APPENDIX I Toxicity Identification Evaluation Reports for Chollas ...

APPENDIX I 

Toxicity Identification Evaluation 

Reports for Chollas Creek and 

Sweetwater River

Toxicity Identification Evaluation (TIE) of 

County of San Diego and Copermittees 

Chollas Creek Stormwater Sample 

Prepared For: 


August 2006



Chollas Creek Stormwater Sample 

Prepared For: 


Prepared By: 

Weston Solutions, Inc. 

2433 Impala Drive 

Carlsbad, California 92010 

August 2006

Toxicity Identification Evaluation of Chollas Creek 

Stormwater using Hyalella azteca 

Table of Contents 

August 2006 

TABLE OF CONTENTS 

1. Executive Summary ..........................................................................................................1 

2. INTRODUCTION...............................................................................................................4 

2.1 Background and History ............................................................................4 

2.2 Toxicity Identification Evaluation (TIE) Testing..........................................5 

2.3 Initial Toxicity Testing Summary For Chollas Creek Stormwater...............7 

3. MATERIALS AND METHODS ..........................................................................................8 

3.1 Test Procedures ........................................................................................8 

3.1.1 Toxicity Test Using Hyalella azteca........................................................... 8 

3.2 Test Solution Preparation ..........................................................................8 

3.3 Water Quality.............................................................................................8 

3.4 Sample Receipt .........................................................................................8 

3.5 Phase I TIE Methods .................................................................................9 

3.5.1 Protocol Modifications................................................................................ 9 

3.5.2 Baseline Tests ........................................................................................... 9 

3.5.3 Ethylenediaminetetraacetic Acid (EDTA) Tests......................................... 9 

3.5.4 Sodium Thiosulfate (STS) Tests.............................................................. 10 

3.5.5 Aeration Tests.......................................................................................... 10 

3.5.6 Filtration Tests ......................................................................................... 11 

3.5.7 C 18 Solid Phase Extraction and Methanol Add-Back Tests ..................... 11 

3.5.8 Graduated pH Tests ................................................................................ 12 

3.5.9 Piperonyl Butoxide (PBO) Tests.............................................................. 12 

3.5.10 Carboxyl Esterase Tests.......................................................................... 12 

3.6 Statistical Analysis...................................................................................13 

4. Results ............................................................................................................................14 

4.1 TIE Tests on Chollas Creek Stormwater Using Hyalella azteca..............14 

4.1.1 Results of TIE Performed on the October 18, 2005 Sample ................... 14 

4.1.2 Summary of TIE Performed on the October 18, 2005 Stormwater 

Sample .................................................................................................... 16 

4.1.3 Results of Special Testing Performed on November 11, 2005 

Sample .................................................................................................... 17 

4.1.4 Summary of Specialized Testing Performed on November 11, 2005 

Sample .................................................................................................... 18 

4.1.5 Results of Phase I TIE Performed on January 2, 2006 Sample .............. 18 

4.1.6 Summary of TIE performed on January 2, 2006 Sample ........................ 20 

4.1.7 Results of Phase I TIE Performed on February 19, 2006 Sample........... 22 

4.1.8 Summary of TIE performed on February 19, 2006 Sample..................... 24 

4.2 Chemical Analyses of Chollas Creek Stormwater ...................................24 

5. DISCUSSION..................................................................................................................28 

6. REFERENCES................................................................................................................31 


i



Table of Contents 

August 2006 

LIST OF TABLES 

Table 1. Triad Definitions for San Diego Storm Water Monitoring Program. ................................4 

Table 2. Toxicity Identification Evaluation (TIE) Manipulations.....................................................6 

Table 3. Percent Survival of Hyalella azteca in TIE Tests performed on the Chollas Creek 

Stormwater Sample Collected on October 18, 2005. ..........................................................14 

Table 4. Percent Survival of Hyalella azteca in Special Tests performed on the Chollas Creek 

Stormwater Sample Collected on November 11, 2005........................................................17 


Stormwater Sample Collected on January 2, 2006 .............................................................19 


Stormwater Sample Collected on February 19, 2006..........................................................22 

Table 7. Chemical Analyses of Chollas Creek Stormwater Samples Collected During the 2005- 

2006 Monitoring Season......................................................................................................26 


ii



Acronyms and Abbreviations 

August 2006 

ACRONYMS AND ABBREVIATIONS 

BSA 

COC 

CDPR 

DO 

ECp 

EDTA 

EI 

GC-MS 

HCl 

KCl 

K ow 

LC 

LC 50 

MDL 

MEC 

MeOH 

MLS 

NaOH 

NICI 

NOEC 

NPDES 

PBO 

pH 

RL 

STS 

SPE 

TIE 

TRE 

TSS 

TUc 

USEPA 

Weston 

Bovine serum albumin 

Constituent of concern 

California Department of Pesticide Regulation 

Dissolved oxygen 

Estimated concentration causing an effect on p% of the population 

Ethylenediaminetetraacetic acid 

Electron Ionization 

Gas chromatography – mass spectrometry 

Hydrochloric acid 

Potassium Chloride 

Octanol – Water Partition Coefficient 

Lethal concentration 

Median Lethal Concentration 

Method detection limit 

MEC Analytical, Inc. 

Methanol 

Mass loading stations 

Sodium hydroxide 

Negative Ion Chemical Ionization 

No Observable Effect Concentration 

National Pollutant Discharge Elimination System 

Piperonyl butoxide 

Hydrogen ion concentration 

Reporting limit 

Sodium Thiosulfate 

Solid Phase Extraction 

Toxicity Identification Evaluation 

Toxicity Reduction Evaluation 

Total Suspended Solids 

Toxic Unit Chronic = 100/NOEC 

United States Environmental Protection Agency 



iii



Units of Measure 

August 2006 

UNITS OF MEASURE 

°C degree(s) Celsius 

> greater than 

< less than 

µg/L microgram(s) per liter 

µm micron(s) 

g 

gram 

L 

liter 

M 

moles 

mg/L milligram(s) per liter 

mL 

milliliter(s) 

mm millimeter(s) 

ng/L nanogram(s) per liter 

ppt 

parts per thousand 

% percent 

Weston Solutions, Inc. 1


Stormwater Using Hyalella azteca 

Executive Summary 

August 2006 

1. EXECUTIVE SUMMARY 

As part of the San Diego municipal stormwater monitoring permit (National Pollutant Discharge 

Elimination System [NPDES] Order 2001-01) for the San Diego, California region, stormwater 

runoff from Chollas Creek is evaluated during the wet weather season for chemical constituents, 

toxicity to test organisms, and health of the benthic community. A decision matrix based on 

these three lines of evidence is then used to determine whether Toxicity Identification 

Evaluations (TIEs) will be initiated on Chollas Creek samples to identify the causative agents of 

toxicity in stormwater samples collected during major storm events. The decision to initiate a 

TIE is based on the detection of high frequency contaminants of concern, persistent toxicity, and 

evidence of an impaired benthic community. In the 2004-2005 stormwater monitoring period for 

the San Diego County Municipal Stormwater Copermittees, high frequency constituents of 

concern (COCs), including turbidity, diazinon, total and dissolved copper, and total zinc were 

measured in Chollas Creek stormwater samples. In addition, persistent toxicity was found 

because more than 50% of the toxicity tests conducted to date with Hyalella azteca had a no 

observable effect concentration (NOEC) of less than 100%. The benthic community was rated 

as poor and was determined to be impacted. Based on these findings, it was determined that 

TIEs would be performed on Chollas Creek stormwater samples collected during the 2005-2006 

monitoring period if toxicity was observed in standard toxicity tests. 

During the 2005-2006 monitoring period, samples were collected from three major storm events 

and one minor storm event, and stormwater from the major storm events was analyzed for 

toxicity in tests using Ceriodaphnia dubia, Hyalella azteca, and Selenastrum capricornutum. No 

toxicity was found in tests using either C. dubia or S. capricornutum. Toxicity tests with the 

amphipod H. azteca showed significant toxicity during initial tests performed on samples 

collected October 18, 2005, January 2, 2006, and February 19, 2006. As a result, separate TIEs 

were conducted on stormwater samples from each of these three sampling events in an attempt 

to establish the potential cause or causes of toxicity. There was a minor storm event on 

November 11, 2005; however, at this time there was insufficient sample volume for the 

necessary physical, chemical, and biological analyses. However, this sample was used for 

additional specialized testing to confirm some of our findings from the October 18, 2005 storm 

event, and to compare toxicity of the stormwater sample tested in glass versus plastic 

containers. 

Stormwater toxicity tests and TIEs using the freshwater amphipod Hyalella azteca were 

performed according to a modified version of the USEPA protocol for testing sedimentassociated 

contaminants with freshwater invertebrates (EPA/600/R-99/064). TIEs were 

conducted according to guidelines for characterizing chronically toxic effluents (USEPA, 1991, 

1992, 1993a, and 1993b). Phase I TIEs included the following battery of tests to help establish 

potential causative agents of toxicity in stormwater samples: 

• Baseline tests were performed to benchmark toxicity of the unmanipulated stormwater 

samples run concurrently with the TIE tests for comparative purposes. 

• Filtration tests were performed to establish whether the chemicals causing toxicity were 

particulate bound or freely dissolved. 

• Aeration tests were performed to determine whether volatile chemicals and/or 

surfactants were potential causative agents of toxicity. 

• Graduated pH tests were performed to determine if the causative agents showed pHdependent 

changes in toxicity such as ammonia and aluminum. 

• Ethylenediaminetetraacetic Acid (EDTA) tests were performed to determine whether 

metals were a potential contributor to the toxicity of the sample. 





August 2006 

• Sodium Thiosulfate (STS) tests were performed to determine whether oxidative 

chemicals were contributing to toxicity of the sample. 

• Solid Phase Extraction (SPE) Tests (followed by methanol elution) were performed to 

evaluate whether non-polar organics could be contributing to toxicity of the sample. 

• Piperonyl butoxide (PBO) tests were performed to determine whether organophosphate 

pesticides or pyrethroids could be potential contributors to toxicity. 

• Carboxyl esterase tests were performed to help determine potential contribution of 

pyrethroids to toxicity. 

Results of TIE tests conducted on stormwater samples collected from Chollas Creek during the 

2005-2006 monitoring season provided strong evidence that pyrethroids were the causative 

agents of toxicity. TIE tests including the PBO and the carboxyl esterase tests, indicated that 

pyrethroids were a likely causative agent. PBO treatments led to increased toxicity in the 

stormwater samples, indicative of pyrethroids because PBO is known to potentiate pyrethroid 

toxicity by interfering with key metabolic pathways (e.g., the P-450 mixed function oxygenase 

system) important in the inactivation of pyrethroid compounds. Conversely, carboxyl esterase, 

a known antagonist of pyrethroid toxicity, reduced toxicity in the stormwater samples, also 

indicating pyrethroids as the causative agents. The carboxyl esterase enzyme has a strong 

affinity for pyrethroids and readily metabolizes pyrethroids to less toxic forms. Results of 

filtration tests and a special study performed on the November stormwater sample also provided 

evidence of pyrethroids. In the filtration test, toxicity was completed reduced in filtered 

stormwater samples, suggesting that the causative agents were bound to particulates. 

Pyrethroids are insoluble in water and have high adsorption coefficients, indicating their 

tendency to adsorb to particulates. As part of a special study conducted on the November 

stormwater sample, significantly higher toxicity was demonstrated in glass test chambers 

relative to plastic test chambers. Pyrethroids readily adsorb to plastic, leading to a timedependent 

reduction in toxicity for samples held in plastic containers. Finally, all other TIE 

manipulations indicated that the causative agents of toxicity did not share similar 

physicochemical properties to those of metals, oxidative chemicals, pH-sensitive chemicals, or 

volatile chemicals/surfactants, as demonstrated by the lack of toxicity reduction in the EDTA, 

STS, graduated pH, and aeration tests, respectively. 

Chemical analyses of Chollas Creek stormwater samples indicated that pyrethroids were 

present. Pyrethroids including bifenthrin and permethrin were measured in Chollas Creek 

stormwater samples at levels that exceeded aqueous bifenthrin and permethrin 96 hr median 

lethal concentrations (LC 50 s) for H. azteca. These results indicate that bifenthrin and permethrin 

were likely contributors to the toxicity observed in the Chollas Creek stormwater samples. A 

number of other pyrethroids were detected in stormwater samples including cyfluthrin, 

cypermethrin, and cyhalothrin. While the aqueous LC 50 s for H. azteca exposed to these 

pyrethroids are currently unknown, concentrations of these pyrethroids were comparable to 

those of bifenthrin and permethrin, suggesting that cyfluthrin, cypermethrin, and cyhalothrin 

pyrethroids also may have contributed to the toxicity observed in the Chollas Creek samples. In 

addition, PBO, a chemical used as a synergist in pyrethroid pesticide formulations, was 

detected along with pyrethroids in Chollas Creek stormwater samples. Other chemicals 

including metals were detected in the Chollas Creek stormwater samples; however, the 

detected concentrations were significantly below known effect levels for the test species and 

therefore were unlikely to have contributed to the observed toxicity. 

Results of the TIE manipulations taken in concert with results of chemical analyses indicate that 

pyrethroids are the likely primary cause of toxicity in Chollas Creek stormwater samples. It is 

interesting to note that this finding is consistent with recent changes in residential insecticide 





August 2006 

formulations where pyrethroids (e.g., bifenthrin) have replaced traditional organophosphate 

pesticides (e.g., diazinon and chlorpyrifos). 




Introduction 

August 2006 

2. INTRODUCTION 

2.1 BACKGROUND AND HISTORY 

The San Diego municipal stormwater monitoring permit (National Pollutant Discharge 

Elimination System [NPDES] Order 2001-01) for the San Diego, California region requires 

monitoring of stormwater runoff at ten mass loading stations (MLS) during the wet weather 

season within San Diego’s watersheds. The ten MLS evaluated in this program include: San 

Luis Rey River, Aqua Hedionda Creek, Escondido Creek, San Dieguito River, Penasquitos 

Creek, Tecolote Creek, San Diego River, Chollas Creek, Sweetwater River, and the Tijuana 

River. The determination of when to perform a Toxicity Identification Evaluation (TIE) at a MLS 

is identified by following a triad decision matrix method found in the Watershed Data 

Assessment Framework (MEC Analytical, Inc. [MEC]-Weston Solutions, Inc. [Weston], 2004). 

The triad decision matrix assesses data collected from the program including water chemistry 

and toxicity results from the mass loading stations, and results from benthic community 

structure analysis from rapid stream bioassessment. Definitions that are utilized to determine 

when a TIE is to be performed are listed below in Table 1. 

Table 1. Triad Definitions for San Diego Storm Water Monitoring Program. 

Triad Component 

Definition 

Persistent Exceedance of Water Quality 

Objectives 

Evidence of Persistent Toxicity 

Indication of Benthic Alteration 

A constituent of concern with a high frequency 

of occurrence 1 based on wet and dry weather 

data exceedances compared to established 

list of benchmarks or trigger levels 

More than 50% of the toxicity tests for any 

given species have a NOEC of less than 

100%. 

IBI score indicates a substantially degraded 

community (very poor) 

Source: Watershed Data Assessment Framework (MEC-Weston, 2004) 

Samples from three storm events are analyzed for toxicity using Ceriodaphnia dubia (7-day 

Survival and Reproduction), Hyalella azteca (4-day Survival), and Selenastrum capricornutum 

(4-day Growth). When findings from these toxicity tests at the mass loading stations indicate 

the presence of persistent toxicity, a TIE may be conducted to determine the potential cause or 

causes of toxicity. 

