Metal Loadings to the Hudson River and New York Harbor: How have

they changed?

Robert MasonDepartment of Marine Sciences

University of Connecticut, Groton, CT 06340

Mostly based on data collected and analyzed during a project funded by the Hudson River

Foundation (Fitzgerald and Mason as PI’s)

Outline

• Why we care about metals in the Hudson?

• Sediment Data

• Water Data – Factors controlling distributions in the upper estuary and NY/NJ Harbor

• Water Data – Changes over time

• Mass Balances for the Metals - Sources to the Harbor, and major sinks

Metal Toxicity?

1) Impact is on lower food chainorganisms for most metals asbioaccumulation is greatest for plankton. In the Hudson,Cd is the most important example as there are known point sourcecontamination (e.g. Foundry Cove)that been shown to cause toxicity.However, organisms were able togenerate resistance to the Cd.

2) Toxicity of highly bioaccumlative compounds is manifest in the higher level consumers and toxicity to plankton may not occur. Example is methylmercury which has most impact for higher food chain consumers (humans, mammals and birds) that feed on piscivorous fish

Levington et al 2008

EPA

A number of species exceed the EPA recommended value for mercury (assuming all as methylmercury)

Health of the Harbor report

Gobeille et al., 2006

Gobeille et al., 2006

People consuming local fish have higher mercury levels than those who do not .Levels found in Hudson River similar to that for other ecosystems where people eat local fishLevels below the EPA suggested reference level of 5.8 ng/mL.

Sediment Concentrations

Levels of metals in sediments of the Hudson River are elevated above other US estuaries and higher than background concentrations

Taken from Bopp et al., Ch 24

MetalNY/NJ Harbor

Jamaica Bay

Arthur Kill

Newark Bay

Hacken-sack R

Passaic R

Pb mid-60’s 235 336 398 373 343 568

Pb 1990’s 126 81 211 142 109 307

%Decrease 46 76 47 62 68 46

Cd mid-60’s 5.6 10.9 21.3 11 8.1 12.4

Cd 1990’s 1.1 1.7 2.9 1.95 3 4.2

%Decrease 80 84 86 82 63 66

Cu mid-60’s 268 362 1395 342 240 248

Cu 1990’s 122 93 312 160 132 120

%Decrease 54 74 78 53 45 52

Zn mid-60’s 352 608 1045 717 741 1052

Zn 1990’s 232 210 404 265 327 523

%Decrease 34 65 61 63 56 50

Cr mid-60’s 286 225 325 968

Cr 1990’s 116 102 166

%Decrease 59 55 49

Hg mid-60’s 3.4 2.3 20 10.8 7.77 12.48

Hg 1990’s 1.6 0.45 4.94 3.6 3.64 1.83

%Decrease 53 80 75 67 53 85

Changes in Sediment Metal Conc (ppm) between 1960’s & 1990’s

MetalMain, above Dam

Batten Kill

Main, Lock 2

Mohawk River

Main at Troy

Main, at Kingston

Pb mid-60’s 1560 46 512 68 411 118

Pb 1990’s 69 45 42 9.1 42 44

%Decrease 96 2 92 87 90 63

Cd mid-60’s 115.0 5.7 22.4 0.5 10.6 3.4

Cd 1990’s 30 0.6 <0.5 1.2 2.6 0.03

%Decrease 74 89 >98 Increase 75 99

Cu mid-60’s 121 42 31 132 508 76

Cu 1990’s 31 30 22 43 59 27

%Decrease 74 29 29 67 88 64

Zn mid-60’s 1100 1026 662 - 1056 308

Zn 1990’s 108 265 79 103 224 127

%Decrease 90 74 88 79 59

Cr mid-60’s 1440 26 - - 195

Cr 1990’s 154 21 - 52 74

%Decrease 89 19 62

Hg mid-60’s 8.93 0.45 1.95 0.93

Hg 1990’s 0.56 0.2 0.13 0.42 0.19

%Decrease 94 71 78 80

Changes in Sediment Metal Conc (ppm) between 1960’s & 1990’s

From Bopp et al., Ch 24

Sediment Concentration Changes

• For most metals and locations, there has been a dramatic decrease in concentration between 1960’s and 1990’s

• In mainstem, no trend in concentration with distance

• Changes somewhat less dramatic in the lower estuary

• Concentrations still elevated above background for many metals.

