Sonia Nagorski, University of Alaska Southeast AWRA Conference, - - PowerPoint PPT Presentation

Sonia Nagorski, University of Alaska Southeast AWRA Conference, September 17, 2019

Collaborators

➢ John Hudson, Independent Aquatic Ecologist, Juneau, AK. ➢ Eran Hood and Jason Fellman, University of Alaska Southeast ➢ John DeWild, David Krabbenhoft, and staff at USGS Mercury Research Lab, Middleton, WI ➢ Gina Ylitalo at Northwest Fisheries Science Center, Seattle, WA ➢ Undergraduate research assistants: Chris Salazar, Alex Whitehead, and Alex Botelho (UAS)

Salmon in the trees

Increase

• streamwater nutrient concentrations and

biofilm abundance (Mitchell and Lamberti 2005;

Chaloner et al. 2004, 2007; Tiegs et al. 2011; Hood et al 2019)

• Benthic macroinvertebrate abundance

(Minikawa 1997, Wipfli et al. 1998, Lessard and Merritt 2006)

• fish growth and fat content (Wipfli et al. 2003,

Heinz et al. 2004)

• >4000 salmon-supporting streams in southeast AK
• $1 billion annual industry

Salmon growth

• -Accumulate pollutants
• -Hg projected to double by 2050

Biovectors: Migrating animals may transport and focus pollutants

From Blais et al. Environ. Sci. Technol. 2007

 Long-range atmospheric deposition of contaminants

GRAHM (Global/Regional Atmospheric Heavy Metals Model) simulation – Ashu Dastoor, Meteorological Service of Canada,Environment Canada

Average elemental mercury surface concentrations for July 2001 (ng/m3)

 Chichagof Island lake sediment cores

show 2.9±0.5-fold increase since industrialization

Engstrom et al. ES&T, 2014

Marine Intertidal Freshwater aquatic Atmospheric deposition

What happens to deposited contaminants?

Salmon Wetland Glacier Upland Forest Toxic methylmercury

Geogenic Hg

Hg POPs POPs Hg

Benthic macroinvetebrates above and below waterfall barrier at Peterson Creek

28.4 46.1 34.4 52.5 80.7 57.8 98.5 68.5 68.1 53.5 20 40 60 80 100 120 Baetidae Heptageniida Limnephilidae Ephemerelidae Chloroperlidae Total Hg (ng/g) Peterson A Peterson B

Purpose of this study

 To investigate the relationship between salmon spawner density and

contaminant levels in streams

❖ To assess concentrations of marine-derived pollutants in various aquatic components (water, sediment, biofilm, macroinvertebrates, juvenile fish) ❖ Measure upstream (salmon absent) vs. downstream (salmon present) ❖ Compare across streams with varying spawnerdensity

Study sites: 5 Juneau- area watersheds

Bridget Cove Cr. Peterson Cr. Shrine Cr. Fish Cr. Salmon Cr.

Stream water: filtered and particulate fractions Streambed sediment Benthic macroinvertebrates Juvenile/resident fish Biofilm on incubated leaves

Sampling for mercury and POPs

Hg Hg + POPs

+ Fish density counts 1-3x/ week

Results

Leaf incubation Water samples Sediment samples Fish, BMI samples

 1. Salmon spawnerdensities varied among streams

• 2. Contaminant concentrations were higher in the lower reaches where salmon spawners

were present (one-way paired t-test, p<0.05) for:

• %methyl-Hg in filtered water and in biofilm
• methylmercury in Heptageniidae mayfly larvae
• Methylmercury in streambed sediments

No spawners----------------------------→highest spawner density

Methyl-Hg in Heptageneiid larvae

S a l m

• n

F i s h P e t e r s

• n

B C C S h r i n e

Methyl-Hg (ng g

• 1 dw)

20 40 60 80 100 120 140

Bed sediment Methyl-Hg

Salmon Fish Peterson BCC Shrine

Methyl Hg in bed sediment (ng g-1 dw)

0.1 0.2 0.3 0.4 0.5 0.6 0.7

Upstream (Site A) Downstream (Site B)

Biofilm+leaves %Hg as MeHg

Salmon Fish Peterson BCC Shrine

% methyl-Hg of total Hg in biofilm+leaves

5 10 15 20

No spawners--------------------------→highest spawner density No spawners---------------------------→-highest spawner density No spawners---------------------------→-highest spawner density

%Filtered Hg as MeHg

Salmon Fish Peterson BCC Shrine

5 10 15 20 25 30 35 Percent methyl-Hg

Upstream (Site A) Downstream (Site B)

Results

• Several other POPs (dieldrin, oxychlordane, and ΣBDEs), where present, were higher at the lower sites.
• The POPs data indicate a stronger marine-derived influence than the Hg, which has geogenic sources as

well.