As listed in Table 1 above, toxicity test results are reported as the no observable effects 

concentration (NOEC). The NOEC is the lowest concentration at which there is no statistical 

difference from the control. Therefore, a concentration of less than 100% is considered to have 

some degree of toxic effect. Persistent toxicity is evident when more than 50% of the toxicity 

tests conducted to date for any given species at a specific site have a NOEC of less than 100%. 

The results of this determination are then combined with the high frequency constituents of 

1 A more detailed definition of high frequency occurrence of constituents may be found in Section 3 of the 

San Diego County Municipal Copermittees 2005-2006 Urban Runoff Monitoring Final Report 




Introduction 

August 2006 

concern (chemistry data) and benthic data in the Triad Decision Matrix to determine the actions 

to be taken. 

Results from chemistry, toxicity and relative benthic community health were assessed together 

using the triad approach to determine what short and/or long term actions are appropriate in a 

watershed. This approach examines persistence of toxicity using several indicators to provide 

an indication of an ecological concern. When persistence is found, this triggers the initiation of 

short term actions such as a TIE to identify the constituents of concern (COCs) in the watershed 

that may be responsible for stormwater toxicity and/or benthic community degradation). 

In the 2004-2005 stormwater monitoring period for the San Diego County Municipal Stormwater 

Copermittees, high frequency COCs including turbidity, diazinon, total and dissolved copper, 

and total zinc were measured in Chollas Creek stormwater samples. In addition, persistent 

toxicity was found because more than 50% of the toxicity tests conducted to date with this 

species had a NOEC of less than 100%. The benthic community was rated as poor and was 

determined to be impacted. Thus, it was determined that TIEs would be performed at Chollas 

Creek if toxicity was observed during the standard toxicity testing of stormwater samples during 

the 2005-2006 monitoring period. 

2.2 TOXICITY IDENTIFICATION EVALUATION (TIE) TESTING 

The United States Environmental Protection Agency (USEPA) has issued TIE testing guidelines 

for characterizing chronically toxic effluents (USEPA, 1991, 1992, 1993a, and 1993b). These 

guidelines are often effective for effluents that have similar toxic constituents to those identified 

in the model effluents used to develop the TIE guidelines. A Toxicity Reduction Evaluation 

(TRE) is an evaluation which involves the identification of toxicants, location of the source, 

and treatment of the causative agents to a less toxic form; the ultimate goal of a TRE is to 

reduce toxicity associated with contaminated water (or sediment). Thus, TIEs are important 

tools used in a TRE to initially help with the identification of toxicants, such that the source of the 

toxicant can be determined. 

The TIE typically consists of three test phases. Phase I of a TIE involves procedures designed 

to provide information for identifying the class of the toxic constituents within an effluent 2 sample 

based on their chemical characteristics (e.g., volatility, ionization state, degree of adsorption to 

particulates, polarity, oxidative state, pH sensitivity, and interaction with synergistic and 

antagonistic compounds). These classification characteristics are examined by comparing the 

results of tests conducted on raw effluent samples to effluent samples that have been physically 

or chemically manipulated. Phase I testing involves manipulating the sample at the effluent’s 

initial pH using the manipulations shown in Table 2 below. 

The goal of Phase II TIE testing is to identify the toxicants in the sample, and Phase III methods 

are used to confirm that the suspected toxicants are the true cause of toxicity in the effluent 

samples (USEPA, 1993a and 1993b). It should be noted that the boundaries between Phases I, 

II, and III are not distinct and there may be cases where it is appropriate for their respective 

procedures to overlap because confirmation information can be obtained during Phases I and II. 

2 The USEPA protocol is designed for performing TIEs on effluent samples, however, modifications have 

been made for performing TIEs on stormwater samples. 




Introduction 

August 2006 

TIEs are initiated when standard toxicity testing demonstrates toxicity in an effluent or 

stormwater sample that has previously been toxic to test organisms. Standard toxicity test 

methods sometimes rely on sublethal endpoints, such as C. dubia reproduction as indicators of 

chronic toxicity, and require substantially more time and resources to evaluate than methods 

that rely exclusively on a mortality endpoint. In addition, the USEPA guidelines do not provide 

specific guideline for test procedures using each toxicity test species, given the large number of 

test organisms. Therefore, the USEPA’s TIE documents are used as guidance for conducting 

TIEs because it may not be possible or cost-effective to strictly adhere to these protocols. In 

addition, the USEPA protocols were initially designed for TIEs on whole effluent samples, and 

not the more variable, temporally distinct, and less predictable stormwater samples. Specifically, 

in contrast to effluent samples, chemical constituents found in stormwater may vary extensively 

between storm events due to seasonal chemical applications associated with non-point source 

pollution (i.e., pesticide applications associated with seasonal crops), duration and extent of 

rainfall, and other water quality measures such as total suspended solids (TSS) and 

temperature. Thus, modifications for efficiently conducting TIEs on stormwater samples using 

specific test organisms, and for specific site conditions may sometimes include the following: 

changes in test volumes, test duration, replicate number, number of test concentrations, and 

reduction in frequency of test solution renewal. 

Phase I test procedures are designed to identify obvious alterations in effluent toxicity, which 

may be achieved using modified chronic test methods. 

Table 2. Toxicity Identification Evaluation (TIE) Manipulations 

Physical and Chemical Manipulations 

(Tests) on Stormwater Samples 

Filtration 

Aeration 

Graduated pH Adjustment 

Ethylenediaminetetraacetic Acid (EDTA) 

Addition 

Sodium Thiosulfate (STS) Addition 

Purpose of Test 

Detects filterable compounds (e.g., TSS related) 

Detects volatile, oxidizable, sublatable, or 

spargeable compounds 

Detects pH dependent chemicals (e.g., ammonia 

and sulfides) 

Detects cationic metals (e.g., cadmium) 

Detects oxidative compounds (e.g., chlorine) 

Solid Phase Extraction (SPE) over C 18 

Column, followed by Methanol Elution 

Piperonyl Butoxide (PBO) Addition 

Carboxyl esterase Addition 

Detects non-polar organics and some surfactants 

Detects organophosphate pesticides and 

pyrethroids 

Detects pyrethroids 




Introduction 

August 2006 

2.3 INITIAL TOXICITY TESTING SUMMARY FOR CHOLLAS CREEK STORMWATER 

Chollas Creek persistently exceeds water quality objectives for diazinon, chlorpyrifos, turbidity, 

total and dissolved copper, and total zinc. This site was previously identified as a candidate TIE 

site due to the persistent toxicity to H. azteca in this watershed; however, the cause of toxicity in 

recent years was unclear. In addition, the initial toxicity test from the October 18, 2005 storm 

event demonstrated that Chollas Creek stormwater caused significant toxicity. Survival in the 

6.25, 12.5, 25, 50, and 100 percent sample concentrations was 97.5, 87.5, 82.5, 42.5, and 

7.5%, respectively. The median lethal concentration (LC 50 ) at 96 hours was estimated to be 

44.1% of the stormwater sample, while the NOEC was 25%. Weston initiated Phase I TIE 

testing on October 26, 2005 with this sample utilizing the manipulations listed above. 

On November 11, 2005 an additional sample was collected from this site. Due to insufficient 

sample volume, this sample was not used to perform the standard stormwater toxicity tests with 

H. azteca or other species. However, this sample was used for additional specialized testing to 

confirm some of our previous findings, and to compare toxicity of the stormwater sample tested 

in glass versus plastic containers. 

Initial toxicity tests from the January 2, 2006 storm event also demonstrated that stormwater 

from Chollas Creek caused significant toxicity. Survival in the 6.25, 12.5, 25, 50, and 100 

percent sample concentrations was 100, 82.5, 92.5, 90, and 25%, respectively. The LC 50 at 96 

hours was estimated to be 76.6% of the stormwater sample, while the NOEC was 50%. Weston 

initiated Phase I TIE testing on January 11, 2006 with this sample. Additional treatments were 

added (i.e. carboxyl esterase) to help confirm whether pyrethroids were the likely causative 

agent of toxicity in this stormwater sample. 

Similar to previous stormwater events, initial toxicity tests from the February 19, 2006 storm 

event also demonstrated that stormwater from Chollas Creek caused significant toxicity. 

Survival in the 6.25, 12.5, 25, 50, and 100 percent sample concentrations was 100, 97.5, 92.5, 

55, and 2.5%, respectively. The LC 50 at 96 hours was estimated to be 76.6% of the stormwater 

sample, while the NOEC was 50%. Weston initiated Phase I TIE testing on March 2, 2006 with 

this sample. Similar treatments to those used on the January 11, 2006 TIE were used on this 

stormwater sample. 




Materials and Methods 

August 2006 

3. MATERIALS AND METHODS 

3.1 TEST PROCEDURES 

Bioassay methods for the species H. azteca test are from the USEPA guidance manual, 

“Methods for Measuring the Toxicity and Bioaccumulation of Sediment-associated 

Contaminants with Freshwater Invertebrates” (USEPA, 2000). 

3.1.1 Toxicity Test Using Hyalella azteca 

Stormwater tests for acute toxicity using the freshwater amphipod H. azteca were performed 

according to a modified version of the USEPA protocol for testing sediment-associated 

contaminants with freshwater invertebrates (EPA/600/R-99/064). This protocol provides test 

methods for measuring acute and chronic toxicity in Hyalella exposed to freshwater sediments, 

as well as a test method for conducting a water-only acute reference toxicant test. The 

reference toxicant test protocol was modified to conduct the toxicity testing on samples collected 

from the mass loading stations. The test solution was prepared and 250-mL aliquots were 

placed into four replicate test chambers. Clean sand was placed as a thin “monolayer” in the 

bottom of the test chamber. Temperature, pH, dissolved oxygen, and salinity were measured at 

test initiation prior to adding organisms. Ten organisms per replicate were added. Tests were 

run at 23 ± 1°C under a 16 hour light: 8 hour dark photoperiod. Water quality was performed 

daily on a surrogate chamber. The animals were exposed for four days, and fed on day 0 and 

2. At the end of the test, the survivors were removed from the sand and counted. Prior to 

analysis of the data, test acceptability was determined by evaluating the response of the control 

organisms. The test was considered invalid if survival of control animals was less than 90%. A 

reference toxicant test was conducted using copper sulfate with concentrations of 62.5, 125, 

250, 500, and 1000 µg Cu 2+ /L to establish the sensitivity of test organisms used in the 

evaluation of the Chollas Creek stormwater. 

3.2 TEST SOLUTION PREPARATION 

Control and dilution water for the H. azteca tests was Evian mineral water that was diluted 

with deionized water to achieve a moderate hardness (80-100 mg/L as CaCO 3 ). This water 

source has been used successfully on numerous similar bioassay testing programs conducted 

by Weston and others. Extensive testing with a variety of species and biannual chemical 

analysis of this water type has shown that this water source provides for good survival in 

laboratory controls with little to no measurable levels of contaminants. 

3.3 WATER QUALITY 

Water quality was monitored daily as appropriate for each test, and data were recorded on data 

sheets. Dissolved oxygen and temperature was measured using Orion Model 840 oxygen 

meters and probes; pH was measured using Orion Model 230A pH meters and probes. 

Conductivity was measured with Orion Models 142 conductivity/salinity meters. Ammonia 

was analyzed using an Orion 720 digital ion analyzer with a three-point calibration curve (1, 

10, and 100 mg/L). Hardness and alkalinity were measured utilizing LaMotte titration kits. 

3.4 SAMPLE RECEIPT 

The stormwater samples were composited at the laboratory and stored at 4°C. A chain-ofcustody 

was completed for all samples received. Before samples were used in the tests, initial 





August 2006 

water quality measurements were taken. These measurements included temperature, total 

chlorine, total ammonia, pH, dissolved oxygen, and salinity. 

3.5 PHASE I TIE METHODS 

3.5.1 Protocol Modifications 

Measures to conserve time and resources required to conduct TIE testing have been developed 

and approved by the USEPA (USEPA, 1992). This study incorporated modifications which 

allowed for the reduction in number of test concentrations, replicates, organisms per replicate, 

and volumes for all TIE tests conducted, relative to test conditions used in standard toxicity 

tests. The test concentrations used for the TIE tests performed in October and November of 

2005 included 50, 75, and 100 percent sample concentrations. In January and February of 

2006, only the 100 percent sample concentration from Chollas Creek was tested due to 

limitations in sample volume. For the H. azteca toxicity test, three replicates with five organisms 

per replicate in a volume of 50 mL were used in TIEs performed in October, November, and 

January. In the February TIE, the replicate number was increased to five replicates with five 

organisms per replicate in a volume of 50 mL, to decrease variability in the test results, and 

increase our power to detect differences among treatments. Treatment blanks were created for 

each TIE test to determine the effects of the manipulation on laboratory dilution water. The 

results of these blanks were used to determine if any changes in toxicity of the control (dilution 

water) were impacted by the chemical or physical manipulation of the sample. 

3.5.2 Baseline Tests 

The baseline test assesses the toxicity of the unmanipulated sample run concurrently with the 

TIE tests. This test confirms the presence of toxicity in the stormwater sample, and benchmarks 

the toxicity for comparison to toxicity in TIE treatments. 

On Day 1 of the October 18, 2005 Chollas Creek toxicity test, the dissolved oxygen level of the 

water quality surrogate for the 100 percent sample concentration briefly dropped below the 

protocol limit of 4.0 mg/L. The test was aerated and dissolved oxygen levels returned within 

range. To avoid a contribution of toxicity in the Baseline tests due to low dissolved oxygen 

levels, an Aerated and a Non-aerated Baseline test were performed. The Aerated Baseline test 

received head space aeration, at low air flow rates, for 5 minutes each day of high grade 

oxygen. It should be noted that this Aerated Baseline test differs from the aerated test below, in 

which the sample was aerated at high flow rates for an hour prior to test initiated. 

The Baseline tests conducted on the November 11, 2005 Chollas Creek sample were 

performed in both plastic and glass test chambers. This comparison was done because it was 

hypothesized that pyrethroids could be the causative agent for the toxicity measured in samples 

from this site. Specifically, pyrethroids are known to adsorb to plastic, leading to a timedependent 

reduction in toxicity (Wheelock et al., 2005). 

3.5.3 Ethylenediaminetetraacetic Acid (EDTA) Tests 

EDTA tests were performed to determine whether metals were the causative agent of toxicity in 

the Chollas Creek stormwater sample. Specifically, EDTA binds certain ionic metals, making 

them biologically unavailable to the test organisms. A reduction in the toxic response of a 

sample treated with EDTA may indicate the presence of divalent (metal) cation toxicity. 





August 2006 

Two series of EDTA tests were conducted with concentrations of 3.0 and 8.0 mg/L on the 

October 18, 2005 stormwater sample. First, a stock solution of 2.5 g/L of EDTA was prepared. 

EDTA treatments were prepared by diluting the EDTA stock solution to final concentrations of 

both 3.0 and 8.0 mg/L in stormwater (50, 75, and 100 percent sample concentrations) and 

dilution water. The EDTA was allowed to react with the sample for a minimum of 2 hours prior 

to the addition of the test organisms. Prior to test initiation, the pH was adjusted to the initial pH 

of the stormwater sample using 0.12 and 2.0 N hydrochloric acid (HCl) and 0.5 M sodium 

hydroxide (NaOH). 

