Scientists completed a massive data compilation effort in order to quantify mercury loading and accumulation in watersheds of the Northeast.

This section presents information regarding:
1. Deposition of mercury from the air;
2. Accumulation of mercury in sediment; and
3. Concentrations of mercury in water and fish.

Mercury deposition (Papers 4 , 5 and 6)

Mercury travels for days to years after it is emitted to the air and event ually settles out onto the landscape. This settling process is called “deposition” and includes dry gases and particles, as well as rain and snow. Mercury deposition in wet forms (such as rain and snow) is measured by the national Mercury Deposition Network (MDN). There are 13 MDN sites wit hin the northeastern United States that have been operating since 1996 and most are located in rural and semi-rural areas. In addition to MDN, there is a comprehensive mercury monitoring site at the Proctor Maple Research Center in Underhill, Vermont (operating since 1993) and ot her sites near Boston operated by the Northeast States for Coordinated Air Use Management (NESCAUM) and the U.S. Geological Survey. Scientists have used the MDN and Underhill data to estimate the changes in wet mercury deposition with time and to map mercury deposition across the landscape.

In reviewing the mercury deposition data, scientists found t hat the annual deposition of wet mercury ranged from 3.1 to 9.5 micro-grams per meter-squared (µg-m2) in 2002. Seasonal patterns show that the concentration and amount of wet mercury deposited was greatest in the spring and summer months. Much of the wet mercury deposited by precipitation at the MDN sites arrived during specific storm events (20 to 60 percent of the total annual loading). The time period covered by the MDN data is too short to determine whether a trend exists in t he amount of mercur y deposited for the period 1996-2002. However, the number of weeks with very high mercury deposition decreased markedly in 2001 and 2002 compared to previous years. These high deposition periods may have ecological importance. The “fresh” new mercury deposited on the surface of a lake is more rapidly converted to toxic methylmercury than pre-existing mercury in the water. This conversion process is most pronounced during the summer growing period when high deposition events are more likely to occur.

Like at the MDN sites, wet forms of mercury deposited in precipitation at the Underhill site did not show a clear trend from 1993 to 2003, despite the decrease in emissions from nearby sources. This may be due to the impact of large sources to the west and southwest of the Underhill site.

The data from these monitoring sites were used as part of a larger effort to map the estimated total (wet and dry) mercur y deposition across the region. While this analysis was limited by the number and location of monitoring sites, the final map depicts higher mercury inputs in some areas of the Northeast than previously estimated. This is because all of the major deposition pathways were included in the model for the first time.

The mercury deposition model includes two important pathways for the dry deposition of mercury and highlights the important effects of forest cover and elevation on mercury deposition. According to these new model estimates, the greatest amount of mercury is deposited in forested and mountainous terrain (41.0 µg/m2/yr) and grades to lower amounts in flat northern landscapes (3.0 µg/m2/yr) (Figure 4). The new model also estimates that total mercury deposition is likely two to three times greater than wet mercury deposition that is currently measured by the national Mercury Deposition Network.

In addition to providing initial estimates of deposition, this map draws attention to the ecological importance of mercury uptake and release in forests. Forests enhance mercury deposition by “scavenging” mercury out of the air with their rough foliage. It is also thought that trees may assimilate mercury through gas exchange sites on the foliage known as stomata. For example, research has shown that tree leaves contain a higher proportion of mercury in the bioavailable methyl form than once thought. While it is not yet understood how this methylmercury is produced, it is reasonable to expect that once the leaves fall from the trees, the mercury can be ingested by insects, which are then eaten by amphibians, reptiles, birds or mammals. Methylmercury in leaves may also wash to streams as water flows over the forest floor during snowmelt and thereby serves as an important mercury input to nearby surface waters.

Last, the mercury deposition map points out the difficulty in estimating mercury deposition in urban areas or areas affected by point sources. More monitoring sites are needed to better depict this variation across the landscape and more accurately assess mercury exposure risks.

Mercury in lake and river sediments (Papers 7 & 8)

Scientists use cores of lake and river sediments to document changes in mercury deposition over time and to provide a baseline against which to measure future changes in mercury loading. By comparing the amount of mercury in sediments to mercury emissions, these data illustrate the connection bet ween airborne mercury and mercury in lakes.

