Chem 11:503-511 .

Lucu C, and Flik G . 1999. Am J Physiol 276:R490-R499 .

McMahon, RF, and Bogan, AE . 2001. Mollusca: Bivalvia, in JH Thorp and AP Covich (eds

.)

Ecology and Classification of North American Freshwater Invertebrates, 2 nd Edition

.

Academic Press, San Diego, CA . pp 352-353

.

Mount DR, Gulley DD, Hockett JR, Garrison TD, and Evans JM . 1997. Environ Toxicol Chem

16:2009-2019 .

Pic P, and Maetz J. 1981 . J Comp Physiol B 141 :511-521

.

Soucek, DJ. 2004. Effects of hardness, chloride, and acclimation on the acute toxicity of sulfate

to freshwater invertebrates . Final Report to: Illinois Environmental Protection Agency

and Illinois Coal Association

.

Smith ME, Lazorchak JM, Herrin LE, Brewer-Swartz S, and Thoeny WT . 1997. Environ

Toxicol Chem 16:1229-1233

.

U.S. EPA . 1985. Guidelines for deriving numerical water quality criteria for the protection of

aquatic organisms and their uses. PB85-227049. Washington, DC.

U.S. EPA . 1993. Methods for measuring the acute toxicity of effluents and receiving waters to

freshwater and marine organisms, 4 th ed. EPA/600/4-90/027F, Cincinnati, OH .

U.S. EPA. 1994a. Methods for measuring the toxicity and bioaccumulation of sediment

associated contaminants with freshwater invertebrates . EPA/600/R-94/024. U.S .

Environmental Protection Agency, Washington, DC .

U.S. EPA. 1994b. Methods for the determination of metals in environmental samples. EPA/600/R-

4/111 . U.S. Environmental Protection Agency, Cincinnati, OH

.

U.S. EPA. 1999. National Recommended Water Quality Criteria - Correction. EPA 822-Z-99-

001, Washington, DC

.

Zipper CE. 2000. Coal mine reclamation, acid mine drainage and the Clean Water Act, in R

.

Barnhisel, W . Daniels, and R. Darmody (eds) Reclamation of Drastically Disturbed

Lands. American Society of Agronomy, Madison, WI .

---

Effects of Water Quality on Acute and Chronic Toxicity of Sulfate

to Freshwater Bivalves, Ceriodaphnia dubia, and Hyalella azteca

.

Third Quarterly Report

Submitted to :

Edward Hammer and Dertera Collins

United States Environmental Protection Agency

Region 5, Water Division, 77 West Jackson Boulevard

Chicago, Illinois 60604

October 20, 2005

Illinois Natural History Survey, Champaign, IL

U.S. EPA Region 5

INHS

October 20, 2005

Page 1 of 22

---

Background

While there are no Federal water quality criteria (WQC) for the protection of freshwater

life for total dissolved solids (TDS), sulfate, or sodium (U.S. EPA 1999), several states,

including Minnesota, Indiana, and Illinois, are at various stages in the process of

developing standards for sulfate. Water quality standards are developed to protect

designated uses, aquatic life uses in this case, but the economic impacts of these standards

are important considerations as well . For example, after existing sulfate standards were

enacted in Illinois, the Illinois Pollution Control Board adopted exceptions to the

standards to provide relief for a number of coalmines that were enduring severe economic

hardship. In fact, -60% of coalmines have expired permits in Illinois because of

violations of the sulfate standard, and -50% of those have been expired for more than

three

years (T .

Frevert, pers. comm.). The current "General Use" standard of 500 mg/L in

Illinois is based on the value thought to be protective of livestock. Consultation with

appropriate authorities revealed that livestock were capable of tolerating much higher

levels of sulfate. In light of these factors, the Illinois Environmental Protection Agency

(IL EPA) is actively pursuing an update of the sulfate standards based on scientific

research, and is close to proposing an updated standard (R . Mosher, IL EPA, pers

.

comm.) .

Sodium is one of the most common major cations in high TDS effluents, but calcium and

chloride are usually present in mine-impacted waters as well

. While major ion or TDS

toxicity is caused by osmoregulatory stress from the combination of all cations and

anions, chloride standards currently exist, and Illinois plans to additionally regulate for

sulfate in order to address the major non-chloride component of TDS in these waters .

Therefore, studies (funded by IL EPA, and IL Coal Association) were conducted by

Soucek (2004; In Press Environmental Toxicology and Chemistry) to (1) generate LC50s

(lethal concentration to 50% of a sample population) and LC1Os (lethal concentration to

10% of a sample population) for sulfate with selected freshwater invertebrates

(Ceriodaphnia dubia, Chironomus tentans, Hyalella azteca, and Sphaerium simile) in

U.S. Environmental Protection Agency's (US EPA, 1993) moderately hard reconstituted

water (MHRW) and (2) determine the effects of laboratory water composition, water

hardness, and test organism acclimation on the acute toxicity of sulfate to

Ceriodaphnia

dubia and Hyalella azteca (Soucek, 2004). In these previous studies (Soucek, 2004), the

mean LC50s, expressed as mg S04 2-/L, in moderately hard, reconstituted water (MHRW ;

U.S. EPA, 1993) ranged from: 512 to 1,4134 mg/L The LC50 generated for the

amphipod, Hyalella (512 mg/L) was surprisingly low, given that it is known as a

euryhaline organism (Ingersoll et al ., 1992), but as will be discussed below, water quality

data, including other cations and anions present, are critical for predicting the responses

of freshwater organisms (especially Hyalella) to elevated sulfate concentrations

.

INNS

October 20, 2005

Page 2 of 22

---

INHS

October 20, 2005

Page 3 of 22

The composition of the dilution water used during testing in the Soucek (2004) study had

a dramatic effect on the toxicity of sulfate to

Hyalella .

Whereas the 96-hour LC50 in

MHRW was 512 mg/L, the LC50 increased to 2,855 mg/L when using a "Reformulated

Moderately Hard Reconstituted Water" (RMHRW, Smith et al ., 1997). The LC50 for C .

dubia

also increased from 2,050 in MHRW to 2,526 mg/L in RMHRW . Both dilution

waters were similar in terms of hardness (-90-106 mg/L as CaCO3), alkalinity, and pH,

but RMHRW had a higher chloride concentration and different calcium to magnesium

ratio than that in MHRW. An additional experiment conducted at the Illinois Natural

History Survey (INHS) laboratory, but not included in the Soucek (2004) report, indicated

that when sulfate (-2,800 mg/L) and hardness (106 mg/L) were held constant, percent

survival of

H. azteca

was positively correlated with chloride concentration (up to 67 mg

CE/L). These experiments illustrate the need to further characterize the interacting effects

of chloride and sulfate on aquatic organisms .

Another factor that appears to have a strong effect on the toxicity of sulfate is the

presence of other major cations, in this case, calcium and magnesium, measured as

hardness. In the previous study (Soucek 2004), increased hardness reduced the toxicity of

sulfate to

Hyalella

and had a dramatic effect on the 48-hour LC50 for C

.

dubia,

increasing from 2,050 at a hardness of 90 to > 2,900 mg/L at hardness values higher than

194 mg/L as CaCO3. Others have observed reduced toxicity of saline solutions due to

increased hardness as well (e.g ., Dwyer et al. 1992; Mount et al. 1997) .

While a great deal of progress was made in the understanding of sulfate toxicity under

varying water quality conditions, several important data gaps remain . In the previous

studies (Soucek 2004), the fingernail clam,

Sphaerium simile,

had a lower LCIO than that

of C.

dubia,

but because of the temporal nature of its availability, this bivalve was only

tested in MHRW. It remains unclear whether or not a mollusk will have the same

physiological response as two crustaceans to increased chloride or hardness in these

experiments with sulfate. The principal inorganic anion of crustacean blood, or

hemolymph, is chloride, and it has been suggested that low chloride concentrations may

limit the distribution of at least one euryhaline amphipod

(Corophium curvispinum)

in

freshwaters (Bayliss and Harris, 1988). However, in the unionid mussel,

Toxolasma

lexasiensis,

chloride and bicarbonate are equally important anions in the hemolymph (see

McMahon and Bogan 2001). Because bicarbonate is readily available via respiration and

metabolism, this mussel may not depend on external chloride concentrations to the extent

that some crustaceans do . If this is the case, the protective effect of chloride observed for

Hyalella

and

Ceriodaphnia

might not be manifest in some unionoidean bivalves . The

hardness effect observed in the Soucek (2004) study may be more widespread among

aquatic phyla, because calcium simply reduces gill permeability (Lucu and Flik, 1999 ; Pic

and Maetz, 1981). However, McMahon and Bogan (2001) state that unionoideans

"generally lose capacity for osmotic and volume regulation above 3-4 ppt" (salinity)

.

TDS is a rough measure of salinity, and the TDS of a sample of RMHRW with 2000

---

INNS

October 20, 2005

Page 4 of 22

mg/L sulfate is 2.9 g/L or ppt (Soucek unpublished data). Further experiments with

freshwater bivalves are required to determine if there is an absolute TDS level that is

tolerable, or if the limit depends upon water quality characteristics such as chloride

concentration and hardness

.

An additional data gap is the fact that all of the tests conducted in the Soucek (2004)

study were acute exposures of 48 to 96 hours . Sublethal effects of sulfate in longer-term

exposures are unknown to this date. Given that, under the new limit proposed by IL EPA,

a continuous, long-term release at a given concentration will be allowed chronic testing

should be conducted to determine potential sublethal impacts

.

The purpose of the present study is to further provide data to support an appropriate

sulfate criterion for the protection of aquatic life in Illinois . Therefore, the objectives of

the current study are to build on previous studies conducted to support development of a

sulfate criterion for protection of aquatic life by (1) determining the effects of hardness on

toxicity of sulfate to bivalves, (2) determining the toxicity of sulfate to juvenile unionid

mussels, (3) determining the short-term (7 days) chronic toxicity of sulfate to

Ceriodaphnia dubia,

(4) determining the effects of chloride on acute toxicity of sulfate to

Hyalella

and

Ceriodaphnia,

and (5) determining the effects of hardness on toxicity of

sulfate to

Hyalella

at a critical chloride level,

i .e ., the chloride concentration at which

sulfate is significantly less toxic to

Hyalella

as determined in #4 above

.

Project Objectives

Based on existing data gaps described above, and because of the desire to expedite the

process of updating sulfate standards in several states with limits based on scientific

research, the following tasks will be conducted :

Examine hardness and chloride effects on acute sulfate toxicity to bivalves .

The first experiments conducted in this task will include testing the acute toxicity of

sulfate to

Sphaerium simile

in RMHRW (Smith et al ., 1997). The other two organisms

tested in this water had markedly different responses compared to those observed in

MHRW; whereas the 96-hour LC50 in MHRW for

Hyalella

was 512 mg/L, the LC50

increased to 2,855 mg/L in RMHRW. The LC50 for C.

dubia

also increased from 2,050

in MHRW to 2,526 mg/L in RMHRW . However, both of these organisms are

crustaceans. Sulfate LC50s will be generated for S

.

simile

at two chloride concentrations

(5 and 30 mg/L) and three hardness levels (100, 200, 300 mg/L). Because they will be

conducted with organisms collected from the field, some of the tests may be eliminated

depending on availability of test organisms at the collection site. In addition, we will

generate sulfate LC50s using freshwater unionid mussel juveniles as test organisms, to

---

INHS

October 20, 2005

Page 5 of 22

determine if this family is more sensitive to sulfate than the family Sphaeriidae (fingernail

clams) .

Conduct 7-day chronic sulfate toxicity tests withCeriodaphnia dubia .

To test the hypothesis that the acute safe level is similar to the chronic safe level, we will

conduct 7-day chronic, survival/reproduction tests with Ceriodaphnia dubia in both

MHRW and RMHRW. Tests will be conducted according to ASTM methods (2002b),

and endpoints generated will include the number of young (both live and dead recorded

separately) produced by each first generation C. dubia, and survival of first generation C .

dubia .

EC20s and EC50s for survival and reproduction will be generated as will No

Observable Adverse Effects Concentrations (NOAEC) and Least Observable Adverse

Effects Concentrations (LOAEC) for both endpoints

.

Determine the effects of chloride on acute toxicity of sulfate toHvalellaand

Ceriodaphnia .

Chloride standards currently exist in Illinois (500 mg1L, R. Mosher, pers. comm.), and the

State plans to additionally regulate for sulfate in order to address the major non-chloride

component of TDS in these waters . Because IL EPA proposes to eliminate the TDS

standard, opting instead to regulate sulfate and chloride, the interacting toxic effects of

these two anions must be characterized . Soucek (manuscript accepted with minor

revisions) has already shown that incrementally increasing chloride from 5 to 67 mg/L

reduces sulfate toxicity to Hyalella . We will further characterize this interaction by

determining sulfate LC50s over a wide range of chloride concentrations (10, 15, 20, 25,

100, 300, 500 mg/L) with hardness held constant. These tests will be conducted with

bothH. azteca and C. dubia .

Test effects of hardness on toxicity of sulfate to Hvalellaat critical chloride level .

Having determined the critical chloride concentration required for

Hyalella to show some

resistance to sulfate (task 3), we will determine how hardness affects the toxicity of

sulfate to Hyalella at this chloride concentration by generating LC50s at hardness values

of 100, 200, 300, 400, and 500 mg/L (as CaCO3) and a constant chloride concentration .

Methods

Invertebrates selected for testing include Ceriodaphnia dubia, Hyalella azteca, Sphaerium

simile (Pelecypoda, Sphaeriidae), and a juvenile freshwater unionid mussel (species to be

determined). The cladoceran, Ceriodaphnia dubia, was cultured in-house (Soucek

laboratory, INHS) according to U .S. EPA (1993) methods . Prior to testing, C. dubia were

fed a diet ofPseudokirchneriella subcapitata (also known as Raphidocelis subcapitata or

Selenastrum capricornutum) and a Yeast-Cereal Leaves-Trout Chow (YCT) mixture at a

rate of 0.18 ml each per 30-m1 water, daily . Cultures were maintained at 25 °C, and a 16:8

---

INHS

October 20, 2005

Page 6 of 22

(L:D)

photoperiod in Moderately Hard Reconstituted Water (MHRW ; U.S. EPA, 1993) .

Amphipods,

Hyalella azteca, w

---

Effects of Water Quality on Acute and Chronic Toxicity of Sulfate

to Freshwater Bivalves, Ceriodaphnia dubia, and Hyalella azteca .

Third Quarterly Report

Submitted to :

Edward Hammer and Dertera Collins

United States Environmental Protection Agency

Region 5, Water Division, 77 West Jackson Boulevard

Chicago, Illinois 60604

October 20, 2005

Illinois Natural History Survey, Champaign, IL

U.S. EPA Region 5

INHS

October 20, 2005

Page I of 22

Illinois EPA Exhibit No

.

C

---

Background

While there are no Federal water quality criteria (WQC) for the protection of freshwater

life for total dissolved solids (TDS), sulfate, or sodium (U .S. EPA 1999), several states,

including Minnesota, Indiana, and Illinois, are at various stages in the process of

developing standards for sulfate . Water quality standards are developed to protect

designated uses, aquatic life uses in this case, but the economic impacts of these standards

are important considerations as well . For example, after existing sulfate standards were

enacted in Illinois, the Illinois Pollution Control Board adopted exceptions to the

standards to provide relief for a number of coalmines that were enduring severe economic

hardship. In fact, -60% of coalmines have expired permits in Illinois because of

violations of the sulfate standard, and -50% of those have been expired for more than

three years (T. Frevert, pers. comm .). The current "General Use" standard of 500 mg/L in

Illinois is based on the value thought to be protective of livestock. Consultation with

appropriate authorities revealed that livestock were capable of tolerating much higher

levels of sulfate. In light of these factors, the Illinois Environmental Protection Agency

(IL EPA) is, actively pursuing an update of the sulfate standards based on scientific

research, and is close to proposing an updated standard (R . Mosher, IL EPA, pers .

comm.)

.

Sodium is one of the most common major cations in high TDS effluents, but calcium and

chloride are usually present in mine-impacted waters as well . While major ion or TDS

toxicity is caused by osmoregulatory stress from the combination of all cations and

anions, chloride standards currently exist, and Illinois plans to additionally regulate for

sulfate in order to address the major non-chloride component of TDS in these waters

.

Therefore, studies (funded by IL EPA, and IL Coal Association) were conducted by

Soucek (2004; In Press Environmental Toxicology and Chemistry) to (1) generate LC50s

(lethal concentration to 50% of a sample population) and LCIOs (lethal concentration to

10% of a sample population) for sulfate with selected freshwater invertebrates

(Ceriodaphnia dubia, Chironomus tentans, Hyalella azteca,

and

Sphaerium simile)

in

U .S. Environmental Protection Agency's (US EPA, 1993) moderately hard reconstituted

water (MHRW) and (2) determine the effects of laboratory water composition, water

hardness, and test organism acclimation on the acute toxicity of sulfate to

Ceriodaphnia

dubia

and

Hyalella azteca

(Soucek, 2004). In these previous studies (Soucek, 2004), the

mean LC50s, expressed as mg S04 2"/L, in moderately hard, reconstituted water (MHRW ;

U.S. EPA, 1993) ranged from: 512 to 1,4134 mg/L The LC50 generated for the

amphipod,

Hyalella

(512 mg/L) was surprisingly low, given that it is known as a

euryhaline organism (Ingersoll et al ., 1992), but as will be discussed below, water quality

data, including other cations and anions present, are critical for predicting the responses

of freshwater organisms (especially

Hyalella)

to elevated sulfate concentrations .

INNS

October 20, 2005

Page 2 of22

---

INHS

October 20, 2005

Page

3

of 22

The composition of the dilution water used during testing in the Soucek (2004) study had

a dramatic effect on the toxicity of sulfate to Hyalella

. Whereas the 96-hour LC50 in

MHRW was 512 mg/L, the LC50 increased to 2,855 mg/L when using a "Reformulated

Moderately Hard Reconstituted Water" (RMHRW, Smith et al ., 1997). The LC50 for C .

dubia also increased from 2,050 in MHRW to 2,526 mg/L in RMHRW. Both dilution

waters were similar in terms of hardness (-90-106 mg/L as CaCO3), alkalinity, and pH,

but RMHRW had a higher chloride concentration and different calcium to magnesium

ratio than that in MHRW. An additional experiment conducted at the Illinois Natural

History Survey (INNS) laboratory, but not included in the Soucek (2004) report, indicated

that when sulfate (-2,800 mg/L) and hardness (106 mg/L) were held constant, percent

survival ofH. azteca was positively correlated with chloride concentration (up to 67 mg

Cl -/L). These experiments illustrate the need to further characterize the interacting effects

of chloride and sulfate on aquatic organisms .