Fish tissues: sumDDTs

Salmon Fish Peterson BCC Shrine

sum DDTs (ng/g)

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Fish tissues: HCB

Salmon Fish Peterson BCC Shrine

HCB (ng/g)

0.2 0.4 0.6 0.8 1.0 1.2

Fish tissues: sumPCBs

Salmon Fish Peterson BCC Shrine

sum PCBs (ng/g)

1 2 3 4 5 6

Fish tissues: chlordanes

Salmon Fish Peterson BCC Shrine

sum chlordanes (ng/g)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

• 2. Contaminant concentrations were higher in the lower reaches where salmon spawners

were present (one-way paired t-test, p<0.05) for:

• ΣHCBs, ΣDDTS, Σchlordanes, and ΣPCBs in fish tissues

 3. Total and methyl-Hg in biofilm (via

incubated leaf packs) were strongly

correlated with

 aqueous total Hg  aqueous methyl-Hg  spawner density,

 indicating their potential usefulness as a

passive integrator of MeHg and monitoring/assessment tool.

R² = 0.7434 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 2 4 6 8 10 Mean salmon density (fish/m2) Methylmercury in biofilm (ng/g)

Methyl-Hg in biofilm vs avg salmon density

R² = 0.7227 1 2 3 4 5 6 10 20 30 40 Filtered water (ng/L) Total Hg in biofilm (ng/g)

Total Hg in biofilm vs Filtered water

Alder leaf packs, incubated for biofilm growth, resulted in particularly consistent spatial patterns

 4. Other analytes followed this trend but did not pass statistical tests (likely due to

small sample size (n=5)of individual streams.

 For example, unfiltered total and methyl-Hg were consistently higher (up to 20x) in the

lower reaches except for in Salmon Creek (no salmon present)

Total Hg (unfiltered water)

Salmon Fish Peterson BCC Shrine

2 4 6 8 10

Methyl-Hg (unfiltered water)

Salmon Fish Peterson BCC Shrine 0.0 0.5 1.0 1.5 2.0 2.5 3.0

(ng L

• 1)

In the two streams with the highest spawner densities, filtered MeHg was 10 to 11-fold higher in the lower stream reach and made up 5-33% of the total Hg.

Total Hg (unfiltered water)

Salmon Fish Peterson BCC Shrine

2 4 6 8 10

Methyl-Hg (unfiltered water)

Salmon Fish Peterson BCC Shrine 0.0 0.5 1.0 1.5 2.0 2.5 3.0

(ng L

• 1)

30-day standard for fish-eating wildlife National average

• 5. Comparison of concentrations relative to health criteria and other sites nationwide shows:
• exceedance of 30-day fish-eating wildlife criterion for total Hg occurred in 3 of the 5 streams, especially

in the salmon-supporting reaches.

• unfiltered methyl-Hg in water is among the highest in the nation at Shrine B, the reach

with the highest salmon density

2.4 ng/L= Within the top 1% %

• f MeHg in streams and lakes in

the U.S.A. and 12x higher than the national mean (n=336; Scudder et al. 2009). Less than 1 km upstream, with few salmon, the concentration was only 0.18 ng/L).

Methylmercury bioaccumulates and biomagnifies in aquatic ecosystems

Methylmercury concentrations in aquatic organisms increase with increasing methylmercury concentrations in water and with increasing trophic level. Fish at the top of the food web tend to have the highest concentrations of methylmercury. (From: USGS Circular 1395 (Wentz et al. 2014).

Fish tissue-total Hg

Salmon Fish Peterson BCC Shrine

total Hg concentration (ng g-1 dw)

50 100 150 200 250 300 350 400

Concern for fish-eating mammals EPA protection of Human health

• Half the samples from resident/rearing fish exceeded 100 ng/g, which is the level of concern

for fish-eating mammals (Fig.9)

• Only exceedance of human health criteria was for fish tissues in lower Shrine Creek.

Conclusions

• Contaminant loads appear to be measurably influenced by the presence
• f salmon spawners and carcasses
• Mercury sources include a combination of spawner, geogenic, and

atmospheric influences.

• POPs occurrences in fish tissues were consistently enhanced in

lower stream reaches, indicating a dominant marine source

• inconsistent with Hg, which was also present in upper reaches
• Comparisons of concentrations in higher trophic organisms is

challenging due to differences in presence (by species, age) above and below barriers.

• Passive integrators (e.g. incubated leaf packs) should be further explored

as a meaningful monitoring tool in streams

• Implications:
• As salmon accumulate contaminants in the ocean and return

to streams to spawn, they can have a measureable effect on contaminant concentrations in stream ecosystems.

• Contributions by salmon should be better defined and

monitored into the future as marine contaminant levels change

Thank you! Questions?

Acknowledgments

Funding: INBRE

John Hudson, Independent Aquatic Ecologist, Juneau, AK. John DeWild, David Krabbenhoft, and staff at USGS Mercury Research Lab, Middleton, WI Gina Ylitalo and Bernadita Anulacion at Northwest Fisheries Science Center, Seattle, WA Eran Hood and Jason Fellman, University of Alaska Southeast Undergraduate research assistants: Chris Salazar, Alex Whitehead, and Alex Botelho (UAS)