Lower concentrations of EDTA were used on the January 2, 2006 stormwater sample, because 

the higher concentration of EDTA (8.0 mg/L) demonstrated toxicity to H. azteca in the dilution 

water control in the October Phase I TIE. Thus, EDTA tests were conducted with concentrations 

of 0.8 and 3.0 mg/L and were prepared as described above. For the TIE using the February 19, 

2006 stormwater sample, only one concentration of EDTA (3.0 mg/L) was used in the TIE, due 

to high survival in the control (blank) using this EDTA concentration. Specifically, we wished to 

use the highest concentration of EDTA in the test that would not cause toxicity to the test 

organisms. EDTA treatment blanks for all TIEs were prepared identically to the stormwater 

samples described above, using dilution water instead of stormwater. 

3.5.4 Sodium Thiosulfate (STS) Tests 

STS tests were performed to determine whether oxidative chemicals were the causative agent 

of toxicity in the Chollas Creek stormwater sample. Specifically, the addition of STS removes 

the effects of oxidative compounds including compounds such as chlorine, bromine, and ozone. 

This treatment also has success in removing some cationic metals. 

Two series of STS tests were conducted with concentrations of 10 and 25 mg/L. A stock 

solution of 2.5 g STS/L was prepared. STS treatments were then prepared by diluting the STS 

stock solution to final concentrations of 10 and 25 mg/L, separately, in stormwater (50, 75 and 

100 percent sample concentrations). In the TIE using the February 19, 2006 stormwater 

Sample, only one concentration of STS (10 mg/L) was used, due to high survival in the STS 

treatment control (blank) in the January TIE. STS treatment blanks for all TIEs were prepared 

identically to the stormwater samples described above, using dilution water instead of 

stormwater. 

3.5.5 Aeration Tests 

Aeration tests were performed to determine whether volatile chemicals and surfactants were the 

causative agent of toxicity in the Chollas Creek stormwater sample. Aeration helps reduce 

concentrations of volatile chemicals and surfactants. 

The stormwater sample (300 - 750 ml) was aerated with filtered air at a medium to high rate 

(500 mL air/min) for a minimum of one hour at test temperature. After one hour the sample was 

collected by siphoning from the middle of the beaker to avoid surfactants on the side of the 

beaker. This aerated sample was then used to prepare 200 mL dilutions of 50, 75 and 100 

percent samples. Aeration treatment blanks for all TIEs were prepared identically to the 

stormwater samples described above, using dilution water instead of stormwater. 





August 2006 

3.5.6 Filtration Tests 

Filtration tests were performed to determine whether the chemicals causing toxicity were bound 

to particulate matter in the Chollas Creek stormwater sample. A reduction in toxicity following 

filtration indicates that chemicals are particulate bound. 

The October 18, 2005 stormwater sample manipulation was prepared by filtering stormwater (3 

L) through a preconditioned nominal 1 µm groundwater sampling capsule using a peristaltic 

pump. Half of the filtrate (1.5 L) was reserved for the C 18 SPE Test described below. The 

remaining filtrate (1.5 L) was then used to prepare 200 mL dilutions of 50, 75 and 100 percent 

samples. On November 11, 2005 the filtration test was repeated using a 1 µm glass fiber filter 

and a vacuum pump to compare to toxicity resulting after filtration through the groundwater 

sampling capsule. Specifically, it was uncertain whether the groundwater sampling capsule was 

removing just particulate matter (i.e., the capsule may have removed adsorptive chemicals as 

well). The January 2, 2006 stormwater sample manipulation was prepared by filtering 1.25 L of 

stormwater. Approximately 360 mL of the filtrate was reserved for the C 18 SPE Test described 

below. The remaining filtrate (~890 mL) was reserved for the C 18 SPE. Stormwater (1.35 L) 

from the February 19, 2006 event was filtered and approximately 300 mL was used in the 

filtration test, while approximately 1 L was reserved for SPE. The filtration treatment blanks for 

all TIEs were prepared identically to the stormwater sample filtration described above, using 

dilution water instead of stormwater. 

3.5.7 C18 Solid Phase Extraction and Methanol Add-Back Tests 

SPE followed by methanol elution is a test performed to evaluate whether non-polar organics 

were the causative agents of toxicity. Specifically, non-polar organics are initially retained on the 

C 18 column and may be eluted with methanol. In subsequent toxicity tests, toxicity associated 

with the methanol extract is toxicity is indicative of non-polar organics. 

The October 18, 2005 stormwater sample manipulation was prepared by passing 1.5 L of 

filtered stormwater sample through a pre-conditioned 5 g C 18 SPE column at a rate of 5 mL per 

minute using a peristaltic pump. The January 2, 2006 stormwater sample manipulation involved 

extracting 887.6 mL of filtered sample through a pre-conditioned 1 g C 18 SPE column at a rate 

of 5 mL per minute using a peristaltic pump. The February 19, 2006 stormwater sample 

manipulation involved extracting 1.12 L of filtered sample through a pre-conditioned 1 g SPE C 18 

column at a rate of 5 mL per minute using a peristaltic pump. 

The SPE C 18 treatment blanks for all TIEs were prepared identically to the stormwater sample 

extractions described above, using dilution water instead of stormwater. 

The SPE C 18 column was then eluted with 100% reagent grade methanol to extract any 

contaminants bound to the column. To elute the column for the October 18, 2005 stormwater 

event, a total of 4.5 mL of 100% methanol was passed through the 5 g column in three, 1.5 mL 

aliquots, and collected into a test tube. The methanol was then added to clean control water at 

concentrations equaling 1.0, 2.0, and 4.0 times the concentration of the potential contaminants 

in the original stormwater sample. 

To elute the stormwater samples collected from January 2 and February 19, 2006, a total of 10 

mL of 100% methanol was passed through the 1 g column in five, 2 mL aliquots, and collected 

into a test tube. For this treatment, the methanol elution sample was concentrated using 

nitrogen gas to reduce the sample volume such that toxicity associated with methanol, as 





August 2006 

indicated by the October TIE, would be reduced. Specifically, the methanol elution sample was 

placed on a warming plate, and nitrogen gas was blown into the headspace of the vial until the 

volume was reduced to approximately 2-4 mL. The exact concentration was determined and the 

methanol elution was added to clean control water at a concentration of 2.0 times the 

concentration of the potential contaminants in the original stormwater sample. 

3.5.8 Graduated pH Tests 

Graduated pH tests are performed to see if the causative agents are pH-sensitive chemicals 

such as ammonia and aluminum. If changes in pH increase or decrease toxicity, this indicates 

that pH sensitive chemicals are present. 

Aliquots of stormwater (200 - 750 mL) and dilution water (750 mL) were elevated to test 

temperature and then adjusted to a pH of 6.8. These were used to prepare 200 mL dilutions of 

50, 75 and 100 percent samples. The pH was adjusted to the target using 4 N HCl and 1.5 M 

NaOH. Treatment blanks for all TIEs were prepared identically to the stormwater samples 

described above, using dilution water instead of stormwater. 

3.5.9 Piperonyl Butoxide (PBO) Tests 

PBO tests are performed to identify whether the causative agents are organophosphate 

pesticides or pyrethroids. Specifically, PBO blocks specific cytochrome P450 enzymes that are 

involved in metabolizing chemicals such as organophosphates to more toxic metabolites and 

chemicals such as pyrethroids to less toxic metabolites. Thus, if results from this test 

demonstrate increased toxicity in the stormwater sample, this is indicative of chemicals (e.g. 

pyrethroids) that are metabolized to less toxic forms by cytochrome P450 enzymes. In contrast, 

if the results demonstrate decreased toxicity in the stormwater samples, this is indicative of 

chemicals (e.g. malathion, organophosphates) that are metabolized to more toxic forms by 

cytochrome P450 enzymes. 

Two concentrations of PBO tests were used during each TIE. Concentrations of 0.025 and 

0.050 mg/L were used in October, November, and January, whereas concentrations of 0.0007 

mg/L and 0.025 mg/L were used in the February TIE. A stock solution of 2.5 mg PBO/L was 

prepared. PBO treatments were prepared by diluting the PBO stock solution to final 

concentrations of 0.0007, 0.025, and/or 0.050 mg/L, separately, in stormwater (50, 75 and 100 

percent sample concentrations) and dilution water. The initial phase of TIE testing on Chollas 

Creek indicated that pyrethroids could be the cause of concern for toxicity. Because 

pyrethroids can adsorb to plastic (Wheelock et al., 2005), this test was repeated on November 

15, 2005 using a new sample in both glass and plastic test chambers. 

In January, a reference toxicant test was also conducted concurrently with Phase I TIE tests to 

confirm that the concentrations of PBO used in the TIE treatments were at sublethal levels. 

PBO reference toxicant test used concentrations of 0.01, 0.05, 0.1, 0.5, and 1 mg/L. 

3.5.10 Carboxyl Esterase Tests 

Carboxyl esterase treatments are performed to remove toxicity in samples that may contain 

pyrethroids. Carboxyl esterase is an enzyme that degrades type I and type II pyrethroids. Thus, 

recent studies have been performed in which carboxyl esterase has been used to reduce 

pyrethroid-associated toxicity to test organisms such as H. azteca (Wheelock et al., 2004). The 

carboxyl esterase test method used in the January and February TIEs was provided to Weston 





August 2006 

Solutions by Bryn Phillips (personal communication, Granite Canyon Laboratory). In this 

method, the protein Bovine Serum Albumin (BSA) is used as a control for the 

esterase. Specifically, if toxicity is reduced in both the BSA treatment and the carboxyl esterase 

treatment, this indicates that both BSA and esterase are adsorbing the chemicals in the water 

samples. In contrast, if toxicity is only reduced in the carboxyl esterase treatment, this indicates 

that the chemicals in the water sample have been enzymatically altered by the carboxyl 

esterase, and that these chemicals may be pyrethroids due to the affinity of carboxyl esterase 

for these compounds. 

This sample treatment was only performed on the January 2, 2006 and February 19, 2006 

stormwater samples. A stock solution of 2 mg carboxyl esterase/mL was prepared. The 

carboxyl esterase treatment was prepared by diluting the carboxyl esterase stock solution to 

achieve a final concentration of 0.52 μg/mL (or 1.25 Units/ml) in stormwater (100 percent 

sample concentration only). A stock solution of 2 mg BSA/mL was also prepared. The BSA 

treatment was prepared by diluting the BSA stock solution to achieve a final concentration of 

0.52 μg/mL in stormwater (100 percent sample concentration only). 

3.6 STATISTICAL ANALYSIS 

At the conclusion of all tests, test species data were evaluated statistically using ToxCalc to 

determine the ECp, or estimated concentration that causes any effect, either lethal (LC) or 

inhibitory (sublethal; IC), on p% of the test population, and NOEC values. ToxCalc is a 

comprehensive statistical application that follows standard guidelines for acute and chronic 

toxicity data analysis. 

Statistical effects can be measured by the ECp. The LC 50 or LC 25 is the point estimate of the 

concentration at which a lethal effect is observed in 50% or 25% of the test organisms. ECp 

values include 95% confidence limits where available. The NOEC is the highest tested 

concentration at which mortality and other sublethal measured effects are not significantly 

different from the those in the control treatment. All statistics were run against treatment blanks 

to mitigate for any artifactual effect that the treatment had upon the toxicity. 

The non-parametric Wilcoxon Signed Rank test (WSR) was used according to Gilbert (1987) to 

determine whether or not there were statistically significant differences between survival of H. 

azteca in TIE treatments and survival of H. azteca in the Baseline, or untreated stormwater 

sample. Briefly, this method involved the following steps: (1) subtracting the concentrations of 

contaminants (for specific contaminants/classes separately) in sediment within and adjacent to 

the leasehold from the reference level, (2) ranking these differences according to their absolute 

values, and (3) calculating the WSR statistic of the test according to standard procedures 

(Gilbert 1987). The null hypothesis was rejected when the WSR statistic was below the 

Wilcoxon critical value (from a pre-determined table in Zar 1991). 





August 2006 

4. RESULTS 

4.1 TIE TESTS ON CHOLLAS CREEK STORMWATER USING HYALELLA AZTECA 

4.1.1 Results of TIE Performed on the October 18, 2005 Sample 

Results for the TIE performed on the October 18, 2005 sample are summarized in Table 3. 

Table 3. Percent Survival of Hyalella azteca in TIE Tests performed on the Chollas Creek Stormwater Sample 

Collected on October 18, 2005. 

Test 

Control 

(Blank) – 

Dilution Water 

Stormwater Dilution 

50% 

Stormwater 

(1X 1 ) 

75% 

Stormwater 

(2X 1 ) 

100% 

Stormwater 

(4X 1 ) 

NOEC LC 50 

Baseline 86.7 100 80 80 2 100 >100 

Non-aerated 

Baseline 

73.3 93.3 86.7 66.7 100 >100 

EDTA – 3 mg/L 60 66.7 66.7 40 100 >100 

STS - 10 mg/L 86.7 86.7 80 60 75 >100 

Aeration 53.3 93.3 80 33.3 100 97.37 

Filtration 46.7 100 100 100 100 >100 

Solid Phase 

Extraction (C 18 ) 

93.3 100 100 100 100 >100 

Methanol Elution 33.3 86.7 33.3 0 2X 2.2X 

pH 6.8 100 93.3 93.3 86.7 100 >100 

PBO - 0.025 mg/L 86.7 13.3 26.7 0



Results 

August 2006 

in the 100 percent sample concentration) 4 . Survival in the 3.0 mg/L EDTA treatment blank 

(dilution water) was 60%. A higher concentration of EDTA (8.0 mg/L) was also used; however, 

this EDTA concentration demonstrated toxicity in the dilution water (46.7% survival), indicating 

that higher concentrations of EDTA were toxic and thus not appropriate in TIEs with H. azteca. 

The 10 mg/L STS manipulation also did not reduce toxicity, but instead caused a slight increase 

in toxicity in the 100 percent sample concentration (60% survival) relative to toxicity in the 

unmanipulated Baseline test (80% survival in the 100 percent sample concentration). The 

NOEC was 75%. Survival in the 10 mg/L STS treatment blank was 86.7%. A higher 

concentration of STS (25 mg/L) was also used; however, this STS concentration demonstrated 

toxicity in the dilution water (53.3% survival), indicating that at high concentrations, STS may 

cause toxicity. 

The Aeration treatment also did not reduce toxicity in the 100 percent sample concentration 

(33.3% survival) relative to toxicity in the unmanipulated Baseline test (80% survival in the 100 

percent sample concentration). Survival in the Aeration treatment blank was 53.3%. These 

results indicate that volatile chemicals and/or surfactants were not the causative agent of toxicity 

in the stormwater sample. 

The Filtration manipulation using a 1 µm Groundwater Sampling Capsule filter slightly reduced 

toxicity in the 100 percent sample concentration (100% survival) from toxicity in the 

unmanipulated Baseline test (80% survival in the 100 percent sample concentration). Survival 

in the Filtration treatment blank was 46.7%. Similar results were found using a 1 µm Glass 

Fiber Filter with the same sample on November 8, 2005. Toxicity was reduced in the 100 

percent sample concentration (100% survival) relative to toxicity in the unmanipulated Baseline 

test (66.7% survival in the 100 percent sample concentration). Survival in the Filtration 

treatment blank was 86.7%. These results indicate that the causative agent may be associated 

with particulates in the stormwater sample. 

The SPE C 18 reduced toxicity in the 100 percent sample concentration (100% survival) relative 

to toxicity in the unmanipulated Baseline test (80% survival in the 100 percent sample 

concentration). Survival in the SPE C 18 treatment blank was 93.3%. Since this sample is prefiltered 

prior to the C 18 manipulation, the reduction in toxicity was most likely due to pre-filtration 

of the sample prior to C 18 elution. 

