Eco-Logical Webinar: Improving Aquatic Connectivity at the Landscape Scale
in the Southeastern United States

October 13, 2016

Presenters:
Mike Ruth, Federal Highway Administration, Office of Project Development and Environmental Review
Duncan Elkins and Nate Nibbelink, University of Georgia
Evan Collins, U.S. Fish and Wildlife Service
Thomas Prebyl, University of Georgia

Learn more about Eco-Logical at the FHWA website

Images: Volpe, The National Transportation Systems Center logo and the U.S. Department of Transportation's Federal Highway Administration logo
Image: Collage of colored photographs of a bridge, a deer, a fish, and a curved rural road from the cover of the Eco-Logical: An Ecosystem Approach to Developing Infrastructure Projects report

This webinar broadcasts audio over the phone line and through the web room, which can strain some internet connections. To prevent audio skipping or webinar delay we recommend participants:

• Close all background programs
• Use a wired internet connection, if possible
• Do not use a Virtual Private Network (VPN), if possible
• Mute webroom audio (toggle is located at the top of webroom screen) and use phone audio only

• An ecosystem methodology for planning and developing infrastructure projects
• Developed by eight Federal agency partners and four State DOTs
• Collaboration between transportation, resource, and regulatory agencies to integrate their plans and identify environmental priorities across an ecosystem

Images: Logos of the following U.S. agencies: Bureau of Land Management (BLM), Environmental Protection Agency (EPA), Department of Transportation (DOT), National Oceanic and Atmospheric Administration (NOAA), National Park Service (NPS), Army Corps of Engineers (USACE), Department of Agriculture (USDA), Forest Service (USFS), and Fish and Wildlife Service (FWS).

• Roadways have direct and indirect effects on wildlife
• Improving connectivity for aquatic organisms a key goal of improved infrastructure planning
• Most previous work focused on terrestrial connectivity

• Helps different agencies understand importance of connectivity from the outset of the planning process
• Communication tool for engaging stakeholders in discussing barriers to wildlife movement

Image: Photo of a flock of pelicans standing in a wetland

• Mapping Tools key to characterizing resources in preparation for developing Regional Ecosystem Framework (REF)
• Helps identify existing concerns for aquatic species
• Helps identify past impacts at critical locations

Image: Photo of hikers at the edge of a pond, high up in the mountains

• Creating Regional Ecosystem Framework (REF) requires collecting geospatial data on resources and transportation plans
• Creating Planning Scenarios to define footprint of potential projects

Image: Photo of a beach edged by pine trees

October 13, 2016

Presenters:
Mike Ruth, Federal Highway Administration, Office of Project Development and Environmental Review
Duncan Elkins and Nate Nibbelink, University of Georgia
Evan Collins, U.S. Fish and Wildlife Service
Thomas Prebyl, University of Georgia

Learn more about Eco-Logical at the FHWA website

Images: Volpe, The National Transportation Systems Center logo and the U.S. Department of Transportation's Federal Highway Administration logo
Image: Collage of colored photographs of a bridge, a deer, a fish, and a curved rural road from the cover of the Eco-Logical: An Ecosystem Approach to Developing Infrastructure Projects report

• ~2010, Regional connectivity analyses focused on Dams
• 2011 SARP-funded assessment of culvert studies in Southeast
• Summer 2012, FHWA project: Predicting barrier presence and passabilityat road/stream xings
• Summer 2013, SA-LCC project: Expanded region, more GIS tools

Image: Photo of a two-person crew taking culvert measurements under a bridge
Image: Photo of a circular culvert

Freshwater fish account for ≈25% of all vertebrate diversity

Image: Global map that is color-coded to indicate levels of freshwater fish species per ecoregion area

Image: Map of the continental U.S., labeled “Freshwater Fish Richness/863 spp,” color-coded on a scale of 2 to 238
Image: Map of the continental U.S., labeled “Endemism/591 spp,” color-coded on a scale of 1 to 77

Key Threats

• Habitat degradation
• Water pollution
• Overexploitation
• Invasive speciess

Image: Map of North America, color-coded to indicate the number of imperiled taxa in ecoregion

