Aquatic

Expanding South Dakota Aquatic Gap Analysis to the Upper Missouri River Basin

Steven S. Wall, Chad J. Kopplin, Brenda L. Kopplin, Jonathan A. Jenks, and Charles R. Berry, Jr.

South Dakota Gap Analysis Project, Department of Wildlife and Fisheries Sciences and South Dakota Cooperative
Fish and Wildlife Research Unit, South Dakota State University, Brookings

The South Dakota Gap Analysis team has completed both the terrestrial and aquatic gap analysis for South Dakota, and a draft of the final report was submitted to the GAP Operations office in August 2002.  We are now expanding our aquatic gap analysis to the Upper Missouri River Basin, which includes watersheds in Montana, Wyoming, North Dakota, South Dakota, Iowa, Minnesota, Alberta, and Saskatchewan (Figure 1).   We are using fish species distributions to evaluate biodiversity at the watershed (10-digit hydrologic unit) and valley segment scale.  Our methodology is based on that proposed by the Missouri Resource Assessment Partnership (MoRAP), who is working on a similar Aquatic GAP project for the Lower Missouri River Basin.

Figure 1.  Missouri River Basin.  The Upper Missouri River Basin includes portions of Montana, Wyoming, North Dakota, South Dakota, Minnesota, and Iowa, as well as Alberta and Saskatchewan in Canada.

There are essentially four steps to our project: (1) coordinate with various organizations across state and international boundaries to acquire necessary databases, (2) attribute fish species distribution and physical habitat features affecting fish distribution to stream (valley) segments, (3) predict the distribution of fish species at the valley segment and watershed scale based on physical habitat features and water quality, and (4) perform a gap analysis.  Specific objectives are to

1. define range extents for all fish species within 10-digit hydrologic units occurring in the Upper Missouri River Basin based on collection data,

2. determine species richness by 10-digit hydrologic units,

3. define habitat affinities for each fish species occurring within the Upper Missouri River Basin based on literature review,

4. predict occurrence of each fish species in river reaches by similarity of stream properties to habitat affinities and fish collection sites,

5. determine protection offered each fish species by hydrologic unit and river reach using stewardship layers available from states and provinces,

6. coordinate with MoRAP to merge the Upper and Lower Missouri River Basin analyses.

Data Collection

For a gap analysis of this magnitude, over 30 agencies and organizations, some across state and international boundaries, had to be contacted to locate necessary data sets (Table 1).  The completion of the South Dakota Gap Analysis Project made many data sets available and gave us a head start on data processing and analysis.  GAP products from other states within our study area also supplied many data sets.  Our base data sets include the National Hydrography Dataset (NHD), digital elevation models (DEM), surficial geology (1:500,000), 10- and 8-digit hydrologic units and equivalent-sized watersheds from Canada, land cover, land stewardship, ecoregional boundaries, and fish distributions from various state and federal agencies.  We have collected most of the data sets for each state and province.  A digitized data set of 10-digit hydrologic units for North Dakota was incomplete, and we are cooperating with the North Dakota Department of Health and several federal and state agencies to digitize 10-digit hydrologic units for our study area.  Land cover and stewardship maps are not yet complete for North Dakota but will be completed by ND-GAP in time for our project. 

PFRA = Prairie Farm Rehabilitation Administration, Agriculture Canada                          

UND = University of North Dakota

EROS = Earth Resources Observation Systems                                                                 

SDSU = South Dakota State University

EMAP = US Geological Survey Environmental Monitoring and Assessment Program      

SD Coop = South Dakota Cooperative Research Unit

EPA NHD = US Environmental Protection Agency National Hydrography Dataset         

EDWDD = East Dakota Water Development District

MT RIS = Montana Rivers Information Systems                                                               

IA NRCS = Iowa Natural Resource Conservation Service

MT FW&P = Montana Department of Fish, Wildlife and Parks                                        

MN DNR = Minnesota Department of Natural Resources

MT NRCS = Montana Natural Resources Conservation Service                                        

BMNH = Bell Museum of Natural History

MT Coop = Montana Cooperative Fisheries Research Unit                                               

MN NRCS = Minnesota Natural Resource Conservation Service

MT BMG = Montana Bureau of Mines and Geology                                                        

AB ESD = Alberta Environmental Sustainable Development

WY G&F = Wyoming Game and Fish Department                                                             

AB F&W = Alberta Fish and Wildlife Service

WY GIS = Wyoming Geologic Information Systems Center                                               

AB FMIS = Alberta Fisheries Management Information Systems

UWY = University of Wyoming                                                                                          

SK ERM = Saskatchewan Environment and Resource Management

WY NDD = Wyoming Natural Diversity Database                                                             

SK CD = Saskatchewan Conservation Data Center

ND G&F = North Dakota Game and Fish Department                                                       

ISC = Information System Corporation of Saskatchewan

ND DH = North Dakota Department of Health    TM 7 = Thematic Mapper 7 Satellite Imagery

A digitized land cover for areas in Canada was not available at a scale compatible to land cover digitized in the USA.  We produced a digitized map of land cover for our study area in Alberta and Saskatchewan from Landsat 7 Thematic Mapper (TM) imagery and are presently matching this coverage to the digitized land cover from Montana.  The map has five land cover attributes important to aquatic ecosystems; these five categories match well with the trees, cropland, grasslands, water, and urban areas of Montana. 

One problem we encountered when acquiring data sets from Canada was differences in licensing agreements.  Many GIS data sets have been privatized in Canada, and we were unable to secure a licensing agreement that would allow us to possess or redistribute data derived from original data sets.  We also experienced the same problem with data sets from the Canadian government.  Fortunately we were able to negotiate a license agreement that suited our needs for the majority of our project area.  For areas where we were unable to obtain usable data sets, we used topographic maps to digitize hydrography to match the NHD at a scale of 1:100,000.  We also used contour lines to determine elevation and channel slope. 