Another factor that appears to have a strong effect on the toxicity of sulfate is the

presence of other major cations, in this case, calcium and magnesium, measured as

hardness. In the previous study (Soucek 2004), increased hardness reduced the toxicity of

sulfate to Hyalella and had a dramatic effect on the 48-hour LC50 for C

. dubia,

increasing from 2,050 at a hardness of 90 to > 2,900 mg/1- at hardness values higher than

194 mglL as CaCO3. Others have observed reduced toxicity of saline solutions due to

increased hardness as well (e.g., Dwyer et al. 1992; Mount et al . 1997) .

While a great deal of progress was made in the understanding of sulfate toxicity under

varying water quality conditions, several important data gaps remain . In the previous

studies (Soucek 2004), the fingernail clam, Sphaerium simile, had a lower LC10 than that

of C. dubia, but because of the temporal nature of its availability, this bivalve was only

tested in MHRW. It remains unclear whether or not a mollusk will have the same

physiological response as two crustaceans to increased chloride or hardness in these

experiments with sulfate. The principal inorganic anion of crustacean blood, or

hemolymph, is chloride, and it has been suggested that low chloride concentrations may

limit the distribution of at least one euryhaline amphipod

(Corophium

curvispinum) in

freshwaters (Bayliss and Harris, 1988) . However, in the unionid mussel, Toxolasma

texasiensis, chloride and bicarbonate are equally important anions in the hemolymph (see

McMahon and Bogan 2001). Because bicarbonate is readily available via respiration and

metabolism, this mussel may not depend on external chloride concentrations to the extent

that some crustaceans do. If this is the case, the protective effect of chloride observed for

Hyalella and Ceriodaphnia might not be manifest in some unionoidean bivalves . The

hardness effect observed in the Soucek (2004) study may be more widespread among

aquatic phyla, because calcium simply reduces gill permeability (Lucu and Flik, 1999

; Pic

and Maetz, 1981). However, McMahon and Bogan (2001) state that unionoideans

"generally

lose capacity for osmotic and volume regulation above 3-4 ppt" (salinity)

.

TDS is a rough measure of salinity, and the TDS of a sample of RMHRW with 2000

---

INNS

October 20, 2005

Page 4 of 22

mg/L sulfate is 2 .9 g/L or ppt (Soucek unpublished data) . Further experiments with

freshwater bivalves are required to determine if there is an absolute TDS level that is

tolerable, or if the limit depends upon water quality characteristics such as chloride

concentration and hardness.

An additional data gap is the fact that all of the tests conducted in the Soucek (2004)

study were acute exposures of 48 to 96 hours. Sublethal effects of sulfate in longer-term

exposures are unknown to this date . Given that, under the new limit proposed by IL EPA,

a continuous, long-term release at a given concentration will be allowed chronic testing

should be conducted to determine potential sublethal impacts

.

The purpose of the present study is to further provide data to support an appropriate

sulfate criterion for the protection of aquatic life in Illinois . Therefore, the objectives of

the current study are to build on previous studies conducted to support development of a

sulfate criterion for protection of aquatic life by (1) determining the effects of hardness on

toxicity of sulfate to bivalves, (2) determining the toxicity of sulfate to juvenile unionid

mussels, (3) determining the short-term (7 days) chronic toxicity of sulfate to

Ceriodaphnia dubia, (4) determining the effects of chloride on acute toxicity of sulfate to

Hyalella and Ceriodaphnia, and (5) determining the effects of hardness on toxicity of

sulfate to Hyalella at a critical chloride level, i .e., the chloride concentration at which

sulfate is significantly less toxic to Hyalella as determined in #4 above

.

Project Objectives

Based on existing data gaps described above, and because of the desire to expedite the

process of updating sulfate standards in several states with limits based on scientific

research, the following tasks will be conducted :

Examine hardness and chloride effects on acute sulfate toxicity to bivalves

.

The first experiments conducted in this task will include testing the acute toxicity of

sulfate to Sphaerium simile in RMHRW (Smith et al., 1997). The other two organisms

tested in this water had markedly different responses compared to those observed in

MHRW; whereas the 96-hour LC50 in MHRW for Hyalella was 512 mg/L, the LC50

increased to 2,855 mg/L in RMHRW . The LC50 for C. dubia also increased from 2,050

in MHRW to 2,526 mg/L in RMHRW. However, both of these organisms are

crustaceans. Sulfate LC50s will be generated for S. simile at two chloride concentrations

(5 and 30 mg/L) and three hardness levels (100, 200, 300 mg/L). Because they will be

conducted with organisms collected from the field, some of the tests may be eliminated

depending on availability of test organisms at the collection site. In addition, we will

generate sulfate LC50s using freshwater unionid mussel juveniles as test organisms, to

---

INNS

October 20. 2005

Page 5 of 22

determine if this family is more sensitive to sulfate than the family Sphaeriidae (fingernail

clams) .

Conduct 7-day chronic sulfate toxicity tests with

Ceriodaphnia dubia .

To test the hypothesis that the acute safe level is similar to the chronic safe level, we will

conduct 7-day chronic, survival/reproduction tests with

Ceriodaphnia dubia

in both

MHRW and RMHRW. Tests will be conducted according to ASTM methods (2002b),

and endpoints generated will include the number of young (both live and dead recorded

separately) produced by each first generation C.

dubia,

and survival of first generation C.

dubia .

EC20s and EC50s for survival and reproduction will be generated

as will No

Observable Adverse Effects Concentrations (NOAEC) and Least Observable Adverse

Effects Concentrations (LOAEC) for both endpoints .

Determine the effects of chloride on acute toxicity of sulfate to

Hvalella

and

Ceriodaphnia .

Chloride standards currently exist in Illinois (500 mg/L, R. Mosher, pers. comm.), and the

State plans to additionally regulate for sulfate in order to address the major non-chloride

component of TDS in these waters. Because IL EPA proposes to eliminate the TDS

standard, opting instead to regulate sulfate and chloride, the interacting toxic effects of

these two anions must be characterized . Soucek (manuscript accepted with minor

revisions) has already shown that incrementally increasing chloride from 5 to 67 mg/L

reduces sulfate toxicity to

Hyalella

.

We will further characterize this interaction by

determining sulfate LC50s over a wide range of chloride concentrations (10, 15, 20, 25,

100, 300,500 mg/L) with hardness held constant . These tests will be conducted with

both H.

azteca

and C.

dubia .

Test effects of hardness on toxicity of sulfate to

Hyalella

at critical chloride level

.

Having determined the critical chloride concentration required for

Hyalella

to show some

resistance to sulfate (task 3), we will determine how hardness affects the toxicity of

sulfate to

Hyalella

at this chloride concentration by generating LC50s at hardness values

of 100, 200, 300, 400, and 500 mg/L (as CaCO3) and a constant chloride concentration

.

Methods

Invertebrates selected for testing include

Ceriodaphnia dubia, Hyalella azteca, Sphaerium

simile

(Pelecypoda, Sphaeriidae), and a juvenile freshwater unionid mussel (species to be

determined). The cladoceran,

Ceriodaphnia dubia, was

cultured in-house (Soucek

laboratory, INHS) according to U .S. EPA (1993) methods. Prior to testing, C .

dubia

were

fed a diet of

Pseudokirchneriella subcapitata

(also known as

Raphidocelis subcapitata

or

Selenastrum capricornutum)

and a Yeast-Cereal Leaves-Trout Chow (YCT) mixture at a

rate of 0.18 ml each per 30-m1 water, daily. Cultures were maintained at 25 °C, and a 16 :8

---

INHS

October 20, 2005

Page 6 of 22

(L:D)

photoperiod in Moderately Hard Reconstituted Water (MHRW; U.S. EPA, 1993) .

Amphipods,

Hyalella azteca,

were cultured in-house (Soucek laboratory, INHS) according

to U.S. EPA (1994a) methods in RMHRW at 22 °C and a 16:8 (L:D) photoperiod for at

least 7 d prior to testing .

Hyalella

were fed a diet of

Pseudokirchneriella subcapitata

and

a Yeast-Cereal Leaves-Trout Chow (YCT) mixture as well. Sphaeriid clams were and will

be collected from Spring Creek, near Loda, Illinois, (Iroquois County) and acclimated to

MHRW at 22 °C and a 16:8 (L:D) photoperiod for 5-7 d prior to testing . Clams collected

from this site were previously identified to species by Dr. Gerald Mackie, of the

University of Guelph, Department of Zoology, Guelph, Ontario, Canada

.

For toxicity testing, a pure (99%) grade of anhydrous Na2SO4 served as the source of

sulfate . A concentrated solution of this salt as well as a sample of laboratory deionized

water, will be acidified to pH <2.0 and analyzed for priority metal concentrations at the

Illinois State Water Survey (Champaign, IL) using inductively coupled plasma-atomic

emission spectrometry according to U.S. EPA (1994b) methods to determine if samples

are contaminated with trace metals .

For static, non-renewal acute toxicity tests, conducted according to ASTM E729-96

methods (2002a), treatments are comprised of a 75% dilution series (i.e ., the 100%

concentration is serially diluted by 25%), rather than the standard 50%, because major ion

toxicity tests often cause 100% mortality in one concentration and 0% mortality in the

next highest concentration if the spread is too great. For the C.

dubia

and

H. azteca

tests,

five to six concentrations were tested, with four replicates tested per concentration, five

organisms per replicate . Tests with C.

dubia

were conducted for 48 h with a 16:8 (L:D)

photoperiod at 25 °C

.

H. azteca

and S.

simile

were exposed for 96 h at 22 °C and a 16 :8

(L:D) photoperiod .

C.

dubia

and

H. azteca

were exposed in 50-m1 glass beakers with 5

organisms per beaker, and for

H. azteca, 1

g of quartz sand was added to each beaker to

serve as substrate. Clam tests were and will be conducted in 150-m1 glass beakers (no

substrate). All clams used are juveniles .

C.

dubia

used in tests were less than 24-h old,

and

H. azteca

were -third instar (7 - 14 d old). Percent survival in each replicate was

recorded every 24 h and at the end of the exposure period. A dissecting microscope was

used to assess survival of

Hyalella

and

Simile.

For results to be acceptable, controls must

have had at least a 90% survival rate .

Chronic testing will be conducted according to guidelines described in ASTM E 1295-01

(2002). Ten replicates will be used per sulfate concentration with one organism per

replicate. Endpoints will include the number of young (both live and dead recorded

separately) produced by each first generation C

.

dubia,

and survival of first generation C .

dubia .

Standard water chemistry parameters were measured at both the beginning and the end of

each exposure period, including temperature, pH, conductivity, dissolved oxygen,

---

INNS

October 20, 2005

Page 7 of 22

alkalinity and hardness. The pH measurements were made using an Accumet®(Fisher

Scientific, Pittsburgh, PA, USA) model AB15 pH meter equipped with an Accumet ® gel-

filled combination electrode (accuracy c ± 0 .05 pH at 25 °C). Dissolved oxygen was

measured using an air-calibrated Yellow Springs Instruments (RDP, Dayton, OH, USA)

model 58 meter with a self-stirring BOD probe . Conductivity measurements

were

made

using a Mettler Toledo (Fisher Scientific, Pittsburgh, PA, USA) model MC226

conductivity/TDS meter. Alkalinity, and hardness were measured (beginning of tests

only) by titration as described in APHA et al . (1998). Samples from each treatment will

be analyzed to confirm sulfate concentrations by ion chromatography at the INHS Aquatic

Chemistry Laboratory, Champaign, IL

.

LC50 values were calculated using the Spearman-Karber method . To increase confidence

in LC50 values, three assays weretwill be conducted for each objective

(i.e., either

organism or chloride X hardness combination) . This will provide a stronger estimate of

the mean LC50 value for each species .

In experiments testing effects of hardness of sulfate toxicity, hardness will be increased by

adding enough CaSO4and MgSO4 to achieve the nominal hardness values. Then Na2SO4

will be added as was done with the standard MHRW. Whole carboys will be made at each

elevated hardness level and this water will used as both diluent and control ; therefore,

each concentration within a given test will have the same hardness

(i.e., [Ca2r] and [Mg2+ ]

will not change with dilution) . In experiments testing the influence of chloride on the

toxicity of sulfate to

H. azteca,

chloride, as NaCl, was added at appropriate concentrations

to solutions with a hardness of -100 mg/L .

Progress to date :

Hardness and chloride effects on acute sulfate toxicity to bivalves

.

Five toxicity tests were conducted with 4-day-old, juvenile unionid mussels

(fatmuckets,Lampsilis

siliquoidea)

at the U.S. Geological Survey's Columbia Environmental

Research Center. Tests were conducted according to newly developed standard methods (ASTM

2005), and five different diluents were used with varying levels of hardness and chloride

.

Because of constraints with availability of organisms, one test was conducted for each diluent

type. Test water #4 was U.S. EPA's MHRW (USEPA 1993), and the 96-h LC50 in this water

(1,727 mg/L) was fairly similar to, but slightly lower than LC50s for

C. dubia

(2,050 mg/L), and

the fingernail clam, S .

simile

(2,078 mg/L) in MHRW. When Ca:Mg and chloride concentration

were held constant, increased hardness appeared to slightly reduce the toxicity of sulfate to the

juvenile mussels (Table 1), but confidence intervals overlapped . Chloride did not appear to have

a strong effect either, but increasing chloride concentration from 5 to 33 increased the sulfate

LC50 by approximately 100 mg/L (Table 1). Interestingly, while hardness and chloride did not

strongly affect sulfate toxicity to freshwater mussels, Ca:Mg ratio appeared to have a substantial

---

INNS

October 20, 2005

Page 8 of 22

effect. In tests conducted with CaMg = 1 .46, 96-h LC50s ranged from 1,727 to 1,822 mg SO,/L,

while at a ratio of 2.33, LC50s ranged from 3,377 to 3,729 mg SOdL . This trend suggests that

there may be a calcium threshold for unionids, which when reached provides a certain level of

protection against sulfate toxicity, but beyond which, additional calcium provides minimal

additional protection. Further testing should be conducted to verify these trends .

Table 1. Toxicity of sulfate to juvenile fatmuckets (Lampsilis siliquoidea)

.

All tests were conducted at 20 °C for 96

hours with 4-d-old mussels. Mortality was evaluated at 48, and 96 hours. Sulfate LC50 values shown are based on

measured concentrations, and all treatments within a given test had the same chloride concentration and hardness

.

LC50s were generated using the Spearman-Karber method . One test was conducted for each diluent type.

In tests with fingernail clams

(Sphaerium simile),

increasing hardness from -100 to 200 mg/L did

not have a substantial effect on sulfate toxicity, but an additional increase to hardness = 300

mg/L caused a significant increase in the mean 96-h LC50 value (Table 2). When Ca:Mg and

hardness were held constant, increased chloride caused a substantial increase in mean LC50 from

2,186 mg/L (C1= 2 mg/L) to 2,650 (Cl = 33 mg/L) (Table 2)

.

Table 2. Effects of hardness and chloride on sulfate toxicity to fingernail clams (Sphaerium simile) . All tests were

conducted at 22 °C for 96 hours. Sulfate LC50 values shown are based on nominal concentrations, and all

treatments within a given test had the same chloride concentration and hardness . LC50s were generated using the

Examining sulfate toxicity at different hardness levels reveals that hardness appears to ameliorate

sulfate toxicity to a variety of species in at least two different phyla (Arthropoda and Mollusca)

.

Test

Cl

mg/L

Ca:Mg

ratio

hardness

mg/L (CaCO1)

48 h LC50

(95% C.I.)

96 h LC50

(95% C.I.)

1

25

2.33

100

3183(2837-3571)

3377 (3205-3558)

2

25

2.33

300

3523 (unrel. var .)

3525 (3330-3731)

3

25

2 .33

500

3574 (unrel. var.)

3729 (unrel var.)

5

1.46

100

1702 (1598-1812)

1727 (unrel var.)

33

1 .46

100

2661 (2268-3123)

1822 (unrel var.)

Spearman-Karber method

.

Different capital letters following means indicate means are significantly different (p <

0.05) .

Diluent

n

Cl

Ca:Mg

hardness

mean LC50, mg SO4/L

type

mg/L

ratio

mg/L (CaCO I)

(std. dev)

MHRW

3

2

1 .46

84

2186 (388) B

Hard 200 (1 .46)

2

2

1 .46

193

2190 (151) B

Hard 300 (1 .46)

2

2

1 .46

269

2926 (340) A

Hard 300 (2.33)

2

2

2.33

280

2590 (76) AB

RMHRW

2

33

5.4

93

2200 (30) B

CI 33/ hard 100

3

33

1

.46

90

2650 (341) AB

---

INHS

October 20, 2005

Page 9 of 22

In fact, for three of the species tested in this study, the increase in sulfate LC50 from hardness =

100 mg/L to hardness = 300 mg/L was surprisingly consistent at approximately 900 mg/L (Fig.

1) .

3500 -

al 3000 -

A

7

4

2500 -

E

0

2000 -

J

Ceriodaphnia

D hardness = 100

D hardness =

300

∎ hardness = 500

Hyalella

Sphaerium

Species

a

Lampsilis

Figure 1. Comparison of LC50s at different water hardness levels for four different species of

freshwater invertebrates tested in the laboratory.

Effects of chloride on acute toxicity of sulfate to

Hyalella

and

Ceriodaphnia .

Increasing chloride concentration from 5 to 33 mg/L results in a sharp decrease in sulfate toxicity

to

Hyalella

(table 3 and fig. 2). In the Soucek (2004) study, RMHRW had a chloride

concentration of 33, and this water resulted in a mean LC50 of 2,855 mg S04/L. The 33 mg Cl/L

treatment in this study resulted in a mean LC50 of 1,825 mg S0 4/L, but it is important to

remember that the solutions in these tests have a lower calcium concentration than in RMHRW

(23 mg/L compared to 32 mg/L in RMHRW) despite having the same hardness . In the current

study, a Ca:Mg ratio of 2.33:1 (compared to 5 .40:1 in RMHRW) was selected to reflect the

median ratio for streams in Illinois . Calcium is probably more important in mediating sulfate

toxicity than magnesium, thus accounting for the relatively large difference in LC50 values

despite nearly identical hardness values .

