v/EPA An Evaluation of Mercury and Selenium Fish Tissue Monitoring Alternatives:

Fish Biopsy Plug Samples Versus Homogenized Whole Fillets

Leanne Stahl1, Blaine Snyder2, Harry McCarty3, Tara Cohen2, Kenneth Miller3, Mark Fernandez2, and John Healey1

Background

The U.S. Environmental Protection Agency (EPA) and some states have collected and ana-
lyzed fish fillet plugs (i.e., biopsy punch samples) as a more cost-effective alternative for
monitoring mercury levels in fish for human health applications than the routine approach
of removing entire fillets from each whole fish sample and analyzing homogenized fillet
tissue. In 2013, EPA's Office of Water added fillet plug sampling from fish collected at river
and stream sites to its whole fillet sampling approach during the 2013—2014 National Rivers
and Streams Assessment (NRSA). EPA has continued to use fillet plug sampling and whole
fillet sampling for monitoring mercury concentrations in fish during the 2018-2019 NRSA
and the 2015 and 2020 National Coastal Condition Assessments (NCCAs). In addition, EPA
received comments about including fillet plug sampling with respect to implementing the
Agency's 2016 Selenium Criterion.

Study Objectives

More states have been introducing fillet plug sampling into their
fish monitoring programs since 2013 when EPA began applying this
technique in the NRSA and NCCA. However, the question remained
about whether fish fillet plug sampling and analysis can serve as
a reliable alternative for homogenizing and analyzing fillet tissue
from each whole fish sample to monitor mercury and selenium
concentrations. To answer this question, EPA designed and conducted
the Fish Plug Evaluation Study to assess the comparability of mercury
and selenium concentrations in fish fillet plugs vs. homogenized
whole fillet tissue samples in order to meet the following objectives:
(1) to test whether collecting and analyzing fish fillet plug samples can
serve as a reliable alternative for homogenizing and analyzing whole
fillet tissue to monitor mercury concentrations in fish (mercury phase)
and (2) to investigate if it is technically feasible to collect fillet plug
samples and analyze them for monitoring selenium levels in fish to
support implementation of EPA's tissue-based water quality criterion for
selenium (selenium phase).

Fish Plug Evaluation Study Design Summary

Design Element

Mercury
Phase

Selenium
Phase

Description

Waterbody Types

2

2

Great Lakes and East Coast Rivers

Sampling Sites and Fish
Species Collected

6

6

Lake Erie, Walleye; Lake Michigan, Lake Trout; Lake Ontario,
Chinook Salmon; Anacostia River, Blue Catfish; Potomac
River, Largemouth Bass; St. Lawrence River, Smallmouth Bass

Fish Collected per Site

10

5

Each fish sample consisted of a single specimen

Fish Tissue Sample Types

2

2

Fillet plug samples (2 plugs per sample) and Flomogenized
fillet tissue samples

Replicates per Sample
Type

5

4

Number applies to each individual fish sample

Total Fillet Plug Samples
Analyzed

300

120

Sampling sites (6) x Fish collected per site x Replicates per
sample type

Total Flomogenized Fillet
Tissue Samples Analyzed

300

120

Sampling sites (6) x Fish collected per site x Replicates per
sample type

Total Fillet Samples
Analyzed

600

240*

Sampling sites (6) x Fish collected per site x Sample types (2)
x Replicates per sample type

*An additional 120 single-plug fillet samples were analyzed for percent solids to convert wet-weight concentrations to dry-weight concentrations.

Author Affiliation

1 U.S. Environmental Protection
Agency (4305T)

Office of Water, Office of Science
and Technology

1200 Pennsylvania Avenue, NW
Washington, DC 20460

2Tetra Tech, Inc.

