River Restoration From Boundary Layer to Watershed: Integrating Physical and Biological Science Over Space and Time VII Posters
Presiding: J Jack, University of Louisville; A C Parola, Stream Restoration Institute, University of Louisville
NB33E-01 1330h
Evaluation of Cross Vanes as a Stream Restoration Measure, Batavia Kill, Catskill Mountains, New York
A Natural Channel Design (NCD) stream restoration effort incorporates a number of components to improve stream stability, including cross vanes. Cross vanes are rock structures placed in the stream channel designed to maintain stable channel width and depth. An arm, constructed of boulders, extends upstream at bankfull from each bank. The arms are joined by a sill, constructed of boulders, perpendicular to the direction of flow. Design specifications state that the angle of the arm with the adjacent bank should be 20-30 degrees and the arm slope should be 3-7 degrees. Also, each arm and the sill should each occupy 1/3 of the width of the channel. Downstream from the vane is a scour pool. While the design and placement of cross vanes are based on hydraulic considerations, a secondary goal is that they will also improve aquatic habitat and likewise aquatic biodiversity. Even though cross vanes have design specifications, in practice it has been found that the structures fall along a spectrum. Additionally, it has been observed that the downstream scour pool dimensions vary, which may be an early indicator cross vane failure. As part of a stream restoration project twelve cross vanes were constructed along a 3400-foot reach of the Batavia Kill, located in the Catskill Mountains, in Greene County, New York. In order to conduct a post-construction assessment of cross vanes extensive physical surveys of each structure were conducted. Along with the survey data, hydrologic data for the site was obtained, bank erosion hazard data was recorded and macroinvertebrates were collected downstream from the cross vanes. Lastly, an evaluation of this data was conducted to better understand the relationship with scour pool depth.
NB33E-02 1330h
If we Build it Will They Come Back? Linking Physical to Biological Response in Stream Restoration Projects
Ideally, evaluation of habitat modification and restoration projects would include pre- and post-construction data. Given that many projects have not typically included any pre-construction information, and post-construction data is relatively sparse, can an assessment be made of the success of the project. In order to understand the link between the physical structure and biological response I borrowed published data from Index of Biotic Integrity studies. I test the correlation between species richness, key indicator species abundance, and overall community biomass with habitat variables, controlling for watershed position. If habitat condition can adequately predict the in-stream biotic community, this would greatly simplify the task of monitoring the success stream modification projects.
NB33E-03 1330h
The Impact of Restoration Efforts on the Geomorphology, Substrate, and Organic Matter Budgets of Three West Michigan Streams.
The structure and function of stream ecosystems is largely determined by the physical habitat of the stream. Anthropogenic addition of fine sediment from eroding stream banks and improper road stream crossings can have many negative consequences, including a reduction in substrate heterogeneity, masking of fish spawning gravels and reduction of habitat. Our objective is to examine the consequences of stream restoration efforts on the sediment composition of three streams. Specifically this includes an assessment of substrate before and after, and above and below restoration sites as well as changes in substrate composition from headwaters to mouth. Initial results from a subset of 10 restoration sites suggest that fine sediments are more ubiquitous downstream of erosion sources (p<0.05 for 16 mm, 500, 250 and 125 um for weight and/or volume; ANOVA). In addition, there was a trend for restoration efforts in the form of road stream crossing improvement, to increase sand substrate - likely as a result of construction. Although the initial increase in fine sediments downstream from the road stream crossing seems negative it is expected continued monitoring will show an improvement in substrate.
NB33E-04 1330h
Constructed Pools-and-Riffles: Application and Assessment in Illinois.
The diversity of Illinois' streams provides a broad range of conditions, and thus a variety of restoration techniques may be required to adequately compensate for watershed alterations. Resource management agencies and research institutions in the state have collaborated on a variety of applied research initiatives to assess the efficacy of various stream protection and restoration techniques. Constructed pool-and-riffle structures have received significant attention because they tend to address watershed processes (i.e., channel evolution model) and may benefit biotic communities and processes along with physical habitat. Constructed pools-and-riffles have been applied primarily to address geomorphic instability, yet understanding biological responses can provide further rationale for their use and design specifications. In three stream systems around the state, fish were collected pre- and post- installation of structures, using primarily electrofishing techniques (e.g., electric seine & backpack). In general, within the first five years after installation, changes in fish communities have included a shift from high-abundance, small cyprinid-dominated assemblages to low-density Centrarchidae and Catostomidae assemblages. Changes in macro invertebrates at selected sites included increases in filter feeders and sensitive taxa such as the Ephemeroptera, Plecoptera, and Trichoptera (EPT). Ongoing assessments will be critical for understanding long-term influences on stream ecosystem structure and function.