• Hydrologic connectivity - “water-mediated transfer of matter, energy, and/or organisms between elements of the hydrologic cycle”
• Longitudinal
• Lateral
• Vertical

Image: Photo of a cascading waterfall in a lush forest
Image: Map showing aquatic connectivity in the study region

Dams

• Barriers
  • Animal movement
  • Nutrient and Sediment transport
• Fragment/alter habitat

Image: Aerial photo of a large dam and the wetland above it
Image: Map showing aquatic connectivity in the study region

• Barriers
  • Animal movement
  • Nutrient and Sediment transport
• Fragment/alter habitat
• Growing knowledge of their location

Image: Aerial photo of a large dam and the wetland above it
Image: Map showing aquatic connectivity in the study region marked with numerous black boxes representing dams that fragment connectivity

Road Crossings

• Fragment habitat
• Obstruct fish movement
• Passability variability

Image: Photo of two very small culvert openings
Image: Map showing aquatic connectivity in the study region marked with numerous black boxes representing road crossings that fragment connectivity

Image: Map of the southeastern continental U.S., marked and labeled “SLEUTH urbanization potential, 2020-2070, by Clarke and Doato, salcc.databasin.org”

• Attributed to the structure
• Needs to be discussed in terms of a species or group of species

Image: Two photos of fish spawning upstream over a low dam
Image: Photo of a swarm of eels traveling down a river

Images: Photos of eleven southeastern fishes

Images: Three photos of field workers measuring the water depth of three different culverts

Images: Three photos of field workers measuring the water depth of three different culverts

Images: Two photos of field workers measuring the water depth of three different culverts (there are two culverts side-by-side in one of the photos)

1. Can we use spatial modeling to predict impassable or problematic structures on the ground?

Image: Flowchart showing two inputs (“Landscape Gradients” and “Classification Method”) as inputs to “Build Predictive Models,” which then feeds into “Use Predictions to understand cumulative effects of culverts”

• Three watersheds that reflect the geographic diversity of the Southeast
• Sites of previous connectivity studies

Image: Map showing three watersheds: one straddles Tennessee and North Carolina, one is in northern Georgia, and the third is in the Florida Panhandle and extends north into Alabama

• Upstream catchment area attributed with land cover types, uses, and ownership
• Only sites with an upstream catchment area <60 km² were selected

Image: Map that shows the cluster analysis stream line of the upstream catchment area

60 km²?

Images: Two photos of bridges over small rivers

• 95% of all culverts cross a stream with a watershed < 60 km²
• 95% of bridges cross a stream with a watershed area of <84 km²

Image: Vertical bar graph that shows that bridges outnumber culverts in the watershed area by about five to one

Image: Map of the watershed area that straddles Tennessee and North Carolina, marked with many green circles

Image: Map of the watershed area in northern Georgia, marked with many green circles

Image: Map of the watershed area that straddles the Florida Panhandle and Alabama, marked with many green circles

June 2013 - February 2014

• Perch Height
  • To Water
  • To Sediment
• Culvert Length
• Culvert Slope
• Sediment in Culvert
• Scour pool presence

Image: Photo of two field workers taking measurements of two culverts

Image: Diagram of a cross section of a culvert under a road, with the following elements labeled: road surface, culvert channel bottom, culvert inlet P₁, culvert outlet P₂, outlet perch, water surface, and tailwater control P₃

• Passabilityin terms of three families
  • Percidae
  • Cyprinidae
  • Salmonidae

Image: Three photos of three fish

• 52% pipe culverts
• 16% bridges
• 14% box culverts

Image: Vertical bar graph of the Road Crossing Structure Types for Chipola, Etowah, and Nolichucky shows that Circular Culvert is by far the most prevalent, followed by Bridge and Box Culvert, and then No Structure, No Access, Bottomless Box Culvert, and Other

• Mostly passable for all families
• 7 indeterminate culverts (removed from analysis)
• 35 dry crossings (removed from analysis)

Image: Vertical bar graph of the Field Data Summary for Percidae, Cyprindae, and Salmonidae shows that the Salmonidae passes the highest through both passable and impassible watersheds, followed by Cyprindae and then Percidae

