Habitat Corridors and Management

Wildlife Corridors in Ripton, Vermont

Evaluating Wildlife Habitat Management Using GIS and Ground-Truthing

Benefits of Wildlife Corridors

As development and urbanization of previously open lands increases, many potential risks are posed to wildlife in these areas. Specifically, roads and infrastructure can lead to further habitat fragmentation, leaving animals with large ranges isolated from prey, shelter, and one another. Habitat fragmentation can contribute to species population decreases, potential species extinction, and vulnerability to natural disasters and predation (Ecological Society of America). Roads can also be a source of mortality for especially larger species, in some cases the leading cause in areas where roads are being upgraded to accommodate greater traffic volume (Gloyne and Clevenger, 2001). This is why wildlife corridors are so essential for wildlife conservation efforts.

Wildlife corridors have numerous benefits for animals: they allow species to move between previously isolated populations, promote species diversity, and provide areas for juvenile dispersion and seasonal migration (Ecological Society of America). Wildlife corridors are also beneficial for humans, as they can decrease vehicle collisions with large animals and minimize human interactions with predators like wolves and bears (Ecological Society of America). There are many forms of wildlife corridors, the most effective form being bridges over roads (pictured right). However, since this type of infrastructure is expensive, many areas do not have the luxury of these structures. Therefore, wildlife corridors in most areas are simply paths of vegetated areas or habitats away from human development with road crossings at less dangerous points. This project will aim to identify a corridor for wildlife movement in a rural Vermont setting using GIS, which allows us to analyze the "costs" of movement.

Wildlife Corridor over a highway in Banff National Park, Canada

Source: https://conservationcorridor.org/2012/10/banff-national-park/

Overview of Project

For this analysis, I will be assessing potential wildlife corridors in Ripton, Vermont, located in west-central Vermont, about 20 minutes southeast of Middlebury. I used GIS to 1) analyze which sections of roads would be harder for wildlife movement in Ripton, and then, 2) found the least costly route connecting two habitat areas that intersect the town boundaries: the Green Mountain National Forest and the Joseph Battell Wilderness. My entire class also performed this analysis, and after finding the common road crossings we identified, we took part in fieldwork to perform ground-truthing, or using direct observation at a location to perform an accuracy assessment.

In this case, we were looking for signs of animal activity around two areas we identified as the most likely road crossing areas along Vermont Rte. 125 in Ripton. We were hoping to provide more concrete analysis of the current situation at this location and whether it is being used safely. The Critical Paths project, as part of Vermont's State Wildlife Action Plan (SWAP), has been investigating the impact of roads on wildlife habitat, and specifically identifying priority road crossing areas. This area on Rt. 125 near the Breadloaf Wilderness has been identified by the Northeast Regional Center of the National Wildlife Foundation as priority zone 21 (see page 46 of this report)

Data and Methods

To perform this analysis with GIS, it requires data on roads in the study area, habitat areas, and a habitat suitability layer. For more information on habitat suitability layers, specifically for bobcats, look at my project on bobcat habitats in Vermont. Here was the data I used for this project:

Ripton Roads (from VTrans on VCGI)
Habitat areas near Ripton (from USGS Protected Areas Database), selected ones where bobcats were seen
Habitat Suitability Layer (created by Professor Lindsay Dreiss) based on:
- Road density, land use/land cover, elevation, distance from forest edge, snow dept, and distance from roads

1. Wildlife Crossing Values

The first step in my analysis was creating a layer assessing the wildlife-crossing value for each road, similar to analysis by Austin et al. 2005. Wildlife-crossing values (WCV) show the difficulty of crossing a certain road, based on the habitat suitability index. The first step was creating a 100-meter buffer around each road. Then, using zonal statistics, I averaged the HSI values within each 100-meter buffer to get the average habitat suitability around every road. Then, I summarized these values for each road back onto the individual road instead of the 100-meter buffer using raster calculator and converted it back to a polyline feature. This analysis gave me a WCV for each road in Ripton, highlighting where it is easier for animals to cross with lower values and where it is more difficult with higher values.