NB33E-05 1330h
Pre-restoration Evaluation of Obion Creek, Kentucky
The objectives of this project were to determine the baseline water quality of a channelized stream prior to restoration and to determine the habitat parameters that could be used as restoration goals. Macroinvertebrates, habitat, temperature, D.O., %Saturation, pH and specific conductance were sampled at four locations during spring from 2000 to 2003. Data analysis included between-site comparisons, measures of community diversity and similarity, and relationships between the macroinvertebrate, habitat and physicochemical data. Significant (p ≤ 0.1) between-site differences were observed in %EPT, %Clingers, EPT, mHBI, average tolerance value, bank stability, channel attributes (flow status, alteration, sinuosity), epifaunal substrate, pool variability, riparian zone width, sediment deposition, total habitat score, D.O., %Saturation, and specific conductance. Significant (p ≤ 0.1) correlations were observed in 42 of the metric/parameter combinations. The first 4 CCA axes accounted for 53.3% of the variation in the macroinvertebrate communities, with Axis 1 accounting for the majority (16.5%). With few exceptions, all sites had fair water quality, and diversity and evenness values indicated highly diverse and evenly distributed communities at all sites. Total habitat values at unchannelized sites were indicative of higher quality habitat. Historical land-use in the project watershed will have long-term detrimental effects, regardless of restoration activities.
NB33E-06 1330h
Management and Monitoring of Sediment Impacts From River Restoration Activities
Despite countless regulations, BMP's, and sedimentation/erosion control measures, sediment continues to significantly impact surface waters and biological communities. A Natural Channel Design project (Snow Creek, Stokes Co, NC) incorporated the first application of Streamside Systems' passive sediment collector technology to minimize downstream sediment impacts. Collectors below the construction reach selectively removed targeted particle sizes moving as bedload and filtered the output for upland disposal. Initial construction activity caused high transport of TSS and organic debris; only 2 yd3 of bed-migrating sand was removed in two months while the channel aggraded. Progressive stabilization and coarsening of substrates later increased sand transport and removal rates to 20 yds3/wk. Such technology is critical for assessing and controlling impacts of in-channel activities. Streamside Collectors can monitor transport rates of targeted sizes of bed sediments in addition to preventing their downstream habitat impacts. Independent Performance Testing by the Hydraulics Laboratory at Colorado State University (preliminary data) show removal efficiency as high as 90 percent. Collectors may also be used in series to increase removal efficiency, and for fine sediments such as iron floc (Koski 2004). Regulators and responsible contractors have a new tool for monitoring and restoring sediment-impacted habitats.
http://www.streamsidesystems.com
NB33E-07 1330h
Dominant Discharge Analysis of Ecological Processes in Streams
Here we apply the concept of "dominant discharge" from fluvial geomorphology to analyze the influence of flow regime on stream ecology. Quantitative metrics of effective discharge (Qeff) and functionally-equivalent discharge (Qfed) were developed to quantify how discharge drives organic matter transport, algal growth, nutrient retention, macroinvertebrate disturbance, and habitat availability. We quantify the magnitude of discharge at which a variable of interest is maximized over the range of discharges experienced by the stream (Qeff), and the single discharge that recreates the long-term average condition of the variable (Qfed) using a meta-analysis of published studies and modeling. Results suggest that a range of discharges are important for different ecological processes in an individual stream. All discharges are not equally important; instead, clusters of dominant discharge values exist that correspond to near modal flows and moderate floods for the 5 variable sets examined. We suggest four types of ecological response to discharge variability: discharge as a transport mechanism, a regulator of habitat, a process-modulator, and as a disturbance. Dominant discharge analysis will perform well when there is a unique, essentially instantaneous relationship between discharge and an ecological process, and poorly when effects of discharge are delayed or confounded by legacy effects.