Above 33 mg CI/L, LC50 values rose again slightly in the 100 mg CUL treatment but then

dropped in the 300 and 500 mg Cl/L tests (Table 3 and Fig . 2). If these data are viewed in a

---

slightly different way, another interesting trend appears (Fig. 3). I compared total dissolved

solids (TDS, calculated) at LC50 concentrations for the different chloride treatments, and found

that for the lower concentrations, TDS of the LC50 concentration increased with increasing

chloride, as was the case for sulfate concentration, until a threshold was reached . Then, at the

higher chloride concentrations (100 to 500'mg/L), no further benefit was provided by chloride

above 100 mg/L. In fact, it appears that at 100 mg Cl/L, a threshold of -3 .1 g/L TDS is reached .

At higher chloride concentrations, less sulfate is required to reach the critical TDS value, so

LC50s in terms of sulfate decrease at 300 and 500 mg CUL . In other words, it appears that for

Hyalella,

in the range of 100 to 500 mg Cl/L, toxicity is reached at a fixed TDS and if there is

more chloride, less sulfate is required to reach this threshold and vice versa

.

Table 3. Influence of chloride concentration on toxicity of sulfate to Hyalella azteca.

All tests were conducted at 22

°C for 96 hours. Chloride and sulfate values shown are nominal concentrations, and all treatments within a given test

had the same chloride concentration and hardness . Mean hardness for all tests was -100 mg/L CaCO,, and the

Ca:Mg used was 2.33 :1 (mg/L:mg:L) to reflect the median Ca:Mg ratio in Illinois (Clark Olson, IEPA, pers. com.)

LC50s were generated using the Spearman-Karber method

.

2500

2000

A

7

1500

N

oa

1000-

U

',

500

0

0

100

200

300

400

500

600

S

M

chloride (mg/L)

INNS

October 20, 2005

Page 10 of 22

Chloride (mg/L)

mg/L

number of tests

Mean LC50

conducted

mg sulfate/L

10

3

1,387

15

3

1,563

20

3

1,562

25

3

1,854

33

3

1,799

100

3

1,938

300

3

1,691

500

3

1,470

---

a

rn

2.50

2.30

2.10

1.90

1.70

1 .50

0

100

200

300

400

500

600

chloride

INHS

October 20, 2005

Page 11 of 22

Figure 2. Influence of chloride concentration on toxicity of sulfate to Hyalella azteca.

See table 3 for details on test

conditions .

Figure 3. Relationship between chloride concentration and total dissolved solids at the LC50 concentration in

toxicity tests sulfate to Hyalella azteca .

Total dissolved solids values were calculated using nominal concentrations

of all ions present in solution (excluding H+ and OH-) at LC50 concentrations

.

The trend of decreased toxicity with increasing chloride at the lower end of the range as seen for

Hyalella is not as clear with Ceriodaphnia (Table 4 and Fig. 4) .

It appears that at chloride

concentrations between 10 and 100 mg/L, sulfate toxicity is fairly constant (with some noise in

the data), with LC50 values between 2,200 and 2,500 mg S04/L (nominal concentrations). In the

Soucek (2004) study, Ceriodaphnia did not respond as strikingly as Hyalella when tested in

RMHRW compared to MHRW so these results may not be entirely surprising

.

As was the case with Hyalella, LC50 values in terms of sulfate were lower at the 300 and 500 mg

Cl/L concentrations compared to the lower chloride concentrations (Table 4 and Fig . 4) .

Comparing TDS (calculated) at LC50 concentrations for the different chloride treatments at the

lower range of chloride concentrations, TDS values at LC50 concentrations were higher for

Ceriodaphnia (3 .5 - 3.8 g/L) than they were for Hyalella (2.4 - 3.1 g/L). The overall trend was

different for Ceriodaphnia as well with an overall linear trend of decreasing TDS at LC50 with

increasing chloride concentration (Fig . 5)

.

3.50

3.30-

3.10 -

2.90

y = 0.2175Ln(x) + 1 .8951

270

-

a

R' = 0.7317

---

Table 4. Influence of chloride concentration on toxicity of sulfate to Ceriodaphnia dubia. All tests were conducted

at 25 °C for 48 hours . Chloride and sulfate values shown are nominal concentrations, and all treatments within a

given test had the same chloride concentration and hardness . Mean hardness for all tests was -100 mg/L CaCO 3 , and

the Ca:Mg used was 2.33 :1 (mg/L:mg:L) to reflect the median Ca :Mg ratio in Illinois (Clark Olson, IEPA,pers .

com.) LC50s were generated using the Spearman-Karber method .

3000-

2500

C

w

2000

7

y

E

1500

v

1000

a

500 -

0

0

100

200

300

400

500

600

M

chloride (mg/L)

Figure 4. Influence of chloride concentration on toxicity of sulfate to

Ceriodaphnia dubia.

See table 2 for details on test conditions

.

INHS

October 20, 2005

Page 12 of 22

Chloride (mg/L)

mg/L

number of tests

Mean LC50

conducted

mg sulfate/L

10

3

2,469

15

3

2,289

20

3

2,419

25

3

2,272

100

3

2,417

300

3

1,914

500

3

1,496

---

0

100

200

300

400

500

600

chloride

INHS

October 20, 2005

Page 13 of 22

Figure 5. Relationship between chloride concentration and total dissolved solids at the LC50 concentration in

toxicity tests sulfate to Ceriodaphnia dubia .

Total dissolved solids values were calculated using nominal

concentrations of all ions present in solution (excluding H+ and OH-) at LCSO concentrations

.

Effects of hardness on toxicity of sulfate to

Hvalella

at critical chloride level

As was seen with

Ceriodaphnia dubia

in the Soucek (2004) study, a strong linear trend of

decreased sulfate toxicity with increased hardness was observed with

Hyalella

(Fig. 6, 7). LC50

values increased from less than 2,000 mg/L at hardness = 100 mg/L, to greater than 4,000 mg/L

at a hardness of 500 mg/L. The mean LC50 value at 600 mg/L hardness was lower than that at

500 mg/L hardness, as was the case for C

.

dubia .

It remains unclear how the trend will continue

with increasing hardness above 600 mg/L and further testing should be conducted with harder

waters .

3 .9

3 .7

-

3.5 - N

0

a

3.3

A

3 .1

2.9

2.7

y = -0.001x + 3.6964

R2 = 0.628

2.5

---

Figure 6. Influence of hardness on toxicity of sulfate to

Hyalella azteca

at chloride = 25 mg/L

INNS

October 20, 2005

Page 14 of 22

Figure 7. Influence of hardness on toxicity of sulfate to

Hyalella azteca

and C.

dubia. C. dubia

data from Soucek

2004)

.

---

7-day chronic sulfate toxicity tests with

Ceriodaphnia dubia .

All of the tests for this objective have been completed (three chronic tests in each dilution water),

and results were quite consistent from test to test . As depicted in figures 8-11, chronic toxicity

in

terms of survival and reproduction was less in RMHRW compared to MHRW . Table 5

summarizes the Least Observable Adverse Effects Concentrations (LOAEC) and No Observable

Adverse Effects Concentrations (NOAEC) for each dilution water. While LOAECs for

reproduction are quite low, it should be noted that I have had a continuous, self sustaining,

reserve culture of

Ceriodaphnia dubia

in MHRW spiked with 1,000 mg SOIL since at least

August of 2004. Organisms used in chronic testing for this study were cultured in MHRW or

RMHRW as appropriate

.

Table 5. Summary of mean NOAEC and LOAEC concentrations for survival and reproduction of

Ceriodaphnia

dubia

in 7-day static-renewal three brood toxicity tests with sulfate . Values are in terms of mg sulfate/L . All tests

were conducted at 25 °C. Chloride and sulfate values shown are nominal concentrations. MHRW = Moderately

Hard Reconstituted Water (USEPA 1993). RMHRW = Reformulated Moderately Hard Reconstituted Water (Smith

et al., 1997)

Dilution water

number of tests

conducted

Survival Reproduction

NOAEC

LOAEC

NOAEC

LOAEC

MHRW

RMHRW

3

3

1,727

2,264

2,273

3,000

780

906

934

1,195

INNS

October 20, 2005

Page 15 of 22

Figure 8. Mean percent survival of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests with sulfate

in Moderately Hard Reconstituted Water (MHRW) . Bars and error bars indicate means and standard deviations for

three separate tests

.

---

INHS

October 20, 2005

Page 16 of 22

Figure 9. Mean reproduction of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests with sulfate in

Moderately Hard Reconstituted Water (MHRW) . Bars and error bars indicate means and standard deviations for

three separate tests

.

Figure 10. Mean percent survival of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests with

sulfate in Reformulated Moderately Hard Reconstituted Water (RMHRW). Bars and error bars indicate means and

standard deviations for three separate tests

.

---

INNS

October 20, 2005

Page 17 of 22

Figure 11. Mean reproduction of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests with sulfate

in Reformulated Moderately Hard Reconstituted Water (RMHRW). Bars and error bars indicate means and standard

deviations for three separate tests.

Comparison of LC50s for fed and unfed organisms

As an additional analysis, I calculated 48-h LC50s using the data generated in the 7-day chronic

tests to determine if feeding of test organisms has an effect on sulfate toxicity . Others have

shown that feeding algae to cladocerans during toxicity tests may reduce the toxicity of metals

because negative charges on algal cells bind and reduce the bioavailability of positively charged

metals (e.g ., Taylor et al. 1998). In this study, sulfate toxicity to C .

dubia,

was significantly

reduced when organisms were fed in both MHRW and RMHRW (Fig . 12). At this point, it is

unknown if this observed effect is due to binding of sodium to algal cells, or increased robustness

or lowered stress of test organisms that were fed during testing . The most likely explanation is

that of increased health of test organisms because sodium and sulfate are highly stable as dissolve

ions, but this is conjecture at this point

.

---

INNS

October 20, 2005

Page 18 of 22

unfed MHRW

treatment

fed MHRW

unfed RMHRW

fed RMHRW

Figure 12. Effects of presence of food on toxicity of sulfate to

Ceriodaphnia dubia

in two types of test waters

.

Different capital letters indicate that means are significantly different (p<0 .05) .

LCIO/LC50

LC10s were calculated for all the acute tests conducted for this study, and LCIO values were

divided by their respective LC50s to determine the steepness of the toxicity curves for sulfate for

the organisms tested, The mean quotients varied from species to species, but generally were

quite high, ranging from 0 .72

0

0.85 (Fig. 13) .

3500

3000

2500 -

C

2000

-

U

1500 -

a

s

00

1000

-

v

500

---

1.00

0.95

0.90

-

0

0.85

-

U

0

0.80

-

r.

U

a

0.75

-

0.70

-

0.65 -

0.60

A

f

0.792

BA

0.717

B

A

f

0.849

C. dubia

H. azteca

S. simile

C. tentans

INNS

October 20, 2005

Page 19 of 22

Figure 13. Mean quotients of LCIO/LC50 for four species exposed to sulfate in this study . Different capital letters

indicated that means are significantly different (p<0 .05) .

Multiple linear regression analysis

Multiple linear regression analysis was used to generate equations to predict LC50s and LC 1Os

based on the hardness and chloride concentrations of potential receiving waters . Before analyses

were conducted, data at hardness greater than 490 and data for

Hyalella

at chloride less than 25

were deleted, because of the uncertainty in trends above hardness =490 and below Cl =25. In

order to avoid the assumption that all of the species have the same sensitivities, linear regression

with covariance was used

(i.e ., "species" was introduced as a categorical variable); this model

assumes that the species have the same slopes for LC50 vs. hardness and for LC50 vs, chloride,

but do not necessarily have the same sensitivities to sulfate . Equations were generated both

including and excluding the two data points for C.

tentans,

which had drastically different values

compared to the other three species included in the models

.

Equations along with R2 values and intercept and slope estimates are given in appendices 1 and

2. Further analysis and interpretation of these results will be included in the final report for this

project .

Remaining tasks- Approximately 50% of the water samples from tests conducted for this study

have been analyzed by ion chromatography to verify nominal concentrations. When analyses are

completed, all toxicity endpoints will be recalculated using actual concentrations . Samples

analyzed to data are generally within 5% of nominal values so toxicity data are not expected to

change substantially. In addition, further analysis and interpretation of observed trends will be

included in a final report

.

---

Appendix 1.

Results of multiple linear regression analysis to generate equations to

predict LC50 based on hardness and chloride of a water body . Output were generated

using JMP-IN software

LC50 Results without "species" variable with all data points except C. tentans (n = 87)

LC50 Results with "species" variable with all data pointss except C. tentans (n = 87)

LC50 Results without "species" variable with all data points including C. tentans (n = 89)

LC50 Results with "species" variable with all data points including C. tentans (n = 89)

INHS

October 20, 2005

Page 20 of 22

R-squared = 0.964607

Parameter

Intercept

Adjusted R-squared = 0.962475

Estimate

4825.2033

Prob . > t

<0.0001

Species (C . dubia)

-2887.746

<0.0001

-2887.746

Species (H. azteca)

-3070.209

<0.0001

-3070.209

Species (S. simile)

-2979.367

<0.0001

-2979.367

Hardness

3.986

<0.0001

Chloride

-1 .4286

<0.0001

R-squared = 0.73229

Parameter

Intercept

Adjusted R-squared = 0 .719231

Estimate

1846.096

Prob. > t

<0.0001

Species (C. dubia)

91.360858

0.0958

Species (H. azteca)

-91 .1015

0.1311

Hardness

3 .986

<0.0001

Chloride

-1 .4286

<0.0001

R-squared = 0.083351

Adjusted R-squared = 0.062034

Parameter

Estimate

Prob. > t

Intercept

2589.4876

<0.0001

Hardness

2.004

0.2214

Chloride

-3 .028

0.0425

R-squared = 0.717915

Adjusted R-squared = 0.711198

Parameter

Estimate

Prob. > t

Intercept

1876.7114

<0.0001

Hardness

3 .9108

<0.0001

Chloride

-1 .5046

<0.0001

---

Appendix 2. Results of multiple linear regression analysis to generate equations to predict LCI0

based on hardness and chloride of a water body . Output were generated using JMP-IN software

LC I O Results without "species" variable with all data points except C . tentans (n = 87)

INHS

October 20, 2005

Page 21 of 22

R-squared = 0.572252

Adjusted R-squared = 0 .562068

Parameter

Intercept

Hardness

Chloride

Estimate

Prob. > t

1478

<0.0001

2.7286

<0.0001

-1.202874

0.0001

LCIO Results with"species" variable with all data points except C. tentans (n = 87)

R-squared = 0.63676

Adjusted R-squared = 0.619041

species

C. dubia

H. azteca

S. simile

estimate

1587.37

1277.442

??

Parameter

Intercept

Hardness

Chloride

C. dubia

H. azteca

Estimate

Prob. > t

1426

<0.0001

2.9485

<0.0001

-1.857

0.0003

161 .37

0.0022

-148.558

0.0097

LCIO Results without "species" variable with all data points including C

. tentans (n = 89)

R-squared =

Parameter

Intercept

Hardness

Chloride

0.069571

Adjusted R-squared = 0.047933

Estimate

Prob. > t

2056.012

1 .185

-2.4361

<0.0001

0.3757

0.0459

LCIO Results with"species" variable with all data points including C. tentans (n = 89)

R-squared = 0.952131

Adjusted R-squared = 0 .949247

Parameter

Estimate

Prob. > t

Intercept

3843.3786

<0.0001

Hardness

2.848

<0.0001

Chloride

-1.0857

0.0003

species

estimate

C. dubia

-2255.93

<0.0001

C. dubia

1587 .4486

H. azteca

-2565 .864

<0.0001

H. azteca

1277 .5146

C. tentans

7251 .9177

<0.0001

C. tentans

11095.2963

S. simile

??

---

IN}IS

October 20. 2005

Page 22 of 22

Literature cited

American Public Health Association (APHA), American Water Works Association, Water Environment

Federation. 1998. Standard methods for the examination of water and wastewater, 20th ed

.

American Public Health Association, Washington, DC

.

ASTM (American Society for Testing and Materials) . 2002a. Standard guide for conducting acute toxicity

tests on test materials with fishes, macroinvertebrates, and amphibians . E729-96. American

Society for Testing and Materials, Philadelphia, PA, USA

.

ASTM (American Society for Testing and Materials) . 2002b. Standard guide for conducting three-brood,

renewal toxicity tests with

Ceriodaphnia dubia

.

E1295-01 . American Society for Testing and

Materials, Philadelphia, PA, USA

.

ASTM (American Society for Testing and Materials) . 2005. Standard guide for conducting laboratory

toxicity tests with freshwater mussels . E2455-05. American Society for Testing and Materials,

Philadelphia, PA, USA

Bayliss D, Harris RR. 1988. J Comp Physiol 158:81-90

Dwyer FJ, Burch SA, Ingersoll CG, and Hunn JB . 1992. Environ Toxicol Chem 11 :513-520 .

Ingersoll CG, Dwyer FJ, Burch SA, Nelson MK, Buckler DR, and Hunn JB . 1992. Environ Toxicol

Chem 11:503-511

.

Lucu C, and Flik G. 1999. Am J Physiol 276:R490-R499

.

McMahon, RF, and Bogan, AE. 2001 . Mollusca: Bivalvia, in JH Thorp and AP Covich (eds .) Ecology

and Classification of North American Freshwater Invertebrates, 2" d Edition. Academic Press, San

Diego, CA . pp 352-353 .

Mount DR, Gulley DD, Hockett JR, Garrison TD, and Evans JM . 1997. Environ Toxicol Chem 16 :2009-

2019 .

Pic P, and Maetz J. 1981 . J Comp Physiol B 141 :511-521 .

Soucek, DJ. 2004. Effects of hardness, chloride, and acclimation on the acute toxicity of sulfate to

freshwater invertebrates . Final Report to: Illinois Environmental Protection Agency and Illinois

Coal Association.

Smith ME, Lazorchak JM, Herrin LE, Brewer-Swartz S, and Thoeny WT . 1997. Environ Toxicol Chem

16:1229-1233 .