2. Cost Surface

The purpose of a cost surface is to have a specific value for every point that demonstrates the "cost" of a particular activity or object being there. In this case, the cost surface is meant to show how likely it is that a certain species would be at specific location. This is easily calculated using the available data by inverting the habitat suitability layer, since this shows how suitable each specific location is for species. This involved subtracting the values in the HSI from 5 using raster calculator to invert the values. This cost surface is based on the same weighted criteria as the HSI.

3. Cost Distance and Cost Back Link

Using this cost surface, it was possible to find the cost distance and cost back link from each habitat area. Cost distance shows the least accumulative cost distance for each cell to a specific location, in this case from one habitat area to the other. Cost back link, on the other hand, defines the direction of the neighbor that is the next cell on the least accumulative cost path. The ArcGIS tool, simply named Cost Distance, creates these two layers using just the cost surface and the habitat patches. This tool was run twice, once finding the least accumulative cost path from Joseph Battell Wilderness to the Green Mountain National Forest, and vice versa.

4. Corridor Analysis and Cost Path Analysis

The final step in the GIS analysis portion included running a Corridor analysis and a Cost Path analysis. Corridor analysis takes two least accumulative cost layers, as created in the previous step running both directions, and basically calculates the sum to find areas most suitable for movement, in this case for animals from one habitat to the other. I ran this tool to get a sense of where in Ripton are the most suitable corridors for species to move through. The final step was running a Cost Path analysis, which actually calculates the least-cost path from a source to a destination. I decided to calculate the least-cost path from the Joseph Battell Wilderness to the Green Mountain National Forest, but it could have been done either way. This was the layer that was used to find the areas along Rte 125 in Ripton that we identified as most suitable for species movement.

5. Fieldwork: Ground-Truthing in Ripton

In order to both confirm whether the corridor we created using GIS was being used by animals and evaluate whether the area is in need of corridor management, we spent an afternoon in late October doing fieldwork near the Breadloaf campus in Ripton, VT in areas we identified. Before heading out, we decided on a sampling design for our data and what attributes would be recorded. For the sampling design, since we were doing field work near the Breadloaf/Rikert trails, we took samples at equal intervals horizontally (5 meters) on either side of a trail on each side of the road, and at equal intervals vertically (5 meters). This allowed us to cover significant areas systematically and make sure that the error of our GPS devies (~3-5 meters) would not impact our results.

Using the Avenza Maps app on personal mobile devices, we recorded a GPS location every 5 meters along our transect and recorded presence/absence of evidence of wildlife presence, and if there was evidence, what the evidence was and what species it could be. We modeled our fieldwork after the work of Shepherd and Whittington (2006), although they were looking for tracts in the winter for a longer period of time. The results of this fieldwork are summarized below by fieldwork site. I personally worked on the Steam Hill Rd fieldwork site along the trails on either side of the road.

Results

The least cost path and corridor analysis that I conducted show a clear correlation with each other. In terms of the fieldwork matching with these assessments, the conclusions are not as clear. While there was ample evidence of wildlife presence in both of these areas, which were identified in the analysis, this was a severely limited study, as explained below. However, I think we can conclude from this work that these two road crossings are well-populated areas for animals seeking to move between habitat areas. Since I worked on the Steam Hill Rd site, I can speak better to that site specifically. Steam Hill Rd is a dirt road that sees very minimal car traffic. While we were there on a Monday afternoon we only witnessed one other car drive through the area we were analyzing. So already, it is clear that this would be a safer area for animals to cross. And since there is not much development in this area, there is not many barriers to species living there. This would explain why we recorded many pieces of evidence of both small mammals and deer.

At both field sites, the evidence of wildlife is present on both sides of the road, which strongly suggests that neither road is a barrier to crossing. Route 125 is a more major road, so it is interesting that the analysis identified this crossing. While I did not get to personally walk around this area, the other group was taking data points around trails in the area which might explain why this area sees animal evidence. Trails for human use, while not entirely desirable because humans are present, also are surrounded by unfragmented forest which is valuable to animals. However, the reason that animals must use this area to cross is that there is no built wildlife corridor present near this road and it fragments habitats. Overall, the fieldwork results confirm that there is at least some use of the identified corridors using GIS, but also show that there may be more corridor management needed to make movement and crossing even easier for animals.

