Grantee Research Project Results
2005 Progress Report: A Watershed Classification System and Geomorphic Tool to Predict Habitat Variables in the Western Allegheny Plateau Ecoregion: Toward Refined Biocriteria and Stressor Identification of Impaired Streams
EPA Grant Number: R831365Title: A Watershed Classification System and Geomorphic Tool to Predict Habitat Variables in the Western Allegheny Plateau Ecoregion: Toward Refined Biocriteria and Stressor Identification of Impaired Streams
Investigators: Rankin, Edward , Stuart, Ben , Carlson, Bill , Verb, Bob , Harris, Breanna , Doyle, Brian , Bussee, Brianne , Kinney, Chad , Yoder, Chris , Meyer, Christine , Larkins, Clayton , Lopez, Dina , Hollingsworth, Emily , Springer, Gregory , Dyer, James , Zalack, Jason , Johnson, Kelli , Johnson, Kelly , Amaning, Kwarteng , Rose, Levi , King, Lisa , Stoertz, Mary , Clust, Mike , Vis-Chiasson, Morgan , Smucker, Nate , Morman, Tracy
Current Investigators: Rankin, Edward , Stuart, Ben , Carlson, Bill , Verb, Bob , Doyle, Brian , Kinney, Chad , Yoder, Chris , Meyer, Christine , Larkins, Clayton , Lopez, Dina , Hollingsworth, Emily , Springer, Gregory , Dyer, James , Renner, Jared , Zalack, Jason , Johnson, Kelli , Johnson, Kelly , Amaning, Kwarteng , King, Lisa , Stoertz, Mary , Clust, Mike , Vis-Chiasson, Morgan , Smucker, Nate , North, Sheila , Morman, Tracy
Institution: Ohio University - Main Campus
EPA Project Officer: Packard, Benjamin H
Project Period: October 1, 2005 through September 30, 2007 (Extended to September 30, 2009)
Project Period Covered by this Report: August 1, 2004 through July 31,2005
Project Amount: $869,440
RFA: Development of Watershed Classification Systems for Diagnosis of Biological Impairment in Watersheds and Their Receiving Water Bodies (2003) RFA Text | Recipients Lists
Research Category: Watersheds , Water
Objective:
The ultimate goal of this research is the development of a watershed classification system for gauging stream health in terms of refined biocriteria, stressor diagnosis, and geomorphology. Specifically, we are working to develop a refined watershed classification system that incorporates variations in physical and chemical conditions as they influence biological assemblages at wadeable reference streams of the Western Appalachian Plateau (WAP) ecoregion. We are examining factors that include geographically dependent and geographically independent variables (e.g., Level IV ecoregion and gradient). If this refined classification is biologically important we will develop refined biological criteria from these reference sites. More refined goals (biocriteria) will lead to more accurate identification of impaired waters and will enhance our ability to diagnose which stressors are responsible for impairment. We are also working to test a model to predict habitat quality from geomorphic attributes.
Once the classification and refined biocriteria are complete, we will work to identify major stressors and their threshold levels using post hoc analysis. Our plan is to demonstrate the classification system, refined biocriteria, and stressor diagnosis methodology on a monitored watershed and a non-monitored watershed in the WAP ecoregion by comparing expected conditions to those identified as impaired. We will also be working with Ohio EPA and Ohio DNR to integrate the applications resulting from this research into existing water resource management programs in the WAP ecoregion and disseminate information about them outside the WAP.
Progress Summary:
Work Status: This is a relatively large project involving multiple co-investigators across a wide range of disciplines, so it required a substantial planning/coordination effort. It also requires data sharing across these disciplines, so we have spent a substantial effort to design databases that will be: 1) geographically based, 2) compatible and complementary to Ohio EPA’s E3 database and the U.S. EPA STORET database, 3) Web-based, and 4) ultimately useful to the investigators on this project and the intended target groups and agencies including Ohio EPA, Ohio DNR, and watershed groups. We also developed a web site (http://ohioair.info/star/ Exit ) and an ftp site to help distribute information, manage the project and facilitate communication, especially during the field portions of this project with teams from each of the major disciplines. The goals and aims of this project have not changed from the initial proposal.