Taylor G, Baird DJ, Snares AMVM . 1998 . Environ Toxicol Chem 17 :412-419 .

U.S. EPA. 1985. Guidelines for deriving numerical water quality criteria for the protection of aquatic

organisms and their uses. PB85-227049. Washington, DC .

U.S. EPA. 1993. Methods for measuring the acute toxicity of effluents and receiving waters to freshwater

and marine organisms, 4" ed. EPA/600/4-90/027F, Cincinnati, OH

.

U.S. EPA. 1994a. Methods for measuring the toxicity and bioaccumulation of sediment associated

contaminants with freshwater invertebrates. EPA/600/R-94/024. U.S. Environmental Protection

Agency, Washington, DC .

U.S. EPA. 1994b. Methods for the detemtination of metals in environmental samples . EPA/600/R-4/1 11 . U.S .

Environmental Protection Agency, Cincinnati, OH

.

U.S. EPA. 1999. National Recommended Water Quality Criteria-Correction . EPA 822-Z-99-001,

Washington, DC .

Zipper CE. 2000. Coal mine reclamation, acid mine drainage and the Clean Water Act, in R. Bamhisel, W

.

Daniels, and R. Darmody (eds) Reclamation of Drastically Disturbed Lands. American Society of

Agronomy, Madison, WI

.

---

--DRAFT--

Effects of Water Quality on Acute and Chronic Toxicity of Sulfate

to Freshwater Bivalves,

Ceriodaphnia dubia,

and

Hyalella azteca.

Final Report

Submitted to

:

Edward Hammer and Dertera Collins

United States Environmental Protection Agency

Region 5, Water Division, 77 West Jackson Boulevard

Chicago, Illinois 60604

By

David J. Soucek, Ph.D .

Illinois Natural History Survey

Champaign, Illinois 61820

April 20, 2006

U.S. Environmental Protection Agency Grant # CP96543701-0

INHS

April 20, 2006

Illinois EPA Exhibit No .

---

INNS

April 20, 2006

Background

While there are no Federal water quality criteria (WQC) for the protection of freshwater life for

total dissolved solids (TDS), sulfate, or sodium (U .S

. EPA 1999), several states, including

Minnesota, Indiana, and Illinois, are at various stages in the process of developing standards for

sulfate. The current "General Use" standard of 500 mg/L in Illinois is based on the value thought

to be protective of livestock, but the Illinois Environmental Protection Agency (IL EPA) is

actively pursuing an update of the sulfate standards based on scientific research, and is close to

proposing an updated standard (R. Mosher, IL EPA, pers. comm.)

.

Sodium is one of the most common major cations in high TDS effluents, but calcium and

chloride are usually present in mine-impacted waters as well. While major ion or TDS toxicity is

caused by osmoregulatory stress from the combination of all cations and anions, chloride

standards currently exist, and Illinois plans to additionally regulate for sulfate in order to address

the major non-chloride component of TDS in these waters . Therefore, studies were conducted by

Soucek (2004; also published as Soucek and Kennedy 2005) to (1) generate LC50s (lethal

concentration to 50% of a sample population) and LCIOs (lethal concentration to 10% of a

sample population) for sulfate with selected freshwater invertebrates

(Ceriodaphnia dubia,

Chironomus tentans, Hyalella azteca,

and Sphaerium simile) in U.S. Environmental Protection

Agency's (US EPA, 1993) moderately hard reconstituted water (MHRW) and (2) determine . the

effects of laboratory water composition, water hardness, and test organism acclimation on the

acute toxicity of sulfate to Ceriodaphnia dubia and Hyalella azteca (Soucek, 2004). In these

previous studies (Soucek, 2004), the mean LC50s, expressed asmg 5042,/L, in moderately hard,

reconstituted water (MHRW; U.S. EPA 2002) ranged from: 512 to 14,134 mg/L. The LC50

generated for the amphipod, Hyalella (512 mg/L) was surprisingly low, given that it is known as

a euryhaline organism (Ingersoll et al., 1992), but as will be discussed below, water quality data,

including other cations and anions present, are critical for predicting the responses of freshwater

organisms (especially Hyalella) to elevated sulfate concentrations .

The composition of the dilution water used during testing in the Soucek (2004) study had a

dramatic effect on the toxicity of sulfate to Hyalella . Whereas the 96-hour LC50 in MHRW was

512 mg/L, the LC50 increased to 2,855 mg/I. when using a "Reformulated Moderately Hard

Reconstituted Water" (RMHRW, Smith et al ., 1997). The LC50 for C. dubia also increased

from 2,050 in MHRW to 2,526 mg/L in RMHRW . Both dilution waters were similar in terms of

hardness (-90-106 mg/L asCaCO;), alkalinity, and pH, but RMHRW had a higher chloride

concentration and different calcium to magnesium ratio than that in MHRW. An additional

experiment, not included in the Soucek (2004) report, indicated that when sulfate (-2,800 mg/L)

and hardness (106 mg/L) were held constant, percent survival ofH. azteca was positively

correlated with chloride concentration (up to 67 mg Cl -/L; Soucek and Kennedy, 2005). These

experiments illustrated the need to further characterize the interacting effects of chloride and

sulfate on aquatic organisms.

---

Another factor that appears to have a strong effect on the toxicity of sulfate is the presence of

other major cations, in this case, calcium and magnesium, measured as hardness . In the previous

study (Soucek 2004), increased hardness reduced the toxicity of sulfate to Hyalella and had a

dramatic effect on the 48-hour LC50 for C. dubia, increasing from 2,050 at a hardness of 90 to >

2,900 mg/L at hardness values higher than 194 mg/L as CaCO3. Others have observed reduced

toxicity of saline solutions due to increased hardness as well

(e.g ., Dwyer et al. 1992; Mount et

al. 1997)

.

While a great deal of progress has been made in the understanding of sulfate toxicity under

varying water quality conditions, several important data gaps remained. First, quantification of

the effects of hardness on sulfate toxicity to Hyalella was needed to determine if the previously

observed phenomenon was specific to Ceriodaphnia . Another research need was quantification

of the effects of a wide range of chloride concentrations on sulfate toxicity to both Hyalella and

Ceriodaphnia

.

In addition, previous studies (Soucek 2004) indicated that the fingernail clam,

Sphaerium simile, had a lower LC 10 than that of C. dubia, but because of the temporal nature of

its availability, this bivalve was only tested in MHRW. It remained unclear whether or not a

mollusk will have the same physiological response as two crustaceans to increased chloride or

hardness in these experiments with sulfate . Another data gap was the fact that all of the tests

conducted in the Soucek (2004) study were acute exposures of 48 to 96 hours . Sublethal effects

of sulfate in longer-term exposures were unknown . Therefore, the objectives of this study were

to build on previous studies conducted to support development of a sulfate criterion for

protection of aquatic life by (1) determining the effects of hardness on toxicity of sulfate to

bivalves, (2) determining the toxicity of sulfate to juvenile unionid mussels, (3) determining the

short-term (7 days) chronic toxicity of sulfate to Ceriodaphnia dubia, (4) determining the effects

of chloride on acute toxicity of sulfate to Hyalella and Ceriodaphnia, and (5) determining the

effects of hardness on toxicity of sulfate to Hyalella at a critical chloride level,

i.e., the chloride

concentration at which sulfate is significantly less toxic to Hyalella as determined in #4 above

.

Methods

General culturing and testing methods

Invertebrates selected for testing include Ceriodaphnia dubia, Hyalella azteca, Sphaerium simile

(Pelecypoda, Sphaeriidae), and a juvenile freshwater unionid mussel

(Lampsilis siliquoidea).

The cladoceran, Ceriodaphnia dubia, was cultured in-house (Soucek laboratory, Illinois Natural

History Survey) according to U.S. EPA methods (2002). Amphipods, Hyalella azteca, also were

cultured in house according to U.S. EPA methods (2000) in a "Reformulated Moderately Hard

Reconstituted Water" described in Smith et al. (1997). Sphaeriid clams were collected from

Spring Creek, near Loda, Illinois, (Iroquois County) and acclimated to MHRW at 22 °C and a

16:8 (L:D) photoperiod for 5-7 d prior to testing . Clams collected from this site were previously

identified to species by Dr . Gerald Mackie, of the University of Guelph, Department of Zoology,

Guelph, Ontario, Canada.

INHS

April 20, 2006

---

INNS

April 20, 2006

For toxicity testing, a pure (99%) grade of anhydrous sodium sulfate (Na2SO4) (CAS No. 7757-

82-6) was obtained from Fisher Scientific (Pittsburgh, PA, USA) to serve as the source of sulfate

.

Previous experiments indicated that the salts and deionized water sources used for our

experiments had low to undetectable levels of trace metal contaminants (Soucek 2004)

.

For definitive static, non-renewal toxicity tests, conducted according to American Society for

Testing and Materials (ASTM) E729-96 methods (2002a) treatments were comprised of a 75%

dilution series (i.e., the 100% concentration was serially diluted by 25%), rather than the standard

50%, because major ion toxicity tests often cause 100% mortality in one concentration and 0%

mortality in the next highest concentration if the spread is too great . Five to six concentrations

were tested in addition to controls with four replicates tested per concentration . Tests with C.

dubia

were conducted for 48 h with a 16:8 (L:D) photoperiod at 25 °C, and

H. azieca

and S .

simile

were exposed for 96 h at 22 °C and a 16:8 (L:D) photoperiod. Both crustaceans were

exposed in 50-m1 glass beakers with 5 organisms per beaker, and for H .

azteca, I

g of quartz

sand was added to each beaker to serve as substrate. Clam tests were conducted in 150-m1 glass

beakers (no substrate). All clams used are juveniles . Only one of the 63 tests was fed, and that

fed test had a median LC50 value compared to two other tests conducted with the same organism

in the same water type

.

Ceriodaphnia dubia

used were less than 24-h old, and

H. azteca

were

-third instar (7 - 14 d old) . Percent survival in each replicate was recorded every 24 h and at the

end of the exposure period. A dissecting microscope was used to assess survival of

H. azreca.

Chronic testing was conducted according to guidelines described in ASTM E 1295-01 (2002b) .

Ten replicates were used per sulfate concentration with one organism per replicate. Endpoints

included the number of young (both live and dead recorded separately) produced by each first

generation C.

dubia,

and survival of first generation

C.

dubia .

Standard water chemistry parameters were measured at both the beginning and the end of each

exposure period, including temperature, pH, conductivity, dissolved oxygen, alkalinity and

hardness. The pH measurements were made using an Accumet® (Fisher Scientific, Pittsburgh,

PA, USA) model AB15 pH meter equipped with an Accumet® gel-filled combination electrode

(accuracy < ± 0.05 pH at 25 °C). Dissolved oxygen was measured using an air-calibrated Yellow

Springs Instruments (RDP, Dayton, OH, USA) model 58 meter with a self-stirring biochemical

oxygen demand probe. Conductivity measurements were made using a Mettler Toledo® (Fisher

Scientific, Pittsburgh, PA, USA) model MC226 conductivityiTDS meter . Alkalinity, and

hardness were measured (beginning of tests only) by titration as described in American Public

Health Association (APHA) et al. (1998). Samples from each treatment were analyzed to

confirm sulfate concentrations by ion chromatography at the Illinois Natural History Survey

Aquatic Chemistry Laboratory, Champaign, IL, USA .

All LC50 values were calculated based on both measured sulfate concentrations and measured

specific conductivity values for each test concentration using the Spearman-Karber method, and

sulfate LC IOs, LCSs, and LC ls were calculated using probit analysis . To increase confidence in

LC50 values, three to five assays were conducted with each organism for each water quality

---

INHS

April 20, 2006

combination. This provided a stronger estimate of the mean LC50 value for a given set of water

quality parameters for each species . In all, 63 new LC50s were generated . After LCXs were

calculated using probit analysis, the following quotients were calculated: LCIO/LC50,

LC5/LC50, and LC1/LC50. Then geometric mean quotients for each species were compared

using Student's T-test .

Toxicity testing with freshwater mussels (USGS)

Test conditions and procedures for conducting toxicity tests with newly-transformed juvenile

mussels of fatmucket (Lampsilia siliquoidea) were in accordance with the recommended test

conditions outlined in ASTM E2455-05 (2005, and see Appendix 1) . Five 96-h sulfate toxicity

tests with juvenile fatmuckets were conducted at three hardness levels (100, 300, and 500 mg/L

as CaCO3with Ca:Mg ratio of 2.33) and two chloride concentrations (5 and 33 mg/L) at one

hardness (100 mg/L with Ca :Mg ratio of 1 .46) at the U.S. Geological Survey's Columbia

Environmental Research Center. A preliminary range-finding test with juvenile fatmuckets

indicated that the concentration of 5000 mg sulfate/L resulted in more than 50% mortality and

therefore was chosen as the highest concentration for the 96-h toxicity tests. Each of the five tests

was conducted with five concentrations of sulfate in a 50% serial dilution and a control with four

replicates. Test water and solution were provided by Illinois Natural History Survey, Champaign,

IL. 1-d-old juvenile fatmuckets were obtained from laboratory cultures at Southwest Missouri

State University, Springfield, MO . When juvenile mussels were received, the water temperature

was gradually adjusted to the test temperature (20°C). Shipment water was gradually replaced

with test water over a 48-h acclimation period . Five juveniles exhibiting foot movement were

impartially transferred into each of 50-m1 glass beakers with about 30-ml test solution. Water

quality characteristics including pH, conductivity, hardness, and alkalinity in each exposure

treatment were determined at the beginning and the end of each test (Table 2) and were close to

the nominal. Dissolved oxygen was above 8.2 mg/L during all tests. Sulfate concentrations were

measured at INHS as described above. After 48- and 96-h exposure, survival of juvenile mussels

was determined. Individuals which exhibited foot movement within a 5-min observation period

were classified as alive (ASTM 2005). Median effective concentrations (EC50s) were

determined by the Trimmed Spearman-Karber method (TOXSTAT 3.5; WEST 1996). Measured

sulfate concentrations were used for EC50 calculation .

Influence of chloride on the toxicity of sodium sulfate

In these experiments, we tested the toxicity of sulfate (with sodium

as the major cation) to H .

azteca and C. dubia in freshwater solutions having nominal chloride concentrations of 1 .9, 10,

15, 20, 25, 33 (H. azteca only), 100, 300 and 500 mg Cl/L. Chloride, as NaCl (CAS No.7647-

14-5, Fisher Scientific Cat. # AC42429-0010) was added at appropriate concentrations to a

solution with a hardness of -100 mg/L (molar ratio of Ca :Mg = 1 .41 ; 2.33 in terms of mass) .

Whole carboys were made for each elevated chloride level, and this water was used

as both

diluent and control; therefore, each concentration within a given test had the same chloride

concentration (i.e., [Cl'] did not change with dilution) . The only parameters that varied within a

particular test were sodium, sulfate and conductivity. At least three tests were conducted for

---

each hardness level to provide a mean LC50 value and standard deviation . Exposures were

conducted using the same laboratory and calculation methods described above

.

After LC50s were calculated as described above, regression analysis was conducted using JMP®

software (Sall and Lehman,

1996)

to determine the relationship between chloride concentration

and sulfate LC50 for each species. Mean LC50 values for each chloride concentration were used

in these analyses, and two separate analyses were conducted for each species : one for the range or

5 to 25 mg Cl/L and one for the range of 25 to 500 mg Cl/L . Then, multiple regression analysis

with covariance was conducted for the same data ranges using all individual data points to

generate an equation for both species, and to determine if the curves were significantly different

for the two species .

Influence of hardness on the toxicity of sodium sulfate

In these experiments, we tested the toxicity of sulfate (with sodium as the major cation) to H

.

azteca

in six freshwater solutions having nominal hardness values of <100, 200, 300, 400, 500,

and

600

mg/L (as CaCO3). Hardness was increased by adding enough CaSO 4(CAS No .

7778-

18-9)

and MgSO4(CAS No .

7487-88-9),

at a set molar ratio (Ca :Mg = 1 .41 ; 2.33 in terms of

mass) to achieve the nominal hardness values . The Ca:Mg value was chosen because it is the

median value for water bodies sampled in Illinois (R . Mosher, IL EPA, pers. com.). A chloride

concentration of 25 mg/L was used for all tests investigating the effects of hardness on sodium

sulfate toxicity to

H.

azteca

based on results from above described tests investigating the effects

of chloride on sodium sulfate toxicity to

H.

azteca

and C.

dubia .

Whole carboys were made for

each elevated hardness level, and this water was used as both diluent and control ; therefore, each

concentration within a given test had the same hardness

(i.e ., [Ca

2

*] and [Mg2'] did not change

with dilution). The only parameters that varied within a particular test were sodium, sulfate and

conductivity. At least three tests were conducted for each hardness level to provide a mean LC50

value and standard deviation. Exposures were conducted using the same laboratory and

calculation methods described above . LC50 values for H.

azteca

were compared to previously

generated LC50s for C.

dubia

(Soucek and Kennedy 2005), which were conducted in solutions

having a Ca:Mg molar ratio of

0.88

and [Cl] of

1.9

mg/L.

After LC50s were calculated as described above, regression analysis was conducted using JMP®

software (Sail and Lehman,

1996)

to determine the relationship between hardness and sulfate

LC50 for each species. Mean LC50 values for each hardness level were used in these analyses

.

Then, multiple regression analysis with covariance was conducted for the same data ranges using

all individual data points to generate an equation for both species, and to determine if the curves

were significantly different for the two species

.

INHS

April 20, 2006

---

INHS

April 20, 2006

Relationship between sulfate LC50s and conductivity LC50s

To investigate variability in conductivity at sulfate LC50 concentrations, linear regression

analysis was used, and

C.

dubia

data from Soucek and Kennedy (2005) were included in the

analyses. Three data ranges were used to compare sulfate and conductivity LC50 relationships

:

Cl = 5 to 100 mg/L, Cl = 5 to 300, and Cl = 5 to 500 mg/L .

Comparison of test results with STR model predictions

To compare results from the present study to toxicity predictions generated by the STR model

(Mount and Gulley, 1993), we calculated nominal concentrations of all constituent ions (except

H+ and OH", which are not required by the model) at observed mean sulfate LC50 levels for each

test solution type (Cl = x, hardness = y, and Ca:Mg = z). Ions required by the model include

Cat+ , Mgt+ , Na', K', Cl - , HCO3 , and S04'-- These calculations were possible because we

generated all test solutions using deionized water with known salt concentrations added . In

addition, confidence in the use of nominal concentrations for all ions is provided by the fact that

the average of the absolute value of % difference between nominal and measured sulfate

concentrations was 2.082% .

