Ecological Effects of Small Dams
There are over 3000 dams in Pennsylvania, and over 100,000 throughout the United States. The great majority of these dams are small (<15 ft. height) and many of these small dams are obsolete and in poor repair. A growing number of small dams have been removed in Pennsylvania, Wisconsin, and many other states during the last two decades to reduce safety and liability risks (Bednarek 2002). Although safety and liability concerns are usually the primary factor influencing decisions to remove dams, some government agencies and environmental organizations have also proposed dam removal as a method of restoring fish passage and improving the health of stream and river ecosystems. This idea is derived in part from the extensive scientific literature documenting various effects of large (>30 ft. height) dams, along with the notion that dam removal might reverse some of these effects.
Large dams are known to impact river systems by altering several key parameters including: flow regimes and physical habitats, channel shape, sediment transport, water temperature and chemistry, and populations of algae, benthic macroinvertebrates, riparian vegetation, and resident and migratory fish (Poff and Hart 2002). The nature and magnitude of these effects are likely to depend, however, on dam size and other stream and watershed characteristics, so it is unclear whether the existing information on large dam effects is applicable to smaller dams. A better understanding of the effects of dams —particularly across a range of dam sizes— is needed to guide management decisions and maximize the effectiveness of river restoration projects.
With funding from the Pennsylvania Growing Greener Program and the Patrick Center’s Endowment for Innovative Research, scientists from the Patrick Center and the University of Delaware are studying the potential ecological effects of dam removal across a range of dam and stream/watershed characteristics using an approach based on ecological risk assessment. The study involves a regional assessment of the effects of different-sized dams on important components of the ecosystem including river channel structure, water quality, algae, benthic macroinvertebrates, fish, and streamside vegetation.
Ecological Risk Assesment
In ecological risk assessment, ecological effects are characterized by determining the potential effects imposed by a stressor, linking these effects to assessment endpoints, and evaluating how effects change with varying stressor levels (Hart et al. 2002). This basic framework can be used to evaluate the effects of dam removal by considering dams to be stressors, dam size (or a related measure accounting for stream/watershed characteristics) to be a measure of stressor level, and determining how the effects of different dams (varying across a range of stressor levels) affect stream condition. For any negative ecological impacts of a dam, the maximum potential benefit of a dam removal would be a return of the stream to pre-dam conditions.
In the language of ecological risk assessment, the relationship between stream condition (response) and dam size (stressor) is called a stressor-response relationship. When a reference curve is also shown, the maximum potential benefit of a dam removal can be illustrated as the difference between the stream response level and reference condition at any point along the stressor-response relationship. In our study, we are determining the effects of dams of varying sizes on stream condition by comparing a stream reach immediately below a dam with an undammed reference reach (either an upstream reach or an adjacent, undammed stream depending on the characteristic being evaluated). For each dam, the degree to which the dam affects stream condition can be measured as the difference between the dammed reach and the reference reach. This difference is then used to compare dams. Below is the hypothetical form of the stressor response relationship for dams, with larger dams hypothesized to have a greater impact on stream condition.
Application of this framework allows an assessment of the potential benefits of dam removal across a range of dam and stream/watershed conditions in the context of specific watershed management goals (e.g. cold water fisheries, nutrient/sediment reduction). This information would be very useful to help select and prioritize dam removal projects, and thereby maximize the effectiveness of dam removal in river restoration.
