Annotated Bibliography

A collection of research and tools for using GIS in river restoration decision making.

GIS can be a fast, objective and cost-effective tool for professionals in the field of river restoration.  Applications of GIS include identifying and prioritizing conservation and restoration sites, quantifying potential impacts of restoration activities, establishing baseline conditions of sites, conducting site monitoring and project evaluation and finally, sharing this information with collaborators.  What follows is an annotated bibliography of research and tools that have used GIS to facilitate river restoration.  While these tools were developed and applied to specific watersheds, they can be modified for use in any watershed, incorporating local ecological and socio-economic conditions.

Clarke, Sharon E., Kelly M. Burnett, and Daniel J. Miller, 2008. Modeling Streams and Hydrogeomorphic Attributes in Oregon From Digital and Field Data. Journal of the American Water Resources Association (JAWRA) 44(2):459-477.

This article addressed the challenges of creating detailed stream maps covering large areas. Field surveys produce detailed data yet are costly so typically cover only small areas.  Regional digital data available through GIS covers large areas yet is often not accurate on a fine scale.  Authors combined data from field inventories and regional digital data in a model to create stream maps of the Oregon Coastal Province. Attributes developed by the model included watershed size, precipitation, flow, likelihood of perennial flow, channel gradient, channel width and depth and floodplain boundaries.   Model results of these attributes were cross-checked with field data and most indicated a high level of correlation. This model can be repeated in other watersheds and might be used to map riparian areas, identify headwater streams and determine sites suitable for salmonid habitat restoration.

Map comparing distributions of coho salmon based on modeled channel gradient and as estimated by ODFW

Feist, B., E. Steel, D. Jensen and D. Sather.  2010. Does the Scale of our Observational Window Affect our Conclusions about Correlations Between Endangered Salmon Populations and Their Habitat? Landscape Ecology 25: 727-743.

Scale can affect the strength of apparent relationships between animals and habitat conditions.  In stream ecosystems, impacts on habitat in one reach of a stream might be the result of conditions in a different part of the watershed.  The importance of scale is especially true for anadromous salmon whose life cycle covers broad geographic areas.  To analyze the influence of scale on salmon habitat studies, authors examined relationships between Pacific salmon (Oncorhynchus spp.) and landscape data from GIS for three spatial scales:  local (stream reach), intermediate and basin-wide. Authors found that correlation between landscape condition and salmon response is influenced by scale.  Results however, suggest that there is no ideal scale for which to examine the significance of habitat attributes.  Authors therefore recommend that river restoration projects should consider landscape conditions at multiple scales. Reach, riparian area and greater basin land use/land cover, all impact salmon use of a site.  To maximize return on restoration projects, resource managers can use GIS to examine the influence of each of these 3 scales on salmon habitat use.

Fig. 2 Box and whisker plots comparing AIC for various fixed- and mixed-extent models as a function of extent, by subbasin and species (steelhead and Chinook)

Fullerton, A., T. Beechie, S. Baker, J. Hall and K. Barnas.  2006. Regional Patterns of Riparian Characteristics in the Interior Columbia River Basin, Northwestern USA: Applications for Restoration Planning. Landscape Ecology 21: 1347-1360.

Functioning riparian areas are an integral component of salmon habitat.  Authors used GIS to identify stream reaches in the interior Columbia River basin that are accessible to salmon and are lacking in native riparian cover.  They then assessed riparian areas using remotely sensed data, aerial photographs and some field observations.  Based on these assessments authors identified riparian areas that were inadequate, prioritizing these sites for salmon habitat restoration.  The majority of these sites occurred in low-gradient streams, with floodplains, in semi-arid ecoregions.  Authors also found that the number of restoration projects varied greatly from one sub-basin to the next and that restoration sites were often determined by opportunity (landownership, funding)  rather than ecological need.  This technique of using GIS to identify restoration opportunities provides a cost-effective means for resource managers to consider ecological needs while still working within the confines of social constraints.  Authors recognize the restrictions of identifying restoration sites on a coarse scale and recommend local field surveys to further identify restoration priorities.

Estimated % of riparian areas >22.8 m wide for streams accessible to anadromous salmonids in the interior Columbia River basin.

Gillenwater, D., T. Granata and U. Zika.  2006.  GIS-based Modeling of Spawning Habitat Suitability for Walleye in the Sandusky River, Ohio, and Implications for Dam Removal and River Restoration.  Ecological Engineering 28:311-323.