Limitations and Sources of Error

There are a number of limitations to my approach that most likely caused error and does not allow us to conclude much about this area based on the study:

The ground-truthing was conducted on one afternoon in late October in a limited area, with leaves covering the ground and no snow. Therefore, tracks were much harder to see due to the lack of snow and droppings and other evidence could have been covered by fallen leaves on the ground. While we saw some evidence of animals, to be able to conclude more definitively if there is significant activity in these areas, a study would need to be carried out over a larger area, over a longer period of time, with more advanced techniques, such as a wildlife camera.
Due to the fact we were using our mobile cellphones to acquire locations for evidence or absence instead of a more advanced GPS device, there is a 3-5 meter source of error on each point. This is why we made a 5+ meter distance between points, but this is still a limitation of my approach.
Inverting the HSI is one simple approach to calculating a cost surface for this type of analysis, and was done here due to time constraints, but it is not the most accurate model. To create a cost surface raster with ArcGIS, you can also use variables that are specific that negatively impact species movement, instead of using the same ones as the HSI.
The data we used for the habitat areas was taken as only areas where bobcat had been identified, but we did not see any evidence explicitly of bobcats. If we had used other habitat areas, that may be used by other animals that more often leave droppings or evidence of presence, and then did the analysis on these with ground-trothing, the results might have been better.

Management Implications

Based on my analysis, it is clear that my study areas, and locations like it across the State of Vermont, are used often by animals attempting to move between habitats or access prey elsewhere. This type of GIS analysis and fieldwork is important for identifying areas of specific need and seeing what potential infrastructure improvements could be made. Work is already being done by the State of Vermont and conservation organizations to analyze and improve road crossings for animals and deal with increased habitat fragmentation.

The Northeast Regional Center of the National Wildlife Foundation have been working on the Critical Path project, which is focused on improving wildlife and public safety around roads through minimizing habitat fragmentation and creating wildlife corridors. This project has been necessary in the pursuit of making Vermont a safer area for all species. This project has identified vehicle-wildlife collisions and habitat fragmentation as the major issues in the state and points of emphasis. There are also other organizations, like the Staying Connected Initiative, that serve the greater Northeast region and are focused on improving landscape connection for both wildlife and human benefit. Organizations like these are important for managing the issues arising from human-wildlife interaction in Vermont and elsewhere.

Moose and calf crossing a road near an oncoming car.

Source: https://www.nwf.org/en/Magazines/National-Wildlife/2015/JuneJuly/Conservation/Wildlife-Corridors

I would recommend several future steps Vermont could take to improve the management of habitats and road crossings in the state:

Increase awareness of popular animal crossings along major roads with road signs or an ad campaign. Making Vermont's drivers more aware of the possible presence of wildlife near the road will make these areas safer for humans and wildlife
Invest in research about the possible construction of vegetated animal bridges over major roadways that could allow animals to cross without danger of cars. While this would be a major expense, other places have implemented them and had good results pertaining to safety and wildlife management.
Investment in research and tracking studies that can continually offer new data and habitat models for experts to use to define areas of need and potential further infrastructure improvement. Using more advanced techniques, as used by Austin et al. (2005), allows researchers to gather more accurate and specific data for the area of interest.

Sources

Austin, J. M., K. Viani, F. Hammond, and C. Slesar. 2005. A GIS-based identification of potentially significant wildlife habitats associated with roads in Vermont.

Ecological Society of America, "Habitat corridors help preserve wildlife in the midst of human society," August 2, 2011 by Terence Houston. https://www.esa.org/esablog/ecology-in-policy/habitat-corridors-help-preserve-wildlife-in-the-midst-of-human-society/

Gloyne, C.C. & Clevenger, A.P. 2001: Cougar Puma concolor use of wild-life crossing structures on the Trans-Canada highway in Banff National Park, Alberta. - Wildl. Biol. 7: 117-124.

“Northeast Regional Center | National Wildlife Federation.” The National Wildlife Federation, www.nwf.org/Northeast.

Shepherd, B., and J. Whittington 2006. Response of wolves to corridor restoration and human use management. Ecology and Society 11(2): 1.