The methanol add-back manipulation resulted in 0% survival in the 4X methanol add-back 

concentration. Survival was 33.3% in the Methanol treatment blank indicating that the methanol 

concentration used in this manipulation likely contributed to the observed toxicity. 

The pH 6.8 manipulation did not result in a significant change in toxicity in the 100 percent 

sample concentration (86.7% survival) relative to toxicity in the unmanipulated Baseline test 

(80% survival in the 100 percent sample concentration). Survival in the pH 6.8 treatment blank 

was 100%. This indicates that the causative agent was likely not a pH sensitive chemical such 

as ammonia. 

4 It should be noted that toxicity in the initial toxicity test performed upon sample collection demonstrated 

significantly lower survival in the 100 percent sample concentration (7.5% survival, with an LC 50 of 44.1%) 

relative to toxicity in the Baseline test (80% survival). 




Results 

August 2006 

The 0.025 mg/L PBO treatment increased toxicity in the 100 percent sample concentration (0% 

survival) relative to toxicity in the unmanipulated Baseline test (80% survival in the 100 percent 

sample concentration). The NOEC was less than 50%, and the LC 50 was 32.5%. Survival in 

the 0.025 mg/L PBO treatment blank was 86.7%. The 0.050 mg/L PBO treatment also 

increased toxicity in the 100 percent sample concentration (20% survival) relative to toxicity in 

the unmanipulated Baseline test (80% survival in the 100 percent sample concentration). The 

NOEC was less than 50%, and the LC 50 was 27.27%. Survival in the 0.050 mg/L PBO 

treatment blank was 80 percent. The increase of toxicity in these manipulations suggests that 

pyrethroids may be a causative agent in this stormwater sample. Specifically, it is well-known 

that sublethal concentrations of PBO potentiate the toxicity of pyrethroids (Budavari, 1989). 

A copper sulfate reference toxicant was tested at nominal concentrations of 62.5, 125, 250, 500 

and 1000 µg Cu 2+ /L. The calculated 96-hour LC 50 (189.41 µg Cu 2+ /L) was within two standard 

deviations of the laboratory mean (365.19 µg Cu 2+ /L) at the time of testing. This indicates that 

the sensitivity of H. azteca used in this evaluation fell within the normal range. 

Survival of H. azteca in three of the treatment blanks (Aeration, Filtration, and Methanol Elution 

tests) was lower than expected. While reference toxicant tests demonstrated adequate 

sensitivity of these test organisms, these tests use a higher volume of water (200 mL) than in 

the TIE tests (50 mL), which likely is preferred by these test organisms. It is also possible that 

test organisms used in the October TIE tests were more sensitive. More specifically, lower 

tolerance to test conditions and lower survival in three of the treatment blanks may be related to 

the source and age of the test organisms. Due to the last second nature of the October 

stormwater sampling event, there was not sufficient time as to allow for the ordering of test 

organisms from the standard source (i.e. Aquatic Indicators). Thus the organisms used in the 

October TIE were those cultured according to standard protocols at Weston. As a result, slight 

differences in age and source of organisms may explain the low tolerance to TIE test conditions, 

and slightly reduced survival in the three treatment blanks. 

4.1.2 Summary of TIE Performed on the October 18, 2005 Stormwater Sample 

The results from the Phase I TIE performed on the Chollas Creek stormwater sample collected 

in October indicated that pyrethroids may be the cause of the toxicity observed in the initial 

toxicity tests. First, in the Baseline test, toxicity of H. azteca exposed to diluted and undiluted 

stormwater (80 to 100% survival) was significantly lower than that in the initial toxicity test (7.5 

to 42.5% survival). This indicates that during the time elapsed between the initial and baseline 

toxicity tests, the causative agent in the stormwater sample had degraded, or had significantly 

adsorbed to the plastic container in which the sample was held. Pyrethroids are known to 

adsorb to plastic, thus leading to a time-dependent reduction in toxicity (Wheelock et al., 2005). 

Second, the lack of toxicity reduction in the EDTA, STS, graduated pH, and aeration treatments 

indicates that the causative agent was likely not a metal, an oxidative chemical, a pH-sensitive 

chemical, or a volatile chemical or surfactant, respectively. However, the reduction in toxicity of 

H. azteca following filtration of the stormwater sample indicates that the causative agent was 

highly bound to particulates in the sample. Pyrethroids have physicochemical properties that 

match the results of the present TIE; pyrethroids are insoluble in water but soluble in solvents 

(have high K ow s), have low vapor pressures (indicating low volatility), and have high adsorption 

coefficients, indicating their tendency to adsorb to particulates (Kidd and James, 1991). In 

addition, toxicity was significantly increased in the diluted and undiluted PBO-treated stormwater 

samples, as compared to toxicity in the baseline test (untreated diluted and undiluted 

stormwater). These results also indicate pyrethroids as the causative agents because pyrethroid 

toxicity is potentiated by PBO (Budavari, 1989). 




Results 

August 2006 

4.1.3 Results of Special Testing Performed on November 11, 2005 Sample 

On November 11, 2005 a water sample was collected from Chollas Creek; however, there was 

not enough rain to consider this sampling event a storm event. Thus, there was not a large 

enough volume of stormwater to perform the standard toxicity tests using Hyalella azteca, 

Selenastrum capricornutum, and Ceriodaphnia dubia and to perform complete TIEs, as done in 

October. However, using the collected water, some TIE manipulations were performed using H. 

azteca as the test organism. The results of these tests are presented in 

Table 4. 

Table 4. Percent Survival of Hyalella azteca in Special Tests performed on the Chollas Creek Stormwater Sample 

Collected on November 11, 2005 

Test 

Control (Blank) 

– Dilution Water 


50% 

75% 

Stormwater Stormwater 

100% 

Stormwater 

NOEC LC 50 

Baseline (Glass) 80 86.7 66.7 40 75 98.44 

Baseline (Plastic) 93.3 100 100 80 100 >100 

PBO - 0.025 mg/L 

(Glass) 

73.3 26.7 0 0



Results 

August 2006 

The Plastic 0.025 mg/L PBO treatment led to increased toxicity in the 100 percent sample 

concentration (33.3% survival) relative to toxicity in the unmanipulated Plastic Baseline test 

(80% survival in the 100 percent sample concentration). The NOEC was 100%, and the LC 50 

was 70.8%. The Plastic 0.050 mg/L PBO treatment led to increased toxicity in the 100 percent 

sample concentration (0% survival), relative to toxicity in the unmanipulated Plastic Baseline 

test (80% survival in the 100 percent sample concentration). The NOEC was less than 50%, 

and the LC 50 was 36.1%. Survival in the Plastic 0.025 mg/L PBO treatment blank was 66.7% 

and survival in the Plastic 0.050 mg/L treatment blank was 86.7%. 





4.1.4 Summary of Specialized Testing Performed on November 11, 2005 Sample 

Results from the specialized tests performed on the Chollas Creek stormwater sample collected 

in November provided additional evidence that pyrethroids were a likely causative agent. In the 

baseline test, the higher toxicity in the glass test chambers relative to the plastic test chambers 

suggests that the causative agent had adsorbed to the plastic, thus leading to a reduced toxicity 

to organisms in the plastic test chambers. In addition, not only did PBO potentiate the toxicity, 

further indicating that the causative agent was pyrethroid-like chemicals, but there was also 

higher toxicity in the PBO treatments using glass containers relative to those treatments in 

which plastic was used. These results indicate that the causative agents have similar 

physicochemical properties to those of pyrethroids; the chemicals causing toxicity adsorbed to 

plastic, and their toxicity to H. azteca was potentiated by PBO. 

4.1.5 Results of Phase I TIE Performed on January 2, 2006 Sample 

Results for the TIE performed on the January 2, 2006 sample are summarized in Table 5. 

In the Baseline test, toxicity of H. azteca exposed to the 100 percent sample concentration 

(53.3% survival) was slightly lower than in the initial toxicity test (25% survival) started on 

January 2, 2006. The NOEC was 100%, and the LC 50 was greater than 100%. Survival in the 

dilution water control was 86.7%. 

The 0.8 mg/L EDTA manipulation did not reduce toxicity in the 100 percent sample 

concentration (40% survival) relative to toxicity in the unmanipulated Baseline test (53.3% 

survival in the 100 percent sample concentration). Survival in the 0.8 mg/L EDTA treatment 

blank was 93.3%. Higher concentrations of EDTA (3 mg/L) used as part of this manipulation, 

also did not reduce toxicity in the 100 percent sample concentration (26.7% survival). Survival 

in the 3 mg/L EDTA treatment blank was 86.7%. These results indicated that metals were likely 

not responsible for the observed toxicity. 

The 10 mg/L sodium thiosulfate (STS) manipulation did not reduce toxicity in the 100 percent 


(53.3% survival in the 100 percent sample concentration). Survival in the 10 mg/L STS 

treatment blank was 86.7%. A higher concentration of STS (25 mg/L) was also used in this 

manipulation; however, this concentration of STS caused toxicity in the dilution water control 

(60% survival), indicating that higher concentrations of STS are not appropriate for TIEs with H. 

azteca. 




Results 

August 2006 

Table 5. Percent Survival of Hyalella azteca in TIE Tests performed on the Chollas Creek Stormwater Sample 

Collected on January 2, 2006 

Test 




50% 

Stormwater 

75% 

Stormwater 

100% 

Stormwater 

(2X 1 ) 

Baseline 86.7 80 53.3 53.3 

Aeration 93.3 40 

pH 6.8 100 46.7 

Filtration 100 93.3 


(C 18 ) 80 100 

Methanol Elution 73.3 46.7 

STS - 10 mg/L 86.7 46.7 

EDTA - 3 mg/L 86.7 26.7 

PBO - 0.025 mg/L 86.7 0 

PBO - 0.05 mg/L 66.7 13.3 

Carboxyl esterase (0.52 

μg/mL) 80 86.7 

BSA (esterase control, 

0.52 μg/mL) 100 66.7 

1 

2X concentration is associated with Methanol Elution of C- 18 column only. 

The Aeration treatment did not reduce toxicity in the 100 percent sample concentration (40% 

survival) relative to toxicity in the unmanipulated Baseline test (53.3% survival in the 100 

percent sample concentration). Survival in the Aeration treatment blank was 93.3%. 

The Filtration manipulation reduced toxicity to organisms in the 100 percent sample 

concentration (93.3% survival) relative to toxicity in the unmanipulated Baseline test (53.3% 

survival in the 100 percent sample concentration). Survival in the Filtration treatment blank was 

100%. These results indicate that the causative agent is likely associated with particulates in 

the stormwater sample. 

In the SPE C 18 manipulation, no toxicity was observed in the 100 percent sample concentration 

(100% survival). Because this sample was filtered prior to the C 18 extraction, it is likely that the 

lack of toxicity observed was primarily due to removal of particulates by filtration. Nonetheless, a 

small fraction of toxicity may be explained by the C 18 extraction; there was slightly higher 

survival in the 100 percent sample concentration from the SPE C 18 manipulation (100% survival) 

relative to the 100 percent sample concentration from the filtration manipulation (93.3% 

survival). Survival in the SPE C 18 treatment blank was 80%. 

The methanol add-back manipulation resulted in 46.7% survival at the 2X methanol add-back 

concentration. Because this level of survival was similar to that found in the 100 percent sample 

concentration from the Baseline test (53.3% survival), this indicates that there was likely some 

elution of a toxic chemical off of the C 18 column. Survival in the Methanol treatment blank was 

73.3%. Results of the SPE C 18 manipulation together with methanol extraction indicate that the 

causative agent was likely not a non-polar organic compound. 




Results 

August 2006 

The pH 6.8 manipulation did not reduce toxicity in the 100 percent sample concentration (46.7% 

survival) relative to toxicity in the unmanipulated Baseline test (53.3% survival in the 100 

percent sample concentration). Survival in the pH 6.8 treatment blank was 100%. This 

indicates that the causative agent was not a pH sensitive chemical such as ammonia. 

The 0.025 mg/L PBO treatments led to increased toxicity in the 100 percent sample 


survival in the 100 percent sample concentration). Survival in the 0.025 mg/L treatment blank 

was 86.7%. The 0.050 mg/L PBO treatments also led to increased toxicity in the 100 percent 


(53.3% survival in the 100 percent sample concentration). Survival in the PBO 0.050 mg/L 

treatment blank was 66.7%. The increase of toxicity in these manipulations suggests that 

pyrethroids may be a causative agent in this stormwater sample. Specifically, it is well-known 

that sublethal concentrations of PBO potentiate the toxicity of pyrethroids (Budavari, 1989). 

A reference toxicant test was conducted using PBO with concentrations of 0.01, 0.05, 0.1, 0.5, 

and 1 mg/L to verify that the concentrations of PBO used in the Phase I TIE were at sublethal 

levels. The calculated 96-hour LC 50 was 0.41 mg/L. The NOEC (0.1 mg PBO/L) was above both 

concentrations of PBO manipulations used to evaluate the stormwater sample. This indicates 

that PBO alone is most likely not a factor in any of the observed toxicity. 

The carboxyl esterase treatment reduced toxicity in the 100 percent sample concentration 

(86.7% survival), relative to toxicity in the unmanipulated Baseline test (53.3% survival in the 

100 percent sample concentration). Survival in the carboxyl esterase treatment blank was 80%. 

In contrast, the BSA treatment did not reduce toxicity in the 100 percent sample concentration 

(66.7% survival), relative to toxicity in the unmanipulated Baseline test (53.3% survival in the 

100 percent sample concentration). Survival in the BSA treatment blank was 100%. These 

data indicate that the chemical(s) responsible for the observed toxicity were not just adsorbing 

to binding sites on proteins (BSA or carboxyl esterase), but were enzymatically altered by the 

carboxyl esterase. 





4.1.6 Summary of TIE performed on January 2, 2006 Sample 

Results from the TIE performed on the Chollas Creek stormwater sample collected in January 

provided additional evidence that pyrethroids were likely responsible for the observed toxicity. 

In accordance with the October TIE, there was a lack of toxicity reduction in the EDTA, STS, 

graduated pH, and aeration treatments, which indicates that the causative agent was likely not a 

metal, an oxidative chemical, a pH-sensitive chemical, or a volatile chemical or surfactant, 

respectively. However, the reduction in toxicity of H. azteca following filtration of the stormwater 

sample indicates that the causative agent was highly bound to particulates in the sample. 

These results support the evidence provided by the October TIE which indicate that the 

causative agents have similar physicochemical properties to those of pyrethroids. Also similar 

to results of the TIEs performed in October and November of 2005, PBO treatments led to 

increased toxicity in the undiluted stormwater sample. The repeated potentiation of toxicity using 

PBO in all recent TIEs provides strong evidence that pyrethroids are the causative agents, 




Results 

August 2006 

because pyrethroid toxicity is well-known to increase following the addition of PBO (Budavari, 

1989). The carboxyl esterase treatment caused a reduction in toxicity in the undiluted 

stormwater sample, whereas toxicity was not removed in the BSA treatment, used as a control 

for the carboxyl esterase. These results indicate that the chemical(s) causing toxicity in the 

stormwater sample was enzymatically degraded, and not just adsorbed to binding sites on these 

proteins (BSA and esterase). These data further support the idea that pyrethroids were the 

causative agent in the Chollas Creek stormwater sample because carboxyl esterase is an 

enzyme that metabolizes pyrethroids to less toxic forms. 




