Grantee Research Project Results
Final Report: Linking Watershed Characteristics with Flow Regime and Geomorphic Context to Diagnose Water Quality Impairment at Multiple Spatiotemporal Scales
EPA Grant Number: R831367Title: Linking Watershed Characteristics with Flow Regime and Geomorphic Context to Diagnose Water Quality Impairment at Multiple Spatiotemporal Scales
Investigators: Poff, N. LeRoy , Bledsoe, Brian P. , Dean, Denis
Institution: Colorado State University
EPA Project Officer: Packard, Benjamin H
Project Period: November 1, 2003 through October 31, 2007
Project Amount: $897,798
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:
There were 3 broad objectives for this research project. These represented a combination of GIS computer work, field surveys and field experiments.
Objective 1: Quantify interaction between watershed characteristics, flow regime components, and geomorphic settings that explain variation in sediment and biological indicators of water quality impairment in EMAP datasets.
Objective 2: Create a national classification of water quality impairment in terms of flow regime components and land use.
Objective 3: Regionally test flow regime-water quality linkages by conducting field studies in Rocky Mountain streams.
Summary/Accomplishments (Outputs/Outcomes):
Our first research objective was to develop models that could explain variation in benthic macroinvertebrate community composition across western EMAP sites in terms of multi-scale environmental metrics that define both the static habitat conditions of collection points and the dynamic variation at those sites. We developed tools to estimate 106 hydrologic metrics of ecological importance at each of >200 ungauged EMAP sites in a GIS environment using coverages of hydroclimatic (e.g. PRISM), geomorphic, and land use descriptors. Similarly, we developed and tested several GIS-based tools for predicting geomorphic features of stream sites and valley bottoms that provide multi-scaled descriptions of habitat for benthic macroinvertebrate samples. These included metrics describing climate, hydrology, geomorphology, geology, land use, riparian condition and local channel and substrate characteristics.
Macroinvertebrate data were acquired from R/EMAP data archives. Data were summarized in terms of community structure (e.g., diversity, proportion of EPT taxa) and “functional” organization (e.g., proportion of taxa sharing similar biological attributes such as thermal tolerance, disturbance tolerance, feeding strategy, etc.). We developed a series of models to relate the among-site variation in these biological metrics to a robust set of multi-scaled environmental variables across the western EMAP watersheds. The models used included ANN (Artificial Neural Networks), CART (Classification and Regression Tree) and MLR (Multiple Linear Regression).
We found that geographically independent stratification (e.g. snowmelt, snow/rain, rain, and variable flow regime types) within the confines of broad climatic regions is an appropriate strategy for modeling suites of streamflow metrics. Linking the hydrologic classification with GIS-derived geomorphic information allowed us to show that among-site variation in benthic macroinvertebrate data was better explained using the site-specific (and geographically-independent) hydrogeomorphic classification rather than a more commonly used ecoregional classification of sites.
Classification trees developed using both hydrologic and geomorphic metrics generally outperformed and were more consistent in attaining high classifications strengths than models using either type of metric alone. In general, this work provides a methodology for predicting ecologically relevant hydrogeomorphologic units from digital elevation models at the network scale. Mapping of network or regional scale distributions of stream physical habitat could benefit from refinements of this approach. These results also demonstrate that geospatial data can be used to create a geographically-independent classification of benthic macroinvertebrate community types that partitions biological variability better than ecoregions with the same number of classes. Further, predicting ecologically significant streamflow metrics at ungauged sites advances our ability incorporate these important variables into the design or stratification of biomonitoring sites for water-quality assessment. This research also demonstrates the possibility of combining readily predicted flow regime metrics and geomorphic attributes to explore and quantify linkages between specific aspects of the flow regime and aquatic community structure.
We also examined how well the derived hydrogeomorphic metrics for each site could explain biological variation across the western EMAP sites using regression techniques. First, we developed single-scale, multi-scale and hierarchical multiscale artificial neural network models relating EPT (Orders: Ephermeroptera, Plecoptera, Trichoptera) richness to 33 interpretable environmental variables describing habitat conditions at the watershed (n=15), valley (n=7) and reach scales (n=11). Models based on multiple spatial scales greatly outperformed single-scale analyses (mean R = 0.74 vs. R = 0.51). This result highlights the importance of incorporating environmental hierarchies to better understand and predict local patterns of macroinvertebrate assemblage structure in stream ecosystems.