The model output includes equivalents of cations and anions, and requires that, to have

confidence in model output, the difference between the two be less than 15% . The average (t

SD) % difference between cations and anions for out inputs was 0 .09 (± 0.01)%, indicating

excellent agreement between cation and anion equivalents. Other model outputs included

calculated TDS, a "NUMCAT" value, LC50 in terms of % of solution, and % survival in 100%

of solution. Because all of our inputs were concentrations at observed LC50, the observed %

survival in 100% solution was always 50% . To examine the effectiveness of the predictive

ability of the STR model over the range of solutions tested, we created scatter plots of predicted

% survival versus either chloride concentration or water hardness as appropriate for each species

tested. Model outputs included toxicity predictions for C . dubia, Daphnia magna, and

Pimephales promelas . Because Hyalella was most similar in sensitivity to C . dubia, we

compared observed results for

Hyalella

to predicted results for C .

dubia .

Predictive equations for standard development

A subset of the available data (from this study) was used to calculate slopes describing the effects

of hardness and chloride on sulfate toxicity . In consultation with Charles Stephan (USEPA), it

was decided that only tests conducted with a Ca:Mg molar ratio of 1.41 (2.33 in terms of mass)

would be included for standard development . Further more, because consistent linear

relationships between hardness and sulfate were observed for both

C.

dubia

and H.

azteca

in the

range of -100 to -500 mg/L as CaCO3 (see data below), only tests conducted within this range

were included for standard development . Finally, because the relationship between chloride

concentration and sulfate toxicity was inconsistent below 25 mg/L (see data below), only tests

conducted at chloride concentrations of 25 mg/L or higher were included for standard

development. This left 42 tests (14 for C.

dubia

and 28 for

H. azteca)

for inclusion in calculation

of slopes describing the effects of hardness and chloride on sulfate toxicity. To calculate slopes,

multiple linear regression analysis with covariance was conducted with LC50 (based on

---

INHS

April 20, 2006

measured sulfate concentrations) as the dependent variable and chloride, hardness, and species as

independent variables .

RESULTS and DISCUSSION

Influence of chloride on the toxicity of sodium sulfate

Chloride had variable effects on sodium sulfate toxicity to C .

dubia

andH.

azteca

over the range

of 5 to 500 mg CC/L. For

Hyalella

in particular, two different linear trends were observed

depending on the chloride range (Fig . 1a, b). Increasing chloride concentration from 5 to 25

mg/L resulted in increasing S04" LC50s (R2 = 0.8503, p = 0.0258) for

Hyalella

(Fig. I a), while

for C.

dubia,

a mildly positive trend was observed, but the relationship was not significant over

this chloride concentration range (R 2 = 0.4906, p = 0.1877). In addition, the LC50s for C.

dubia

were higher than those for

Hyalella

for each chloride concentration over this range. When using

a combined data set of individual test LC50s for C

.

dubia

and

Hyalella

over this chloride range

and at hardness = 100 (n = 33) in a simple linear regression analysis with covariance (with

species as a treatment effect and chloride concentration as continuous effect), a strong positive

relationship was observed (R 2 = 0.7900, p < 0.0001) with both the chloride and treatment

(species) effects being significant (p < 0.0001, Table 1) .

While a positive relationship between chloride concentration and SO 42 " LC50 was observed for

Hyalella

over the range of 5 to 25 mg CDL, a significantly negative trend (R 2 = 0.875, p =

0.0195) was observed over the range of 25 to 500 mg Cl'/L . An even stronger negative

relationship (R2 = 0.9493, p = 0.0257) was observed for C .

dubia

over the same chloride range

.

When using the combined data set of individual test LC50s for C

.

dubia

and

Hyalella

over this

chloride range and at hardness = 100 (n = 30) in a simple linear regression analysis with

covariance as described above, a negative relationship was observed (R

2 = 0.6539, p < 0.0001)

with both the chloride and treatment (species) effects being significant (p < 0.0001 and p =

0.0003, respectively, Table 1)

.

Hyalella

appears to require a minimal amount of chloride for effective osmoregulation . While

there are several different osmoregulatory strategies used by freshwater organisms, most

freshwater amphipods and daphnid cladocerans regulate hypertonically with respect to the

surrounding medium, and this is achieved by active transport of ions into the hemolymph

(Dorgelo 1981, Aladin and Potts 1995, Greenaway 1979, Schmidt-Nielsen 1997) . The principal

inorganic anion of crustacean hemolymph is chloride, and it has been suggested that low chloride

concentrations may limit the distribution of at least one euryhaline amphipod

(Corophium

curvispinum)

in freshwaters (Bayliss and Harris 1986) . Even among amphipods, there is a wide

range of sodium and chloride influx rates and integument permeabilities which determine

osmoregulatory effectiveness (Bayliss and Harris 1986, Taylor and Harris 1986) ; therefore, it is

not surprising that the responses of H .

azteca

and C.

dubia

to sodium sulfate were quite different

over the lower range of chloride concentrations. While Borgmann (1996) suggested that under

low salinity conditions, bromide was required but chloride was not needed by

H. azteca

for

survival, growth and reproduction, data from this study suggest that the chloride is quite

---

INHS

April 20, 2006

important in determining the response of that organism to elevated levels of sodium sulfate

.

Laboratory deionized water and concentrated sodium sulfate solutions were analyzed previously

for bromide, and levels were below detection limits (Soucek 2004)

.

Over the higher range of chloride concentrations (25 to 500 mg/L), a different trend was

observed than that over the lower concentrations . While the slopes of the lines for the two

crustacean species were different, there was a negative correlation between chloride

concentration and sulfate LC50 for both species. The trend was stronger for C.

dubia,

with a

more negative slope (-2.2) compared to H.

azteca (-0.875),

although R2 values were high and

relationships statistically significant for both. These data suggest that, over this range of chloride

concentrations, chloride and sodium sulfate toxicity are additive . Chloride LC50s (as NaCl) for

C. dubia

generally range from 900 to 1,200 mg CE/L (e.g ., Mount et al. 1997), and so the highest

two chloride concentrations in this study were likely to cause some toxicity without sulfate

present .

Table 1. Results of multiple regression analysis with covariance for three different subsets of

data. Individual LC50s were used as data points

.

Data for both species were included

.

ICI -1 range = 5-25 mpJL, hardness -1.00 mzJL

R2 = 0.7900, n =33

fCll range = 25-500 mg/L . hardness -100 mg/LL

R2 = 0.6539, n =30

Hardness range = 100-600, Cl = 25 mg/L

P

<0.0001

<0.0001

0.9046

R2 = 0.5 177, n = 38

Term

Estimate

Intercept

1969.38

Hardness

3.15

Species

-10.38

Term

Estimate

P

Intercept

1270.23

<0.0001

chloride

35.14

<0.0001

Species

449.68

<0.0001

Term

Estimate

a

Intercept

2189.48

<0.0001

chloride

-1.46

<0.0001

Species

178.92

0.0003

---

S

r

0

e

1500

-

O

u

1000

500

3000

-

2500

2

O

y

E!

1500

0

j 1000

500

0

5

chloride (mg/L)

y = -2.2023x + 2530 .7

R2 = 0.9493, p = 0.0257

y = -0.875x + 1904 .4

R2 =0.875,p=0.0195

•

CC dubia

O H. azteca

0

100

200

300

400

500

600

chloride (mg/L)

Figure 1. Influence of chloride concentration over two ranges, 5 to 25 mg/L (A) and 25 to 500 mg/L (B),

on toxicity of sodium sulfate to Ceriodaphnia dubia and Hyalella azteca . Hardness was -100 mg/L for all

tests and Ca:Mg molar ratios were 1 .41 except for the tests at 5 mg CI7L (0.88)

.

INHS

April 20, 2006

3000

-

y = 13.868x + 2085 .1

2500 -

R' = 0.4906, p = 0.1877 1

2000

-

y = 57.791x + 445.59

R' = 0.8503, p = 0.0258

•

C. dubia

O H. aZteca

10

15

20

25

30

---

INHS

April 20, 2006

Effects

of hardness on toxicity of sulfate to Hyalella at chloride = 25 mg/L

When chloride was maintained at 25 mg/L, a strong linear trend of decreased sulfate toxicity with

increased hardness (R2 = 0.7092, p = 0.0354) was observed for Hyalella (Fig. 2). LC50 values

increased from less than 1,900 mg/L at hardness = 100 mg/L, to greater than 4,000 mg/L at a

hardness of 500 mg/L. The mean LC50 value at 600 mg/L hardness was lower than that at 500

mg/L hardness. It remains unclear how the trend will continue with increasing hardness above

600 mg/L. When using the combined data set of individual test LC50s for C

. dubia andHyalella

over this hardness range (n = 38) in a simple linear regression analysis with covariance as

described above, a positive relationship was observed (R 2 = 0.5177, p < 0.0001, Table 1). The

hardness effect was observed to be significant (p < 0 .0001), but the treatment (species) effect was

not (p = 0.9046, Table 1) .

Hardness had a strong influence on sulfate acute toxicity that was similar for both crustacean

species. A number of studies have provided evidence that increasing hardness ameliorates

toxicity of waters with high dissolved solids concentrations (Kennedy et al . 2003, 2005, Dwyer et

al. 1992, Mount et al. 1997, Latimer 1999) and Soucek and Kennedy (2005) showed

quantitatively that, in a sodium-dominated system, sulfate toxicity to C. dubia is reduced as

hardness progressively increases. In this study, the results of multiple linear regression analyses

indicated no difference between the responses of the two species over the hardness range of 100

to 600

mgf L

as CaCO3. This was in contrast to the results of the tests in which chloride was

varied, where the two species had different responses (slopes) over both ranges of chloride

concentrations examined. In addition, these results are notable because nearly identical slopes

were observed for the two species despite the fact that the waters for tests conducted with C .

dubia had a different chloride concentration (5 mg/L) and Ca :Mg molar ratio (0.88) than those

used for tests with H . azteca (25 mg CI'/L, and 1 .41 Ca:Mg molar ratio). Soucek and Kennedy

(2005) proposed as an explanation for this phenomenon of hardness ameliorating sulfate toxicity

that increased calcium concentrations decrease the passive permeability of epithelial cells to

water and ions in various aquatic organisms (Lucu and Flik 1999, Pic and Maetz 1981), reducing

passive diffusion and the energy required to osmoregulate, and accounting for the decrease in

toxicity. Calcium can mitigate hydrogen ion toxicity to aquatic organisms by decreasing

membrane permeability to H` and stimulating active Ne uptake (see Havas and Advokaat 1995) ;

however, Potts and Fryer (1979) found that calcium had little effect on sodium loss in Daphnia

magna. While data from the present study support this hypothesis, other explanations are

possible and empirical work is needed to determine the mechanism behind the phenomenon

.

---

y = 3.6716x + 1822.1

R2 = 0.7092, p = 0.0354

Hyaletla

only

•

C. dubia*

o

H. azteca

0

100

200

300

400

500

600

700

hardness (mg/L)

Figure 2. Influence of hardness on toxicity of sulfate to Hyalella azteca and Ceriodaphnia dubia. C .

dubia data are from Soucek and Kennedy (2005) .

Chloride concentration for all H. azteca tests was -25

mg/L, and Ca:Mg molar ratio was 1 .41

.

Relationship between sulfate LC50s and conductivity LC50s

Conductivity LC5Os ranged from 2,650 to 8,449 µmhos/cm, while LC50s based on sulfate ranged

from 1,116 to 4,345 mg SO 21L (Fig. 3). For tests with Cl <_ 100, sulfate LC50s were strongly

correlated with conductivity LC50s (R 2 = 0.9769, p < 0.0001, Fig.3a), while the relationship

weakened slightly for Cl <_ 300 (R2 = 0.9522, p < 0.0001, Fig.3b), and Cl <_ 500 (R2

=

0.8987, p <

0.0001, Fig.3c). Thus, LC50s in terms of conductivity were highly correlated with LC50s in

terms of sulfate for both species except when extremely high chloride concentrations were used

(300 to 500 mg/L). The plots of conductivity LC50s and sulfate LC50s clearly illustrate the

contention that knowledge of the contribution of various major ions is critical to effectively

managing "produced waters" or effluents with high concentrations of dissolved solids (Ho et al

.

1997). Not only did sulfate LC50s range from 1,200 to 4,345 mg/L, but conductivity LC50s

ranged from 2,650 to 8,449 µmhos/cm. These wide ranges were observed for just two species

with relatively similar sensitivity . Clearly, any attempt at water quality standard development,

whether based on TDS, conductivity, sodium, or sulfate, should incorporate the fact that the

water quality parameters like hardness and chloride strongly regulate the toxicity of high TDS

solutions. Finally, the conductivity/sulfate plots provide further evidence that chloride and

sulfate toxicity are additive . When chloride was less than or equal to 100 mg/L, sulfate toxicity

was strictly related to conductivity, but when 300 and 500 mg CE/L solutions were tested, sulfate

LC50s were lower than would be predicted by LC50s based on conductivity

.

INHS

April 20, 2006

5000

4500

4000

3500

0

rn

cu

E

3000

0

n

U

2500

2000

1500

1000

---

9000

8000

0

U

_ 7000

A 0

6000

L

o

a

5000

e

s

400°

3000

2000

1500 2000

2500

3000

3500

4000

4500

5000

sulfate at LC50 (mg/L)

B .

1000 1500 2000 2500

3000 3500

4000

4500

5000

sulfate at LC50 (mgfL)

C

.

sulfate at LC50 (mg/L)

Figure 3. Relationship between LC50s in terms of sulfate (mg/L) and LC50s in terms of conductivity (µmhos/cm)

including tests with chloride ranging from 5 to 100 mg/L (A), 5 to 300 mg/L (B), and 5 to 500 mg/L (C) . In addition

to the 63 new tests generated for this study, 19 tests from Soucek and Kennedy (2005) with C. dubia were included .

Hardness values ranged from 100 to 600 mg/L and two Ca :Mg molar ratios (0.88 and 0.41) in parts A, B, and C.

INHS

April 20, 2006

---

1

0.9

-

0.8

-

0.7

0.6

0.5

0.4

0.3

-

0.2

-

0.1

-

0

0 C . dubia

p H. azteca

L

LC1O/LC50

LC5/LC50

LC1/LC50

Figure 4. Comparison of geometric means of probit-calculated LCI0/LC50, LC5/LC50 and LC1/LC50

quotients for Hyalella azteca (n = 33) and Ceriodaphnia dubia (n = 18). Only new tests generated for this

study were included. Measured chloride concentrations ranged from 11 to 525 mg/L, and hardness ranged

from 92 to 604 mg/L. Ca:Mg molar ratio for all tests was 1 .41 . Means with different capital letters are

significantly different (p < 0.01) .

INHS

April 20, 2006

Mean LCX quotients for C . dubia and H. azteca

For 12 of the 63 tests conducted, probit analysis would not calculate LCX values because of the

distribution of the mortality data. The geometric mean (n = 51) of the LC10/LCSO quotients for

both species combined was 0 .730 ± 0.092, while the geometric means of LC5/LC50 and

LC1/LC50 quotients for both species were 0.668 ± 0.106 and 0.565 ± 0.125, respectively . Using

the combined data sets (n = 51) for C

. dubia andH. azteca, each quotient (LC IO/LC50,

LC5/LC50, LC1/LC50) was used as a dependent variable with hardness, chloride, and species as

explanatory variables in a simple linear regression analysis with covariance as described above

.

In all three cases, only the species effect was found to be significant (p = 0 .0187, p = 0.0189, and

p = 0.0196, respectively). In pair-wise comparisons,

C.

dubia had significantly higher geometric

means than H. azteca did for all three quotients (p = 0 .0014, p = 0.0014, and p = 0.0015,

respectively, Fig. 4), indicating that the dose-response curves for the two crustaceans were

significantly different .

Ceriodaphnia dubia

had significantly higher geometric mean

LC1OILC50, LC5 .LC50, and LC1/LC50 quotients than did H. azteca, suggesting that more

mortality was seen at sulfate concentrations lower than the LC50 in the H. azteca sub-

populations than in those of C. dubia . It remains unclear whether this observation was due to a

difference in the physiological responses or tolerances of the two organisms, or whether the trend

was simply a result of the difference in test duration for the two species (48 h for C

. dubia and 96

h forH. azteca)

.

Results of the multiple linear regression analysis indicated that hardness and

chloride concentration did not have a significant effect on LCx/LC50 quotients

.

---

INHS

April 20, 2006

Comparison of test results with STR model predictions

All ion concentrations used as input for the STR model were concentrations at the observed

mean sulfate LC50 levels for each test solution type (Cl = x, hardness = y, and Ca :Mg = z), so

"observed" percent survival in each case was 50% . For tests with Ceriodaphnia in which

hardness was fixed at -100 mg/L and chloride varied from 5 to 500 mg/L, the STR model

predicted % survival values ranging from 69 .0 to 48.4% (Fig. 5a). Most predictions were greater

than 50%, and thus the model slightly under-predicted toxicity in most cases. STR does not

predict toxicity for Hyalella, so for Hyalella test inputs we used the Ceriodaphnia predictions

from STR. For tests with Hyalella in which hardness was fixed at -100 mg/L and chloride

varied from 5 to 500 mg/L, predicted % survival varied more widely than for Ceriodaphnia with

values ranging from 62.5 to 97.10% (Fig. 5a). However, as chloride concentration increased, the

STR model predictions became closer to the observed 50% survival values . In fact, there was a

significant negative relationship between chloride concentration and predicted % survival (R 2

=

0.9045, p = 0.0003, n = 8)

.

For tests with Ceriodaphnia in which chloride was fixed at -25 mg/L and hardness varied from

100 to 600 mg/L, the STR model predictions were highly variable, ranging from 4.1 to 82.9%

survival (Fig. 5b). Only the hardness = 100 mg/L prediction was greater than 50% (82 .9%),

while for hardness values of 200 to 600 mg/L, toxicity was strongly over-predicted, with %

survival predictions of 4.1 to 21 .8. For tests with Hyalella in which chloride was fixed at -25

mg/L and hardness varied from 100 to 600 mg/L, a similar patter was observed with and under-

prediction of toxicity at hardness = 100 (88 .6% survival) and over-prediction of toxicity at

hardness values of 200 to 600 mg/L (42 .3 to 0.7% survival, Fig . 5b) .