GIS was used to assess the potential for Walleye to spawn in areas upstream of a dam that was slated for removal.  Authors combined a hydraulic river model with a GIS to calculate depth, velocity and spawning habitat suitability in a river reach. Model accuracy was then checked with field measurements.  While the model accurately predicted depths, velocity predictions were mostly lower than those measured in the field.  However, egg densities measured in the stream were positively correlated with the habitat suitability index calculated by the model (R² = 0.19, P=0.036), suggesting that the model may predict approximate spawning habitat in the river.  Study findings indicate that habitat suitability is heavily correlated with discharge.  Discharges of 20-25m³/s are most appropriate for suitable habitat.  This type of model can be useful in weighing the ecological impacts of dam removal.  Authors note that accuracy of models is particularly essential in highly contentious restoration projects such as the removal of Sandusky River dam.

Habitat Suitability Index maps of the study area (shown in the inset photograph) produced by the model for a range of discharges.

Hulse, D., S. Gregory, J. Baker. (Eds). 2002. Willamette River Basin Planning Atlas: Trajectories of environmental and ecological change. (2nd edition), Oregon State University Press, Corvallis, Oregon 97333. p. 180.

The Willamette River Basin Planning Atlas uses GIS to describe physical, biological and human aspects of the Willamette River basin.  It is predicted that the population of the Willamette Valley will increase by 1.7 million people by the year 2050.  Authors predict changes to the Willamette Basin which might occur as a result of this population increase.  Changes in land use and land cover are depicted under 3 different future scenarios: a conservation scenario, a development scenario and a scenario which reflects no change in current land and water use policies.  Using GIS, these 3 scenarios are depicted spatially throughout the Willamette River basin.  Finally, considering the human pressures that will ensue with population growth, authors identify suitable sites for restoration along the Willamette River.  Authors recommend that restoration options which have maximal ecological benefit and minimal social constraints be pursued.  These sites can be identified using social and ecological GIS data represented over time.

Willamette River basin trajectories of change

Institute for a Sustainable Environment. SLICES Introduction.   Accessed November 26, 2011: http://ise.uoregon.edu/slices/Main.html.

The Slices Framework is an open source tool for those interested in restoration and conservation within the Willamette River floodplain.  Slices represents the Willamette River and adjacent areas as a series of 1km slices that lie perpendicular to the Willamette floodplain.  While the river channel, floodplain vegetation and land use changes throughout time, the floodplain remains relatively constant.  By using the floodplain as a reference point, Slices allows us to track ecological changes to a single location over time.  Current scientific studies in the Willamette basin gather data on channel complexity, floodplain forests, cold water refuges, fish assemblages, flood levels and flood storage.  This data will is stored in a GIS and layered over the Slices Framework.  Currently the Slices Framework consists of pre-made interactive maps, attribute tables and GIS files. These resources provide access to data useful in identifying potential restoration sites and predicting restoration impacts through time.

Willamette River floodplain slices

Lunetta, R., B. Cosentino, D. Montgomery, E. Beamer and T. Beechie. 1997. GIS-Based Evaluation of Salmon Habitat in the Pacific Northwest. Photogrammetric Engineering and Remote Sensing 63: 1219–1229.

Using GIS, authors categorized streams in western Washington according to their potential as salmon habitat. Channel slope and forest seral stage were used to roughly estimate stream condition.  High quality salmon habitat was defined as being low gradient and existing in later seral stage forests.  Older forests are a source for large wood which creates pool-riffle habitat used by salmon.  Using 30-meter DEM data, low gradient reaches (<4% slope) were accurately identified (96% accuracy). About 23% of reaches had slopes less than 4%.  Of these 8.7% were in late-seral and 20.7% in mid-seral stage forests.  Authors found GIS analyses to be less accurate in predicting channel morphology.  However, overall findings from the GIS-based analyses were similar to those from field based assessment.  For example, both the GIS analysis and field research found that the greatest habitat losses have occurred in the Skagit River floodplain and delta. GIS was determined to be a quick, objective and cost-effective mechanism of identifying priority sites for salmon habitat restoration.

Mollot, L.,and R. Bilby. 2008. The Use of Geographic Information Systems, Remote Sensing, and Suitability Modeling to Identify Conifer Restoration Sites with High Biological Potential for Anadromous Fish at the Cedar River Municipal Watershed in Western Washington, U.S.A. Restoration Ecology 16: 336-347.