• How does this impact the ecosystem?

MetalMain, above Dam

Batten Kill

Main, Lock 2

Mohawk River

Main at Troy

Main, at Kingston

HRM ~195 HMR ~165 HRM 150 HRM ~90

Hg mid-60’s 8.93 0.45 1.95 0.93

Hg 1990’s 0.56 0.2 0.13 0.42 0.19

%Decrease 94 71 78 80

Relative change in fish concentration less than that of sediment for Hg. Also fish do not track sediment concentration

Levinton & Pochron, 2008

Data summarized was derived from a variety of literature sources and databases. Evaluated for appropriateness of the analytical methods and complete presentation of quality assurance/quality control (QA/QC) procedures and results.

Large databases included CARP (Contaminant Assessment and Reduction Project), EMAP (Environmental Monitoring and Assessment Project), and data from regulatory agencies (e.g., NYSDEC State Waters Monitoring Section [SWMS] and Division of Water).

Most published studies have reported surface and deep water concentrations of metals (Hg, Cd, Cu, Fe, Ni, Zn, Ag, and Pb) in the lower Hudson (south of Newburg; ~25 to 100 km north of the Battery) and the estuarine turbidity maximum (ETM; south of the Harlem River; ~ 4 to 25 km).

Sources of DataAssessment of Water Column Data

• Kingston-Poughkeepsie

• Glenmont

• Cohoes• Waterford

0 10 20 30 Km

Hudson RiverWatershed

NY/NJ Harbor

Hudson River Estuary MetalAtomic Mass

1 ug/L = x nM 1ppm = x nmol/g

Ni 58.7 17.0Cu 63.5 15.7Zn 65.4 15.3Cd 112.4 8.9Hg 200.6 5.0Pb 207.2 4.8

Water - Middle & Upper Hudson(SWMS; collected 1993-2006; mean ± SD)

Water - Middle & Upper Hudson(CARP; collected 1999-2001; mean ± SD)

TSS - Upper Hudson/Sources of Metals(SWMS; collected 1993-2006; mean ± SE)

TSS - Upper Hudson Sources of Hg

(CARP & SWMS; collected 1999-2001; mean ± SE)

Spatial/seasonal MMHg distributions

(CARP; collected 1999-2000; mean ± SE)

Average total (filtered + particulate) concentrations of Hg (58 to 130 pM), Cu, Ni, Zn, and total suspended solids (TSS) in surface waters are highest at Poughkeepsie.

Cd (0.97 to 1.3 nM) and Pb were more uniformly distributed in the middle and upper Hudson.

Fe-normalized suspended particle concentrations of Hg, Cd, Pb, Cu, Ni, and Zn were elevated in the upper Hudson at Waterford and Cohoes as compared to Poughkeepsie, indicating watershed sources of trace metals.

MMHg in the middle and upper Hudson is likely derived/connected to net production in the Hudson River watershed.

Summarymiddle and upper Hudson

Hudson River - Water(Lower Hudson and ETM)

(CARP; collected 1999-2001; mean ± SE)

In the estuarine turbidity maximum (ETM), average total Cd concentrations (0.74 to 1.2 nM, surface waters) are similar to the middle and upper Hudson sites, while average total Hg (54 to 1110 pM, surface and bottom waters) is elevated as compared to the rest of the upper Hudson.

The filtered fraction of Cd increases dramatically with salinity (83 to 120% of total Cd) in the lower Hudson River, and higher total Hg and Cd levels are associated with spring high flow periods.

Average surface water suspended particle concentrations of Hg (2.20 to 3.54 nmol g-1) and Cd (5.45 to 12.6 nmol g-1) are largely overlapping between the ETM and lower Hudson.

ETM and upper Hudson suspended particle metals concentrations are elevated as compared to the middle Hudson (Poughkeepsie), indicating sources of metals in both regions.