In a companion analysis, we characterized macroinvertebrate species in terms of their biological attributes (traits) that are related directly to environmental variables; this affords a more “mechanistic” basis for prediction of how species change along environmental gradients. Using regression tree analysis, we found that environmental variables at all three spatial scales were important in explaining variation in proportion of taxa at a site having “disturbance-tolerant” traits (68% variation explained) or thermally tolerant traits (69% of variation). Models for “silt tolerance” were less successful (<50% variation). These analyses provide a more general approach to predict biological response to environmental alteration, and they suggest GIS-derived hydrogeomorphic variables can be used to assess where water quality impairment is likely across the landscape. This has broad implications for biomonitoring across broad geographic areas where differences in taxonomic identity of species limit spatial generalization.
Our second research objective was to develop some national-scale evaluation of hydrologic alteration of streams and rivers and diagnose the relative influences of human land use change versus dams on this alteration. Using a national database, we characterized the composition of land cover of 158 watersheds for small streams (contributing area < 282 km2) in terms of aggregate measures of agriculture, urbanization, and least disturbed (“natural”). We calculated hydrologic alteration using 10 ecologically-relevant hydrologic metrics that describe magnitude, frequency, and duration of flow for 158 watersheds within the Southeast (SE), Central (CE), Pacific Northwest (NW), and Southwest (SW) hydrologic regions of the United States. We found that regional context influences hydrologic response to particular kinds of human alteration. For example, in an analysis of flow alteration along gradients of increasing proportion of urban land cover, peak flows increased (SE and CE), minimum flows increased (CE) or decreased (NW), duration of near-bankfull flows declined (SE, NW) and flow variability increased (SE, CE, and NW). The effects of dams on flow alteration were more consistent across regions. These results focus underscore the importance of interpreting the effects of land use types in the context of region-specific histories, hydrologic processes, and channel sensitivities.
In a further, specific analysis of the effects of dams on flow regimes, we tested the hypothesis that dams have homogenized historically distinctive variation in natural flow regimes at large regional scales. Using national databases on daily discharge and dam coverage, we identified 186 long-term streamflow records on intermediate-sized rivers with long term pre- and post-dam records that had adequate pre- and post-dam flow records. We showed that dams have indeed homogenized the flow regimes on 3rd through 7th order rivers in 16 historically-distinctive hydrologic regions over the course of the 20th Century. Such homogenization was not found for unregulated, reference streams over the same period, even though precipitation intensity has changed over the 20th Century. These results have important implications for understanding the cumulative, regional-scale impacts of hydrologic alteration by dams and suggest that conservation strategies for native biodiversity will have to incorporate regional scale indicators of environmental change.
Our third research objective was to conduct field studies and field experiments to link biological measures of water quality impairment to water quantity. We conducted a field experiment in a single gravel-bed stream we installed 120 colonization boxes in 12 flow troughs containing five levels of fine sediments (0-14%, by mass) that span a range considered to be nonimpairing and impairing of biological water quality in field settings. We also exposed them to four levels of streamflow (10, 33, 67, 100% of ambient, in-channel flow) to capture a range of abstraction that occurs in human-altered streams. Macroinvertebrate community data were collected after 4 and 8 weeks and analyzed in relation to levels of flow and fine sediment. Model selection results showed greatest support for an additive effect between reduced flow and fine sediment deposition for several measures of biotic condition; however, there was also some evidence to suggest that these two stressors interact synergistically in some cases, with fine sediments causing a greater decline in biotic condition at higher flow velocities than at lower flow velocities. We took these experimental results and compared them to R/EMAP sites from the Southern Rockies Ecoregion of Colorado. A conditional probability approach based on the benthic macroinvertebrate index of EPT richness was employed to express the likelihood of a reduced number of sensitive taxa given an exceedance of reach-scale percent sand and fines (≤ 2 mm), and departure from context-specific expected levels of deposited fine sediments. Combined results from the ecoregion analysis and from our field experiment suggest that the most pronounced response of EPT taxa richness to the departure from expected surface coverage of sand and fines occurs at 10 to 30 percent in excess of predicted ambient conditions. In a natural experiment conducted as a survey across 14 Rocky Mountain streams in Colorado and Wyoming we found streamflow diversion was associated with large changes in riffle habitat but not pools. Generally, we observed a strong shift in taxonomic and trait composition at diversion levels exceeding 90%. In particular there was a strong decline in current-loving (rheophilic) taxa (mostly mayflies) and an increase in depositional types of taxa (chironomids). Filter-feeding taxa also were preferentially reduced as diversion intensity increased. This survey showed that the removal of base flow by diversion dams has significant negative effects on stream habitat and benthic macroinvertebrates. Overall, the experimental and survey projects show that flow and sediment interact and that more research is needed to advance our quantitative understanding of the interaction between low flow and sediment deposition into criteria for water quality impairment.