Dr. Rebecca Brown, Project Leader
Dr. James Pizzuto, Principal Investigator (University of Delaware)
Katie Skalak, Geologist (University of Delaware)
Dr. Thomas Johnson, Watershed Hydrologist
Jamie Carr, Aquatic Biologist
Dr. David Velinsky, Principal Investigator
Paul Kiry, Senior Chemist
Nate Saxe, Chemist
Quill Bickley, Research Assistant
Lara Jarusewic, Research Assistant
Dr. Donald F. Charles, Principal Investigator
Diane Winter, Algal Biologist
Erin Hagan, Algal Biologist
Mark Schadler, Co-ordinator
Lont Marr, Algal Biologist
Marina Potapova, Data Analysis
Tim Nightengale, Principal Investigator
Dr. Richard J. Horwitz, Principal Investigator
Amanda Kindt, Fisheries Biologist
Kevin O'Donnell, Fisheries Biologist
Paul Overbeck, Fisheries Biologist
Josh Collins, Fisheries Biologist
Dr. Heidi Hertler, Aquatic Biologist
Roger Thomas, Aquatic Biologist
Matt Wilhelm, Chemist
To predict how streams and watersheds are likely to respond to the potential removal of small dams, we believe that it is important to first understand the ecological effects of existing small dams. Accordingly, the central goal of our study is to quantify how dams of different size influence various physical, chemical, and biological properties of Pennsylvania streams. To gain an understanding of how dams of different size influence stream characteristics, we are studying how the ecological effects of dams vary along a continuum of dam characteristics, including dam height and storage volume.
Assessments of physical habitat, river channel structure, water quality, periphyton (attached algae), benthic macroinvertebrates, fish, and riparian vegetation have been made downstream of each dam and at an upstream reference location.
The quantity and quality of available stream habitat has a significant effect on biological communities. Dams can potentially alter flow and sediment transport, and thus can impact channel shape and physical habitat. Visual habitat assessments were made following EPA rapid assessment protocols, and include estimates of biological cover, substrate embeddedness, water velocity and depth, pool variability, sediment deposition, riffle frequency, channel modifications, and bank stability. Quantitative measurements of physical habitat also included channel dimensions and bed substrate size distributions.
Geomorphology and Sediment Characterization
Dams can cause significant changes in both the volume and size distribution of the sediment along a stream channel. Because dam removal can likely mobilize a substantial fraction of the sediment stored behind a dam, any risk assessment of the effects of dam removal must consider sediment-related impacts. To determine the effect of the dam on sediment size and distribution, pebble counts were conducted at both upstream and downstream reaches.
To determine the effect of the dam on channel morphology, surveys of each site were conducted using a Total Station surveyor. A mid-channel longitudinal profile was constructed along the entire length of both the upstream and downstream reaches to quantify potential differences in slope or pool/riffle structure, which may result from the presence of the dam. Three cross-sections of the channel were also surveyed at both the upstream and downstream reaches to determine the effect of the dam on channel width and depth. The size distribution and thickness of sediment fill behind a dam are being assessed at a select number of sites using a chirp sonar system, which can provide a high-resolution image of the sediment surface and subsurface material.
Water quality is an important aspect of stream ecological integrity. Samples were collected upstream and downstream of dams during low-flow periods to characterize dam-related water quality impacts. Water temperature, pH, dissolved oxygen and conductivity were measured at all study sites, including within the impoundment, using a YSI DM 6000 multiprobe meter. Water samples are being analyzed for total suspended matter, soluble reactive phosphorus (SRP), nitrite+nitrate (NO2+NO3- N), dissolved organic phosphorus (DOP), dissolved organic nitrogen (DON), total phosphorus (TP), total nitrogen (TN), dissolved organic carbon (DOC), and dissolved silicate. In addition, particles within each water sample are being analyzed for suspended chlorophyll a, particle size distribution, particle biochemistry, and bacteria cell concentration. Water quality data will be analyzed with respect to several potential effects of the dams including the type of dam release (surface or bottom) and residence time of the water within the impoundment. In addition, elemental carbon, nitrogen and phosphorus ratios will be used to assess within-impoundment transformations and the degree to which nutrient concentrations potentially limit primary productivity (i.e., the growth of algae, including diatoms).