SummaryLower Hudson and ETM

Major Rivers – NY/NJ Harbor(CARP; mean ± SE; collected 1999-2001)

Average total Hg (145 and 433 pM, surface and bottom waters) and Cd (0.97 and 1.35 nM) concentrations at the mouth of the Passaic River were elevated as compared to the Hudson ETM, while only total Hg (109 pM) was elevated at the mouth of the Hackensack River.

Concentrations in the lower East River and the Raritan River were similar to levels in the Hudson ETM.

Average suspended particle Hg (9.8 to 20.8 nmol g-1) and Cd (9.6 to 49.2 nmol g-1) concentrations were elevated in both the Hackensack and Passaic as compared to the Hudson, although suspended particle Cd was less near the mouths of rivers.

Summary - Rivers

NY/NJ Harbor - Water(CARP; collected 1999-2001; mean ± SD)

NY/NJ Harbor Water(Balcom et al. 2008; collected 2002-2003; mean ± SD)

NY/NJ Harbor - Water

(Paulson, 2005; collected 1999; mean ± SD)

Temporal Comparison Water

Klinkhammer and Bender (1981)collected 1974-1975

NY Bight

Paulson (2005) collected 1999

Rockaway-Sandy Hooktransect

(mean ± SD)

Sañudo-Wilhelmy and Gill (1999)reported a similar decrease indissolved metals for the HudsonETM and Lower Harbor between1974/1975 and 1995/1997.

NY/NJ Harbor – Recent Sediment(CARP, collected 1999-2001;

EMAP, surface sediment collected 1998; mean ± SD)

Circles are data not included in the averages as considered outliers

NY/NJ Harbor - Recent Sediment(Balcom et al., 2008 and Hammerschmidt et al., 2008

suspended particles and surface sediment collected 2002-2003; mean ± SD)

Circles - data not included in the averages

NY/NJ Harbor – Recent Sediment(Paulson, 2005, suspended particles collected 1999;

EMAP, surface sediment collected 1998 and 1993-1994; mean ± SD)

Circles - data not included in the averages

ETM and NY/NJ Harbor - Ratios(Paulson, 2005 and Feng et al., 2002 - suspended particles collected 1994-1999 ;

EMAP, surface sediment collected 1998 and 1993-1994; mean ± SE)

ETM and NY/NJ Harbor - Ratios(Paulson, 2005 and Feng et al., 2002 - suspended particles collected 1994-1999 ;

EMAP, surface sediment collected 1998; mean ± SE)

ETM (10 km)

ETM (4 km)

Upper Harbor

Raritan Bay

Lower Harbor

Rock.-Sandy Hook

NY Bight

Hg

/Fe

an

d H

g/A

l (x

10

-4)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

PO

C (

µM

) an

d s

iltc

lay

(%)

0

20

40

60

80

100

TO

C (

µm

ol

g-1

)0

500

1000

1500

2000

2500

water POC

sediment siltclay

sediment TOC

Hg

ETM (10 km)

ETM (4 km)

Upper Harbor

Raritan Bay

Lower Harbor

Rock.-Sandy Hook

NY BightC

d/F

e a

nd

Cd

/Al

(x 1

0-4

)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

suspended particle metal/Fe sediment metal/Fesuspended particle metal/Alsediment metal/Al

Cd

Regional average trace metal concentrations of Hg (3.2 to 166 pM) and Cd (0.19 to 0.90 nM) in NY/NJ Harbor were generally elevated and more variable near the input of major rivers in Newark Bay and the Upper Harbor, roughly corresponding with elevated TSS levels.

Harbor Hg and Pb were mainly in the particulate phase, while Cd, Cu, and Ni have more metal in the dissolved fraction at elevated Harbor salinities.

Over 25 years (1974 to 1999), average dissolved metals concentrations (surface and deep) in the Lower Harbor and NY Bight region were reduced by 85 to 90%.

Summary – NY/NJ Harbor

There was good agreement between average suspended particle concentrations of Hg (1.1 to 8.7 nmol g-1), Cd (3.1 to 13.2 nmol g-1), and other metals with surface sediment concentrations in the Harbor, while suspended particle concentrations in the ETM are known to be elevated as compared to sediments (Feng et al., 2002).