Results 

August 2006 

4.1.7 Results of Phase I TIE Performed on February 19, 2006 Sample 

Results for the TIE performed on the February 19, 2006 sample are summarized in Table 6. 


Stormwater Sample Collected on February 19, 2006 

Test 




100% 

Stormwater 

(2X 1 ) 

Baseline 100 28 

Aeration 84 36 

pH 6.8 100 16 

Filtration 100 100 


(C- 18 ) 

92 100 

MeOH Eluate 84 52 

STS - 10 mg/L 100 12 

EDTA - 3 mg/L 96 40 

PBO - 0.025 mg/L 100 4 

PBO - 0.0007 mg/L 100 28 

Carboxyl esterase (0.52 

μg/mL) 

96 60 

BSA (esterase control, 

0.52 μg/mL) 

100 20 

Carboxyl esterase plus 

PBO (0.025 mg/L) 

96 40 

In the Baseline test, toxicity of H. azteca exposed to the 100 percent sample concentration 

(28.0% survival) was slightly lower than that in the initial toxicity test (2.5% survival) started on 

February 19, 2006. The NOEC was 25%, and the LC 50 was greater than 51.6%. Survival in the 

dilution water control was 95%. 

The 3.0 mg/L EDTA manipulation did not significantly reduce toxicity in the 100 percent sample 


survival in the 100 percent sample concentration). Survival in the 3 mg/L EDTA treatment blank 

was 96%. 

The 10 mg/L sodium thiosulfate (STS) manipulation did not reduce toxicity in the 100 percent 

sample concentration (12% survival) relative to toxicity in the unmanipulated Baseline test (28% 

survival in the 100 percent sample concentration). Survival in the 10 mg/L STS treatment blank 

was 100%. 

The Aeration manipulation did not reduce toxicity in the 100 percent sample concentration (36% 





Results 

August 2006 

sample concentration). Survival in the Aeration treatment blank was 84%. These results 

indicate that the causative agent was not a volatile and/or a surfactant. 

The filtration manipulation significantly reduced toxicity to organisms in the 100 percent sample 

concentration (100% survival) relative to toxicity in the unmanipulated Baseline test (28% 

survival in the 100 percent sample concentration). Survival in the Filtration treatment blank was 

100%. These results indicate that the causative agent is likely associated with particulates in the 

stormwater sample. 

In the SPE C 18 manipulation, no toxicity was observed in the 100 percent sample concentration 

(100% survival). Because this sample was filtered prior to the C 18 extraction, and that toxicity 

was completed removed followed filtration (100% survival), it is likely that the lack of toxicity 

detected in this treatment was due to removal of particulates by filtration, and not by SPE. 

Survival in the SPE C 18 treatment blank was 92%. 

The methanol elution of the C 18 column resulted in 52% survival at the 2X methanol add-back 

concentration. This level of survival was higher than that found in the 100 percent sample 

concentration from the Baseline test (28% survival), indicating there was likely some elution of a 

toxic chemical off of the C 18 column. Survival in the Methanol test blank was 84%. 

The pH 6.8 manipulation did not reduce toxicity in the 100 percent sample concentration (16% 


sample concentration). Survival in the pH 6.8 treatment blank was 100%. This indicates that the 

causative agent was not a pH sensitive chemical such as ammonia. 

The 0.025 mg/L PBO test led to increased toxicity in the 100 percent sample concentration (4% 


sample concentration). Survival in the 0.025 mg/L PBO test blank was 100%. The 0.0007 mg/L 

PBO test, designed to mimic the levels of PBO detected in Chollas Creek stormwater, did not 

reduce toxicity in the 100 percent sample concentration (28% survival) relative to toxicity in the 

unmanipulated Baseline test (28% survival in the 100 percent sample concentration). Survival 

in the PBO 0.0007 mg/L treatment blank was 100%. The increase of toxicity in the 0.025 mg/L 

PBO manipulation suggests that pyrethroids may be a causative agent in this stormwater 

sample. 

The carboxyl esterase test reduced toxicity in the 100 percent sample concentration (60% 

survival), relative to toxicity in the unmanipulated Baseline test (28% survival in the 100 percent 

sample concentration). Survival in the carboxyl esterase treatment blank was 96%. In contrast, 

the BSA treatment did not reduce toxicity in the 100 percent sample concentration (20% 


sample concentration). Survival in the BSA treatment blank was 100%. These data indicate 

that the chemical(s) responsible for the observed toxicity were not just adsorbing to binding sites 

on proteins (BSA or carboxyl esterase), but were enzymatically altered by the carboxyl 

esterase. 

The carboxyl esterase plus PBO (0.025 mg/L) test demonstrated no significant reduction in 

toxicity (40% survival) relative to toxicity in the unmanipulated Baseline test (28% survival in the 

100 percent sample concentration). In contrast, PBO (0.025 mg/L) alone increased toxicity (4% 

survival), and carboxyl esterase alone reduced toxicity (60% survival). These results indicate 

that the synergistic effects of PBO together with chemicals in the sample are cancelled out by 




Results 

August 2006 

the enzymatic degradation of the chemicals in the sample, and provide more evidence that the 

causative agents of toxicity are pyrethroids. 





4.1.8 Summary of TIE performed on February 19, 2006 Sample 

Results from the TIE performed on the Chollas Creek stormwater sample collected in February 

re-confirmed that pyrethroids were likely responsible for the observed toxicity. In accordance 

with the October and January TIEs, there was a lack of toxicity reduction in the EDTA, STS, 

graduated pH, and aeration treatments, which indicates that the causative agent was likely not a 

metal, an oxidative chemical, a pH-sensitive chemical, nor a volatile chemical or surfactant, 

respectively. However, the reduction in toxicity of H. azteca following filtration of the stormwater 

sample indicates that the causative agent was highly bound to particulates in the sample. 

These results support the evidence provided by the October and January TIEs which indicate 

that the causative agents have similar physicochemical properties to those of pyrethroids. Also 

similar to results of the tests performed in October, November, and January, PBO treatments 

led to increased toxicity in the undiluted stormwater sample. The repeated potentiation of 

toxicity using PBO in all recent TIEs provides strong evidence that pyrethroids are the causative 

agents, because pyrethroid toxicity is well-known to increase following the addition of PBO 

(Budavari, 1989). The carboxyl esterase treatment caused a reduction in toxicity in the undiluted 

stormwater sample, whereas toxicity was not removed in the BSA treatment, used as a control 

for the carboxyl esterase. These results indicate that the chemical(s) causing toxicity in the 

stormwater sample was enzymatically degraded by carboxyl esterase, and not just adsorbed to 

binding sites on these proteins (BSA and esterase), and support the idea that pyrethroids were 

the causative agent in the Chollas Creek stormwater sample. Finally, results of the combined 

carboxyl esterase and PBO test demonstrate toxicity similar to the level of that in the 

unmanipulated Baseline test. Because In contrast, PBO alone increased toxicity while carboxyl 

esterase alone reduced toxicity, the lack of significant change in toxicity measured in the 

combined carboxyl esterase plus PBO test indicates that the synergistic effects of PBO together 

with chemicals in the sample are cancelled out by the enzymatic degradation of the chemicals in 

the sample, and re-confirm that the causative agents of toxicity are likely pyrethroids. 

4.2 CHEMICAL ANALYSES OF CHOLLAS CREEK STORMWATER 

Below is a summary of chemical analyses performed on Chollas Creek stormwater samples 

during the 2005-2006 monitoring season (Table 7). Briefly, with the exception of malathion, all 

organophosphate pesticides were below the detection limit. Malathion was found at 

concentrations ranging from 89 to 345 ng/L across all storm events, all of which were below the 

water quality objective of 430 ng/L. All dissolved metals analyzed were found at concentrations 

in the Chollas Creek water sample that were below their water quality objectives, with the 

exception of dissolved copper, which only exceeded its water quality objective in the February 

stormwater sample. Concentrations of total cadmium, chromium, nickel, and silver were below 

their respective water quality objectives. Concentrations of total copper, lead, and zinc in the 

October, January, and February samples exceeded their water quality objectives. Initial 

analyses of pyrethroids and PBO a chemical used as a pyrethroid synergist in pesticide 

formulations, were not detected in either the October, or the January stormwater samples. 

These samples were initially analyzed using the gas chromatography – mass spectrometry (GC- 




Results 

August 2006 

MS) in Electron Ionization (EI) mode. It was hypothesized that if the pyrethroids were present at 

concentrations below detection limits, the pyrethroid synergist PBO might still be measurable. A 

re-analysis of PBO on the GC-MS demonstrated that concentrations of 630 and 883 ng/L PBO 

were present in the October and January stormwater samples, respectively, from Chollas 

Creek. Further examination and re-analysis of the October and January stormwater samples 

from Chollas Creek on the GC-MS in a different mode (i.e., Negative Chemical Ionization, NCI) 

also demonstrated that a number a pyrethroids including bifenthrin (14-110 ng/L), cyfluthrin (80- 

110 ng/L), cypermethrin (13-53 ng/L), cyhalothrin (48-89 ng/L), and permethrin (116 ng/L) were 

present in these samples. Because the re-analysis of these stormwater samples occurred on 

February 21, 2006, the concentrations of pyrethroids measured in Chollas Creek stormwater 

samples in October of 2005 and January of 2006 may be lower than the actual concentrations 

due to degradation during this holding time. Chemical analysis of the February stormwater 

sample also demonstrated significant concentrations of bifenthrin (112 ng/L), cyfluthrin (185 

ng/L), cyhalothrin (27 ng/L), and permethrin (484 ng/L). 


Toxicity Identification Evaluation of Chollas Creek Stormwater Using Hyalella azteca 

Results 

August 2006 

Table 7. Chemical Analyses of Chollas Creek Stormwater Samples Collected During the 2005-2006 Monitoring Season 

10-18- 

05 

Sample 

10-18-05 Sample Reanalyzed 

on 02-21- 

06 Using NICI Mode 

on GC-MS 

1-2-06 Sample Reanalyzed 

on 2-21-06 

Using NICI Mode on 

GC-MS 

1-2-06 

2-19-06 

Chemical Class Chemical Measurement Fraction Units MDL RL 

Sample 

Sample 

General Chemistry Ammonia-N NA mg/L 0.01 0.05 0.64 0.55 0.48 

General Chemistry Dissolved Orthophosphate as P NA mg/L 0.0075 0.01 0.367 

General Chemistry Nitrate-N NA mg/L 0.01 0.05 2.75 0.82 1.02 

General Chemistry Nitrite-N NA mg/L 0.02 0.05 0.25 0.28 0.3 

General Chemistry Total Hardness as CaCO 3 NA mg/L 1 5 122 49.6 45.4 

General Chemistry Total Orthophosphate as P NA mg/L 0.01 0.01 0.41 

General Chemistry Total Phosphorus NA mg/L 0.016 0.05 0.505 

General Chemistry Total Suspended Solids NA mg/L 0.5 0.5 173 217 139 

Organophosphorus Pesticides Bolstar (Sulprofos) Total ng/L 10 20

Toxicity Identification Evaluation of Chollas Creek Stormwater Using Hyalella azteca 

Results 

August 2006 

Trace Metals Barium (Ba) Dissolved µg/L 0.1 0.5 19.6 

Trace Metals Barium (Ba) Total µg/L 0.1 0.5 110 92.6 

Trace Metals Beryllium (Be) Dissolved µg/L 0.1 0.5



Discussion 

August 2006 

5. DISCUSSION 

TIE test results provide strong evidence that pyrethroids are the causative agent of toxicity in 

the Chollas Creek stormwater samples collected during the 2005-2006 monitoring season. In 

TIEs performed on samples collected in October, November, January, and February, PBO 

treatments led to increased toxicity in the 100% stormwater sample. The repeated potentiation 

of toxicity using PBO in all TIEs performed on the 2005-2006 Chollas Creek stormwater 

samples suggests that pyrethroids may be the causative agents, because PBO is known to 

potentiate pyrethroid toxicity (Budavari, 1989). In the TIEs performed on January and February 

stormwater samples, a carboxyl esterase test was added to the TIE tests as an additional 

method for determining whether pyrethroids were the causative agents of toxicity. The carboxyl 

esterase enzyme caused a significant reduction in toxicity in the stormwater samples, indicating 

that the chemicals causing toxicity in the stormwater sample were enzymatically degraded by 

carboxyl esterase. Because carboxyl esterase enzymes have a strong affinity for pyrethroids 

and are known to metabolize pyrethroids to less toxic forms (Wheelock et al., 2004), these 

results support the idea that pyrethroids were the causative agent of toxicity in the Chollas 

Creek stormwater samples. 

Additional TIE tests also indicated that the causative agents of toxicity in stormwater had similar 

physicochemical properties to those of pyrethroids. Filtration of the stormwater sample 

completely reduced the toxicity to H. azteca in the 100% stormwater samples from October and 

February and almost completely reduced toxicity in the January stormwater samples, indicating 

that the causative agents of toxicity were highly bound to particulates in the stormwater sample. 

It is well known that pyrethroids are insoluble in water but soluble in solvents (have high K ow s), 

have low vapor pressures (indicating low volatility), and have high adsorption coefficients, 

indicating their tendency to adsorb to particulates (Kidd and James, 1991). In the October TIE, 

as well as in TIEs performed on Chollas Creek stormwater samples in previous years, toxicity of 

H. azteca exposed to diluted and undiluted stormwater (80 to 100% survival) was significantly 

lower than that in the initial toxicity test (7.5 to 42.5% survival). This indicates that during the 

time elapsed between the initial and baseline toxicity tests, the causative agent in the 

stormwater sample had degraded, or had significantly adsorbed to the plastic container in which 

the sample was held. Moreover, in the specialized test performed on the November Chollas 

Creek stormwater sample, significantly higher toxicity was demonstrated in glass test chambers 

relative to plastic test chambers further demonstrating that the causative agent had adsorbed to 

the plastic, thus leading to a reduced toxicity to organisms in the plastic test chambers. In 

addition, not only did PBO potentiate the toxicity, but there was also higher toxicity in the PBO 

treatments using glass containers relative to those treatments in which plastic was used. These 

results are not unexpected because in a study by Wheelock et al. (2005), it was demonstrated 

that pyrethroids are chemicals that are well-known to adsorb to plastic, thus leading to a 

significant time-dependent reduction in toxicity to species including C. dubia and H. azteca. 

TIE tests also indicate that the causative agents of toxicity in the Chollas Creek stormwater 

samples did not share similar physicochemical properties to those of many other classes of 

chemicals. The lack of toxicity reduction in the EDTA, STS, graduated pH, and aeration tests in 

TIEs performed on the October, January, and February stormwater samples indicates that the 

causative agent was likely not a metal, an oxidative chemical, a pH-sensitive chemical, or a 

volatile chemical or surfactant, respectively. 

In addition to TIE tests, chemical analyses of Chollas Creek stormwater samples indicate that 

pyrethroids are likely the primary causative agents of toxicity. Pyrethroids including bifenthrin 

and permethrin were measured in Chollas Creek stormwater samples at levels that exceeded 




Discussion 

August 2006 

aqueous bifenthrin and permethrin 96 hr LC 50 values for H. azteca of 9.3 ng/L and 21.1 – 47 

ng/L, respectively (Wheelock et al., 2005; Anderson et al., in press). A number of other 

pyrethroids were detected in stormwater samples including cyfluthrin, cypermethrin, and 

cyhalothrin. While the aqueous LC 50 values for H. azteca exposed to these pyrethroids are 

currently unknown, concentrations of these pyrethroids in Chollas Creek stormwater samples 

were comparable to those bifenthrin and permethrin, suggesting that cyfluthrin, cypermethrin, 

and cyhalothrin pyrethroids may have also contributed to the toxicity observed in the Chollas 

Creek samples. In addition, the chemical PBO was detected along with pyrethroids in Chollas 

Creek stormwater samples. PBO is chemical used as a synergist in pesticide formulations along 

with pyrethroids to enhance toxicity to target pests (USEPA 2006). As a consequence, even 

pyrethroids measured at concentrations below their LC 50 s would demonstrate more severe 

toxicity to test organisms due to the synergistic effects of PBO. 