Periphyton (attached algae)
Periphyton are the major primary producers in most streams, and are extremely useful in assessing their ecological condition. Because periphyton exhibit a wide range of sensitivities to stress and pollution, they are good diagnostic indicators of particular types of disturbance. Periphyton are sensitive to physical and chemical factors that are likely to be influenced by dams, including substrate type and stability, nutrients, and water temperature. Periphyton samples were collected from rock substrates below each dam and at upstream reference points and are being analyzed for measures of biomass (chlorophyll a and ash free dry mass) and species composition. In addition, the percent cover and thickness of algal growth were measured and samples of abundant filamentous algae present in the study reaches were collected for identification.
The use of benthic macroinvertebrates (aquatic insects, snails, mussels, etc.) to assess ecological conditions in streams and rivers is widely accepted among state and federal water resource agencies. Benthic macroinvertebrates include species with a diverse array of habitat preferences, feeding types, and pollution tolerances, thus providing considerable information concerning the nature and cumulative effects of stress. More specifically, benthic macroinvertebrate communities are sensitive to variations in sediment and flow characteristics, concentrations of dissolved oxygen, temperature, quantity and quality of organic matter, and other environmental factors potentially affected by dams. Benthic macroinvertebrate assemblages were sampled at each study site using U.S. EPA rapid bioassessment protocols. The organisms in these samples are being enumerated and identified in the laboratory.
Fishes are commonly used in assessments of the biological integrity of streams. Stream fish communities respond to numerous environmental attributes, including water quality, hydrology, and habitat structure. The effects of dams on migratory fishes and other recreationally important species are of particular interest. Because dams can affect fishes in the free-flowing sections above their impoundment (e.g., by migratory blockage), the upstream reaches are not appropriate reference sites for all aspects of fish community structure. Instead, reference sites on similar streams, some sampled for other studies in the region, are being used as reference sites. Fishes have been sampled below each dam and at comparable reference streams with backpack or tow-barge electroshocking gear. Standard metrics of community structure (e.g., species richness, richness of habitat-specific groups, relative abundance of different trophic groups) are being calculated to compare below-dam and reference sites. In addition, fishes are being sampled above each dam and compared with those recorded downstream to determine whether the dam is blocking fish migration.
Riparian (streamside) Vegetation
Riparian zones (the areas along rivers that are periodically flooded) contain highly diverse plant communities that are structured by flooding, which creates disturbance and acts as a dispersal mechanism for plants. Because dams potentially change the flood disturbance regime and block the downstream dispersal of plant propagules (i.e. seeds), they may cause a downstream change in native and exotic plant species diversity. Dams also drown the riparian zone within their impoundments and cause the riparian ecosystem to become fragmented. To assess the effects of dams on riparian plant communities, vegetation plots were established below each dam and at upstream reference points (unaffected by the dam) to compare native and exotic species diversity in the dammed sites and reference sites. In addition, vegetation plots were established adjacent to each impoundment to document the potential effects of inundation on riparian vegetation.
A total of 15 small dams in southeastern Pennsylvania and eastern Maryland were selected for study. They represent a spectrum of sizes. One dam approchaed 200 ft in height, while four were between 40-70 feet high. About half were less than ten feet high.
We are still analyzing our physical, chemical, and biological data, but one or our initial results is that the impoundments formed by dams have a significant effect on water chemistry. In the example shown below, the removal of silica appears to increase asymptotically with hydraulic residence time. Hydraulic residence time provides an estimate of the amount of time it takes for water to move through the impoundment; it is measured by dividing the impoundment’s water volume divided by the volumetric flow rate of water (or stream discharge) leaving the impoundment.
This relationship may be due to physical and chemical characteristics of the impoundment, which in turn create favorable conditions for the growth of certain types of algae, including diatoms. When diatoms grow, they incorporate silica into their cell walls, and when they die, the silica tends to sink with them to the bottom of the impoundment rather than flowing downstream. In general, bigger dams have longer hydraulic residence times, which results in an increased likelihood of silica deposition and a higher percent removal of silica.