• Landscape characteristics around in the upstream catchment area of a culvert and within a 100 m buffer are likely to influence erosional processes
• Increased erosion at a culvert will cause more scour and increase the perch height of a culvert and prevent a fish from entry

• Percent land cover type
• Percent impervious surface
• Roughness
• Compound topographic index
• Stream power
• Slope position of culvert
• Stream reach gradient
• Road type
• Flow accumulation
• Upstream catchment area
• Discharge for a 5 year flood

Image: Map showing mean impassability per HUC 10 in the Study Area, ranging from 0.510 to 0.588

Image: Map showing Chipola mean impassability per HUC 10 in the Study Area, ranging from 0.387 to 0.319

Image: Map showing Etowah mean impassability per HUC10 in the Study Area, ranging from 0.0551 to 0.346

• Passability can be predicted with landscape characteristics
• Predictive modeling can help us gain a better understanding of where problems may occur
• Predictions can be used to guide prioritizations

Image: Photo of a work crew demolishing a short dam

Thomas Prebyl, Duncan Elkins, Evan Collins, Nathan Nibbelink
University of Georgia Warnell School of Forestry and Natural Resources

Image: Fluvial map showing obstructions to connectivity in the Study Area
Images: Two photos of fish

1. Overview of prioritization problem
2. General workflow and ArcGIS toolbox
3. Application Examples
4. Ongoing work

• How do dams and culverts influence connectivity?
• Which barriers (if removed) would most benefit connectivity.

Image: Fluvial map showing dams and culverts in the study area
Image: Photo of water rushing over a dam
Image: Photo of three square culverts under a bridge

1. Data Sources
2. Data Preparation
3. Network Simplification
4. Prioritization Algorithms
5. Visualizing Results

Image: Cartoon graphic of two toolboxes, one of which is open and contains a python. The image is labeled “Custom ArcGIS Toolbox | ‘Stream Network Tools.’”

• Streams
  • National Hydrology Dataset (USGS)
  • NHDPlus Version 2 (Horizon Systems)
    • 1:100K but many value added attributes
• Dams
  • National Inventory of Dams (NID): US Army Corps of Engineers
  • National Anthropogenic Barrier Dataset (NABD)
  • GeoFIN(USFWS)
  • NHD Dam Events (USGS)
  • Various regional and state datasets (e.g. SARP)
• Roads
  • TIGER/Line (US Census)
• Bridges
  • National Bridge Inventory (FHWA)

• Intersect roads and streams to identify likely culverts
• Remove known bridges
  • NBI and drainage area
• Snap dams to stream network
• Identify disjunct stream segments (tool)

Image: Fluvial map showing dams and culverts in the study area
Image: Small version of the Python Toolbox image from Slide 46

Image: Graphic showing two fluvial maps of the Study Area stream network. The input shows many colors representing different networks with an arrow to the output showing fewer colors, representing simplification
Image: Small version of the Python Toolbox image from Slide 46

Image: Fluvial map showing the stream network in the study area. Triangles dispersed throughout represent dams, circles dispersed throughout represent likely culverts. Different colors represent subnetwork length in kilometers. Most of the subnetwork length is less than 10km.

Image: Graphic of the simplified stream network with an arrow to an output spreadsheet representing stream data

How to prioritize barriers for removal?

Image: Graphic of the stream network, with different colors representing the subnetworks, with triangles representing dams.
Image: The dendritic connectivity index equation

Computation Size:
n = total # of barriers
b = # to remove

     n!      
b!(n-b)!

Image: Fluvial map showing the stream network in the larger study area, with triangles representing barriers

Computation Size:
n = total # of barriers
b = # to remove

     n!      
b!(n-b)!