However, Al- and Fe-normalized suspended particle metal concentrations were reduced in the Hudson ETM and elevated in Raritan Bay and at the Rockaway-Sandy Hook transect suggesting metals enrichment due to association with fine particulate material or particulate organic matter.

Summary – NY/NJ Harbor(continued)

57±26%

26±3% 9±1%4±1%

4±15%

59±24%

8±1%

10±1%

2±0.3%

21±11%

Budget sources – percent inputs(mean ± SE)

2±0.2%

Rivers East River WPCFs Atm. Deposition Benthic flux

69±27%

23±7%5±0.5%

2±4%

Hg - 3100 moles y-1

Cd - 25 K moles y-1 MMHg - 14 moles y-1

Budget sources – percent inputs(mean ± SE)

Pb - 850 K moles y-1

Cu - 2800 K moles y-1

38±21%

25±4%

16±0.4%

2±0.2%

19±27%

Ni - 1700 K moles y-1

39±9%

19±2%

23±0.4%8%

11±28%

Zn - 7100 K moles y-1

Temporal Comparison of Metal Inputs

(mean ± SE or range)

Reference

Sampling

Period

Cd

(K moles y-1)

Ni

(K moles y-1)

Cu

(K moles y-

1)

Zn

(K moles y-1)

This Study ~ 2000 - 2006 25 ± 7.4 1700 ± 580 2800 ± 660 7100 ± 2100

Klinkhammer & Bender (1981)

April 1974 280 - 590 4500 - 95005200 - 10000

11100 - 28300

(enhancement) (11 to 24 x) (2.5 to 5.5 x) (2 to 3.5 x) (1.5 to 4 x)

October 1975 190 - 500 3400 - 7500 4000 - 8400 7000 - 15100

(enhancement) (8 to 20 x) (2 to 4.5 x) (1.5 to 3 x) (1 to 2 x)

Decrease most dramatic for Cd - >90% reduction in inputs. Other metals show more modest reductions over time.

Current annual loadings (~2000 to 2006) to NY/NJ Harbor estimated for Hg (3100 mol y-1), MMHg (37 mol y-1), Cd (25 Kmol y-1), Pb (850 Kmol y-1), Ni (1700 Kmol y-1), Cu (2800 Kmol y-1), Zn (7100 Kmol y-1)

The majority of inputs are from rivers (38 to 88% of totals).

River fluxes of metals were estimated independently using a Hudson sediment delivery flux (8.12 ± 3.25 x 1011 g y-1; Wall et al. [2008] and HydroQual [2003]) and mean suspended particle metals concen-trations in the middle/lower Hudson. These estimates agreed well with the annual loadings shown above.

Total inputs are significantly decreased for Cd (8 to 42x), Ni (2 to 5.5x), Cu (1.5 to 3.5x), and Zn (1 to 4x) since the mid 1970s (25 to 30 years).

The relative contribution of metals from sewage has decreased.

Hg inputs are likely to have declined by about a factor of 3-5 since the mid-1960s.

Summary – Mass Balances

There is a need for high quality time series measurements of filtered and suspended particle metals concentrations, as well as important ancillary parameters (e.g., POC), in the waters of the Hudson River and NY/NJ Harbor.

The decline in metals concentrations in the waters of NY/NJ Harbor since the mid-1970s was established by a comparison between adjacent regions. The time series of measurements is incomplete for any one region of the Harbor.

Limited measurements (1999-2001) of Hg and Cd concentrations in the middle and upper Hudson, major regions of the Harbor, and the East River, with almost no data on suspended particle concentrations.

Conclusions

The sediment-water fluxes of most trace metals has not been measured in NY/NJ Harbor, and are needed considering the importance of this source term in some mass balances (e.g., MMHg)

Bulk (wet and dry) atmospheric deposition measurements are limited for most metals in the Harbor region, and are important to mass balances since atmospheric deposition of metals and subsequent yield from watersheds accounts for a substantial portion of the fluxes of metals in rivers.

We hypothesize that much of the MMHg in NY/NJ Harbor is likely connected to net production in the Hudson River watershed, and that there is limited MMHg input from river sediments.