These data indicate that when chloride was variable and hardness was fixed at -100 mg/L, the

STR model was relatively accurate in predicting toxicity to C . dubia ; predicted survival ranged

from 48 to 69% and observed survival was 50% in each case because ion concentrations at LC50

were used as inputs. With one exception (48%), the model under-predicted toxicity for this data

range. This may be because the STR model is largely based on the results of fed tests, which the

authors acknowledged had a small influence on test results (Mount et al . 1997). However, I

compared 48 h sulfate LC50s in unfed tests using moderately hard reconstituted water (MHRW,

USEPA 2002) and Reformulated MHRW (Smith et al. 1997) as diluents with 48 h sulfate LC50s

obtained from fed, 7-d chronic tests in the same two diluents (Fig . 6). In both cases, average

LC50s for unfed tests were significantly lower than those in fed tests . This factor alone may

represent the discrepancy between predicted and observed results for C

. dubia for these tests

.

Others have shown that feeding algae to cladocerans during toxicity tests may reduce the toxicity

of metals because negative charges on algal cells bind and reduce the bioavailability of positively

charged metals (e.g., Taylor et al. 1998). At this point, it is unknown if this observed effect is

due to binding of sodium to algal cells, or increased robustness or lowered stress of test

organisms that were fed during testing . The most likely explanation is that of increased health of

test organisms because sodium and sulfate are highly stable as dissolve ions, but this is

conjecture at this point .

---

INHS

April 20, 2006

For the same range of variables (Cl = 10 to 500 mg/L, hardness -100 mg/L), the STR model was

less accurate at predicting toxicity to H. azteca . One reason for this is that the model was not

designed using results of tests with

H. azteca

(Mount et al. 1997), and

H. azteca

is generally

slightly more sensitive to sodium sulfate than C

. dubia . When combining all data from this study

and those from Soucek and Kennedy (2005), the range of LC50s for the two species are similar

(512 to 4,345 mg SO42-/L for

H. azteca,

and 1,116 to 4,220 mg SO42-/L for C.

dubia)

over the

range of hardness and chloride concentrations studied . While the range of LC50s for the two

species overlapped,H. azteca's average LC50 (2,158 mg SO42,/L, n = 43) was significantly lower

(p = 0.0316) than C. dubia's (2,562 mg SO427L, n = 42). Thus, as would be expected, the STR

model under-predicted toxicity to

H. azteca

when using the C .

dubia

output; however, a

difference in sensitivity does not entirely explain the trend shown in figure 5a. Observed survival

ofH. azteca got closer to model predictions as chloride increased from 10 to 500 mg/L, whereas

for C. dubia, chloride concentration did not appear to influence the accuracy of the model's

predictions. This is likely the result of the fact that between 5 and 25 mg Cl'/L, H. azteca 's

response to sulfate was strongly influenced by chloride concentration whereas C. dubia's was not

(Fig la). Between 25 and 500 mg Cl7L, the two species have similar responses to sulfate with

increasing chloride (Fig. lb), and the model appears to account for the additive toxicity of the

two ions .

When chloride was held constant (5 mg/L for C . dubia and 25 mg/L for Hyalella) and hardness

was varied from 100 to 600 mg/L, the STRmodel was relatively inaccurate in predicting toxicity

for both species. This finding is in agreement with Kennedy et al. (2005) who found that the

STR model over-predicted toxicity to C . dubia in sodium sulfate dominated coal-processing

effluents with hardness values in the 700 to 800 mg/L range . Furthermore, the trend of under-

prediction at hardness = 100 mg/L followed by increasing degrees of over-prediction at hardness

= 200 to 600 mg/L (see Fig 5b) was observed for both species . This is a reflection of similarity

with which the two species responded to sulfate with increased hardness levels as depicted in

figure 2. These data suggest that the STR model does not account for the protective effect of

hardness on major ion/TDS toxicity, but because of the presence of a pattern in the inaccuracy,

data from this study may be useful in improving the model

.

---

100

90

80

70

60

50

40-

30

I

a

20

10

0

100

90

80

70

60

50

40

30 -

20 -

10-

0

0

0

100

200

300

400

500

600

chloride (mgfL)

B .

0

•

Ceriodaphnia

0

observed % survival

A .

0

observed % survival

0

0

0

100

200

300

400

hardness (mg/L)

Figure 5. Percent survival of Ceriodaphnia and Hyalella at varying levels of chloride (A) and hardness

(B) as predicted by the STR model . Model inputs were ion concentrations at nominal sulfate LC50s, so

observed % survival in each case is 50% . STR does not predict for Hyalella so Ceriodaphnia predictions

were used for Hyalella test inputs .

•

Ceriodaphnia

Hyalella

0 Hyalella

•

Q

b

500

600

700

INNS

April 20, 2006

---

4000

-

3500

-

3000 -

2500

-

O

h

E 2000

-

0

.Vl

1500

1000

-

500

-

024h

048h

>3000

MHRW unfed

MHRW fed

RMHRW unfed

RMHRW fed

Figure 6. Effects of presence of food on toxicity of sulfate to

Ceriodaphnia dubia

in two types of test

waters

.

Hardness and chloride effects on acute sulfate toxicity to bivalves

.

Five toxicity tests were conducted with 4-day-old, juvenile unionid mussels (fatmuckets,

Lampsilis siliquoidea)

at the U.S. Geological Survey's Columbia Environmental Research

Center. When Ca:Mg and chloride concentration were held constant, increased hardness

appeared to slightly reduce the toxicity of sulfate to the juvenile mussels (Table 2), but

confidence intervals overlapped . Test water #4 was U.S. EPA's MHRW (USEPA 2002), and the

96-h LC50 in this water (1,727 mg/L) was fairly similar to, but slightly lower than LC50s for C

.

dubia

(2,050 mg/L), and the fingernail clam, S

.

simile

(2,078 mg/L) in MHRW. Chloride did not

appear to have a strong effect either, but increasing chloride concentration from 5 to 33 increased

the sulfate LC50 by approximately 100 mg/L (Table 2) . Interestingly, while hardness and

chloride did not strongly affect sulfate toxicity to freshwater mussels, Ca :Mg ratio appeared to

have a substantial effect . In tests conducted with Ca:Mg = 1.46, 96-h LC50s ranged from 1,727

to 1,822 mg S04/L, while at a ratio of 2.33, LC50s ranged from 3,377 to 3,729 mg SO

4/L. This

trend suggests that there may be a calcium threshold for unionids, which when reached provides

a certain level of protection against sulfate toxicity, but beyond which, additional calcium

provides minimal additional protection . Further testing should be conducted to verify these

trends .

INNS

April 20, 2006

---

INNS

April 20, 2006

Table 2. Toxicity of sulfate to juvenile fatmuckets (Lampsilis siliquoidea). All tests were conducted at 20 °C for 96

hours with 4d-old mussels. Mortality was evaluated at 48, and 96 hours. Sulfate LC50 values shown are based on

measured concentrations, and all treatments within a given test had the same chloride concentration and hardness

.

LC50s were generated using the Spearman-Karber method . One test was conducted for each diluent type

.

In tests with fingernail clams (Sphaerium simile), increasing hardness from -100 to 200 mg/L did

not have a substantial effect on sulfate toxicity, but an additional increase to hardness = 300

mg/L caused a significant increase in the mean 96-h LC50 value (Table 3) . When Ca:Mg and

hardness were held constant, increased chloride caused a substantial increase in mean LC50 from

2,184 mg/L (CI = 2 mg/L) to 2,598 (C1= 33 mg/L) (Table 3). In the unionid mussel, Toxolasma

texasiensis, chloride and bicarbonate were found to be equally important anions in the

hemolymph (see McMahon and Bogan 2001). Because bicarbonate is readily available via

respiration and metabolism, this mussel may not depend on external chloride concentrations to

the extent that some crustaceans do . If this is the case, the protective effect of chloride observed

forHyalella and Ceriodaphnia might not be manifest in some unionoidean bivalves . The

hardness effect observed in this study may be more widespread among aquatic phyla, because

calcium simply reduces gill permeability (Lucu and Flik, 1999; Pic and Maetz, 1981). However,

McMahon and Bogan (2001) state that unionoideans "generally lose capacity for osmotic and

volume regulation above 3-4 ppt" (salinity). TDS is a rough measure of salinity, and the TDS of

a sample of RMHRW with 2000 mg/L sulfate is 2.9 g/L or ppt (Soucek unpublished data)

.

Further experiments with freshwater bivalves are required to determine if there is an absolute

TDS level that is tolerable, or if the limit depends upon water quality characteristics such

as

chloride concentration and hardness .

Table 3 .

Effects of hardness and chloride on sulfate toxicity to fingernail clams (Sphaerium simile) .

All tests were

conducted at 22 °C for 96 hours. Sulfate LCSO values shown are based on nominal concentrations, and all

treatments within a given test had the same chloride concentration and hardness . LC50s were generated using the

Spearman-Karber method. Different capital letters following means indicate means are significantly different (p <

0.05) .

Diluent

type

MHRW

Hard 200 (1 .46)

Hard 300 (1 .46)

Hard 300 (2.33)

RMHRW

Cl 33/ hard 100

Test

mg/L

Cl

ratio

Ca:Mg

mg/L (CaCO 3)

hardness

(95% C.1 .)

48 h LC50

(95% C .I.)

96 h LC50

1

25

2.33

100

3183(2837-3571)

3377 (3205-3558)

2

25

2.33

300

3523 (unrel. var.)

3525 (3330-3731)

3

25

2.33

500

3574 (unrel . var.)

3729 (unrel var.)

4

5

1 .46

100

1702 (1598-1812)

1727 (unrel var.)

5

33

1 .46

100

2661 (2268-3123)

1822 (unrel var.)

n

Cl

mg/L

Ca:Mg

ratio

hardness

mg/L (CaCO3)

mean LC50, mg S04/1-

(std. dev)

3

4.5

1 .46

99

2184 (419) B

2

6.0

1 .46

205

2244 (204) AB

2

4.3

1 .46

292

2949 (427) A

2

4.4

2.33

285

2640 (129) AB

2

32.7

5.4

106

2178 (210) AB

3

34.4

1 .46

103

2598 (341) AB

---

Examining sulfate toxicity at different hardness levels reveals that hardness appears to ameliorate

sulfate toxicity to a variety of species in at least two different phyla (Arthropoda and Mollusca)

.

In fact, for three of the species tested in this study, the increase in sulfate LC50 from hardness =

100 mg/L to hardness = 300 mg/L was quite consistent at approximately 900 mg/L (Fig. 7) .

Ceriodaphnia

O hardness = 100

0 hardness = 300

N hardness = 500

n/a

Hyalella

Sphaerium

Species

tampsilis

INNS

April 20, 2006

Figure 7. Comparison of LC50s at different water hardness levels for four different species of freshwater

invertebrates tested in the laboratory .

7-day chronic sulfate toxicity tests with Ceriodaphnia dubia

.

Chronic toxicity in terms of survival and reproduction was less in RMHRW compared to

MHRW. The Least Observable Adverse Effects Concentrations (LOAEC) and No Observable

Adverse Effects Concentrations (NOAEC) for both survival and reproduction were influenced by

dilution water with those for RMHRW being higher than those for MHRW (Fig. 8 and 9). While

LOAECs for reproduction are quite low, it should be noted that I have had a continuous, self

sustaining, reserve culture of

Ceriodaphnia dubia

in MHRW spiked with 1,000 mg SO,/L since

at least August of 2004 . Organisms used in chronic testing for this study were cultured in

MHRW or RMHRW as appropriate .

---

s

120

100 -

40

35 -

y 30

-

25 -

3

20

-

9 15-

e 10-

5

0

If

500

1000 1500 2000

2500

3000

3500

sulfate (mg/L)

Figure 8. Mean percent survival of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests

with sulfate in Moderately Hard Reconstituted Water (MHRW). Bars and error bars indicate means and

standard deviations for three separate tests .

•

MHRW,

mean LOAEC = 899 mg/L

O RMHRW,

mean LOAEC= 1236 mg/L

0

500

1000

1500

2000

2500

sulfate (mg/L)

Figure 9. Mean reproduction of

Ceriodaphnia dubia

in 3-brood, static-renewal, chronic toxicity tests with

sulfate in Moderately Hard Reconstituted Water (MHRW) . Bars and error bars indicate means and

standard deviations for three separate tests .

INNS

April 20, 2006

-

R

-

-

•

MHRW,

mean LOAEC=2,216 rng/L

o RMHRW,

mean LOAEC = 3,000 mg/L

i~

---

INNS

April 20, 2006

Predictive equations for standard development

Multiple linear regression analysis with covariance with the sub-dataset (n = 42) described

previously using sulfate LC50 as the dependent variable and hardness, chloride and species as

independent variables resulted in the following equation :

LC50 = 1646+ 5.508(Hardness)

- 1.457(C1), R2 = 0.8408, p < 0.0001

This was the combined model for both species. The intercept for C. dubia was 1,828.07 and for

H. azteca it was 1,463 .93 .

Acknowledgement

This study was funded by the U.S. Environmental Protection Agency Grant # CP96543701-0

.

The unionid mussel toxicity testing was conducted by Dr. Ning Wang, Chris Ivey, and Dr. Chris

Ingersoll of USGS, Columbia Environmental Reasearch Center, Columbia, MO, through a

subcontract. The authors thank Alex Haldeman and Jens Sandberger of the Illinois Natural

History Survey for technical assistance, and Charles Stephan, USEPA, for helpful consultation .

Literature cited

Aladin NV, Potts WTW. 1995. Osmoregulatory capacity of the cladocera

.

J

Comp Physiol B 164:671-683

ASTM (American Society for Testing and Materials) . 2002a. Standard guide for conducting acute toxicity

tests on test materials with fishes, macroinvertebrates, and amphibians . E729-96 . American

Society for Testing and Materials, Philadelphia, PA, USA

.

ASTM (American Society for Testing and Materials) . 2002b. Standard guide for conducting three-brood,

renewal toxicity tests with Ceriodaphnia dubia . E1295-01. American Society for Testing and

Materials, Philadelphia, PA, USA

.

ASTM (American Society for Testing and Materials) 2005. Standard guide for conducting laboratory

toxicity tests with freshwater mussels (ASTM E2455-05). ASTM annual book of standards

volume 11.05, ASTM, West Conshohocken, PA

.

American Public Health Association, American Water Works Association, Water Environment Federation

.

1998. Standard methods for the examination of water and wastewater, 20th ed. APHA,

Washington, DC .

Bayliss D, HarrisRR

. 1988. Chloride regulation in the freshwater amphipod Corophium curvispinum and

acclamatory effects of external Cf

.

J

Comp Physic! B 158:81-90.

Borgmann U. 1996. Systematic analysis of aqueous

ion

requirements of Hyalella azteca : a standard

artificial medium including the essential bromide ion

. Arch Environ Contam Toxicol 30:356-363

.

Dorgelo J. 1981. Blood osmoregulation and temperature in crustacea

. Hydrobiologia 81:113-130 .

---

Appendix 1 . Summary of test conditions for conducting toxicity tests with juvenile mussels in

basic accordance with ASTM (2005a) .

Test species :

Test chemicals :

Test type :

Test Duration :

Temperature :

Light quality :

Light intensity :

Photoperiod :

Test chamber size :

Test solution volume

:

Renewal of solution

:

Age of test organism :

No. organisms per

test chamber :

No. replicate chambers

per concentration

:

Feeding:

Chamber cleaning:

Aeration :

Dilution water:

Dilution factor :

Test concentration

:

Chemical residues

:

Water quality :

Endpoint:

Test acceptability criterion :

Fatmucket

Sodium sulfate

Static

96 h (also check survival at 48 h)

20±1°C

Ambient laboratory light

200 lux

16L:8D

50 ml

30

After 48 h

<5 day old

5

4

No feeding

None

None

Reconstituted water at three hardness levels (100, 300, and 500

mg/L as CaCO3), and at two chloride levels (5 and 33 mg/L) at one

hardness level (100 mg/L as CaCO3)

0.5

Five concentrations and a control

Sulfate concentrations were determined at the beginning and the

end of each test

DO, pH, conductivity, hardness, and alkalinity were determined at

the control, medium, and high concentrations of chemicals at the

beginning and the end of each test

Survival (foot movement)

>90% control survival

INNS

April 20, 2006

---

CK

"w

State Water Survey Division

WATER QUALITY SECTION

AT

PEORIA, ILLINOIS

SWS Contract Report 283

SprIngt,a~ d1 !°'noe3

ACUTE TOXICITY OF CHLORIDES, SULFATES,

AND TOTAL DISSOLVED SOLIDS

TO SOME FISHES IN ILLINOIS

by

Paula Reed and Ralph Evans

Prepared for and funded by the

Illinois Environmental Protection Agency,

Division of Water Pollution Control

September 1981

Illinois Department

of

Energy and Natural Resources

t'?SiLii

"4 '

Y

L j:a:ul JIS~IYI

•irial.:~

Illinois EPA Exhibit

---

CONTENTS

PAGE

Introduction

•

•

. ,

•

•

•

•

•

•

1

Scope of study

.

.

.

•

•

•

•

, , , ,

•

.

•

•

•

•

•

•

.

•

3

Plan of report

•

3

Acknowledgments

.

.

.

•

•

•

•

•

•

.

•

•

•

,

•

.

•

•

3

pment and methods

•

.

•

•

•

•

•

.

•

•

•

•

,

3

Equipment modifications and appurtenances

.

•

.

•

.

.

•

•

•

•

•

•

•

4

Stock solutions and chemical analyses

•

•

•

•

•

•

5

Test specimens

.

.

.

•

. ,

. ,

•

,

•

•

•

•

•

,

.

•

6

Reactions of fishes

•

.

•

6

Chloride

10

Sulfate

12

Results and discussion

•

13

Chloride bioassays

•

15

Sulfate bioassays

•

15

Total dissolved solids

22

Summary and conclusions

•

29

References

30

_.pend~ccs

Ap:~endix A .

Appendix S .

A,;,endix C .

_endix D .

Appendix E .

Appendix F .

