Grantee Research Project Results
2002 Progress Report: An Empirical Evaluation of the Performance of Different Approaches to Classifying Reference Conditions in Streams
EPA Grant Number: R828637Title: An Empirical Evaluation of the Performance of Different Approaches to Classifying Reference Conditions in Streams
Investigators: Hawkins, Charles P. , Stevenson, R. Jan , Roberts, David W.
Institution: Utah State University , Michigan State University
EPA Project Officer: Packard, Benjamin H
Project Period: January 1, 2001 through December 31, 2003 (Extended to December 31, 2005)
Project Period Covered by this Report: January 1, 2002 through December 31, 2003
Project Amount: $1,499,485
RFA: Development of National Aquatic Ecosystem Classifications and Reference Conditions (2000) RFA Text | Recipients Lists
Research Category: Water , Aquatic Ecosystems
Objective:
Assessments of biological integrity usually are based on comparisons with "natural habitats of a region." In theory, measurements made at these natural habitats, or reference sites, serve as a control and allow us to specify the biological conditions that would be expected at sites being assessed (test sites) if no human disturbance had occurred. However, because natural aquatic habitats within a region markedly can differ in their biological attributes-often because of local variation in regulating factors-predicting the biological conditions that should occur at a test site can be difficult.
The objective of our research project is to establish a means of conducting more sensitive biological assessments of stream ecosystems by improving both accuracy and precision in classifying aquatic ecosystems and thus predicting the expected biological conditions of natural habitats. In our research, we are focusing on evaluating classification systems applicable to large, topographically heterogeneous regions such as the Western United States. Our specific objectives are to determine: (1) how the sensitivity of assessments is affected by the approach used to classify ecosystems (a priori regionalization versus predictive statistical model), the autecological characteristics of organisms, and the spatial scale of classification; and (2) if approaches to site classification are transferable among regions.
To evaluate our objectives, we are using data collected from a spatially extensive set of streams in California, Oregon, and Washington to address objective 1 (effects of classifications). We will evaluate objective 2 (transferability) by determining how well methods identified in objective 1 perform in different biogeographical regions: streams of the Intermountain West, Rocky Mountains, and Plains. To examine objective 1, we are exploiting pre-existing data that we have collected in other projects. To examine objective 2, we collected data from additional sites in the interior Western United States.
Progress Summary:
Most of the work conducted during the second year of the project involved processing the invertebrate, algae, and chemistry samples, which were collected at 324 sites during the first year of the study, and integrating data from 464 additional sites into a common database of 788 sites that will be used for most analyses. We also have entered and summarized much of the environmental data that will be used to develop and test classifications, and we are in the process of compiling and integrating data collected by other investigators whose sampling procedures were compatible with ours.
Data Collection
Landscape-Level Environmental Data. We have obtained Digital Elevation Model (DEM) coverages for the entire Western United States, as well as climatic data. We currently are in the process of defining watershed boundaries for each of the 788 sample locations, after which we will extract several watershed attributes needed to classfy sites by different criteria (geology, topography, etc). We have not yet obtained all of the geology coverages we need. That task currently is being addressed. One particularly difficult challenge with geology coverages is that coverages largely have not been standardized between states and the level of resolution in geology classification most relevant to stream biota has not yet been identified. We are deriving several different alternative geology classifications that will be tested for effectiveness in discriminating biota among sites.
Site-Scale Environmental Data. All of the site-scale environmental data, with the exception of water chemistry (see below), have been compiled and summarized for the 324 sites we sampled. We also are in the process of completing the compiling and integrating of site-scale environmental data for the 464 non-Science To Achieve Results (STAR) sites with the data previously entered for the 324 sites that we sampled last year.
Water Chemistry Data. All chemistry samples from 2001 have been analyzed for NO3-, NH3, TN, TP, SRP, Cl-, and Si4-. Major and minor ion analysis by Inductively Coupled Plasma Mass Spectrometry (ICPMS) has been completed for the majority of chemistry samples. ICPMS analyses as well as SO42- analyses for all chemistry samples (2001 and 2002) are expected to be completed within the next month.
Algae Data. Subsampling and processing of all periphyton samples is complete. Permanent slides have been prepared for all diatom samples with the exception of approximately 50 samples collected in 2002. Diatom taxonomy training for the graduate student on the project is complete, and diatoms have been counted for approximately one-half of the 2001 samples. Thousands of images have been taken to photo document diatom taxa and to aid in taxonomic synchronization with other large-scale surveys in the region (National Water Quality Assessment Program [NAWQA] and the Western Environmental Monitoring and Assessment Program [WEMAP]). Counts are scheduled for completion by the end of summer.
