Post on 13-Jan-2016
transcript
Mercury Concentrations in Stream Fish Throughout 12 Western States in the USA
Alan Herlihy and Robert Hughes Dept. of Fisheries & Wildlife, Oregon State University
Spencer Peterson and John Van Sickle EPA NHEERL Corvallis, OR
Questions
• What is the extent of mercury (Hg) contamination in fish tissue across all Western U.S. streams and rivers?
• What are the factors related to mercury levels in fish at this scale?
Environmental Monitoring and Assessment Program (EMAP) Western U.S. Pilot Survey
• Sample sites were selected using the systematic, randomized EMAP sampling design from all perennial western U.S. streams/rivers– Additional hand-picked reference sites
• Samples selected off of the digitized version of the 1:100,000 scale USGS maps
• Inferences to the entire stream network can be made from probability survey data using site inclusion probabilities
Field Methods
• Fish sampled by electrofishing
• Streams: backpack shocker on 40 channel width long sample reach
• Rivers: raft mounted shocker on 100 channel width reach
• Associated measurements of water chemistry, physical habitat, and watershed characteristics
Tissue Samples
• Collect a big and small fish sample at each site if sufficient numbers available
• Big Fish: ≥ 120 mm total length– Up to 3 individuals of 3 different species of
varying size, individuals analyzed separately
• Small Fish: Adults < 120 mm– Single composite sample (50-200 g)
• Samples kept on ice, shipped overnight to laboratory and then frozen until analysis.
Hg Laboratory Analysis
• Whole body analysis (µg Hg/g wet weight)• Fish ground up in blender (bass-o-matic)• Subsampled, frozen with no further sample
preparation • Subsamples thawed, rehomogenized, ≈0.25 g
analyzed by Combustion AAS (Milestone MDA80 direct Hg analyzer)
• Subsamples analyzed in duplicate, – repeat if > 5% difference
• Method Detection Limit = 0.0024 µg Hg/g wet wt.
Most Common Species Analyzed
• Big Fish (2,707 fish, 626 sites)
• Non-Piscivores (85%)– Rainbow, Brown, Brook,
Cutthroat Trout– White, Largescale Sucker– Mountain Whitefish, Carp
• Piscivores (15%)– Smallmouth Bass– Northern Pikeminnow– Walleye, Northern Pike
• Small Fish (386 samples)– Mottled Sculpin– Common Shiner– Redside Shiner– Fathead Minnow– Creek Chub– Speckled Dace– Longnose Dace
Three Aggregate Level III Omernik Ecoregions used in Analysis
Mercury – Fish Length Relationship for Individual Big Fish
Factors to Consider
• Fixed Factors– Ecoregion (Mountains, Xeric, Plains)– Site Disturbance Class (Low, Medium, High)– Trophic Group (Piscivore, Non-Piscivore)
• Covariates– Fish Age (Fish length)
• Analysis of Covariance using individual big fish
Big Fish ANCOVA Results
• Effect F-value P - value df
Fish Length 330 < .0001 1
Trophic Group 120 < .0001 1
Ecoregion 1.8 0.17 2
Site Disturbance 3.7 0.03 2
Trophic x Ecoregion 11 < .0001 2
Non-Piscivores
Piscivores
Fish Tissue Mercury in Trophic x Disturbance Class
Fish Tissue Mercury in Ecoregion x Trophic Classes
Non-Piscivores
PiscivoresNon-Piscivores
Piscivores
Correlation between Hg and environmental variables in Big Fish after partialing out the
effect of fish length • Piscivores (n=358)• Most strongly correlated with
– ANC (r = -0.57)– Conductivity (r = -0.56)– Watershed Slope (r = 0.52)– Precipitation (r = 0.51)– % Forested Watershed (r = 0.5)– Sulfate (r = -0.5)
• Non-Piscivores (n=1,964)– No variables
correlated with r > 0.3
Various Fish Tissue Mercury Criteria Values
• Human Health– 0.5, 0.6, 1.0 µg/g – 0.35 µg/g (Oregon Health Div., 1997)– 0.30 µg/g (EPA, 2001)– 0.10 µg/g (Faroe Island Study, 1998)
• Wildlife protection values - Lazorchak et al. 2003
– 0.10 µg/g (Otter)– 0.07 µg/g (Mink)– 0.03 µg/g (Kingfisher)
Relationship between Fish Tissue Filet and Whole body Hg
Cumulative Distribution Function (CDF) of Mean Site Fish Tissue Mercury
Weighted population estimates from EMAP survey design
Extent of Big Fish Hg Contamination by Family
Total Stream Length=305,000 km; 180,000 km with Big Fish
0 20,000 40,000 60,000 80,000 100,000 120,000 140,000
Salmonid
Ictalurid
Esocid
Cyprinid
Centrarchid
Catostomid
Estimated Stream Length (km)
Not Assessed
<0.1 ug/g
0.1-0.185 ug/g
>0.185 ug/g
Summary
• Fish tissue mercury concentrations were most strongly related to trophic group and fish length not environmental factors, ecoregion, or site disturbance– Regional analyses must
consider trophic group and size
• In terms of % Stream Length exceeding criteria
• Big Piscivores– 93% > 0.1 µg/g – 57% > 0.185 µg/g
• Big Non-piscivores– 26% > 0.1 µg/g– 6% > 0.185 µg/g
Summary (con’t)
• Fish tissue mercury concentrations in Western U.S. streams and rivers were found in a fairly narrow range, always above the detection limit (0.0024)– 90% of the small fish and non-piscivores had Hg
between 0.02 and 0.2 µg/g– High concentration “hot spots” (Hg > 0.5 µg/g) were
very rare (< 2% of stream resource)
• This strongly suggests a broad diffuse source of mercury from atmospheric deposition.
Study was funded by EPA’s Environmental Monitoring and Assessment Program
(EMAP)
EMAP Regional Hg Survey Citations• Oregon Streams: Peterson et al., 2002.
Environ. Toxicol. Chem. 21:2157-2164.
• Mid-Atlantic Streams: Lazorchak et al., 2003. Environ. Toxicol. Chem. 22:545-553.
• Northeast Lakes: Yeardley et al., 1998. Environ. Toxicol. Chem. 17:1875-1874.