An analysis of historical mercury accumulation rates in lake sediments shows a clear and consistent pattern. Mercury accumulation was slow prior to 1850, increased with industrialization and peaked across the region from 1970 to 1980 (Figure 5). Mercury accumulation in sediments has declined since that time, consistent with the decrease in mercury emissions in North America. Even with this reduction, mercury is currently accumulating in lake sediments at a rate two to five times faster than pre-industrial rates.

Researchers also analyzed surface sediments that reflect present-day conditions at more than 570 sites. They found that total mercury concentrations ranged from 0.01 to
3.7 ppm with the highest levels reported in lakes. Methylmercury concentrations in the sediments spanned 0.15 to 21.0 ppb, wit h rivers showing higher proportions of mercur y in the methyl form. Forty-four percent of the waterbodies sampled exceed federal guidelines for the protection of aquatic biota (NOA A 1999). No quantifiable spatial pattern was observed from the data, but high values tended to occur in sediments in lakes in Massachusetts and southeastern New Hampshire.

Mercury in water (Paper 9)

Once mercury is deposited to the landscape, most of it flows into rivers and lakes where it becomes available for fish, wildlife and human consumption. Understanding the levels and patterns of mercury in surface waters is critical to addressing this widespread environmental threat.

Scientists have compiled data for mercury in water from more than 1,000 locations from Massachusetts to Newfoundland. They used this information to determine whether spatial patterns exist and to identify the factors that make a waterbody sensitive to methylmercury loading. The analysis was limited to data that were collected under low flow conditions in order to minimize the effects of seasonal changes associated with periods of high streamflow.

The measurements of total mercury in water ranged from 0.5 to 19.5 ppt, with the highest concentrations found in Nova Scotia, Newfoundland and the Adirondacks of New York (see Figure 6). The waters wit h high mercur y levels were often distant f rom direct point sources and urbanized land use, suggesting airborne mercury as a likely source. However, the data also demonstrate that point sources can have a considerable impact in local areas, as seen at two well known sites in the region. Very high mercury concentrations were detected in surface waters near Portland, Maine, and in the urban corridor of Boston, Massachusetts.

These findings point to the need for a two-pronged approach to address mercury levels in surface waters; reducing mercury emissions to the air and controlling direct mercury discharges to surface waters.

Mercury in freshwater fish (Papers 11 & 13)

Scientists analyzed mercury measurements from 1980 to the present for more than 15,000 fishes, spanning 64 different fish species to assess the extent and nat ure of mercury contamination in the Northeast. This analysis is considered t he first published work to utilize such an extensive dataset to describe fish tissue mercury concentrations at the sub-continental scale.

Mercury levels across all fish species ranged from 0.09 to 1.02 ppm, with the highest concentrations in white perch that reside in reservoirs. Overall, 15 and 42 percent of the waterbodies sampled for brook trout and yellow perch, respectively, had average fish mercur y concentrations (in fillets) above the EPA methylmercury criterion of 0.3 ppm. The scientists also identified specific species that tend to have high mercur y levels; bass species, pike, lake trout, white perch and walleye were highest (Figure 7A). Other factors such as fish length and habitat (lake, river or reservoir) are good predictors of mercury levels (Figure 7B).

Individual waterbody characteristics also strongly influence fish mercury concentrations. A detailed analysis of the conditions that most likely lead to mercur y problems in fish identified several important parameters (see the list that follows). In general, acidic water bodies that have complex food chains and numerous wetlands tend to have fish with high mercuy concentrations.

Attributes of mercury-sensitive surface waters:

Chemical

• High acidity
• Low acid neutralizing capacity
• High sulfate

Physical

• Abundant wetlands (particularly along the shore)
• Small lake with a large watershed area
• Summer water level fluctuations > 6 feet

Biological

• Low zooplankton abundance
• Low nutrient level
• Numerous trophic levels in the food chain

Given the variation in waterbody characteristics across the landscape, no distinct spatial pattern was detected in average fish mercur y concentrations, although some areas had high fish mercury levels compared to others. Overall, the characteristics of a watershed may be as important as the actual deposition in predicting mercury levels in fish. For this reason, it is not possible to pick and choose where to reduce mercury pollution across a region to achieve fish mercury goals. Rather, an approach where reductions occur at all facilities would likely be more effective.