The presence of large wood in streams enhances habitat for many types of salmon.  In most watersheds in the Pacific Northwest, conifer inputs in streams have decreased due to land use conversions.  Consequently, planting conifers in riparian areas is a common restoration practice in the Pacific Northwest.  Conifer restoration sites should be located in areas beneficial to salmonids that can also support conifer trees for long periods of time.  Authors created a GIS model, incorporating remotely sensed data and spatial analysis tools, to identify appropriate conifer restoration sites. Site suitability was weighted according to three criteria:  stream gradient, confinement index and riparian forest composition.  Authors used high resolution data which had not previously been available and model results were accurate as confirmed by Cedar River municipal watershed employees.  The model categorized only 11% of the watershed as highly suitable under all three criteria.  Most potential  restoration sites were located in lower elevations and were characterized by low-gradient, unconfined channels and early seral forests.  This model can be used to plan restoration at large spatial scales.  Applications of the model include restoration site identification, establishment of baseline data and monitoring and evaluation. Finally authors recommend this approach for use in other watersheds and in assessing needs for other species.

Detail maps showing the riparian buffer (left) and the clipped suitability (right). Stream reaches are displayed in blue. The maps show a gray scale gradient of values for suitability levels. Darker shades indicate poor suitability and lighter shades high suitability.

NOAA.  Portland Harbor Watershed Database & Mapping Project: GIS & Mapping. Accessed November 22, 2011: http://response.restoration.noaa.gov/index.php.

There are currently a number of scientific studies being conducted to locate and quantify toxics, natural resources and habitat restoration opportunities in the Portland Harbor Superfund site. Cleanup of the Portland Harbor and subsequent restoration activities rely upon understanding the current conditions of the site.  NOAA is combining findings from these studies with information from public agencies into a GIS and making them available through their decision support tool.  Using this tool, decision makers can analyze and share data at numerous spatial scales.  Data available through the Portland Harbor Database & Mapping Project includes: sediment chemistry, river depth, aerial photographs, effluent inputs, shoreline characteristics, habitat features, land ownership, dredging and cleanup activities.  The decision support tool allows individuals to select layers and conduct queries which address their unique needs then develop a visual means of sharing findings.

A sample of GIS layers available through NOAA's database and mapping project

Oetter, D., L. Ashkenas, S. Gregory and P. Minear. 2004. GIS Methodology for Characterizing Historical Conditions of the Willamette River Flood Plain, Oregon. Transactions in GIS 8:367–383.

Flood plain restoration along the Willamette River has become a priority to absorb flood waters and increase wildlife habitat. To maximize limited resources, authors created a GIS that will allow decision-makers to prioritize floodplain restoration sites. Using historical flood maps, photographs, General Lands Office surveys, U.S. Army Corps of Engineers river maps and digital orthophotography, the GIS maps floodplains, active channels, side channels, islands and tributaries for four different time periods. It also includes historic (pre-European settlement) and current land cover data. Findings demonstrate that channel length has decreased by 26% since 1850 and nearly 58% of former side channels have been disconnected. Finally, due to conversion for agricultural and urban land uses, 72% of the basins’ floodplain forests have been lost. The final GIS is a conglomerate of data that represents spatial and temporal variations. This information can be used to identify sections of the floodplain that are good targets for restoration, those that are not overdeveloped and were historically complex.

Mapping the historical extent of the river channel

Oregon State University, Institute for Natural Resources, Northwest Alliance for Computational Science & Engineering. Willamette Basin Explorer.  Accessed November 21, 2011: http://oregonexplorer.info/willamette.

Willamette Basin Explorer, a subset of Oregon Explorer, is a digital library containing numerous tools for policy makers and community members managing land and water resources in the Willamette Valley.  One of these tools, the advanced mapping tool, allows resource managers to create unique maps of the basin without requiring GIS software.  The advanced mapping tool provides open access to GIS data throughout the Willamette Basin.  GIS layers available in Willamette Explorer include: past and present land use and land cover, restoration and conservation opportunities, floodplain restoration, historic floods and water quality.  Willamette Explorer also uses models to project three future scenarios in the Willamette Basin: conservation trend, development trend and plan trend.  These three scenarios predict the impact that land use will have on natural resources in the Willamette Basin.  Resource managers interested in river restoration can use Willamette Basin Explorer to identify restoration opportunities which are both ecologically and socially feasible.

Using Willamette Explorer to map historic flood extent in the Portland Harbor

Rohde, S., M. Hostmann, A. Peter and K. Ewald. 2006. Room for Rivers: an Integrative Search Strategy for Floodplain Restoration. Landscape and Urban Planning 78: 50–70.

