ATLANTIC salmon (Salmo salar) in the state of Maine represents the last wild population of this species in the US, and its listing as endangered in 2000 underscores the seriousness of its decline. Dam removal has been identified as the most important strategy for restoring salmon populations in Maine and two dams on the lower Penobscot river, the Great Works and Veazie dams, have been designated for removal.

However, dam removal can result in the release of contaminants from riverine sediments into overlying waters, potentially increasing water toxicity to resident species, including fish. Because dams are to be removed as part of the Penobscot River Restoration project, there is a need to evaluate the toxic potential of Penobscot river sediments prior to dam removal. A team of scientists working at the University of Maine are using a simple laboratory-based sediment resuspension design, and two well-established aquatic toxicology models, fathead minnows (Pimephales promelas) and zebrafish (Danio rerio), to evaluate if resuspension of Penobscot river sediment significantly elevates the toxicity of river water and to provide preliminary information on the types of chemicals likely to desorb during resuspension.

Project history

Over the past 150 years, there has been a steady decline in fish populations that migrate between freshwater and saltwater in Maine rivers. This is most likely the result of multitude impacts, including over-fishing and habitat degradation by obstructions (such as dam construction), as well as pollutants introduced by industry, most notably dioxins from pulp and paper mills.

Significant fish populations in Maine whose migratory movements are impeded by dams include alewives, American shad, blueback herring, American eels and Atlantic salmon. Of these, Maine Atlantic salmon are of particular importance because of the fact that they represent the last wild population in the US. Maine Atlantic salmon were also the principal fish species drawing fishermen to Maine in the 1800s, leading to the formation of numerous salmon clubs, including the Penobscot Salmon Club in 1884. Indeed, the fame of Maine salmon led to the annual tradition, begun in 1912, of sending the first Penobscot Atlantic salmon of the year to the President of the US. Atlantic salmon are also woven into the culture of native American tribes in Maine, for whom they hold particular spiritual significance. Yet for more than 100 years, the Penobscot Indian nation has been unable to exercise its tribal fishing rights to catch fish such as Atlantic salmon, because the river is virtually devoid of native sea-run fish above Veazie dam (

The dams

The Great Works dam was built in the late 1800s. A concrete and timber crib construction, it is 311m long (main dam and forebay combined) with 5.8m of head and currently generates 7.9MW. The 8.4MW Veazie dam, located 13km downstream from Great Works, was built in 1910 out of concrete, and is 268m long with a 5.8m head. It is the last physical barrier juvenile salmon must cross before they reach Penobscot Bay and the sea.

Removal of these two dams is possible because the owner, PPL Corporation, has the option to increase electricity generation at six existing dams, allowing it to produce power that will nearly equal that of the current energy output. In conjunction with dam removal, fish passage improvements will be made in at least four other PPL dams, including installation of a state-of-the-art fish passageway at the Milford dam, located 3km upstream of the Great Works dam. Removing Great Works and Veazie dams, and improving fish passage, will significantly improve access to more than 500 additional miles of habitat for Atlantic salmon and ten other species of migratory fish while maintaining nearly all of the power generation (Penobscot River Restoration Trust,

Environmental issues

In its 2004 report ‘Atlantic Salmon in Maine’, the US National Research Council identified dam removal as the most important strategy for restoring Maine salmon populations. Atlantic salmon lay their eggs in shallow, gravel-rich areas of fast-moving streams and rivers than can be hundreds of kilometres upstream from the sea. Despite tremendous innovations in fish passage design, dams continue to present significant barriers to the upstream passage of migratory fish; dam removal remains the most effective means for improving fish access to upstream spawning areas.

Although dam removal improves fish passage, it can also result in the release of contaminants from riverine sediments into overlying waters, potentially increasing water toxicity to resident species. Dam removal leads to sediment resuspension, due both to the release of sediments trapped behind dams and resuspension of sediments deposited elsewhere in the river. Sediments act as repositories for persistent organic pollutants, including organochlorines, aromatic hydrocarbons, organo-metals and pesticides. Sediment resuspension can release these chemicals to overlying waters, leading to changes in their physico-chemical properties, including potential alterations in toxicity.