NB33E-08 1330h
The Affects of Mountain Top Removal Mining on Headwater Streams in Eastern Kentucky
Mountain Top Removal/Valley Fill (MTR/VF) coal mining is a relatively new coal extraction technology that is widely utilized throughout the Appalachian region. During this process, the mountaintop is blasted away, the coal removed and the leftover material (spoil) is then deposited into the surrounding valleys. The potential negative ecological effects of these operations on stream biodiversity has received some attention but there is little available data on how these fills affect stream functions such as litter decomposition rates. We selected 4 streams draining "retired" MTR/VF sites of various ages in eastern Kentucky (USA) and one stream from an actively mined site. We compared leaf mass loss rates, N dynamics, fungal colonization (as measured by ergosterol) and water chemistry parameters in these streams to three unmined reference streams. Leaf litter mass loss was usually higher in the reference streams while water chemistry parameters such as conductivity, nitrate and TDS were often much higher in the MTR/VF streams. Such differences in stream function and water quality should be considered in permitting decisions and in assessing recovery of streams after mining.
NB33E-09 1330h
Fish Community Responses to Stream Restoration
Stream restoration projects are often justified based on expected improvements in habitat and ecosystem services, but few of these restorations have been systematically studied to assess their "success." A channelized section of Wilson Creek (Kentucky, USA) was relocated to a new, meandering channel using a natural channel design approach. Fish communities were sampled before and after the restoration and compared to an upstream site in Wilson and two control streams that were not restored. There were no consistent taxa changes among sites at Wilson Creek between the pre- and post restoration samples. Wilson Creek fish communities were always more diverse than either of the control streams. Kentucky Fish Index of Biotic Integrity (IBI) scores in Wilson were Excellent for the pre-restoration fish community and 4 out of 5 reaches sampled after the restoration retained that classification. The reference streams' IBIs were classified as Good and remained unchanged throughout the study period. We are also conducting a stable isotope analysis of representative trophic groups in Wilson to assess if there have been any changes in food web dynamics post- restoration. More pre- and post restoration studies are needed to help develop success criteria and incorporate "lessons learned" in stream restorations.
NB33E-10 1330h
Leaf Litter Decomposition as a Functional Assessment of a Natural Stream Channel Design Project
In October 2003, a 965m reach of Wilson Creek (Bernheim Research Forest, Kentucky, USA) was relocated, and meanders and riffle-pool sequences were restored, providing a unique opportunity to measure the re-establishment of post-restoration stream functions. Leaf litter bags were placed across riffles in the restored reach, in an upstream reference site and in two reference streams. Bags were collected for nine months, and mass loss, N dynamics and fungal ergosterol were measured. Daily mass loss rates in the restored and reference riffles in Wilson Creek were faster (k= -0.00759 and k= -0.00855, respectively) than those of the two reference streams (k= -0.00511 and k= -0.00308). This is equivalent to litter mean residence times of 132 days for the restored reach in Wilson, 117 days in the upstream reference site, and 196 and 325 days for the reference streams. It appears that the decay rate in the restored reach is similar to the upstream portion of Wilson Creek, indicating rapid mass loss recovery in the restored reach. We also determined that same-stream reference sites are important for evaluating the restoration of stream functions, because of high decay rate variation among nearby streams within the same watershed.
NB33E-11 1330h
Decomposition of Reintroduced Native Cane in a Restored Stream Channel
Stream restorations are often coupled with efforts to restore riparian zones. Native plants are often used in restored riparian areas but their effects on the streams (e.g. as a carbon subsidy) are not often assessed as part of a restoration evaluation. A channelized reach of Wilson Creek (Kentucky, USA) was relocated using natural channel design techniques and its new riparian zone was planted with a variety of native species including the river cane, Arundinaria gigantean, which was once a common riparian species in Kentucky. We installed decomposition packs containing stems or leaves of the cane at five locations in the restored section of Wilson Creek, one in an unrestored upstream site and at two locations in two unrestored reference streams to assess mass loss, N dynamics and fungal colonization. Initial results indicate that regardless of site leaf mass loss was faster in Wilson Creek than in the two reference streams (0.019% per day vs. 0.007%); stalk mass losses show similar trends to date. Research is continuing at Wilson to assess the performance of native vegetation at the site and to understand the role of riparian zone species in the success or failure of stream restorations.