Center for Ecological Sciences

10711 Red Run Boulevard, Suite 105
Owings Mills, MD 21117

3 CSRA, LLC, a General Dynamics
Information Technology company

6361 Walker Lane, Suite 300
Alexandria, VA 22310

FIELD (6 Sampling Sites)

5 Fillet Plug Samples

Collected and Analyzed
from Each Fish

5 Homogenized Fillet Samples

Collected and Analyzed from Each Fish

Mercury Analysis pJ/jn

600 Sample Results

FIELD (6 Sampling Sites)

4 Fillet Plug Samples

Collected and Analyzed
from Each Fish

4 Homogenized Fillet Samples

Collected and Analyzed from Each Fish

Selenium Analysis

240 Sample Results

Analytical and Statistical Methods

•	300 fillet plug samples and 300 homogenized fillet samples were prepared and
analyzed for mercury using Appendix to Method 1631, Total Mercury in Tissue, Sludge,
Sediment, and Soil by Acid Digestion and BrCl Oxidation from Method 1631 Revision B
and Revision E, respectively (USEPA 2001 and 2002).

•	120 fillet plug samples and 120 homogenized fillet samples were prepared and
analyzed for selenium using a modification to Method 200.8, Determination of Trace
Elements in Waters and Wastes by Inductively Coupled Plasma-Mass Spectrometry
(USEPA 1994).

•	Null hypotheses (H0) for mercury and selenium: both methods of collecting samples
(fillet plugs vs. homogenized fillets) would yield equivalent mean concentrations of
mercury and of selenium, respectively, for any given specimen.

•	An analysis of variance (ANOVA) model on log-transformed data averaged across
specimens was used to determine whether there are any significant differences across
the two sampling methods, for each analyte (mercury and selenium), and an alpha
value of 0.05 was used to assess significance.

•	The statistical methods evaluated the potential impact of factors that could affect
results, including waterbody type (lake vs. river), specific waterbody (6 locations), and
fish species (6 species).

Disclaimer: The views expressed in this poster are those of the authors and do not necessarily
represent the views or policies of the U.S. Environmental Protection Agency.

Summary Statistics by Sample Type



Mercury (wet weight)

Selenium (dry weight)

Fillet Plug
Concentration,
ng/g (n=300)

Homogenized

Fillet
Concentration,
ng/g (n=300)

Fillet Plug

Concentration,
ng/g (n=120)

Homogenized

Fillet
Concentration,
ng/g (n=120)

Minimum

44.2

23.0

711.0

750.0

10th percentile

82.4

91.0

859.0

874.5

25th percentile

101.0

105.0

1544.5

1542.0

Median

121.0

143.0

1762.0

1781.5

Mean

155.2

161.4

1922.0

1996.0

75th percentile

171.3

185.3

2101.0

2090.0

90th percentile

271.0

262.0

3402.0

3829.5

Maximum

649.0

556.0

3814.0

4084.0

Standard Deviation

101.6

86.5

810.1

936.2

Relative Standard Deviation

65.5%

53.6%

42.2%

46.9%

Log-transformed Mercury (Hg) and
Selenium (Se) Results by Sample Type

Fillet Plug	Homogenized Fillet

Sample Type

Fillet Plug	Homogenized Fillet

Sample Type

Study Findings

The ANOVA main effects models indicated that for mercury
(p=0.4048) and for selenium (p=0.3786), there was no significant
difference between fish fillet plugs and homogenized whole fillets.

The mercury data and the selenium data were log-normally
distributed. For each analyte, there was a large overlap across the
different fillet tissue sampling methods, and large variability across
waterbodies.

Since there was a large variance across waterbodies, waterbody was
included as a blocking factor in the ANOVA model equation used for
each analyte.

The interaction term (MethodfWaterbody) was not significant
(p=0.9728 for mercury, p=0.6740 for selenium), indicating the effect of
sample types was not impacted by site- or species-specific factors.

For more information, contact either Leanne Stahl (stahl.leanne@epa.gov)
or John Healey (healey.john@epa.gov)


-------