Aquatic organisms are particularly vulnerable to dam removal, being subjected to a multitude of stressors associated with a habitat undergoing dramatic physical, chemical and biological changes. With funding from the George Mitchell Center for the Environment at the University of Maine, the research team is conducting a pilot study to evaluate the potential for river water toxicity to increase following dam removal in the Penobscot river. The goal of the study is to determine if resuspension of sediments following dam removal in the Penobscot river is likely to significantly increase the toxicity of riverine water to early life stage fishes.


The researchers have collected sediment from two sites with known chemical contamination downstream of the two dams. One contains fine-grained sediments contaminated with high levels of mercury, the other is a well-characterised petroleum/tar deposition site. To simulate dam removal, the research team is using a simple laboratory-based design in which sediments are resuspended in pre-extracted river water in pre-cleaned teflon bottles for three days using a magnetic stir bar. This concentration represents an upper limit to environmentally realistic concentrations of suspended sediment caused by such events as storms, dam removals and dredging operations. After settling, the water is clarified by filtration to provide ‘desorption water’. River water stirred for three days without sediment and then filtered will serve as ‘control water.’ ‘Control’ sediments and ‘standardised water’ are not needed for this initial screening as the objective is simply to determine changes in the toxicity of existing river water using relevant Penobscot sediments.

Both ‘control water’ and ‘desorption water’ are characterised for general water quality (pH, hardness, etc) and screened for a variety of organic contaminants, including polynuclear aromatic hydrocarbons (PAHs), selected polychlorinated biphenyl (PCB) congeners and organochlorine pesticides such as DDT, DDD and DDE, by gas chromatography/mass spectrometry.

To evaluate the toxicity of the ‘desorption water’, the research team is using two well-established aquatic toxicology models, fathead minnows (Pimephales promelas) and zebrafish (Danio rerio). Several endpoints are being evaluated using different life stages. To evaluate early life-stage effects, fish are exposed as embryos and survival, hatch success, immune competence, and embryological development are measured. To evaluate whether organic contaminants, such as PAHs and PCBs, are bioavailable in ‘desorption water’, the biomarker enzyme, cytochrome P4501A, will also be measured in early life stage fish. Physiological responses – including immune function, reproductive function, and embryologic development – are predictive of population level effects and are commonly used as indicators of contaminant stress. To evaluate whether bioactive metals and/or endocrine disrupting substances are present, transgenic zebrafish are used. The use of transgenic fish with reporter genes indicative of exposure to xenoestrogens (estrogen-like foreign compounds) and metals provides powerful tools for identifying the potential biochemical mechanisms underlying these effects. Understanding underlying mechanisms provides predictive capabilities for extrapolating results to other species.

Past work in this area

Studies of the bioavailability and toxicity of sediment-associated contaminants are numerous, but there are none specifically tied to dam removal. The few studies examining dam removal effects on biota focus on the effects of higher sediment loads, such as mortality due to burial of benthic communities by large amounts of newly-released sediments, or changes in community structure related to channelisation, loss of deep pools, and/or the creation of shallow riffle zones in areas that were once quiescent ponds. However, studies of chemical desorption from the resuspension of lake sediments during storm events, tidal resuspension in estuaries or the release of chemicals during harbour dredging indicate that sediment resuspension can significantly increase the concentration of dissolved metals and hydrocarbon pollutants in the water column.

Summary of closely related research

Numerous studies have simulated sediment resuspension in laboratory settings to examine the sorption and desorption of particle reactive contaminants, including studies on fluxes of PCBs and PAHs, adsorption/desorption models, and remobilisation of metals from contaminated sediments. The research team from the University of Maine has used resuspended sediments to examine the dynamics of PAH desorption and degradation under different resuspension frequencies, and has used clarified water from resuspension experiments for toxicity tests with the water-flea Ceriodaphnia dubia, and other aquatic test organisms, to infer toxicity from sediment-associated current-use pesticides in sediments from northern California rivers.