NB33E-12 1330h
Speedy Recovery - Stream Macroinvertebrate Communities Show Extraordinary Recovery from Mining-Related Acidification
An area in northern Ontario, Wawa, was severely damaged by a century of iron mining and smelting with exceptional acidification (pH 3-4) and the accumulation of arsenic and other toxins. No formal restoration occurred following cessation of operations in 1998, but natural recovery began. In May 2004 we sampled the benthic macroinvertebrate communities of 20 stream riffles within and around the former fume kill area with the goal of estimating the state of community recovery. Despite watercourse colonization routes being blocked by waterfalls and the short time available for recovery, the macroinvertebrate communities showed remarkable recovery with both taxon richness and abundances being well within the range found in nearby reference streams belonging to the same watershed. Even relatively slow colonizers such as Pisidium bivalve mollusks and Orconectes crayfish, were found in the fume kill area streams. The biological recovery has been matched only by the chemical recovery of the systems. We attribute the rapid recovery firstly to the underlying calcium-rich geology, which apparently led to a quick decrease of acidity, thus facilitating re-colonization of the streams. Secondly, stream orientation relative to the acidic deposition zone facilitated rapid re-colonization from upstream areas.
NB33E-13 1330h
Impact of Stream and Floodplain Rehabilitation on Macroinvertebrate Community Structure and Diversity on the Hammer Creek in Lancaster County, PA
Naturally occurring floodplains act as a barrier to adverse effects from anthropogenic sources, while retaining aquatic organism diversity and potentially increasing stream productivity. The purpose of the study is to examine macroinvertebrate communities in response to stream and floodplain rehabilitation. This is an on-going study initiated with pre-restoration sampling conducted in July/August 2001. Post-rehabilitation sampling began in December 2001 and is continuing until the present. Long-term monitoring is being conducted among five sampling sites: above the restored area (control site), two sites within the restored section of the stream, and two sites 100 and 2500 meters below the impacted reach. Macroinvertebrates were sampled from the sites using a modified Hess sampler (n=6 replicates samples/ site). Macroinvertebrates were identified to generic level and analyzed using several metrics such as, Shannon and Simpson biodiversity indices, percent EPT, Functional feeding group analyses, ratio of scrapers to collector-filterers, and ratio of EPT abundance to Chironomidae. We found that stream restoration "traumatized" the macroinvertebrate community and diversity exhibited a lag-time in recovery. Because of an increase in riffle habitat, a modification of flow regime, and potential for preservation of habitat heterogeneity within these riffle zones, macroinvertebrate diversity may respond according to this improvement in habitat.
NB33E-14 1330h
Long-term Environmental Observatories as Predictors of Regional Environmental Change: An Oregon Example
Predictions of future environmental change often rest on long-term datasets collected in reference areas. However, many current landscape conversions occur on very different portions of the landscape than these sites. Can information from localized reference sites be used to predict regional environmental changes? We compared results from three nested watersheds: a small reference basin, the H.J. Andrews Experimental Forest (a Long-Term Ecological Research site, 64 km2), the McKenzie (3465 km2) and the Willamette (29727 km2) for past (1850, pre-Euroamerican settlement), present, and three alternate futures in 2050. Based on empirical field studies, we modeled changes to aquatic and riparian indicators, such as numbers and potential habitat of cutthroat trout, and riparian cover type. The relatively small, homogenous reference observatory was a poor predictor of more complex landscapes. For example, between 1850 and 1990, cutthroat trout habitat declined 50% in the Willamette, 19% in the McKenzie, but only 10% in the H.J. Andrews. Regional networks of observatories ideally should consist of (1) relatively pristine reference areas used to elucidate fundamental local ecological processes, (2) extensively altered sites, and (3) locations on the periphery of rapidly changing portions of the landscape where future intensive impacts, such as landscape conversion, are probable.