Distribution
of Mercury in the Mobile River Basin in Relation to Land Use
Dr. Kimberley Warner
University of Alabama Department of Biological Sciences
Co-authors: Jean-Claude Bonzongo, Eric Roden, W. Berry Lyons, Milt Ward, Indrajeet
Chaubey, and Hobson Bryan.
In the past decade, mercury
(Hg) concentrations above levels that could pose human health risks have been
measured in predatory fish from many rivers and reservoirs in the southeastern
region of the United States. This region, and mainly its Coastal Plain portion,
may be particularly vulnerable to Hg contamination in aquatic food chains, due
to the coexistence of both natural and human-imposed conditions which favor
the production and accumulation of methyl-Hg. Several specific factors are hypothesized
to contribute to the development of such conditions, including: (1) nutrient
loading from certain land-use activities and increased sedimentation above water
impoundments develop conditions favorable for methyl-Hg production. (2) Increased
sulfate loading from energy resource extraction operations result in increased
methyl-Hg production. (3) Abundant wetlands within the Mobile-Alabama River
Basin (MARB) contribute to methyl-Hg loads downstream and in fish. (4) Fish
tissue levels of methyl-Hg are related to levels of Hg in river waters and to
net rates of methyl-Hg production in sediments.
The objectives of this study were to: (1) to determine levels and speciation
of Hg in different compartments of various aquatic systems in the MARB; (2)
to investigate the linkage between land use types or the presence of wetlands
and microbial processes associated with methyl-Hg production; (3) to use GIS
to represent spatially arranged data and ultimately to predict Hg levels in
fish; and (4) to use a participatory approach to environmental decision-making
to ameliorate conflict, and achieve an effective public understanding and support
for Hg policy. This report will address preliminary results and progress on
the science aspects of this interdisciplinary research project conducted by
the University of Alabama Center for Freshwater Studies.
The first phase of the project involved a wide survey of Hg distributions in
largemouth bass, water and sediment at 52 sites representing various hypothesized
impact factors (e.g. wetlands, agriculture, dams). Various sediment chemistry
and water quality parameters were measured in conjunction. A total of 96 fish
samples were taken from 51 out of 52 sites. While we attempted to collect 2
fish at each site, at least two fish were sampled at 43 sites and only one fish
at the remaining 8 sites. The concentrations of Hg in these 96 fish spanned
over 2 orders of magnitude, from 0.02 to 2.8 ppm (mg kg-1 wet tissue).
Mean and median concentrations were 0.45 and 0.32 ppm, respectively, with the
mean close to that of the National Mercury Survey (0.39 ppm) reported for AL
largemouth bass. Twelve percent (12%) of fish had Hg concentrations >
1 ppm, the level at which consumption advisories are posted in Alabama. Average
fish Hg concentrations were > 0.5 ppm at 21 (41%) of the sites. Coefficients
of variation in fish Hg concentrations from any one site were usually high,
averaging 68%. In many cases, Hg concentration was higher in the smaller of
two fish. Total-Hg and methyl-Hg concentrations in water were low and ranged
from 0.2-3.8 and 0.01-1.5ng L-1, respectively. This results in an
average -log bioconcentration factor of 5.6 for Hg between fish and water. Hg
speciation determinations in sediments are still underway. The annual flux of
aqueous total-Hg from the MARB to Mobile Bay is estimated at 138Kg yr-1.
No obvious trends were apparent between fish Hg concentrations and projected
impact factors, likely due to the large within-site variability noted above.
However, some variability within certain impact factors could be explained by
other factors. For example, fish Hg concentrations in wetland sites were positively
related to watershed area and inversely related to water depth above dams, for
dam impacts. Fish Hg concentrations were also found to be a weak negative function
of water pH. Aqueous total-Hg and many sediment and water quality parameters
were found to be a significant function of land use.
The second phase of the project involved in-depth investigations examining controls
on Hg transformation and bioaccumulation. We focused on one pool (Demopolis)
containing 4 sites with different impact factors: (i) dam, (ii) agricultural,
(iii) wetland, and (iv) open river. Potential rates of microbial Hg methylation
and methyl-Hg demethylation were determined in sediments. The ratio of methylation
to demethylation rates was positively related to the percent of methyl-Hg formed
in native sediments. Concentrations of Hg (127-393ng g dry sediment-1) and methyl-Hg
(0.13-2ng g-1) in native sediments increased in the following order of impacts:
open river < dam < agriculture < wetland. The percentage of methyl-Hg
produced in the sediments was a positive function of sediment iron, organic
matter, and porosity. Aqueous total-Hg increased with aqueous total suspended
solids and iron concentrations. Aqueous methyl-Hg was a positive function of
dissolved organic carbon in water and of methyl-Hg and total iron concentration
in sediments.
Despite differences in net production of methyl-Hg in sediments from the different
sites, the average concentration of Hg in fish tissues among the 4 sites was
consistently rather high. The average concentrations of Hg (ppm±1SD)
in 6 fish from each site were: open river: 0.77±0.45; dam: 0.85±0.35;
agriculture: 0.88±0.35; wetland: 1.7±0.80. Comparing mean fish
Hg concentrations among sites, only the wetland site was found to be significantly
different from the other three sites.
These preliminary results suggest net methyl-Hg production in sediments and
flux to water is greater in environments with organic-rich, slower moving turbid
waters and fine-grained sediments. These, in turn, are related to land use and
hydrological variables, consistent with our hypotheses. However, we found that
fish collected from rivers/stream sites with differing impacts have highly variable
Hg burdens, which may be explained, at least in part, by their mobility or the
mobility of their prey. Therefore, it may not be possible to draw direct quantitative
links between land use types and Hg concentration in fish in physically dynamic
riverine ecosystems.
Complete Abstract List (suitable for printing)