The monooxygenase enzyme cytochrome P4501A (CYP1A) is a well-established biomarker of response to organic toxicants, including those likely to be found in Penobscot river sediments. Exposure to these chemicals significantly induces expression of this enzyme in all fish species tested to date. The research team and others have conducted extensive studies of CYP1A expression in fish, including early life-stages. CYP1A induction has been linked to developmental abnormalities in early life stage fishes in both laboratory and field studies, demonstrating the relevance of both CYP1A and deformities as indicators of physiologically adverse effects of aqueous pollutants.

When pathogens invade an organism, immune cells known as phagocytes literally ‘eat’ and destroy them by producing reactive oxygen species (ROS) such as superoxide and hydrogen peroxide. The amount of ROS produced is an indication of the intensity of the immune response and the general health of the organism. The ‘respiratory burst response’ is a method that has been developed to assess the impact of these pathogens on an organism’s immune system and constitutes an important method for the measurement of the immune health of an organism. For example, the toxic metal arsenic decreased the respiratory burst of macrophages in rat and rabbit lung tissue. Cytokines also play an important role in immune function. One group in the University of Maine research team, the Kim lab, has recently developed a respiratory burst assay for zebrafish. Respiratory burst assays have been used to quantify the effects of exposure to environmental toxins and metals, such as polychlorinated biphenyls, zinc, and copper. The team will measure cytokines, in conjunction with the respiratory burst assay, in early life stage zebrafish to determine if riverine sediment resuspension releases chemicals that affect immune health.

Another group in the University of Maine research team, the Mayer lab, has established two transgenic zebrafish lines for determining whether sediment resuspension releases heavy metals, such as cadmium and mercury, or chemicals that act like the female hormone estrogen. These transgenic lines were produced by inserting ‘transgenes’ into fish cells. These transgenes consist of gene promoters responsive to heavy metals or to estrogen-like substances coupled to a ‘reporter’ gene. In the presence of chemicals that act as metals or estrogens, the reporter gene is synthesised and emits light, signaling the presence of these chemicals. The research team has two transgenic lines, one that recognises both metals and estrogen-like chemicals, and one that recognises only metals. By exploiting the difference in chemical responsiveness of the two transgenic constructs, it can qualitatively differentiate the presence of bioactive heavy metals and bioactive xenoestrogens. These investigations will help ascertain whether metals and/or xenoestrogens (contaminants that act like estrogens) are present in the ‘desorption water’ and give insight into possible avenues of sediment remediation.

Importance of this study to resource managers

This study could provide critical information on the relative importance of sediment resuspension as a source of toxicants following dam removal, provide preliminary information on which chemicals are desorbed, and help managers decide whether river sediments should be remediated or removed prior to dam removal. In addition, if water toxicity cannot be attributed to the suite of priority pollutants measured in this study, it would indicate that further chemical characterisation of ‘desorption water’ is needed. Hopefully, it will provide the basis for larger, more in-depth chemical characterisations of contaminant desorption and toxicity studies using native species (Atlantic salmon, smallmouth bass ). This study could also provide a graduate student with real-world experience in the types of water quality issues we face as we strive to remediate degraded waterways to improve fish habitat.

Author Info:

The author is Adria Elskus, Associate Professor of Biological Sciences and the USGS Fishery Toxicologist for the Eastern Region.

The text is based on a proposal written jointly by Adria Elskus, Larry LeBlanc, a biogeochemist, Research Scientist in the School of Marine Sciences; Carol Kim, an Associate Professor in the Biochemistry, Microbiology and Molecular Biology Department; Rebecca Van Beneden, a Professor in the School of Marine Sciences and in the Biochemistry, Microbiology and Molecular Biology Department; and Gregory Mayer, an Assistant Professor in the Biochemistry, Microbiology and Molecular Biology Department. All authors are at the University of Maine, US.