Our workplan was essentially divided into three phases. Phase 1 is the major classification portion of the project where our goals are to refine the existing Ohio EPA major ecoregion classification system for biocriteria. Phase II is the geomorphic modeling of potential habitat within stream of the WAP ecoregion and Phase III is the examination and diagnosis of stressors along human disturbance gradients in the WAP Ecoregion. The activities of each of these phases are overlapping.
The main focus of the Year-1 phase of this project was to begin development of the new data system to share data across disciplines and to compile reference data within existing data systems to begin the classification phase of the project. The first analytical phase of the project relied primarily on existing data which we obtained from Ohio EPA. The initial phase also included extensive GIS work to obtain land use/land cover data for the refined classification analyses. The GIS work took somewhat longer than expected because of extensive verifications needed for geographic locations assigned to historical data. Much of this location data was collected “pre-GPS” and we needed to perform more verification of locations than expected. We have explore some of the initial classification analyses (cluster analyses) and hope to finish more in depth analyses (canonical correspondence) for fish and macroinvertebrate data now that the GIS layers have been completed for the reference sites.
Reference Sites
We used the existing Ohio EPA least impacted reference sites that form the basis of Ohio’s existing biocriteria as the reference sites to refine the classification, but needed to expand this dataset to accommodate the additional classification variables we want to explore (e.g., level 4 ecoregions, stream gradient, etc). To do this we followed the Ohio EPA definition of “least impacted” which was based on examining adjacent land uses, absence of significant upstream dischargers, and relatively intact instream and riparian habitats. This definition is not considered pristine, but an approximation of the best available given human uses of the landscape in Southeast Ohio. We completed selection of additional reference sites and have a database of approximately 200 sites (some with multiple years of data). This database was randomly divided into an exploratory and a verification database to allow cross-validation of results. Randomization followed several important stratifications including drainage size and habitat that may have an important influence on classifications. Distribution of sites across the Region is illustrated in Figure 1.
Field Season 1 – 2005
We completed the first field season successfully during summer 2005. All of the biological data and most of the other data has been proofed and completed and has been available for investigators and students on the ILGARD ftp site. Each of the Co-Investigators successfully completed data collection and entry and data is currently being analyzed. The biological data was collected via a subcontract with MBI in collaboration with Dr. Kelly Johnson and her students and consisted of fish and macroinvertebrate community data, plus habitat data (QHEI). This was collected using the standard and well-tested Ohio EPA methods used to develop Ohio’s existing biocriteria. Additional data was collected to test another index (MAIS) that has been used in similar areas in the Appalachians. Dr. Vis-Chiasson and her students collected algal community data and are currently comparing the PIBI (Periphyton IBI) with new algal indices. Dr. Stoertz and her students collected hydrologic measures at each site as well as extracted other data from USGS stations and extracted modeled flow data based on USGS methods (see below). Dr. Springer and his students performed a geomorphic assessment at each site. This was done to provide geomorphology data not available with the historic datasets.
Although the end product will be the integration of all of the abiotic data to explore refined classification for the biota at reference sites, graduate students are currently pursuing exploration of constituent abiotic variables independently to provide the best set of abiotic variables for the canonical correspondence analyses we will conduct later. The preliminary results of these analyses have been presented at various scientific conferences (see below).
The 2005 sampling season was rated as abnormally dry to a moderate drought in southern Ohio (http://www.drought.unl.edu Exit ) which influenced some of our field data collections. Although fish collection data was similar to historic data (See Attachment 1), macroinvertebrate collections (Hester-Dendy artificial plates) are more sensitive to low flow and invertebrate index scores were influenced by this condition; at some sites only a qualitative sample was possible. Periodic droughts are expected, however, and extreme conditions actually provide opportunities to examine certain stressors (e.g., hydrology) when they are most severe.