Invertebrate Data. All of the invertebrate samples have been processed, identified (with the exception of chironomid midges), and conducted through quality control assessments. Identification of chironomid midges, which the panel and Program Officer requested, is about 30 percent complete. We also have integrated our data with taxa lists generated from 464 other reference-quality stream sites from different parts of the Western United States. This required that we decide on the level of taxonomic resolution to be used for each taxon in the combined data sets. That process resulted in 276 operational taxonomic units (OTUs); 2 percent were based on species level identifications; 64 percent on genus level identifications; 30 percent on family or subfamily level identifications; and 4 percent on levels of resolution coarser than family. The discrepancy in the level of taxonomic resolution used among taxa is the result of the fact that many taxa groups cannot consistently or confidently be identified to lower levels of resolution. Furthermore, individuals within the same taxon cannot always be identified to a consistent level (e.g., some individuals might be identified to species, others only to genus, others only to family, and still others to even higher taxonomic groups). In such cases, we decided on which level of resolution to use by enumerating what percent of individuals within a taxon were identified to each level and then assigning the OTU as the level of resolution most frequently determined. Individuals identified to coarser levels of resolution than the OTU were ambiguous and were therefore eliminated from analyses. Individuals identified to finer levels of resolution than the OTU were aggregated into the OTU.
Results to Date
Algal-Based Classifications. With the data currently available, several analyses have been conducted, and results have been presented at local and national scientific meetings. Predictive models of algal biomass were presented at the North American Diatom Symposium and the Annual Meeting of the North American Benthological Society in 2001 and 2002, respectively. A project description and preliminary data were presented at a land use poster forum at Michigan State University last year. Along with data from several Environmental Monitoring and Assessment Program (EMAP) datasets, responses of individual diatom taxa along environmental gradients are being examined to determine whether diatom species' responses conform with theoretical predictions. Results of these analyses, along with implications and alternatives for bioassessment, will be presented in May at the 2003 Annual Meeting of the North American Benthological Society.
Invertebrate-Based Classifications. Work this year has focused on two issues: (1) continuing work we started last year to quantify data quality and its effect on aspects of precision and accuracy; and (2) starting initial analyses of classifications based on biotic similarities among samples. Our work on data quality has resulted in several presentations and two manuscripts, one of which is in press. Our approach to quantifying data quality focuses on developing measures of data quality that are relevant to multispecies assemblages, rather than applying more commonly used measures of data quality based on single variables, but which are not appropriate for multivariate data.
Our initial analyses of classifications were based on deriving biologically based classifications from the biotic similarities that occurred among the 788 sites in the reference site database. These analyses revealed that sites assigned to the same biotic class tended to be distributed widely across the Western United States. Although we only have examined relationships between class membership and climatic factors at this time, climatic appears to be an important variable influencing biotic structure at a site. We found that class membership was associated with five climatic measures: (1) mean annual precipitation; (2) number of wet days; (3) mean annual relative humidity; (4) mean annual air temperature; and (5) number of frost-free days.
Future Activities:
We will conduct the majority of data analyses during the upcoming project period. Our main activity will be associated with completing extraction of landscape information (watershed boundaries and geology coverages) and refining a priori geology classifications. We also will complete the development and testing of predictive models for both invertebrate and diatom assemblages.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
Other project views: | All 14 publications | 3 publications in selected types | All 3 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Hawkins C, Cao Y, Roper B. Method of predicting reference condition biota affects the performance and interpretation of ecological indices. FRESHWATER BIOLOGY 2010;55(5):1066-1085. |
R828637 (2002) |
Exit |
|
Mustonen K, Mykra H, Sarremejane R, Veihalainen N, Sippel K, Mutoka T, Hawkins C. Thermal and hydrologic responses to climate change predict marked alterations in boreal stream invertebrate assemblages. GLOBAL CHANGE BIOLOGY 2018;24(6):2434-2446. |
R828637 (2002) |
Exit Exit |
Supplemental Keywords:
EPA Regions 8-10, California, CA, Oregon, OR, Washington, WA, Utah, UT, Nevada, NV, Idaho, ID, Wyoming, WY, Montana, MT, Colorado, CO, Arizona, AZ, New Mexico, NM, classification, indicators, regionalization, bioassessment, Environmental Monitoring and Assessment Program, EMAP., RFA, Scientific Discipline, Geographic Area, Ecosystem Protection/Environmental Exposure & Risk, Aquatic Ecosystems & Estuarine Research, Hydrology, Ecosystem/Assessment/Indicators, Ecosystem Protection, State, Aquatic Ecosystem, Ecology and Ecosystems, Ecological Indicators, EMAP, ecological exposure, bioassessment, streams, classifying reference conditions, Oregon, ecology, classification, New Mexico (NM), stream ecosystems, anthropogenic impact, empirical evaluation, Washington (WA), modeling, environmental indicators, aquatic ecosystems, water quality, biological indicators, ecological classification, Environmental Monitoring and Assessment Program, spatial data bases, California (CA), Montana , ORProgress 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.