Conclusions(continued)

Seasonal distributions of Hg

Distance (km)

0 50 100 150 200 250

To

tal

Hg

(p

M)

0

50

100

150

200

250spring 1999 fall 1999 spring 2000 summer2000

ETMWaterford

Distance (km)

0 50 100 150 200 250

Fil

tere

d H

g (

pM

)

0

5

10

15

20spring 1999 fall 1999 spring 2000 summer2000

ETM Waterford

Total Hg Filtered Hg

(CARP; collected 1999-2000; mean ± SE)

Mass balances - NY/NJ Harbor

Sources

Hg

(moles y-1)

MMHg

(moles y-1)

Cd

(K moles y-1)

Rivers 2100 ± 820 23 ± 9 14 ± 6.4

East River 700 ± 200 2.9 ± 0.5 6.5 ± 0.7

WPCF’s 140 ± 15 3.7 ± 0.4 2.1 ± 0.3

Atmospheric Deposition 60 ± 5 0.9 ± 0.1 1.1 ± 0.2

Benthic Flux 60 ± 130 8 ± 4 1.1 ± 3.7

Total 3100 ± 900 37 ± 10 25 ± 7.4

(moles y-1) (moles y-1) (K mol y-1)

Sinks

Estuarine Exchange 1300 ± 270 14 ± 2.0 18 ± 2.4

Evasion 60 ± 20 -- --

Sediment Burial 1700 ± 9001 4.0 ± 3.0 6.7 ± 8.01

(Photo)demethylation -- 2.0 ± 1.0 --

Biological Processes -- 17 ± 111 --

Total 3100 37 25

(mean ± SE; ~2000 to 2006)

1 closing term

Watershed Hg yields were estimated to range from 0.011 to 0.019 µmol m-2 y-1 (25-30% of atmospheric wet deposition) resulting in delivery of 480 to 780 mol Hg y-1 to the major rivers surrounding the Harbor (Balcom et al., 2008).

Watershed fluxes are estimated to account for 23 to 37% of the Hg flux from rivers to NY/NJ Harbor (2100 ± 820 mol y-1)

Assuming that the watershed yield of MMHg is 3% of the HgT flux (based on atmospheric deposition measurements), watershed MMHg yields are estimated to range from 3.4 to 5.6 x 10-4 µmol m-2 y-1 (Driscoll et al. [1998] reported 8.4 x 10-4 µmol m-2 y-1), resulting in delivery of 14 to 23 mol y-1 to rivers.

Indicates that watershed delivery may account for the majority of the MMHg flux from rivers to NY/NJ Harbor (23 ± 9 mol y-1) estimated in the current study, and that there may be limited net in-situ production and input of MMHg from river sediments.

Watershed Contributions

The Pettaquamscutt River Estuary (PRE; RI) is relatively remote from point sources with a watershed that can be characterized as rural/residential.

Scrupulously dated, varved sediment core (Lima et al., 2003/2005) provides a potentially valuable analog for assessing anthropogenic impact and temporal changes in Hg accumulation/inputs in the Harbor Estuary over the past half century.

Hg flux ratio (actual/preindustrial Hg accumulation) peaked in 1950 and 1960 at 12, but the ratio in 1997 was 4, representing a substantial decline in Hg inputs by about a factor of 3.

Hg Accumulation in NY/NJ Harbor

Sediment Hg accumulation fluxes are not available for the Harbor, but using sediment Hg concentrations (Bopp et al., 2006), approximate accumulation ratios can be estimated.

In the mid-1960s the actual/preindustrial Hg concentration ratio ranged from 13 (Jamaica Bay) to >100 (Arthur Kill), and were reduced to 3 to 27 in 1995/1996.

Therefore, there has been a decline in contaminant Hg inputs to the Harbor Estuary since the 1960s, and although difficult to provide a quantitative estimate, a factor of 3-5 is likely

Although there is substantially more Hg accumulating per unit area, the Harbor is comparable to the PRE where the decline in Hg accumulation was about 3.

Hg Accumulation in NY/NJ Harbor(continued)