This study presents a search strategy for identifying floodplain restoration sites for which the ecological and socio-economic systems favor restoration.   It can be used to identify sites for which there will be both ecological and socio-economic gain.   The search strategy applies 3 filters to identify suitable restoration sites: restoration constraints (filters out slopes >6% and urban areas), ecological suitability and socio-economic suitability.  Using a GIS, indicators for suitability are mapped in a database.  Filter 1 excludes all areas with slopes > 6% and all urban areas.  In filter 2 map layers are weighted according to an ecological restoration suitability index.  Finally, filter 3 layers outputs of filters 1 and 2 over maps representing socio-economic factors.  The final output of the model provides potential floodplain restoration sites to be further examined case by case.  Authors applied this search strategy to a case study in Switzerland.  Authors found that this method provided a cost-effective means of pre-screening potential floodplain restoration sites.  It enables the examination of information from a wide range of sources.  Finally, it establishes a set of habitat indicators and criteria that can be replicated in other basins.  Authors noted data limitations and spatial resolution as constraints of the model, recommending that sites be further examined on a local level prior to implementing restoration activities.  GIS is a powerful tool for pre-screening potential restoration sites, saving time and money.  Advances in GIS data facilitate comprehensive restoration analyses, allowing restoration planners to incorporate ecological and socio-economic factors into decision-making.

Map displaying habitat suitability and public support

Russell, G., C. Hawkins and M. O’Neill. 1997. The Role of GIS in Selecting Sites for Riparian Restoration Based on Hydrology and Land Use. Restoration Ecology 5: 56–68.

Authors assessed potential conservation and restoration sites in the San Luis Rey River watershed based on topography and land cover.  Riparian areas were categorized according to wetness as determined from USGS 30-m digital elevation models which calculated potential flow accumulation.  Land cover was determined from Landsat imagery.  Wetness indices and land use/ land cover data were combined in a GIS to create a spatial representation of potential conservation and restoration sites.  Sites with medium or high wetness and existing vegetation were identified as conservation sites.  Sites with medium to high wetness and with agricultural or bare ground were identified as restoration sites.  About 3.67% of the entire basin fell into one of these two categories. Selected sites were prioritized according to patch size and proximity to functioning riparian habitat.  Authors noted limited digitized data (eg: no data was available for soil type) and failure to consider socio-political factors as constraints of this study.  They concluded that GIS can be a cost effective means of pre-screening for selecting and prioritizing restoration sites, especially as additional and more accurate digitized data become available.

Simenstad, C.A., Burke, J.L., O’Connor, J.E., Cannon, C., Heatwole, D.W., Ramirez, M.F., Waite, I.R., Counihan, T.D., and Jones, K.L., 2011, Columbia River Estuary Ecosystem Classification—Concept and Application: U.S. Geological Survey Open-File Report 2011-1228, 54 p.

Authors used GIS data from State and Federal agencies to develop fine scale ecosystem classifications for the Columbia River estuary.  Past spatial frameworks have categorized ecosystems according to geomorphology, hydrology and salinity.  However most have failed to address geologic history, regional climate, watershed, ocean, river and human development factors which greatly affect estuary ecosystem function.  Simenstad et al. incorporate these factors in their ecosystem classification which divides the Columbia River basin into a series of ecosystems from coarse/regional scale to fine/local scale as follows: ecosystem province, ecoregion, hydrogeomorphic reach, ecosystem complex, geomorphic catena, primary cover class.  The final classification is a visual tool for analyzing the distribution of ecosystem units at a variety of spatial scales.  Authors recommend this technique for individuals pursuing scientific research, monitoring, restoration and management.

Thom, R., E. Haas, N. Evans and G. Williams. 2011. Lower Columbia River and Estuary Habitat Restoration Prioritization Framework. Ecological Restoration 29: 94–110.

Authors present a two-tiered framework for identifying restoration projects that maximize return on investment.  The target area for this study is the historic floodplains of the Lower Columbia River (river kilometer 0 to 235).  In Tier I, a GIS is used to analyze anthropogenic impacts caused by dikes, agricultural activities, overwater structures and flow restrictions.  These impacts are quantified by a formula and then stored in the GIS, which assigns a spatial context to the impacts.  In Tier II pre-screened restoration projects are evaluated based on cost, estimated functional change, site size and probability of success.  The second tier is not GIS based.  This framework allows managers to identify impacted areas, locate potential sites and then prioritize the sites for restoration. Controlling factors or indicators of degradation that were examined in this framework are: hydrology (river, watershed and site scales), sediment quality, water quality, light, sediment dynamics, physical disturbance, depth/slope and non-native species.  Authors note that sufficient data for all of these indicators was not available.  They recommend continuously collecting data, documenting knowledge learned and refining the scoring process as knowledge increases.  This framework allows us to better understand the habitat needs of target species, track restoration progress and share information.

Figure 2. Examples of hydrologic context applied to sites according to presence of at least 50 m Columbia riverfront shoreline and slope class (flat/steep). Width of map ~15km.