Other chemicals were detected in the Chollas Creek stormwater sample; however, there is little 

evidence that any of these chemicals were a major cause of toxicity. All organophosphate 

pesticides were below their detection limits, with the exception of malathion. Although malathion 

was detected in Chollas Creek stormwater samples, concentrations of this chemical were below 

its water quality objective, indicating that alone this chemical would likely not cause adverse 

effects to H. azteca. Moreover, concentrations of malathion measured in Chollas Creek water 

samples are below the (LC 50 ) for malathion that has been determined for a number of aquatic 

invertebrate species (Mayer and Ellersieck 1986). It is possible that metals may have 

contributed to the toxicity observed in the Chollas Creek stormwater sample because 

concentrations of total copper, lead, and zinc were above their respective water quality 

objectives stormwater samples; however, concentrations of dissolved zinc and lead were below 

their water quality objectives. Because dissolved forms of metals take are more bioavailable to 

organisms, it is unlikely that zinc and lead were responsible for the observed toxicity in Chollas 

Creek stormwater samples. In addition, copper concentrations were below the LC 50 for H. 

azteca in either soft water (LC 50 = 56 μg/L; Borgmann et al., 2005), or hard water (LC 50 = 330 

μg/L, Weston Bioassay Lab data), further indicating that copper was not associated with the 

observed toxicity in stormwater samples. Nonetheless, in some studies, lead and zinc LC 50 s for 

H. azteca were above the total lead and zinc concentrations measured in stormwater samples, 

respectively (reviewed by Borgmann et al., 2005). 

Pesticide runoff into Chollas Creek is not surprising, because pesticides including diazinon have 

been frequently measured in Chollas Creek stormwater for a number of years (Weston 2005). 

However, diazinon concentrations in Chollas Creek that exceed the water quality objectives 

have diminished significantly over the last five years, in conjunction with reduced toxicity to the 

test species C. dubia. Nevertheless, the frequency of toxicity tests demonstrating persistent 

toxicity to H. azteca continues to be evident (i.e., more than 50% of the toxicity tests conducted 

to date have a NOEC of less than 100%), while no additional pesticides until now have been 

linked to this persistent toxicity. 

A re-examination of TIE results conducted on Chollas Creek stormwater, together with 

information on pyrethroid use in California further indicates the likelihood of pyrethroids as the 

causative agents of toxicity. In 1999, Chollas Creek stormwater caused significant mortality to 

C. dubia and TIEs provided evidence that organophosphates were the causative agents of 

toxicity (Schiff et al., 2000). During subsequent monitoring periods for the San Diego County 

Municipal Stormwater Copermittees (2000 to 2005), high frequency COCs included turbidity, 

diazinon, total and dissolved copper, and total zinc were measured in Chollas Creek stormwater 

samples. However, diazinon use and levels in stormwater diminished significantly during this 

time period, first dropping approximately 10 fold from 2001 to 2002, and then to below detection 




Discussion 

August 2006 

limits by 2005, likely due to the recent removal of pesticide formulations in the U.S. containing 

chlorpyrifos and diazinon. In contrast, pyrethroid use in California has increased over the same 

time period, and is used residentially in insecticides that previously had organophosphates such 

as diazinon and chlorpyrifos as the active ingredients (California Department of Pesticide 

Regulation, 2004). While TIEs performed in recent years on Chollas Creek stormwater samples 

were inconclusive, a re-analysis of TIE data from 2002-2005 demonstrated that may have been 

the causative agent of toxicity in these samples. Specifically, similar to the present study, PBO 

strongly enhanced the toxicity observed in the TIEs performed on samples collected in 2002- 

2003, indicating that pyrethroids were a possible cause of toxicity. Moreover, similar to the TIE 

performed on the October Chollas Creek stormwater sample in 2005, samples in TIEs 

performed in 2002 to the spring of 2005 demonstrated decreased toxicity after a short holding 

time in the laboratory. These results are likely because the causative agent of toxicity was 

strongly adsorbing to the plastic cubitainers, in which the samples were held. 

Results of this study provide strong evidence that pyrethroids are the causative agent of toxicity 

in Chollas Creek stormwater samples. TIE tests indicated that the causative agent(s) of toxicity 

shared all of the physicochemical properties of pyrethroids, and lacked properties that 

characterize other classes of chemicals. Moreover, using the NICI mode on the GC-MS, 

analytical results demonstrated the presence of pyrethroids including bifenthrin and permethrin, 

measured at concentrations in Chollas Creek stormwater samples that exceed the LC50 

concentrations for H. azteca. Finally, increased use of pyrethroids, as demonstrated by the data 

presented by the California Department of Pesticide Regulation, to replace the 

organophosphates previously used in residential insecticides also indicate the likelihood that 

pyrethroids are the causative agents in these stormwater samples. 




References 

August 2006 

6. REFERENCES 

Anderson, B.S., Phillips, B.M., Hunt, J.W., Connor, V., Richard, N., and R.S. Tjeerdema. In 

press. Identifying primary stressors impacting macroinvertebrates in the Salinas River 

(California, USA): relative effects of pesticides and suspended particles. Environmental 

Pollution. 

Borgmann, U., Couillard, Y., Doyle, P., and D.G. Dixon. 2005. Toxicity of sixty-three metals and 

metalloids to Hyalella azteca at two levels of water hardness. Environmental Toxicology and 

Chemistry, 24:641-652. 

Budavari, S., ed., 1989. The Merck Index. Merck & Co. Rahway, NJ, USA. 

California Department of Pesticide Regulation, 2004. http://www.cdpr.ca.gov/. Sacramento, 

California. 

Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution. John Wiley and Sons, Inc., 

New York. 

Kidd, H. and D.R. James, (Eds.). 1991. The Agrochemicals Handbook, Third Edition. Royal 

Society of Chemistry Information Services, Cambridge, UK, pp. 2-13. 

Mayer, F.L., and M.R. Ellersieck. 1986. Manual of Acute Toxicity: Interpretation and Data Base 

for 410 Chemicals and 66 Species of Freshwater Animals. United States Department of the 

Interior, U.S. Fish and Wildlife Service, Resource Publication 160. 

Schiff, K.C., Bay, S.M., and C. Stransky. 2000. Characterization of stormwater toxicants from an 

urban watershed to freshwater and marine organisms. Southern California Coastal Water 

Research Project. 1999-2000 Annual Report. 

United States Environmental Protection Agency (USEPA). 1991. Methods for Aquatic Toxicity 

Identification Evaluations. Phase I Toxicity Characterization Procedures. EPA/600/6-91/003. 

EPA Office of Research and Development. Second Edition. February 

United States Environmental Protection Agency (USEPA). 1992. Toxicity Identification 

Evaluation. Characterization of Chronically Toxic Effluents, Phase I. EPA/600/6-91/005F. 

EPA Office of Research and Development. May. 

United States Environmental Protection Agency (USEPA). 1993a. Methods for Aquatic Toxicity 

Identification Evaluations. Phase II Toxicity Characterization Procedures for Samples 

Exhibiting Acute and Chronic Toxicity EPA/600/R-92/080. EPA Office of Research and 

Development. September. 

United States Environmental Protection Agency (USEPA). 1993b. Methods for Aquatic Toxicity 

Identification Evaluations. Phase III Toxicity Characterization Procedures for Samples 

Exhibiting Acute and Chronic Toxicity EPA/600/R-92/081. EPA Office of Research and 

Development. September. 

United States Environmental Protection Agency (USEPA). 2000. Methods for Measuring the 

Toxicity and Bioaccumulation of Sediment-associated Contaminants with Freshwater 

Invertebrates. EPA/600/R-99/064. EPA Office of Water. March. 




References 

August 2006 

United States Environmental Protection Agency (USEPA). 2002. Short-Term Methods for 

Estimating the Chronic Toxicity of Effluents and Receiving Waters to Freshwater Organisms. 

EPA-821-R02-013. EPA Office of Water. Fourth Edition. October. 

United States Environmental Protection Agency (USEPA). 2006 (updated). Available Online. 

http://www.epa.gov/pesticides/health/mosquitoes/pyrethroids4mosquitoes.htm. 

Weston Solutions, Inc. 2005. San Diego County Municipal Copermittees 2004-2005 Urban 

Runoff Monitoring. Final Report. December. 

Weston, D.P., Holmes, R.W., You, J. and M.J. Lydy. 2005. Aquatic toxicity due to residential 

use of pyrethroid insecticides. Environmental Science and Technology, 39:9778-9784. 

Wheelock, C.E., Miller, J.L., Miller, M.J., Phillips, B.M., Gee, S.J., Tjeerdema, R.S., and B.D. 

Hammock. 2005. Influence of container adsorption upon observed pyrethroid toxicity to 

Ceriodaphnia dubia and Hyalella azteca. Aquatic Toxicology, 74:47-52. 

Wheelock, C.E., Miller, J.L., Miller, M.J., Gee, S.J., Shan, G., and B. Hammock. 2004. 

Development of toxicity identification evaluation procedures for pyrethroid detection using 

esterase activity. Environmental Toxicology and Chemistry, 23:2699–2708. 

Zar, J.H. 1999. Biostatistical Analysis. 4th Edition. Prentice Hall, New Jersey. 




Sweetwater River Stormwater Sample 

Prepared For: 


August 2006



Sweetwater River Stormwater Sample 

Prepared For: 


Prepared By: 


2433 Impala Drive 

Carlsbad, California 92010 

August 2006

Toxicity Identification Evaluation of Sweetwater 

River Stormwater Using Selenastrum 

capricornutum August 2006 

TABLE OF CONTENTS 

1. EXECUTIVE SUMMARY ..................................................................................................1 

2. INTRODUCTION...............................................................................................................3 

2.1 Toxicity Identification Evaluation (TIE) Testing..........................................4 

2.2 Initial Toxicity Testing Summary for Sweetwater Stormwater....................5 

3. MATERIALS AND METHODS ..........................................................................................5 

3.1 Test Procedures ........................................................................................5 

3.1.1 Toxicity Test Using Selenastrum capricornutum ....................................... 5 

3.2 Test Solution Preparation ..........................................................................6 

3.3 Water Quality.............................................................................................6 

3.4 Sample Receipt .........................................................................................6 

3.5 Phase I TIE Methods .................................................................................6 

3.5.1 Protocol Modifications................................................................................ 6 

3.5.2 Baseline Tests ........................................................................................... 6 

3.5.3 Ethylenediaminetetraacetic Acid (EDTA) Tests......................................... 6 

3.5.4 Filtration Tests ........................................................................................... 7 

3.5.5 C 18 Solid Phase Extraction and Methanol Add-Back Tests ....................... 7 

3.5.6 Piperonyl Butoxide (PBO)Tests................................................................. 7 

3.6 Statistical Analysis.....................................................................................7 

4. RESULTS..........................................................................................................................8 

4.1 Sweetwater River Selenastrum capricornutum..........................................8 

4.1.1 Results of TIE performed on October 18, 2005 Sample............................ 8 

4.1.2 Summary of TIE Performed on October 18, 2005 Sample ...................... 10 

4.1.3 Statistical Analysis: Differences in Growth Among Stormwater 

Dilutions................................................................................................... 10 

5. DISCUSSION..................................................................................................................11 

6. REFERENCES................................................................................................................13 

LIST OF TABLES 

Table 1. Triad definitions for San Diego Storm Water Monitoring Program.................................3 

Table 2: Toxicity Identification Evaluation Procedures .................................................................4 

Table 3: Summary of Growth in TIE tests on Sweetwater Stormwater Sample Collected on 

October 18, 2005.................................................................................................................10 

LIST OF FIGURES 

Figure 1. Growth of S. Capricornutum exposed to multiple dilutions of a Sweetwater River 

stormwater sample. Different letters indicate statistically significant differences among 

treatments where B>C>A>D................................................................................................10 


i




ACRONYMS AND ABBREVIATIONS 

°C Degrees Celsius 

EDTA 

Ethylenediaminetetraacetic Acid 

g 

Grams 

> Greater Than 

< Less Than 

L 

Liter 

LADPW 

Los Angeles Department of Public Works 

MeOH 

Methanol 

µL Microliter 

mg/L 

Milligram Per Liter 

mL 

Milliliter 

M 

Molar 

MLS 

Mass Loading Station 

NOEC 

No Observed Effect Concentration 

PBO 

Piperonyl butoxide 

ppt 

Parts Per Thousand 

% Percent 

KCl 

Potassium Chloride 

K ow 

Octanol – Water Partition Coefficient 

STS 

Sodium Thiosulfate 

SPE 


TSS 

Total Suspended Solids 

TDS 

Total Dissolved Solids 

TUc 

Toxic Unit Chronic = 100/NOEC 

USEPA 

United States Environmental Protection Agency 

Weston 



ii




1. EXECUTIVE SUMMARY 

As part of the San Diego municipal stormwater monitoring permit (NPDES Order 2001-01) for 

the San Diego, California region, stormwater runoff from Sweetwater River is evaluated every 

stormwater monitoring period for chemical constituents, toxicity to test organisms, and health of 

the benthic community. A decision matrix based on these three lines of evidence is then used to 

determine whether Toxicity Identification Evaluations (TIEs) will be initiated at Sweetwater River 

to identify the causative agents of toxicity in stormwater samples collected during major storm 

events. This decision matrix includes detection of high frequency contaminants of concern, 

persistent toxicity, and evidence of an impaired benthic community. In the 2004-2005 

stormwater monitoring period for the San Diego County Municipal Stormwater Copermittees, 

total dissolved solids (TDS) exceeded water quality objectives measured in Sweetwater River 

stormwater samples (Weston 2005). While no other contaminants exceeded water quality 

objectives in the 2004-2005 monitoring period, the organophosphate pesticide diazinon 

exceeded water quality objectives in six out of the nine storm events prior to the 2004-2005 wet 

weather season. In addition, persistent toxicity was found because more than 50% of the toxicity 

tests conducted to date with this species had a NOEC of less than 100%. The benthic 

community was determined to be impacted. Based on these findings it was determined that 

TIEs would be performed at Sweetwater River if toxicity was observed during the standard 

toxicity testing of stormwater samples during the 2005-2006 monitoring period. 

During the October 18, 2005 storm event, Sweetwater River water caused significant toxicity, 

measured as reduced growth, to S. capricornutum. The IC 50 at 96 hours was estimated to be 

95.7% of the stormwater sample concentration. The NOEC was 50% of the sample 

concentration. Weston initiated TIE testing on November 1, 2005 with this sample using the 

manipulations described below. 

All Selenastrum capricornutum baseline tests and TIE tests following sample manipulation 

followed general methods described in the U.S. EPA guidance manual, “Short-term Methods for 

Estimating the Chronic Toxicity of Effluents and Receiving Waters to Freshwater Organisms” 

(USEPA, 2002). TIEs were conducted according to guidelines for characterizing toxic effluents 

(USEPA, 1991, 1992, 1993a, and 1993b). Phase I TIEs included the following battery of tests to 

help establish potential causative agents of toxicity in stormwater samples: 

• Baseline tests were performed to benchmark toxicity of the unmanipulated stormwater 

samples run concurrently with the TIE tests for comparative purposes. 