Journal Articles on this Report : 13 Displayed | Download in RIS Format
Other project views: | All 47 publications | 15 publications in selected types | All 13 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Bledsoe BP, Brown MC, Raff DA. GeoTools: a toolkit for fluvial system analysis. Journal of the American Water Resources Association 2007;43(3):757-772. |
R831367 (Final) |
Exit |
|
Craig LS, Palmer MA, Richardson DC, Filoso S, Bernhardt ES, Bledsoe BP, Doyle MW, Groffman PM, Hassett BA, Kaushal SS, Mayer PM, Smith SM, Wilcock PR. Stream restoration strategies for reducing river nitrogen loads. Frontiers in Ecology and the Environment 2008;6(10):529-538. |
R831367 (2006) R831367 (Final) X3832206 (2006) |
Exit Exit |
|
Flores AN, Bledsoe BP, Cuhaciyan CO, Wohl EE. Channel-reach morphology dependence on energy, scale, and hydroclimatic processes with implications for prediction using geospatial data. Water Resources Research 2006;42:W06412, doi:10.1029/2005WR004226. |
R831367 (2005) R831367 (2006) R831367 (Final) |
Exit |
|
Olden JD, Poff NL. Ecological processes driving biotic homogenization: testing a mechanistic model using fish faunas. Ecology 2004;85(7):1867-1875. |
R831367 (Final) |
Exit Exit |
|
Olden JD, Poff NL. Long-term trends of native and non-native fish faunas in the American Southwest. Animal Biodiversity and Conservation 2005;28(1):75-89. |
R831367 (Final) |
Exit Exit |
|
Olden JD, Poff NL, McKinney ML. Forecasting faunal and floral homogenization associated with human population geography in North America. Biological Conservation 2006;127(3):261-271. |
R831367 (Final) |
Exit Exit |
|
Olden JD, Poff NL, Bledsoe BP. Incorporating ecological knowledge into ecoinformatics: an example of modeling hierarchically structured aquatic communities with neural networks. Ecological Informatics 2006;1(1):33-42. |
R831367 (2005) R831367 (2006) R831367 (Final) |
Exit Exit |
|
Olden JD, Poff NL, Bestgen KR. Life-history strategies predict fish invasions and extirpations in the Colorado River Basin. Ecological Monographs 2006;76(1):25-40. |
R831367 (Final) |
Exit Exit |
|
Poff NL, Olden JD, Pepin DM, Bledsoe BP. Placing global stream flow variability in geographic and geomorphic contexts. River Research and Applications 2006;22(2):149-166. |
R831367 (2005) R831367 (2006) R831367 (Final) |
Exit Exit |
|
Poff NL, Bledsoe BP, Cuhaciyan CO. Hydrologic variation with land use across the contiguous United States: geomorphic and ecological consequences for stream ecosystems. Geomorphology 2006;79(3-4):264-285. |
R831367 (2005) R831367 (2006) R831367 (Final) |
Exit Exit |
|
Poff NL, Olden JD, Merritt DM, Pepin DM. Homogenization of regional river dynamics by dams and global biodiversity implications. Proceedings of the National Academy of Sciences 2007;104(14):5732-5737. |
R831367 (2006) R831367 (Final) |
Exit Exit Exit |
|
Sanborn SC, Bledsoe BP. Predicting streamflow regime metrics for ungauged streams in Colorado, Washington, and Oregon. Journal of Hydrology 2006;325(1-4):241-261. |
R831367 (2004) R831367 (2006) R831367 (Final) |
Exit Exit |
|
Wohl E, Angermeier PL, Bledsoe B, Kondolf GM, MacDonnell L, Merritt DM, Palmer MA, Poff NL, Tarboton D. River restoration. Water Resources Research 2005;41:W10301, doi:10.1029/2005WR003985. |
R831367 (2005) R831367 (2006) R831367 (Final) |
Exit |
Supplemental Keywords:
aquatic environment, ecology, watersheds, indicators, EPA Regions 8, 9, 10, channel-reach morphology, scale, hydroclimatology, flow regime, stratification, regression analysis, cluster analysis, discriminant analysis, CART, artificial neural networks, RFA, Scientific Discipline, Water, Hydrology, Water & Watershed, Environmental Monitoring, Ecology and Ecosystems, Watersheds, ecosystem modeling, watershed classification, hydrogeomorphic categories, geomorphic, GIS, water quality, spatial & temporal scalingRelevant Websites:
North American Benthological Society Meeting Abstracts
American Geophysical Union Meeting Abstracts
http://rydberg.biology.colostate.edu/~poff/
http://www.engr.colostate.edu/~bbledsoe/
Progress 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.