Image: Fluvial map showing the stream network in the larger study area, with triangles representing barriers

Use heuristics to limit search

• Connect high-weight streams
• Identify paths where a large cumulative improvement to passability is possible
• Progressively expand search
  • Path, Neighborhood, Full Network

Image: Fluvial map showing the stream network in the larger study area, marked with circles representing barriers
Images: Python and igraph logos
Image: Python toolbox graphic from Slide 46

Evaluation

• Compared to exhaustive searches
• 300 node network
• Removing 3 barriers
• > 4 million possible combinations

Image: Graph of the downward sloping curve representing exhaustive and heuristic removal, which shows a decrease in dendritic connectivity index and an increase in combination rank
Images: Python and igraph logos
Image: Python toolbox graphic from Slide 46

Image: Chart showing results from the study
Image: Fluvial map showing the stream network in the larger study area, with circles representing barriers: five circles are colored turquoise, representing priority removal sites
Image: Python toolbox graphic from Slide 46

DCI in the Etowah River Watershed

Image: Vertical bar graph showing DCI for dams only at 90.877, maximum passability for DS pass 1 at 28.523, maximum passability at 16.555, median passability at 11.252, and minimum passability at 8.251

Distribution of Fragment Lengths in the Etowah River Watershed

Image: Vertical bar graph showing the distribution of fragment lengths in the Etowah River Watershed. Percent of total length is on the Y axis from 0-1.0 and fragment length in kilometers is on the x axis. The highest percentage for dams and culverts is between 0 and 10 km; the highest percentage for dams is greater than 500km. The lowest percentage for dams and dams and culverts is between 100 and 250 km.

Evaluating Alternatives

Image: Graph that shows evaluation of alternatives for Tallapoosa. The graph shows the relationship between barriers removed and percent increase in connectivity.

Compare scenarios

• Add Barriers
• Make barriers ‘un-removable’
• Alter passability assumptions
• Alter stream weight (habitat quality)

Image: Fluvial map showing the stream network in the larger study area

South Atlantic LCC Report

• Assessing the influence of road-stream crossings

Image: Pair of maps showing the assessment of influence of road-stream crossings in the South Atlantic LCC. Left map shows Anderson 3-level with half as 0.55 mean probability impassable. Right map shows Coffman Cyprinid with half showing 0.2 mean probability impassable.

Web-application of aquatic connectivity tools

• Partnership with CSU, UGA, UMASS, & Army COE

Image: Logo: eRAMS | Share your geographic perspective…

• Step 1: Build and strengthen collaborative partnerships and vision
  • Shared understanding of issues surrounding aquatic connectivity
• Step 2: Characterize resource status and integrate natural environment plans
  • Compile data and identify current conditions “such as stream crossings”
• Step 3: Create a regional ecosystem framework (conservation strategy + transportation plan)
  • Inform scenarios to help define footprint for proposed projects

Images: Three photos of forests and mountains

• Step 4: Assess effects on conservation objectives
  • Help create a regional-scale picture of potential and cumulative impacts
  • Scenarios can help identify and quantify mitigation needs

Image: Photo of a rushing stream in a forest

• Step 4: Assess effects on conservation objectives
  • Help create a regional-scale picture of potential and cumulative impacts
  • Scenarios can help identify and quantify mitigation needs
• Step 5: Establish and prioritize ecological actions
  • Tool can help prioritize conservation/restoration actions
  • Spatial location and type of impacts can help identify potential lead agencies

Image: Photo of a rushing stream in a forest
Image: Photo of a forested valley between forested mountains

• Step 4: Assess effects on conservation objectives
  • Help create a regional-scale picture of potential and cumulative impacts
  • Scenarios can help identify and quantify mitigation needs
• Step 5: Establish and prioritize ecological actions
  • Tool can help prioritize conservation/restoration actions
  • Spatial location and type of impacts can help identify potential lead agencies
• Step 6: Develop crediting strategy
  • Identify measurements for mitigation goals - e.g. connectivity indices
  • Help establish off site mitigation opportunities

Image: Photo of a rushing stream in a forest<
Image: Photo of a forested valley between forested mountains
Image: Close-up photo of a pine tree branch with forested mountains in the background

Presenters

• Mike Ruth, Federal Highway Administration, Office of Project Development and Environmental Review
• Duncan Elkins, University of Georgia
• Evan Collins, U.S. Fish and Wildlife Service
• Thomas Prebyl, University of Georgia
• Nate Nibbelink, University of Georgia

Learn more about Eco-Logical at the FHWA website

Image: Collage of colored photographs of a bridge, a deer, a fish, and a curved rural road from the cover of the Eco-Logical: An Ecosystem Approach to Developing Infrastructure Projects report