Observations of percent bass mortality,

chloride bioassays

•

•

•

•

•

•

•

•

•

•

•

-

Observations of percent bluegill mortality,

chloride bioassays

•

•

.

•

.

•

•

.

.

.

•

.

•

•

•

•

•

•

Observations of percent catfish mortality,

chloride bioassays

•

.

•

42

Observations of percent bass mortality,

sulfate bioassays

•

.

•

•

.

•

•

•

•

•

•

•

•

•

•

•

•

•

Observations of percent bluegill mortality,

sulfate bioassays

•

.

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

Observations of percent catfish mortality,

sulfate bioassays

•

•

•

•

•

.

.

.

•

•

•

•

•

•

•

•

•

•

33

39

44

47

49

---

ACUTE TOXICITY OF . CHLORIDES,

SULFATES,

AND TOTAL DISSOLVED SOLIDS

TO SOME FISHES IN ILLINOIS

by Paula Reed and Ralph Evans

INTRODUCTION

This report ?resents the results of a study undertaken to assess the

acute toxicity to certain fishes of various concentrations of chloride, sul-

[ate

and resultant total dissolved solids . A review of the results of the

. :jtot

::uality monitoring program developed by the Illinois State Water Survey

in cooperation with the U .S . Geological Survey during the period 1945-1971

an .. : Teoorted on by Larson and Larson (1957) , Harmeson and Larson (1969) ,

and

; ;_,:meson et al . (1973) suggests that chlorides, sulfates, and total dissolved

solids are not significant sources of pollution. After an evaluation of the

Mater survey's water quality data, Nienkerk and Flemal (1976) concluded that

the statewide discharge-weighted mean concentrations for these constituents

are as follows

:

Chloride :

25 mg/l

Sulfate :

70 mg/l

Total dissolved solids : 303 .mg/l

In li_;ht

o_f the rules governing maximum permissible concentrations of

-,,use substances in the waters of Illinois these mean concentrations are min-

.

. ..a :,

1o.ever Nienkerk and Flemal (1976) suggest that sulfate and chloride

are a-

-,;

-:nose :-ineral constituents most influenced by anthrouogenic proces-

,Ls

.

:.lhough

they speculate that a major source of sulfate in the waters of

nor

.e.ister. Illinois may be atmospheric fallout and a major source of chloride

rn the waters of southeastern Illinois may be the excessive seepage of saline

groundwater, the,.: nevertheless conclude that the principal causes of sulfate

ant -chloride concentrations exceeding background levels are such activities

ai : the use ofstreet de-icing salt, waste disposal, coal mining, and oil

:hv

:ork of Butts et al . (1976) confirmed that high chloride content in

":

:noes streams can be related to oil production and groundwater seepage

.

Tney found for some streamss of the Saline River basins that the chloride con-

tent exceeded 500 mg/1 about 10 to 45 percent of the tine. At the same stream

locations the total dissolved solids exceeded 1000 mg/I about 30 to 60 percent

of the time

.

More recently Toler (1980) reported that a reconnaissance of 50 stream

sampling sites on much of the surface-mined area in Illinois revealed sulfate

concentrations ranging from 25 to 4100 mg/1 . Indeed, sulfate was the major

mineral constituent in the samples from all sites . On the basis of comparisons

with streams having little or no upstream mining activities he concluded that

.

---

concentrations

of sulfate in excess-of-100 mg/l in base stream flow are prob-

ably attributable to drainage from mine spoils

.

The Illinois Pollution Control Board (1977, with amendments through 1979)

recognized the likelihood that excess mineral contributions from human activi-

ties are superimposed upon the background concentrations of certain minerals

in the state's surface waters . The limitations promulgated by the Board for

the three constituents (in milligrams per liter) are

:

In addition to the general stream standards and the public water supply

limitations the Board established the following rule regulating the total dis-

solved solids concentrations in effluent discharges

:

Total dissolved solids shall not be increased more than 750 mg/1

above background concentration levels unless caused by recycling or

other pollution abatement practices, and in no event shall exceed

3500 mg/l at any time ; provided, however, this Rule shall not apply

to any effluent discharging to the Mississippi River, which, after

mixing as set forth in Rule 201, meets the applicable water quality

standard for total dissolved solids

.

In this case the background concentration is that of the production water

.

And although an effluent can contain up to 3500 mg/l of total dissolved solids

(more where discharge is to the

Mississippi

River) the rule does not permit a

violation of the general stream quality standard of 1000 mg/l

.

The Board's regulations also stipulate, in part

:

Any substance toxic to aquatic life shall not exceed 1/10 of the

96-hour median tolerance limit (96-hr .-TL ) for native fish or

essential fish food organisms

.

m

The median tolerance limit (TL

)

is the concentration at which 50 percent

of the test specimens survive . It

ms

also referred to as TLSO, which is the

J,

•s

irnaticn used in this report. A 96-hour bioassay is a desirable minimum

lenith . Djring this study, an exposure time of 14 days (336 hours) was used

.

Of pertinent interest to this study is the validity of the maximum per-

missible concentrations of chloride (500 mg/1), sulfate (500 mg/1), and total

disselved solids (1000 mg/1) permitted in Illinois water in accordance with

the general stream quality rule . The intent of the rule, among others, is to

protect the state's waters for aquatic life . This study is also part of a

continuing effort to develop information useful to persons and agencies whose

activities relate to the enhancement of water quality in the streams and lakes

of Illinois

.

2

Chloride

Sulfate

Total

dissolved

solids

General stream quality

500

500

1000

Public water supplies

250

250

500

---

scope of Study

As part of this investigation certain

fishes native to Illinois lakes and

streams were exposed to varying concentrations of chloride, sulfate, and resul-

tant total dissolved solids in an effort to ascertain acute toxicity effects

.

The fishes used as test specimens were largemouth bass fingerlings, bluegill

fry, and channel catfish fingerlings . Thirty-three bioassays were performed -

requiring the use of 3360 test specimens

.

The bioassays were of 14-day durations and were performed with various

fish sizes and water temperatures . The dilution water was high in the salts

of calcium and magnesium with correspondingly high alkalinity .

Plan of Report

The report contains a description of the equipment and methods used for

all bioassays ; a two-part description of the observed reactions of fishes to

chloride and sulfate ; and a three-part discussion of the results concerning

chlorides, sulfates, and total dissolved solids . All data developed from the

bioassays are included in the appendices

.

Acknowledgments

This study was conducted under the general supervision of Stanley A

.

Changnon, Jr., Chief, Illinois State Water Survey, and Dr . William C . Ackermann,

Chief Emeritus, Illinois State Water Survey. Many persons of the Water Quality

Section assisted in the study . Dave Hullinger and Dana Shackleford provided

guidance and assistance in the analysis of chloride, sulfate, and total dis-

solved solids . Laurie Hebel, Lew Hoffman, and Rick Twait performed analyses,

lent direction to the operation of the dilution apparatus, and occasionally

maintained continuous 24-hcur observations of aquaria . Mr . Maurice Whitacre

of the Department of. Conservation offered advice on the maintenance of test

specimens and supplied many of them. Linda Johnson typed the original manu-

script, and Gail Taylor edited it . Illustrations were prepared under the super-

vision of John W . Brother, Jr .

EQUIPMENT AND METHODS

A modification of a proportional dilutor developed by Mount and Brungs

(1967) was used . Water flow was provided through 12 glass test chambers

.

Each chamber had a volume of 22 liters, and the flow rate, 113 milliliters per

minute (ml/min), produced a 95 percent volume displacement every 10 hours .

The apparatus permitted the flow of five different concentrations of toxicant

into duplicative test chambers, with two chambers available for control pur-

poses . All tests were performed for at least 14 days

.

3

---

Equipment Modifications andAppurtenances

Previous work by the Water Survey, involving stucjies of the acute toxic-

ity

to fishes of residual chlorine

and ammonia

(Roseboom and Richey, 1977),

copper (Richey and Roseboom, 1978), and zinc (Reed et al., 1980), relied on

a syringe style pipettor to inject an exact amount of toxicant from the con-

tainer of a stock solution to the mixing bowl of the dilutor apparatus. This

toxicant feed system is satisfactory when dealing with toxicant concentra-

tions of small magnitude. Since this study involved the use of toxicants gen-

erally exceeding 10,000 mg/l in test tanks, another method of delivery had to

be devised. The dilution apparatus used consisted of a chemical metering pump

supplied by Fluid Metering, Inc ., which derives its feed of stock solution

from a 200-liter container. The system operates in the following manner

.

During the cycling of the dilutor, the timer activates the water solenoid

valve to open and begin filling the dilution water chambers as it simultaneous-

ly engages the chemical metering pump to start pumping toxicant from the stock

solution container into the toxicant bowl. As water from the dilution water

chambers overflows into the water bucket, the bucket fills and descends, there-

by engaging the switch and breaking the electrical current . This shuts off

the water solenoid valve and the chemical metering pump . As dilution water and

toxicant combine in the mixing chambers, the water bucket arm rises to complete

the electrical circuit. Then the cycle repeats itself. The advantages of this

system are an easily adjustable volume and rate of feed at the pump, a fail-

safe design directly timed by dilutor function, an ability to maintain high

concentrations of toxicant in a flow-through unit, and a relatively low price

for a system comprising a timer, a chemical metering pump, and a water solenoid

.

A well on the laboratory site, in the same aquifer as the municipal wells,

was the source of water for the dilution apparatus

.

Two header boxes were used . The first one is a polyethylene plastic bar-

rel equipped with a thermoregulator which can be set at a desired temperature

.

Significant cooling from the pre-set water temperature energizes a relay which

activates a solenoid-controlled valve on a hot water line . Water flows from

the first polyethylene plastic barrel to a second polyethylene plastic header

box, where air agitation keeps the contents mixed and provides a sustained dis-

solved oxygen level .

The following characterize the dilution water used in the bioassays (all

values except pH are in milligrams per liter)

:

4

Chemical oxygen demand

Not detected

Magnesium

25 .3

Ammonia-N

0 .09

Iron

0.11

Nitrate-N

3 .6

Zinc

.07

Phosphate-P

0.20

pH

8.33

Sulfate

183

Hardness

412

Chloride

87

Alkalinity

291

Copper

.008

Cadmium

.004

Fluoride

0.79

Lead

< .08

---

Stock Solutions and Chemical Analyses

The sodium chloride stock solutions were prepared by dissolving technical

grade sodium chloride

in dilution water. Due to the rather low toxicity of

sodium chloride to fish, large quantities of toxicant were used daily in the

dilutor

. To accommodate the preparation of the toxicant and to assure its

thorough mixing, a circulating pump was used .

At least once during the first 24 hours of each bioassay, and generally

daily thereafter, chloride analyses were made by removing a sample from the

middle of each test chamber . All chloride determinations were performed in

accordance with the argentometric method. Results are expressed in mg/l chlo-

ride (C1

)

.

The sodium sulfate stock solutions were prepared by dissolving technical

azaJ2

sodium

ssulfate in dilution water . Due to the low solubility of sodium

~ul ; ate in

20

C dilution water, it became necessary to use dilution water heat -

' d to 30-35 C to achieve the desired stock concentration . Since sodium sulfate

is relatively low in toxicity, large volumes of toxicant were also used daily

in the proportional dilutor

.

AA circulating pump was utilized to facilitate

the preparation of the toxicant and to assure thorough mixing of the sodium

sulfate and dilution water. During the winter months it became necessary to

use

a

submersible thermostat heater and to supply aeration by means of air

stones in the stock solution container because the sulfate' stock solution

had a tendency to stratify

.

;1t least once during the first 24 hours of each bioassay, and generally

datl';

•

thereafter, sulfate determinations were made by removing a sample from

middle of each test chamber. All sulfate analyses were performed in ac-

with

with the turbidimetric method . A Bausch and Lomb Spectronic 20 was

,-,, .-;_for all absorbance readings . All results are expressed as mg/l sulfate

All

analyses were performed as outlined in

Standard Methods for the Ex-

.i,,:n.it:on of Water and Wastewater (American Public Health Association, 1975)

.

Hardness and alkalinity were determined in one control chamber and two

,

.

.

rst chambers on three occasions during each bioassay . Analyses for

mn:iucted on the same three occasions, but samples were taken from

cha .:burs rather than three. Dissolved oxygen levels, measured by a

1 :sw :rrings Instrument Model 57 oxygen meter, were recorded daily from

all test chambers . Water_ temperature also was measured daily by a standard

ara,?

•u

iatcd centigrade thermometer. Hardness determinations were by the EDTA

tirrinetric method with Eriochrome Black T as an indicator . Alkalinity and

p11 were determined by a Metrohm Herisau pH meter, Model 58B, with 0 .02

t1

H,SC4 as a titrant for alkalinity

.

Salinity and conductivity measurements of all test chambers were recorded

generally on a daily basis with a Yellow Springs Instrument S-C-T meter, model

33 . Analyses for total dissolved solids (TDS) were generally determined daily

5

---

m

Table

1

.

Test Conditions

for Chloride Bioassays

fish

weight

(grams)

Avoraye

fish

length

(cm)

Range

chloride

(mg/1)

Range

total

diss

.

solids

(mg/1)

Range

pH

(units)

Average

alkalinity

(mg/1)

Bass

8-6-79

4 .1

6460-9718

8.40-8.57

184

8-8-79

2 .8

9665-9713

8.45-8.46

184

8-9-79

3 . 1

9587

8.43-8.43

8-13-79

3 .1

10199-10947

8.52-8.54

201

8-27-79

3 .9

6119-9493

8 .45-8.59

210

10-22-79

10490-14075

18192-21313

11-5-79

2 .11

5 .1

5898-15308

9951-24529

8.30-8.48

271

11-12-79

2 .01

5 .2

9647-9847

16111-16178

8.22-8.38

298

1-21-80

3 .75

6.6

5358-11067

9741-19158

8.18-8.40

291

2-4-80

4 .38

6.8

6247-11371

10520-18869

8.19-8.39

293

11-10-80

1 .92

5 .6

5968-14126

10617-23289

8.20-8 .70

286

12-2-80

2 .26

5 .6

6237-14432

10981-23437

8.40-8.61

322

Bluegill

7-10-79

2 .64

5 .6

6825-10690

8 .42-8.70

211

7-16-79

4 .51

6 .6

6775-10704

8.28-8.62

221

7-24-79

7 .24

7 .3

5971-9161

8 .08-8 .54

186

11-15-79

2 .24

5 .3

11446-11646

19036-19161

8 .21

298

11-26-79

2 .31

5 .3

5277-11546

9434-18549

8.20-8.40

298

12-10-79

0 . 33

2 .8

5105-11231

9378-19143

8 .22-8.41

294

Catfish

8-18-80

1 .54

5 .6

5175-13783

8899-21265

7 .13-8.42

253

9-2-80

2 .37

6.4

5185-13151

8951-20618

8.32-8 .48

257

9-9-80

3 .51

7 .1

13340-13592

21287-21303

8.32-8 .33

267

---

J

Table

1

.

Concluded

Percent

Range

Avenr7e

dissolved

Average

species

Bass

8-6-79

hardness

(mg/1)

493

oxygen

saturation

92

temperature

( °C)

21

Range

salinity

conductivity

(micro-MHOS)

8-8-79

480

90

21

8-9-79

--

92

21 .7

8-13-79

480

97

20.8

8-27-79

514

93

21 .4

10-22-79

--

--

20.6

11-5-79

407

92

20.5

11-12-79

416

93

20 .3

--

--

1-21-80

415

85

20 .3

9 .2-20 .8

14000-29800

2-4-80

424

81

20 .1

9 .2-19 .0

13800-27200

11-10-80

515

83

19 .4

10 .7-23 .1

15800-31800

12-2-80

527

81

18.6

10.8-24 .3

15500-33200

Bluegill

7-10-79

365

86

22 .4

7-16-79

393

77

22 .7

7-24-79

429

80

21 .4

11-15-79

416

--

--

--

--

11-26-79

428

94

20.6

8 .8-19.5

13200-27800

12-10-79

416

96

20.4

9 .0-19.9

13800-28200

Catfish

8-18-80

383

86

17 .8

8.8-21 .9

12700-30300

9-2-80

387

81

19.9

8.7-21 .6

13000-30500

9-9-80

405

87

20.1

21 .0-21 .2

30500

---

from all test aquaria using filtration and residue on evaporation at 103 to

1050C . Some ranges and averages of these analyses along with other pertinent

data representing test conditions during each bioassay are included in tables

1 and 2 . Illumination for the 16-hour photoperiod was furnished by a combina-

tion of Duro-test and Wide Spectrum Gro-lux fluorescent lighting in circuit

with a timer

.

Test Specimens

Three native Illinois fishes were selected as test specimens for the chlo-

ride and sulfate bioassays . They were largemouth bass

(Micropterus salmoides),

bluegill

(Lepomis

macrochirus), and channel catfish

(Ictalurus punctatus)

.

Table 3 lists the type and number of fishes used, average weight of the fishes,

and sources of the fishes for each of the bioassays

.

All test specimens were acclimated to the 20

0

0 dilution water for a min-

imum of 10 days . When necessary, the temperature was increased 1

0

C per day

and maintained at the desired temperature for 10 days . Holding tanks were

continually flushed with dilution water to eliminate any metabolic waste

.

At the beginning of each bioassay, the temperature, salinity, conductiv-

ity, and toxicant concentration for each test chamber were determined. One

fish at a time was randomly placed in the different aquaria until each of the

12 chambers held 10 fish . Because of rapid mortality at high concentrations,

each test chamber was continuously monitored the first 32 hours, and the exact

time of each mortality was recorded . Appendices A, B, C, D, E, and F provide

the exact mortality times for largemouth bass, bluegill, and channel catfish .

After death, the fish were thoroughly blotted to remove excess moisture, and

their lengths and weights were determined

.

REACTIONS OF FISHES

It is customary to record the behavior of fishes exposed to toxicants

durinq the performance of bioassay work at the Water Survey . This is done for

several reasons . A principal one is the desire to develop information useful

to !."r:nrne1 in Illinois who have the responsibility for investigating fish

Y . :1l, ar determining the likely causes of fish mortality. Observations under

.

.

.:ro11 r: conditions of such factors as behavior during stress, sites of

^-orrnaging, changes in pigmentation, and body configuration may make it

pos-

sible to interpret similar observations under field conditions

.