• Ethylenediaminetetraacetic Acid (EDTA) tests were performed to determine whether metals 

were a potential contributor to the toxicity of the sample. 

• Solid Phase Extraction (SPE) Tests (followed by methanol elution) were performed to 

evaluate whether non-polar organics could be contributing to toxicity of the sample. 

• Piperonyl butoxide (PBO) tests were performed to determine whether organophosphate 

pesticides or pyrethroids could be potential contributors to toxicity. 

Results of TIE tests conducted on stormwater samples collected from Sweetwater River during 

the 2005-2006 monitoring season provided some evidence that one or more ions comprising 

TDS were the causative agents of toxicity. The results of the 8 mg/L EDTA manipulation 

demonstrated that the 8.0 mg/L EDTA test significantly reduced toxicity in the 100% sample 

concentration (106.96 RFU Chlorophyll) relative to toxicity in the unmanipulated Baseline test 

(69.70 RFU Chlorophyll in the 100% sample concentration). This indicated that the causative 

agent of toxicity was a cation, such as a free metal ion (e.g., cadmium, mercury, lead) or a 





dissolved inorganic salt (e.g., sodium, calcium, magnesium), or a cationic organic compound 

such as a cationic surfactant, because EDTA is a chelating agent that is well known to bind 

inorganic and organic cationic chemicals (USEPA 1991). Chemical analyses of Sweetwater 

stormwater samples demonstrated that TDS measured in the Sweetwater sample was 2640 

mg/L, a concentration that exceeds the water quality objectives for this measure, or 1500 mg/L, 

in the lower Sweetwater River (RWQCB 1994). In contrast, no organophosphate pesticides 

were detected in stormwater samples and the only metals detected were arsenic, nickel, and 

zinc, all of which were found at low concentrations, or concentrations far below water quality 

objectives. In addition, the causative agents of toxicity in the Sweetwater River stormwater 

samples did not share similar physicochemical properties to those of many other classes of 

chemicals. The lack of toxicity reduction in the SPE and PBO tests indicate that the causative 

agent was likely not a non-polar organic, or a common-use pesticide, respectively. 

It is well known that elevated concentrations of ions comprising the TDS may be toxic to aquatic 

organisms. Thus, the results of this TIE provide some evidence that the causative agent of 

toxicity may have been due to elevated TDS levels in the Sweetwater River sample collected in 

October 18, 2005. No additional toxicity was found in Sweetwater River samples in subsequent 

standard S. capricornutum tests during the 2005-2006 monitoring period, indicating that the 

causative agent was not persistent during this monitoring period. Additional studies would be 

necessary to confirm whether elevated TDS in Sweetwater River water were the causative 

agent of toxicity to S. capricornutum. 





2. INTRODUCTION 

The municipal stormwater monitoring permit for the San Diego, California region requires the 

monitoring at ten mass loading stations (MLS) during the wet weather season. The ten MLS 

evaluated in this program include: San Luis Rey River, Sweetwater River, Tijuana River, 

Tecolote Creek, Aqua Hedionda Creek, San Dieguito River, San Diego River, Chollas Creek, 

Penasquitos Creek, and Escondido Creek. Areas requiring TIEs are identified by integrating the 

triad of data collected in the program including toxicity and water chemistry from the mass 

loading stations and benthic community structure analysis from rapid stream bioassessment 

and ambient bay and lagoon surveys (Table 1). 

Table 1. Triad definitions for San Diego Storm Water Monitoring Program. 

Triad Component 

Definition 

Persistent Exceedance of Water Quality A constituent of concern with a high 

Objectives 

frequency of occurrence based on wet and 

dry weather data exceedances compared to 

established list of benchmarks or trigger 

levels 

Evidence of Persistent Toxicity 

More than 50% of the toxicity tests for any 

given species have a NOEC of less than 

100%. 

Indication of Benthic Alteration 

IBI score indicates a substantially degraded 

community (very poor) 

Source: Watershed Data Assessment Framework (MEC-Weston, 2004) 

Samples from three storm events are analyzed for toxicity using Ceriodaphnia dubia, Hyalella 

azteca, and Selenastrum capricornutum. When findings from these toxicity tests at the mass 

loading stations indicate the presence of persistent toxicity, a TIE is to be conducted to 

determine the potential cause or causes of toxicity. 

As listed in Table 1 above, toxicity test results are reported as the no observable effects 

concentration (NOEC). The NOEC is the lowest concentration at which there is no statistical 

difference from the control. Therefore, a concentration of less than 100% is considered to have 

some degree of toxic effect. Persistent toxicity is evident when more than 50% of the toxicity 

tests conducted to date for any given species at a specific site have a NOEC of less than 100%. 

The results of this determination are then combined with the high frequency constituents of 

concern (chemistry data) and benthic data in the Triad Decision Matrix to determine the actions 

to be taken. 

Results from chemistry, toxicity and relative benthic community health were assessed together 

using the triad approach to determine what short and/or long term actions are appropriate in a 

watershed. This approach examines persistence of toxicity using several indicators to provide 

an indication of an ecological concern. When persistence is found, this triggers the initiation of 

short term actions such as a TIE to identify the constituents of concern (COCs) in the watershed 

that may be responsible for stormwater toxicity and/or benthic community degradation). 

In the 2004-2005 stormwater monitoring period for the San Diego County Municipal Stormwater 

Copermittees, TDS exceeded water quality objectives measured in Sweetwater River 

stormwater samples (Weston 2005). While no other contaminants exceeded water quality 

objectives in the 2004-2005 monitoring period, diazinon exceeded water quality objectives in six 





out of the nine storm events prior to the 2004-2005 wet weather season. In addition, persistent 

toxicity was found because more than 50% of the toxicity tests conducted to date with this 

species had a NOEC of less than 100%. The benthic community was determined to be 

impacted. Based on these findings it was determined that TIEs would be performed at 

Sweetwater River if toxicity was observed during the standard toxicity testing of stormwater 

samples during the 2005-2006 monitoring period. 

2.1 TOXICITY IDENTIFICATION EVALUATION (TIE) TESTING 

The United States Environmental Protection Agency (USEPA) has issued toxicity identification 

evaluation (TIE) testing guidelines for characterizing chronically toxic effluents (USEPA 1991, 

1992, 1993a, and 1993b). These guidelines are often effective for effluents that have similar 

toxic constituents to those identified in the model effluents used to develop the TIE guidelines. 

A Toxicity Reduction Evaluation (TRE) is an evaluation which involves the identification of 

toxicants, location of the source, and treatment of the causative agents to a less toxic form; the 

ultimate goal of a TRE is to reduce toxicity associated with contaminated water (or sediment). 

Thus, TIEs are important tools used in a TRE to initially help with the identification of toxicants. 

The TIE typically consists of three test phases. Phase I of a TIE involves procedures designed 

to provide information for identifying the class of the toxic constituents within an effluent 1 sample 

based on their chemical characteristics (e.g., volatility, ionization state, degree of adsorption to 

particulates, polarity, oxidative state, or pH sensitivity). These classification characteristics are 

examined by comparing the results of tests conducted on unmanipulated effluent samples to 

effluent samples that have been physically or chemically manipulated. Phase I testing involves 

manipulating the sample at the effluent’s initial pH using some of the manipulations shown in 

Table 3 below. 

Table 2: Toxicity Identification Evaluation Procedures 

Sample Manipulation 

Purpose 

Ethylenediaminetetraacetic Acid (EDTA) Addition 

Detects certain cationic metals (e.g. cadmium) 

C 18 Column Extraction with Methanol Elution 

Detects non-polar organics and some surfactants 

Piperonyl Butoxide Treatment 

Detects organophosphate pesticides and pyrethroids 

The goal of Phase II TIE testing is to identify the toxicants in the sample, while Phase III 

methods are used to confirm that the suspected toxicants are the true cause of toxicity in the 

effluent samples (USEPA, 1993a and 1993b). It should be noted that the boundaries between 

Phases I, II, and III are not distinct and there may be cases where it is appropriate for their 

respective procedures to overlap because confirmation information can be obtained during 

Phases I and II. 

1 The USEPA protocol is designed for performing TIEs on effluent samples. However, modifications have been made 

for testing stormwater samples. 





TIEs are initiated when standard toxicity testing demonstrates toxicity in an effluent sample that 

has previously been toxic to test organisms. Standard toxicity test methods sometimes rely on 

sublethal endpoints, such as C. dubia reproduction as indicators of chronic toxicity, and require 

substantially more time and resources to evaluate than methods that rely exclusively on a 

mortality endpoint. In addition, the USEPA guidelines do not provide specific guideline for test 

procedures using each toxicity test species, given the large number of test organisms. 

Therefore, the USEPA’s TIE documents are used as guidance for conducting TIEs because it 

may not be possible or cost-effective to strictly adhere to these protocols. In addition, the 

USEPA protocols were initially designed for TIEs using whole effluent samples, and not the 

more variable and less predictable stormwater samples. Thus, modifications for efficiently 

conducting TIEs on stormwater samples using specific test organisms, and for specific site 

conditions may sometimes include the following: changes in test volumes, test duration, 

replicate number, number of test concentrations, and reduction in frequency of test solution 

renewal. 

Phase I test procedures are designed to identify obvious alterations in effluent toxicity, which 

may be achieved using modified chronic test methods. 

2.2 INITIAL TOXICITY TESTING SUMMARY FOR SWEETWATER STORMWATER 

Sweetwater River has been identified as a TIE site due to the evidence of benthic community 

impacts and the persistent toxicity to Selenastrum capricornutum. During the October 18, 2005 

storm event, stormwater from this site caused significant toxicity measured as reduced growth 

of S. capricornutum. The IC 50 at 96 hours was estimated to be 95.7% of the sample 

concentration. The NOEC was 50% of the sample concentration. Weston initiated TIE testing 

on November 1, 2005 with this sample utilizing the manipulations listed above. 

During the January 2, 2006 and February 19, 2006 storm events, stormwater from this site did 

not cause toxicity to S. capricornutum. The IC 50 at 96 hours was estimated to be greater than 

100% of the stormwater sample. The NOEC was 100%. Because there was no toxicity in these 

stormwater samples TIE testing was not performed. 

3. MATERIALS AND METHODS 

3.1 TEST PROCEDURES 

All Selenastrum capricornutum baseline tests and TIE tests following sample manipulation 

followed general methods described in the U.S. EPA guidance manual, “Short-term Methods for 

Estimating the Chronic Toxicity of Effluents and Receiving Waters to Freshwater Organisms” 

(USEPA, 2002). 

3.1.1 Toxicity Test Using Selenastrum capricornutum 

Stormwater tests for chronic toxicity using the freshwater algae S. capricornutum were run 

according to the USEPA protocol (EPA/821/R/02/013). Briefly, the sample and the control water 

were spiked with equal amounts of nutrients and subsequently filtered to 0.45 µm to remove any 

unicellular algae that might be present prior to test initiation. The concentration series was 

prepared, and 50-mL aliquots were placed into four replicate test chambers. Approximately 

10,000 cells per mL were added to the test chamber and placed in random order under highintensity 

24-hour light for four days. Temperature, conductivity, pH, hardness, and alkalinity 

was measured at test initiation. Temperature and pH was measured on days 1 to 4. The test 

chambers were shaken twice a day and randomized once a day. At the end of the test period, 

chambers were analyzed for chlorophyll. Test acceptability was determined by evaluating the 





response of the control organisms. The test was considered invalid if the criterion of a cell 

density of 200,000 cells per mL in the control was not met. Variability between the control 

replicates should not have exceeded 20%. A reference toxicant test was conducted using 

copper sulfate with concentrations of 7, 14, 28, 56, and 112 µg Cu 2+ /L to establish the sensitivity 

of test organisms used in the evaluation of the Sweetwater stormwater. 

3.2 TEST SOLUTION PREPARATION 

Control and dilution water for the S. capricornutum tests was synthetic moderately-hard water 

prepared with Epure to achieve a hardness of 80-100 mg/L as CaCO3. This water source 

has been used successfully on numerous similar bioassay testing programs conducted by 

Weston and others. Extensive testing with a variety of species and biannual chemical analysis 

of this water type has shown that this water source provides for good survival in laboratory 

controls with little to no measurable levels of contaminants. 

3.3 WATER QUALITY 

Water quality was monitored daily as appropriate for each test, and data were recorded on data 

sheets. Dissolved oxygen and temperature was measured using Orion Model 840 oxygen 

meters and probes; pH was measured using Orion Model 230A pH meters and probes. 

Conductivity was measured with Orion Models 142 conductivity/salinity meters. Ammonia 

was analyzed using an Orion 720 digital ion analyzer with a three-point calibration curve (1, 

10, and 100 mg/L). Hardness and alkalinity were measured utilizing LaMotte titration kits. 

3.4 SAMPLE RECEIPT 

The stormwater samples were composited at the laboratory and stored at 4°C. A chain-ofcustody 

was completed for all samples received. Before samples were used in the tests, initial 

water quality measurements were taken. These measurements included temperature, total 

chlorine, total ammonia, pH, dissolved oxygen, and salinity. 

3.5 PHASE I TIE METHODS 

3.5.1 Protocol Modifications 

Measures to conserve time and resources required to conduct TIE testing have been developed 

and approved by the USEPA (USEPA 1992). This study incorporated modifications allowed for 

reduction in number of test concentrations, and replicates for all TIE tests conducted, relative to 

test conditions used in standard toxicity tests. The concentrations of stormwater samples used 

for the TIE test in this investigation included the 50, 75, and 100% dilutions of the stormwater 

samples. Three replicates were used in each stormwater sample treatment. Treatment blanks 

were created for each TIE test to determine the effects of the manipulation on laboratory dilution 

water. The results of these blanks were used to determine if any changes in toxicity of the 

control (dilution water) were impacted by the chemical or physical manipulation of the sample. 

3.5.2 Baseline Tests 

The baseline test assesses the toxicity of the unmanipulated sample run concurrently with the 

TIE tests. This test would confirm the presence of toxicity in the stormwater sample, and toxicity 

in TIE treatments are compared to toxicity in the baseline test. 

3.5.3 Ethylenediaminetetraacetic Acid (EDTA) Tests 

Two series of EDTA tests were conducted with concentrations of 3.0 and 8.0 milligrams per liter 

(mg/L). A stock solution of 2.5 grams (g) EDTA/L was prepared. EDTA treatments were 

prepared by diluting the EDTA stock solution to final concentrations of 3.0 and 8.0 mg/L in 

stormwater (50, 75, and 100% sample concentrations) and dilution water. The EDTA was 





allowed to react with the sample for a minimum of 2 hours prior to addition of the test 

organisms. EDTA binds certain ionic metals, making them biologically unavailable to the test 

organisms. A reduction in the toxic response of a sample treated with EDTA may indicate the 

presence of divalent (metal) cation toxicity. 

3.5.4 Filtration Tests 

This test was not performed on the species S. capricornutum, as the sample was already 

filtered to 0.45 µm as a standard part of the test protocol. Filtration identifies chemicals 

associated with the particulate fraction. 

3.5.5 C 18 Solid Phase Extraction and Methanol Add-Back Tests 

The sample manipulation was prepared by passing 1 L of filtered sample through a preconditioned 

5 g C 18 Solid Phase Extraction (SPE) column at a rate of 5 mL per minute. The 

extracted sample was then used to prepare 300 mL dilutions of 50, 75 and 100% samples. This 

treatment mainly removes non-polar organic compounds from the sample, but in some cases 

may partially remove surfactants and metals. 