Stream Habitat Attributes

We are using the valley segment as our base stream unit for modeling fish distributions (Sowa 1998).  A valley segment is a length of stream (typically 3 to 30 km long) that is relatively homogeneous with respect to hydrogeomorphic features such as hydrology, geology, and elevation.  We are attributing valley segments with ten physical habitat affinities that affect fish distribution, including temperature, stream size, flow regime, channel gradient, size discrepancy, floodplain interaction, geology, elevation, stream connectivity, and groundwater input.  Valley segment attribution and delineation are based upon a hierarchical classification system (Lammert et al. 1996) and procedures outlined by Sowa (1998).  We are using GIS tools and procedures developed by The Nature Conservancy Freshwater Initiative (2002) and MoRAP for stream habitat classification, as well as our own innovations.

The NHD stream reach files were preprocessed to remove braided and ponded reaches not previously accounted for in the NHD data set.  Stream segments within the USA have been attributed with all the above habitat affinities, with exception of groundwater input.  We have added attributes to most stream segments in Canada with exception of floodplain influence, flow regime, and groundwater input.  We are using the Darcy groundwater model (Baker et al. 2000) developed by the Michigan Rivers Inventory (MRI) to attribute groundwater delivery to streams within the glaciated landscapes of our study area.  For unglaciated landscapes, we are using GIS coverages of springs to estimate potential for groundwater delivery to streams based upon spring density and spring surface-depth.  When we complete the attributing of habitat features, we will begin grouping features into classifications that represent the diversity of the entire region.

The stream data for South Dakota were processed using River Reach 3 hydrography (RF3) before the NHD data (which is now the national standard hydrography layer) became available.  Transfer of RF3 data to NHD data with the help of procedures developed by The Nature Conservancy is 95% complete. 

Fish Distribution

Fish location data have been obtained for all states and provinces within our study area.  Wyoming and Montana supplied fish locations attributed to NHD, and other states and provinces provided point locations, which saved us much time.  We have fish location data attributed to NHD reaches for all the states, with exception of fish locations for the Missouri River in North Dakota.  We are also working on transferring fish location data to stream reaches in Canada.  Known fish locations will be used in the production of species distributions by 10-digit watershed and in the valley segment-scale models.

Field Surveys

In the summer of 2000 two graduate students began a survey of an 8-digit hydrologic unit in North Dakota and Wyoming.  They also will survey 8-digit units in Canada and Montana in the summer of 2003.  Students stratified their effort by stream size (headwater, creek, small river, and large river) to survey a minimum of four sites of each stream size class and were able to survey about 16 sites in each 8-digit unit.  Students also have collected water quality data and land cover data for each of the sites.  This information will be used in our accuracy assessment of the predictive models by fish species.  These and other data collected for the state of South Dakota (new data since South Dakota Aquatic GAP) will provide a separate data set not used in the production of our models to perform the accuracy assessment.

Analysis 

Gap analysis of aquatic diversity has been completed for South Dakota.  We modeled distributions of 116 fish species across South Dakota and assessed biodiversity of fish species and habitat in relation to land conservation.  We found that the fishes of South Dakota have more protection than terrestrial animals.  A few areas proposed for protection of mammals have the potential to provide additional protection for many aquatic species.  However, the state is far from meeting any acceptable level of protection of all species and habitats for the entire state needed to conserve biodiversity.

Valley segment, species, and aquatic richness by 10-digit hydrologic unit have been completed for South Dakota Aquatic GAP (Figure 2).  The valley segment richness map represents the number of unique valley segments by 10-digit hydrologic unit.  The fish richness map represents the number of fish species predicted to be in each 10-digit hydrologic unit.  These two richness maps were summed to produce the aquatic richness map that displays aquatic biodiversity in South Dakota.  We plan to expand these procedures to map aquatic biodiversity for the entire Upper Missouri River Basin by 10-digit hydrologic unit.  Our Web page located at http://wfs.sdstate.edu/sdgap/aquaticgap.htmhas links to fish distributions by 10-digit hydrologic units, fish habitat affinities, and fish-habitat models by stream reach for South Dakota.

Figure 2.  An example of valley segment (habitat) and fish species richness by 10-digit hydrologic unit for South Dakota from SD-GAP Project.  Valley segment richness and fish species richness were combined to produce an aquatic richness map by 10-digit hydrologic unit showing aquatic biodiversity across the state.

Future Plans

We plan to complete attribution of physical habitat features and fish locations to individual valley segments by the end of March 2003.  We will then complete a quality check and begin fish-habitat modeling.  We are working with The Nature Conservancy to delineate ecoregional drainage units based upon ecoregions and major drainages to further classify fish distributions.

Literature Cited

Baker, M.E., M.J. Wiley, and P.W. Seelbach.  2000.  Spatially-explicit models of groundwater loading in glaciated landscapes: Considerations and development in Lower Michigan.  Michigan Department of Natural Resources, Fisheries Division, Lansing, Michigan. 

Lammert, M., J. Higgins, D. Grossman, and M. Bryer.  1996.  A classification framework for freshwater communities.  Proceedings of The Nature Conservancy’s Aquatic Community Classification Workshop; New Haven, Missouri; April 9-11, 1996.  The Nature Conservancy, Arlington, Virginia. 

Sowa, S.P.  1998.  Gap analysis in riverine environments.  Gap Analysis Bulletin 7:18-20.

The Nature Conservancy Freshwater Initiative.  2002.  GIS tools for aquatic macrohabitat classification.  Internet: http://www.freshwaters.org/info/large.shtml#gis, January 14, 2002.