A

control group of fish was maintained with each bioassay at the ratio

of 20 control fish to 100 test fish . The control fish were kept under exactly

tie same conditions as the test fish in all respects except for the addition

of the toxicant . There was never any occurrence of a mortality in the control

tanks at any time during the bioassays . All fish behaved normally and eagerly

accepted food .

a

---

Table

2

.

Test Conditions

for Sulfate Bioassays

Percent

Range

Average

dissolved

Average

species

hardness

oxygen

temperature

range

conductivity

(np/l)

saturation

('C)

salinity

(micro-MHOS)

Catfish

9

Average

fish

weight

(grams)

- -

Average

fish

length

(cm)

Range

sulfate

(mg/1)

Range

total

diss

.

solids

(mg/1)

Range

pH

(units)

Average

alkalinity

(mg/1)

Bass

9-22-60

1 .24

4 .8

7556-17484

13321-25469

8 .39-8 .53

267

9-30-80

1 .26

4 .7

8627-18868

12001-27277

8.39-8 .55

265

30-6-50

1 .31

4 .8

9953-14567

16104-23666

8 .44-8 .52

291

10-22-80

1 .45

5 .1

11201-18989

16306-29573

8 .60-8 .64

305

10-27-00

1 .77

5 .4-

10323-14907

17183-25986

8.41-8 .57

316

'-lcec_11

--!9-00

0

.67

3 .5

9801-17483

15400-26024

8 .50-8 .60

302

6-2-9D

0 .59

3 .5

9418-18009

15460-26611

8 .50-8 .65

299

c-,90

1 .09

4 .1

13483-13644

.21467-21456

8 .55-8 .59

--

ish

-.6-40

1 .01

4 .7

8845-18205

--

8 .48-8 .60

--

.--3-00

1 .27

4 .9

9n32-19245

13877-25968

8 .49-8 .63

296

-7-BD

1

.55

5 .2

6769-14564

10722-20440

8 .41-8 .60

262

-26-80

1 .85

5 .6

7019-15584

11052-22954

8 .46-8 .51

257

bass

416

86

20 .2

7 .0-19 .9

10500-28100

3-50-90

420

83

20 .3

6 .9-16 .8

10300-24300

4.51

84

20.0

9 .3-18 .3

13700-26500

'J J

461

83

20.2

9.0-19 .0

13300-26500

493

83

19 .9

9.0-19 .2

13600-28100

89

21 .0

9 .5-17 .2

15000-26000

.512

84

21 .7

8 .6-19 .1

13500-28100

81

20.8

12 .8-15 .5

19200-23100

507

Bo

20.9

7.9-17 .0

12000-25000

4A6

84

20.8

8 .8-16.2

13500-23700

440

82

21 .0

6 .3-15 .6

9800-23500

317

87

19 .4

7 .3-14 .1

10900-21100

---

Table 3 .

Types, Numbers, Weights,

Bass and bluegill used

in chloride bioassays

fell

into several distinct weight groups, as indicated

IDOL = Illinois Department of Conservation

Chloride

At high chloride concentrations,

channel catfish exhibited numerous symp-

toms of stress . At the beginning of each bioassay the fish experienced a def-

inite loss of equilibrium. This was accompanied by respiratory difficulty

;

opercular movement was rapid and shallow. Many individuals swam frantically

at the water surface . As time progressed, the eyes appeared glazed and respira-

tion became increasingly labored . In addition, the catfish assumed a variety

of positions in the water column . Some performed short bursts of swimming in

a ziczag fashion at the surface of the water . Others lay on their sides on the

bottom of the tank . Some of the fishes underwent a stiffening of their bodies

and maintained a position perpendicular to the bottom of the tank . Certain

individuals hung at the surface in this rigid position while others stood on

their tails

.

A few channel catfish experienced muscle spasms and twitching along with

tail chasing . Afterwards their bodies became rigid, and death soon followed

.

Certain physical characteristics that accompanied the catfish mortalities were

produced by chloride . They included hemorrhaging in the gills, in the brain,

and at the base of the pectoral fins . Curvature of the body was a common reac-

tion to the toxicant . Death was determined by lack of reaction to prodding

and the cessation of gill movement

.

The appetite of the channel catfish during the bioassay was a function of

the concentration of the chloride . In concentrations above 10,000 mg/1 chlo-

ride, the fishes completely ignored food . In the moderate range of approximate-

ly 7500-9000 mg/l chloride, their appetites fluctuated . Initially, the chloride

produced

a

suppression of the appetite. Later, after perhaps some acclimation

to the chloride, there was a slight improvement in appetite . Concentrations at

or below 5000 mg/1 chloride slightly decreased the appetite of the catfish ini-

tially, but after awhile all fish eagerly accepted food

.

10

and Sources of Fish Used in Bioassays

Bioassay

Type of

fish

No . of

fish

Average wt

.

of fish (grams)'

Sources of fish

Chloride

Bass

1200

2 .08

IDOC, Spring Grove ; Opel's Fish Hatchery, Worden, IL

4 .08

Bluegill

560

0 .33

IDOC, Spring Grove; Opel's Fish Hatchery, Worden, IL

2 .40

5 .9

catfish

260

2 .47

Seven Springs Fish Farm, Evansville, IL

i

Sulfate

Bass

600

1 .41

IDOL, Spring Grove; National Fish Hatchery, Hebron, Ohio

Bluegill

260

0.78

Fender's Fish Hatchery, Baltic, Ohio

Catfish

480

1.42

Seven Springs Fish Farm, Evansville, IL

---

The stress patterns of the bluegill exposed

to chloride concentrations

in

excess of 10,000 mg/l were similar to those of the channel catfish .

Initially

respiration was sluggish and there was a general darkening of body color .

The fish experienced a

loss of equilibrium,

lying on their sides at the surface

and floating sideways

.

Others attempted short dives downward in the water col-

umn and later floated back to the top .

Coughing and regurgitation were expe-

rienced by some bluegill in distress

.

As time progressed,

some of the fishes underwent a frenzied, convulsive

type of activity

.

Other bluegill became rigid and maintained a vertical posi-

tion in the water . The eyes appeared glazed . Death usually occurred within

nine hours and produced certain distinctive features, including flared gills,

severe curvature of the spine, and hemorrhaging in the gills and at the pec-

toral fins

.

At concentrations less than 10,000 mg/l chloride, the same stress pat-

terns occurred as noted before, but with less severity . Deaths seemed to

occur more quietly. There was apparent hemorrhaging at the gills as well as

the tail, at the base of .the dorsal fin, and in the head

.

The appetites of the bluegill exposed to chloride varied inversely with

the concentrations . In the higher concentrations, the fishes ignored food

completely. In lesser chloride concentrations, the bluegill initially would

refuse to eat, but as time continued there was a gradual improvement in appe-

tite from a poor to fair status . At concentrations of less than 5000 mg/l

chloride, all bluegill ate normally

.

The largemouth bass exposed to chloride concentrations in excess of 9000

mg/l revealed stress behavior patterns similar to those of the bluegill and

channel catfish . At the beginning of each bioassay, the fish would hover at

the water surface with respiratory problems . They exhibited a loss of equi-

librium by lying on their sides in the water column . Some bass attempted to

right themselves by diving down towards the bottom of the aquarium, but they

nearly always rose back to the surface . Certain individuals reacted to the

chloride through spinal curvature ; in a couple of severe cases, the body was

almost L-shaped and there was evidence of internal hemorrhaging .

Otherr signs of distress included coughing, regurgitation, and gulping of

water .

As

the fish neared death, respiration became more labored. Some ex-

pcrienced tremors or muscle spasms resulting in rapid bends or flips

.

Upon expiration the largemouth bass exhibited certain distinctive char-

acteristics as a result of their exposure to chloride . These included flared

gills, gaping mouth, loss of pigmentation, and hemorrhaging in the gills,

mouth, head, and at the base of the pectoral and caudal fins .

At chloride concentrations less than 9000 mg/l the stress symptoms were

the same as those at the higher concentrations, but they generally took longer

to occur and were less severe . The appetites of the largemouth bass also were

inversely correlated to the concentration of chloride . At the high concentra-

11

---

tions, the chloride suppressed all appetites, but as the percent of toxicant

present decreased, there was an initial absence of eating and then a gradual

improvement in their eating habits . Lower concentrations of chloride did no

adversely affect the appetites of the largemouth bass . Most ate well from

the beginning to the end of the bioassay . In fact some were eating as well

as the controls. This might indicate an acclimation to chloride .

Sulfate

Bluegill exhibited numerous symptoms of stress when exposed to sulfate

concentrations in excess of 15,000 mg/l . Typically there was an immediate

loss of equilibrium and general body control . Some fishes were observed ly-

ing on their sides at the surface, others were doing "barrel rolls," and

still others were seen diving to the bottom of the tank and floating .back

to the too . All were experiencing respiratory difficulty as they rapidly

beat their pectoral fins . As the bioassay continued, many bluegill preferred

to stay near the bottomm of the aquarium and exhibited very little movement

.

Breathing became more labored and sluggish .

Some noticeable symptoms of distress from the sulfate toxicant included

spinal curvature, tremors, flared gills, gaping mouth, and hemorrhaging in

the gills and head . Most bluegill underwent a change in pigmentation. Some

experienced a darkening of body color, while others were pale in color upon

death. in one instance, a fish displayed dark vertical bands above the lat-

eral line and light ones below . Spiny rayed fins were erect . In sulfate

concentrations greater than 10,000 mg/l but less than 15,000 mg/l, the stress

behavior was similar to that in the higher sulfate concentrations . Upon

introduction to the toxicant, many exhibited disorientation and visited the

surface briefly . Some were seen swimming sideways . Respiration was sluggish

and was accompanied by a rapid beating of the pectoral fins . There was a

change in pigmentation, with some becoming darker and others becoming lighter

in color . Apparently the sulfate solution irritated the muscle and nerve

tissues of certain bluegill to such an extent that they reacted by twitching

and trembling. As they neared death and were severely distressed, the fishes

stayed on the bottom of the tank .

:.t

concentrations less than 10,000 mg/l sulfate there was a drastic de-

crease in mortalities . Apparently after the initial shock was over, the blue-

gill

:radually acclimated to the toxicant . All mortalities involved distress

charatterintics exactly like those which occurred at the higher concentrations

.

Channel catfish appeared to react to the sulfate toxicant in a manner

.

similar to the bluegill . At the onset of each bioassay, there was a loss of

equilibrium. Some were seen stiffening their bodies and hanging vertically in

the water column at the surface. Opercular movement was rapid and shallow as

the catfish tried to compensate for the shock and introduction into a dif-

ferent fluid medium. Some fishes were so distressed by the sulfate toxicant

that they vomited. As time progressed respiration became increasingly,dif-

ficult and many rested on the bottom of the tank . Schooling behavior was

somewhat erratic at this point .

12

---

Certain distressed individuals underwent a tail chasing phenomenon

and

death tremors

.

Upon their expiration, many catfish displayed an open or

gaping mouth,

flared gills,

erect spiny-rayed

fins, curvature of the body,

and hemorrhaging at the base

of

the pectoral,

dorsal,

and caudal fins and in

the head .

In sulfate concentrations in excess of 15,000 mg/l, the largemouth bass

exhibited stress symptoms similar to those-of the bluegill and channel cat-

fish. Initially they hovered at the water surface with breathing difficul-

ties. There was a rapid fluttering of the pectoral fins as they tried to ad-

just to the toxicant . All experienced a loss of balance as they entered the

sulfate solution . Many rolled back and forth in a barrel roll fashion or

simply lay on their sides at the surface . Later it was noted that some fish

had spinal curvature . Muscle twitching was also displayed by a few individ-

uals . Generally, most mortalities occurred within 12 hours at the higher

concentrations . Many bass revealed gaping mouths, flared gills, and hemor-

rhaging at the head and operculum

.

At sulfate concentrations in the moderate range, between 10,000 mg/l

and 15,000 mg/1, the same stress symptoms were observed but appeared to be

less severe . As usual, the fishes experienced breathing difficulty at the

beginning of each bioassay. A loss of equilibrium followed, with some indi-

viduals lying on their sides . Several were observed swimming upside down

.

Man ,., , bass appeared darker in color as the bioassay continued . Death in the

moderate sulfate range was accompanied by distress characteristics similar

to those in the higher concentrations . These included curvature of the body,

an erect dorsal fin, open mouth, flared gills, and hemorrhaging at the oper-

culum and in the gills . Most mortalities occurred within 24-45 hours . The

a;.petites of the largemouth bass exposed to these sulfate concentrations

were: non-existent or very poor . Many completely ignored food or consumed a

little food now and then .

At loss than 10,000 mg/l sulfate, the bass appeared to be okay and acted

nor^.all

•;

after an initial adjustment period . Appetites were usually good, and

in fact many were eating as well as the controls . This might indicate an

acclimation to the sulfate toxicant at this level

.

RESULTS AND DISCUSSION

To estimate the median lethal time -- the time at which 50 percent .mor-

talit

• :

will occur in a particular test chamber -- the percent mortality for

that chamber and its duplicate is plotted against the observed time of mor-

tality . Figure 1 illustrates the procedure, showing that 50 percent mortality

occurred in duplicate chambers in 329 minutes (the median lethal time) at the

chloride concentration of about 10,900 mg/1 . In this manner median lethal

times and corresponding chloride concentrations have been determined for each

bioassay. An acute toxicity curve can then be developed by plotting the me-

dian lethal times against the corresponding chloride concentrations, as shown

13

i

V

---

1 4

99.9

99.8

99.5

99

98

95

90

80

Q

70

I-

60

0

50

1-

40

z

30

Cr

20

CL

10

5

2

1

0.5

0.2

0.1

100

1z

urc

1

.

Percent mortality for channel catfish

(Cl

)

-

f

1

I I

I

(I I

II

I

I

W

-

MEDIAN LETHAL. TIME

-

CHANNEL CATFISH - 9-2-80

o

TANK 1 AT 11,389 mg/I CHLORIDE

TANK 10 AT 10,508mg/I CHLORIDE

ESTIMATED MEDIAN LETHAL

_

TIME IS 329 MINUTES

I I l I I I I I

i I I

500

1000

TIME, minutes

5000

---

in figure

2

.

The arrow in figure 2 represents

the condition developed from

figure 1 .

If less than

50

percent mortality occurred in a test chamber with-

in 14 days, the time selected for representing the median lethal time is 14

days . For the purposes of this study, 24-hour and 96-hour designations are

alsoo included in addition to 14-day times

.

From the acute toxicity curves the

TL50

value is determined. The

TL50

is that concentration at which the curve becomes asymptotic to the time axis

.

As

mentioned previously, the waterr pollution regulations in Illinois re-

quire an application factor of 1/10 to the

TL50

for determining the maximum

permissible concentration of any substance toxic to aquatic life . Because

the

TL50

concentration is derived here from the acute effects of the substance

on fishes it is assumed that an allowable concentration of 1/10 the

TL50

con-

centration in Illinois waters will minimize chronic effects related to growth,

reproduction, and genetic characteristics of aquatic organisms. Nevertheless

the uniform application of the factor (1/10) for

all toxic substances is a

questionable practice without adequate substantiation for Illinois conditions

.

Under present conditions, however, the 1/10 factor is required and shall re-

main so until evidence has been developed to justify a reevaluation of its

usefulness

.

Chloride Bioassays

The reactions of catfish, bass, and

bluegill. to concentrations of chloride

are shown in figures 2, 3, and 4, respectively . It is apparent from these

figures that all three species of fish exhibit a similar sensitivity to chlo-

ride at water temperatures of about 20°C. The

TL50

concentrations range from

8000

to

8500

mg/l chloride with the bass appearing to be slightly more tolerant

to chloride thann the other two species

.

The figures also suggest that there is not a perceptible difference between

TLSO concentrations for bioassays with time lengths of 24 hours, 96 hours, or

14 days

.

S

•u

lfatn Pioansavs

Th.' reactions of catfish, bass, and bluegill to concentrations of sulfate

are ';hewn in figures

5,

6, and 7, respectively . Here also it is apparent that

all three species are similarly sensitive to sulfate at water temperatures of

about 20°C. The

TL50

concentrations at 14 days range from 10,000 to 11,000

mq/1 . Of the three species, bass is the least sensitive to sulfate

.

The figures also show that the

TL50

concentrations will differ depending

on the time length of the bioassay . Generally, the shorter the time length

of the bioassay (24 hours versus 96 hours versus 14 days), the higher the re-

sultant TL50, as shown in the figures . A summary of TL50s for figures 5, 6,

and 7 appears on page 22

.

15

---

1 6

50,000

10,000

100

1000

10,000

CHLORIDE, mg/I

I

I

I

1

11111

I

I

-

CATFISH

-

NaCl

14 DAYS

-At- AD

=96 HOURS

_24 HOURS

Ao

-

•

8.18.80

1

-

A 9-2-80

-

0 9-9-80

j F

14 day TL

50= 8000

mg/I CI' p

I I I 11111 \ I I

50,000

-

`

-

::_rura

Acute toxicity curve for channel catfish (Cl-)

---

I

50,000

10,000

1000

1

1579

2-4 .80

8.13-79

1-21-80

12 .2-80

11-5-79 -01~ 10-22-79

100-

14 day TL50 = 8500 mg/ICI' \-11'10

.80

1 i i

I

1

1

I

I

11 .5-79

1 -

1000

10,000

50,000

CHLORIDE, mg/I

Figure 3. Acute toxicity curve for Zargemouth bass (CZ

)

1 7

I

I

I

I

I

I

II

I

-

LARGEMOUTH BASS

NaCl

1-21-80

1-21-80

14 DAYS

\•

/,11 .10-80

1-12-2-80

.-1 .21-80

1 .21.80

11-10-80

:

12-2-80

I

=96 HOURS

-24 HOURS

-12.2 .80

2-4-80

_

8-6.79

2-4-80

-

11 .12.79

~

•

8-27-79

//11-10.80

_

8.9-79-

s

•

12-2-80

8.8.79

11 .10-80

-

8-27-79

10-22-79-0

10.22.79

---

18

14 day TL50 = 8000 mg/I CI

-

100 I I I I

IIIII

A

1

1 I I

1000

10,000

50,000

CHLORIDE, mg/I

Figure 4. Acute toxicity curve for bluegiZZ (CZ - )

50,000

I

1

I

I

1

I I l

BLUEGILL

NaCl

14 DAYS-_a

•a