The C 18 SPE column was then eluted with 100% reagent grade methanol to recover any 

contaminants bound to the column. To elute the column, a total of 15 mL of 100% methanol 

was passed through the column in two, 7.5 mL aliquots, and collected into a test tube. The 

methanol was then added to clean control water at concentrations equaling 1.0, 2.0, 4.0, and 

8.0 times the concentration of the potential contaminants in the original effluent sample. The 

methanol elution tests are used to confirm the toxicity of the compounds retained on the C 18 

column. 

3.5.6 Piperonyl Butoxide (PBO)Tests 

Two series of PBO tests were conducted with concentrations of 0.025 and 0.050 mg/L. A stock 

solution of 2.5 mg PBO/L was prepared. PBO treatments were prepared by diluting the PBO 

stock solution to final concentrations of 0.025 and 0.050 mg/L in stormwater (50, 75 and 100% 

sample concentrations) and dilution water. PBO blocks specific cytochrome P450 isozymes 

that are involved in metabolizing chemicals such as organophosphates to more toxic 

metabolites and chemicals such as pyrethroids to less toxic metabolites. Thus, if results from 

this test demonstrate increased toxicity in the stormwater sample, this is indicative of chemicals 

(e.g. pyrethroids) that are metabolized to less toxic forms by cytochrome P450 enzymes. In 

contrast, if the results demonstrate decreased toxicity in the stormwater samples, this is 

indicative of chemicals (e.g. malathion, organophosphates) that are metabolized to more toxic 

forms by cytochrome P450 enzymes. 

3.6 STATISTICAL ANALYSIS 

At the conclusion of all tests, test species data were evaluated statistically using ToxCalc to 

determine ECp, NOEC, and TUc values. ToxCalc is a comprehensive statistical application 

that follows standard guidelines for acute and chronic toxicity data analysis. 

Statistical effects can be measured by the ECp, the estimated concentration that causes any 

effect, either lethal (LC) or sublethal (IC), on p% of the test population. The IC 50 or IC 25 is the 

point estimate of the concentration at which an inhibitory effect in a sublethal parameter (e.g. 

growth, reproduction) is observed in 50% or 25% of the organisms. ECp values include 95% 

confidence limits where available. The NOEC (No Observable Effect Concentration) is the 

highest tested concentration at which mortality and other sublethal measured effects are not 

significantly different from the same parameters in the control. TUc (Chronic Toxicity Units) are 





calculated as 100%/NOEC. All statistics were run against treatment blanks to mitigate for any 

artifactual effect that the treatment had upon the toxicity. 

In addition to standard toxicity-related statistics performed by ToxCalc, results of the TIE were 

analyzed using a two-way, fixed factor ANOVA and Dunnett’s multiple comparisons to 

determine whether there were differences in growth among treatments and among stormwater 

sample concentrations(Zar 1999). In addition, a one-way, fixed factor ANOVA and Dunnett’s 

multiple comparisons were used to examine the differences in growth of S. capricornutum 

among the different dilutions of stormwater for all treatments combined. 

4. RESULTS 

4.1 SWEETWATER RIVER SELENASTRUM CAPRICORNUTUM 

4.1.1 Results of TIE performed on October 18, 2005 Sample 

Results for the TIE performed on the October 18, 2005 sample are summarized in Table 3. 

In the Baseline test, toxicity of Selenastrum capricornutum, measured as growth or relative 

fluorescence units (RFU) of chlorophyll, exposed to the 100% sample concentration (69.70 RFU 

Chlorophyll) was slightly lower than that in the initial toxicity test (38.88 RFU Chlorophyll) started 

on October 18, 2005. The NOEC was 75% of the sample concentration, and the IC 50 was 

greater than 100% of the sample concentration. Growth in the dilution water control was 113.87 

RFU Chlorophyll. 

The 3.0 mg/L EDTA manipulation did not significantly increase growth (i.e. reduce toxicity) in the 

100% sample concentration (85.47 RFU Chlorophyll) relative to growth in the unmanipulated 

Baseline test (69.70 RFU Chlorophyll in the 100% sample concentration). Growth in the 3.0 

mg/L EDTA treatment blank was 92.03 RFU Chlorophyll. Statistical analysis showed that 

growth in the 50, 75, and 100% sample concentrations of the 3.0 mg/L EDTA treatment were 

not significantly different from the 3.0 mg/L EDTA treatment blank. The 8.0 mg/L EDTA 

significantly reduced toxicity in the 100% sample concentration (106.96 RFU Chlorophyll) 

relative to toxicity in the unmanipulated Baseline test (69.70 RFU Chlorophyll in the 100% 

sample concentration). Growth in the 8.0 mg/L EDTA treatment blank was 83.91 RFU 

Chlorophyll. Statistical analysis showed that while there was no significant difference in growth 

in the 100% sample concentration of the 8.0 mg/L EDTA treatment relative to the treatment 

blank; however, growth was significantly elevated in the 50 and 75% sample concentrations as 

compared to the 8.0 mg/L EDTA treatment blank. Both treatments produced a NOEC of 100% 

sample concentration and a IC 50 of greater than 100% sample concentration. 

The C-18 SPE treatment did not affect growth in the 100% sample concentration (56.8 RFU 

Chlorophyll) relative to growth in the unmanipulated Baseline test (69.70 RFU Chlorophyll in the 

100% sample concentration). Growth in the C-18 SPE treatment blank was 117.4 RFU 

Chlorophyll. Statistical analysis showed that while there was no significant difference in growth 

in the 50 and 75% sample concentrations of the C18 extraction treatment, there was significant 

reduction in growth in the 100% sample concentration compared to the C18 extraction treatment 

blank. 

The methanol add-back manipulation increased toxicity in the 8X methanol add-back sample 

concentration (10.12 RFU Chlorophyll). Growth of the methanol treatment blank was 8.23 RFU 

Chlorophyll. The significantly impaired growth measured in each treatment indicates that the 





methanol used in this manipulation was likely responsible for the toxicity produced by this 

treatment. 

The 0.025 mg/L PBO treatment did not significantly affect growth in the 100% sample 

concentration (83.64 RFU Chlorophyll) relative to growth in the unmanipulated Baseline test 

(69.70 RFU Chlorophyll in the 100% sample concentration) at P=0.05. Growth in the 0.025 mg/L 

PBO treatment blank was 87.07 RFU Chlorophyll. Statistical analysis showed that while there 

was no significant increase in growth in the 75 or 100% sample concentration of the 0.025 mg/L 

PBO treatment, growth was significantly elevated in the 50% sample concentration compared to 

growth in the 0.025 mg/L PBO treatment blank. The 0.050 mg/L PBO treatment did not 

significantly affect growth in the 100% sample concentration (71.34 RFU Chlorophyll) relative to 

growth in the unmanipulated Baseline test (69.70 RFU Chlorophyll in the 100% sample 

concentration). Growth in the 0.050 mg/L PBO treatment was 93.16 RFU Chlorophyll. 

Statistical analysis showed that while there was no significant reduction in growth in the 75 or 

100% sample concentration of the 0.050 mg/L PBO treatment, there was significantly elevated 

growth in the 50% sample concentration compared to the 0.050 mg/L PBO treatment blank. 

The NOEC was 100% of the sample concentration and the IC 50 was greater than 100% for both 

treatments. 

A copper sulfate reference toxicant was tested at nominal concentrations of 7, 14, 28, 56 and 

112 µg Cu 2+ /L. The calculated 96-hour IC 50 (68.36 µg Cu 2+ /L) was within two standard 


the sensitivity of S. capricornutum used in this evaluation fell within the normal range. 





4.1.2 Summary of TIE Performed on October 18, 2005 Sample 

Table 3: Summary of Growth in TIE tests on Sweetwater Stormwater Sample Collected on October 18, 2005 

Test 



Mean Growth – Chlorophyll (RFU) 

50% 75% 

Stormwater Stormwater 

100% 

Stormwater 

NOEC LC 50 

Baseline 113.87 131.75 112.54 69.70 75 >100 

C-18 117.4 122.37 86.09 56.8 50 97.32 

EDTA 3 92.03 116.02 112.92 85.47 100 >100 

EDTA 8 83.91 131.37 131.28 106.96 100 >100 

PBO 0.25 87.07 119.17 109.66 83.64 100 >100 

PBO 0.50 93.16 137.63 116.7 71.34 100 >100 

4.1.3 Statistical Analysis: Differences in Growth Among Stormwater Dilutions 

Data was evaluated using one-way, fixed factor ANOVA and Dunnett’s multiple comparisons to 

examine the differences in growth of S. capricornutum among the different dilutions of 

stormwater. It was determined that regardless of the TIE treatment, significant differences in 

growth of S. capricornutum were detected among the different dilutions of stormwater (i.e., 0, 

50, 75 or 100%) to which S. capricornutum were exposed (ANOVA, P=0.008). Specifically, as 

depicted in Figure 1, growth of S. capricornutum from highest to lowest in the different 

stormwater dilutions was as follows: 50%, 75%, 0% (control), and 100%. 

Growth of S. Capricornutum (RFU) 

160 

140 

120 

100 

80 

60 

40 

B 

C 

A 

D 

0 50 75 100 

Stormwater Dilution (%) 

Figure 1. Growth of S. capricornutum exposed to multiple dilutions of a Sweetwater River stormwater 

sample. Different letters indicate statistically significant differences among treatments where B>C>A>D. 





5. DISCUSSION 

Results of this TIE indicate that one or more cations that comprise the total dissolved solids 

(TDS) found in stormwater samples from Sweetwater River may be responsible for the reduced 

growth of S. capricornutum. First, the results of the 8 mg/L EDTA test allowed for the increased 

growth of S. capricornutum or demonstrated that toxicity was reduced through the addition of 

EDTA to the 100% stormwater sample. This indicates that the causative agent of toxicity was a 

cation, such as a free metal ion (e.g., cadmium, mercury, lead), a dissolved inorganic salt (e.g., 

sodium, calcium, magnesium), or a cationic organic compound such as a cationic surfactant, 

because EDTA is a chelating agent that is well known to bind inorganic and organic cationic 

chemicals (USEPA 1991). 

In addition to TIE tests, chemical analyses of Sweetwater stormwater samples indicate that 

cations comprising the TDS may be the causative agent(s) of toxicity. Concentrations of TDS 

measured in the undiluted stormwater sample were 2640 mg/L (see results in complete 

stormwater report), a concentration that exceeds the water quality objectives for this measure, 

or 1500 mg/L, in the lower Sweetwater River (RWQCB 1994). In contrast, no organophosphate 

pesticides were detected in stormwater samples and the only metals detected were arsenic, 

nickel, and zinc, all of which were found at low concentrations, or concentrations far below water 

quality objectives. 

TIE test results also indicate that the causative agents of toxicity in the Sweetwater River 

stormwater samples did not share similar physicochemical properties to those of many other 

classes of chemicals. The lack of toxicity reduction in the SPE and PBO tests indicates that the 

causative agent was likely not a non-polar organic, or a common-use pesticide (e.g. pyrethroid 

or organophosphate), respectively. 

Statistical analyses of TIE results also provide some evidence that the causative agent could be 

a ion imbalance; growth was diminished in the undiluted 100% stormwater sample, relative to all 

treatments while growth was elevated in the 50% and 75% stormwater sample relative to the 

dilution water (0% stormwater) sample. This result indicates that the causative agent was toxic 

in the undiluted stormwater sample but significantly less toxic in the diluted stormwater samples. 

In contrast, in the dilution water control, the lack of ions may have led to diminished growth 

relative to the diluted stormwater samples. This result is not surprising because others have 

shown that elevated growth in TDS-enriched natural waters as compared to TDS-poor synthetic 

mediums (LeBlond and Duffy 2001). 

It is well known that elevated concentrations of ions comprising the TDS, or an imbalance of the 

ions comprising the TDS may be toxic to aquatic organisms; however, the concentration of TDS 

that causes toxicity is highly dependent on the species and types of ions comprising the TDS. 

While most species are tolerant of concentrations of TDS that exceed 1000 mg/L, some 

spawning or larval fishes are highly sensitive to TDS. Concentrations of 350 mg/L TDS reduced 

spawning of striped bass (Morone saxatilis) in the San Francisco Bay-Delta (Kaiser Engineers, 

1969). In contrast the LC 50 for TDS for fathead minnows (Pimphales promelas) has been shown 

to range from 2,000 mg/L to 15,000 mg/L in natural waters, depending in part on the types of 

ions comprising the TDS (Mount et al. 1997). To the best of our knowledge, little is known about 

the effects of TDS on S. capricornutum. 





In summary, the results of this TIE provide some evidence that the causative agent of toxicity 

may have been due to elevated TDS levels in the Sweetwater River sample collected in 

October, 2005. Because no additional toxicity was found in Sweetwater River samples in S. 

capricornutum tests conducting during storm events in January and February of 2006, it is likely 

that the causative agent was reduced in this sample during the second and third stormwater 

sampling events. Additional studies would be necessary to confirm whether an ion imbalance or 

TDS were the causative agent of toxicity to S. Capricornutum. 





6. REFERENCES 

California Regional Water Quality Control Board (RWQCB). 1994. Water Quality Control Plan 

for the San Diego Basin. September. 

Kaiser Engineers. 1969. Final Report to the State of California, San Francisco Bay-Delta Water 

Quality Control Program, State of California, Sacramento, CA. 

LeBlond, J.B., and Duffy, L.K. 2001. Toxicity assessment of total dissolved solids in effluent of 

Alaskan mines using 22-h chronic Microtox® and Selenastrum capricornutum assays. The 

Science of the Total Environment, 271:49-59. 

Mount, D.R., Gulley, D.D., Hockett, J.R., Garrison, T.D., and Evans, J.M. 1997. Statistical 

models to predict the toxicity of major ions to Ceriodaphnia dubia, Daphnia magna and 

Pimephales promelas (fathead minnows). Environmental Toxicology and Chemistry, 

16:2009-2019. 

United States Environmental Protection Agency (USEPA). 1991. Methods for Aquatic Toxicity 

Identification Evaluations. Phase I Toxicity Characterization Procedures. EPA/600/6-91/003. 

EPA Office of Research and Development. Second Edition. February. 

_________. 1992. Toxicity Identification Evaluation. Characterization of Chronically Toxic 

Effluents, Phase I. EPA/600/6-91/005F. EPA Office of Research and Development. May. 

_________. 1993a. Methods for Aquatic Toxicity Identification Evaluations. Phase II Toxicity 

Characterization Procedures for Samples Exhibiting Acute and Chronic Toxicity EPA/600/R- 

92/080. EPA Office of Research and Development. September. 

_________. 1993b. Methods for Aquatic Toxicity Identification Evaluations. Phase III Toxicity 

Characterization Procedures for Samples Exhibiting Acute and Chronic Toxicity EPA/600/R- 

92/081. EPA Office of Research and Development. September. 

_________. 2002. Short-Term Methods for Estimating the Chronic Toxicity of Effluents and 

Receiving Waters to Freshwater Organisms. EPA-821-R02-013. EPA Office of Water. 

Fourth Edition. October. 

Weston Solutions, Inc. (Weston). 2005. San Diego County Municipal Copermittees 2004-2005 

Urban Runoff Monitoring. Final Report. For the County of San Diego, San Diego, California. 

December. 

Zar, J.H. 1999. Biostatistical Analysis. 4 th Edition. Prentice Hall, New Jersey.

APPENDIX I Toxicity Identification Evaluation Reports for Chollas ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?