Budget: Funds for the first year of this grant funding principal investigators, students and field work as stated in the grant budget. During the period covered in this annual report period, $246,140.58 was received, and $249,078.29 was disbursed. Summaries are attached.
Quality Assurance Plan: All data collection followed our QA/QC plan we developed at the beginning of the study. Each Co-Investigator was responsible for the day-to-day QA/QC tracking and we utilized a message board on our Web site to facilitate communication between investigators relative to QA/QC and other issues that could affect data quality.
Results: We hope to have the first phase of the project totally complete by early summer which we eventually be developing into papers to be submitted to the primary ecological literature relative to each discipline. Preliminary analyses of geomorphology data, hydrology data, and water chemistry data by graduate students did not find strong patterns with reference assemblages across the WAP ecoregion. Further analyses are proceeding; however, the initial work suggests that these classification variables within the WAP ecoregion are not exerting a strong influence on variation in reference assemblages. In later phases of this project we will be examining these variables as they contribute to variation in response to anthropogenic stress in this ecoregion.
Future Activities:
Plans for 2006: We will be into the second and third phases of this project during 2006. We will be examining historic data along a gradient of human disturbance to identify the predominant stressors in the WAP ecoregion. Biological, water chemistry, habitat and other data has already been compiled for sites across the region. GIS layers are being compiled for a randomly selected subset of sites across the ecoregion. These sites will also be divided into an exploratory and verification dataset.
We will be identifying the predominant stressors across the region using the Water Environment Research Foundation’s (WERF) Integrated Impact Analysis (IIA) methodology to discriminate among the most important stressors within each of the classification strata we have created for the WAP ecoregion with the exploratory. The technique uses a number of statistical methods—Chi-square Automatic Interaction Detection (CHAID), All-Possible Regressions, and Principal Components Analysis PCA)—to isolate the stressors likely to contribute the most to the observed biological impairment. The strength of this approach is that it examines many (sometimes obscure) potential associations or possibilities for impairment. Then, rather than relying on statistical inference, IIA employs a formal validation process on held-back data (verification data set). This dataset will focus on biological condition indices (IBI, ICI) and several major metrics (number of sensitive taxa) as response variables as well as on the suite of abiotic measures mentioned above, including chemistry (water column and sediment), habitat, LULC measures, and others. The LULC metrics will be important because they will include some estimate of the scale of anthropogenic changes as will some summarized measures.
We are also planning our second season of field work. The objective of this data collection effort will be to apply the results of our refined classification and stressor identification efforts in an under-sampled watershed (likely Shade River watershed) and to examine data along a strong identified stressor gradient to allow more testing and verification of our geomorphic tools and field verification of our stressor identification efforts. We have begun discussions with Ohio EPA and Ohio DNR related to how these results can be integrated into their ongoing TMDL and AMDAT programs. A later effort will be workshops with these agencies and watershed groups related to integration of this data and methods into active management efforts.
Appendix 1. Comparison of trends in IBI data (top) and QHEI data (middle) between historic and 2005 sampling periods at a subset of reference sites in the WAP ecoregion used in this project. The bottom graph illustrates regression curves of QHEI vs. IBI for historic data (blue) and 2005 data (red). Trends in regressions are similar. R-square values are low because of the relative small variation in habitat at reference sites; addition of habitat stressed sites generally results in much stronger relationships.
Journal Articles:
No journal articles submitted with this report: View all 20 publications for this projectSupplemental Keywords:
RFA, Scientific Discipline, Water, ECOSYSTEMS, Ecosystem Protection/Environmental Exposure & Risk, Water & Watershed, Monitoring/Modeling, Terrestrial Ecosystems, Environmental Monitoring, Ecology and Ecosystems, Watersheds, ecosystem modeling, watershed classification, wetlands, aquatic ecosystem, monitoring, biodiversity, natural resource management, watershed, conservation, geomorphic tool, integrated watershed model, aquatic ecosystems, ecosystem restoration, ecological indicators, ecology assessment models, watershed assessment, ecological models, predicting habitat variables, biological impairmentProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.