The landscapes ecosystem team now has recommendations for revisions to your cooperative’s landscape indicators. The landscape indicators focus on connections across all terrestrial habitats. These changes are now open for comment until June 19th. You can either comment at the bottom of this post or email me directly (rua@southatlanticlcc.org)
Here are the proposed changes to the landscapes indicators:
Index of structural connectivity
Provides an indicator of how well natural habitat is connected in general without using species specific thresholds
Proposed change: None
Index of functional connectivity
Provides an indicator of connectivity from the perspective of multiple species with different home range sizes and connectivity requirements
Proposed change: Remove this as an Indicator and consider adding it back in the future as more refined information is available. There is an existing project targeting this measure but results were not ready in time for this revision cycle. Functional and structural connectivity are expected to be highly correlated.
Acres of interior natural habitat in low road density areas
Provides an indicator of large patches of natural habitat
Proposed change: Change to a more easily modeled metric for large patches of natural habitat (“Acres in low road density areas” instead of “Acres of interior natural habitat”). Only paved roads are considered for this revised metric of low road density areas. Modeling interior natural habitat directly was extremely sensitive to GIS misclassification between natural and non-natural habitat classes.
Acres of resilient biodiversity hotspots in natural condition
Provides an indicator of landscape condition for rare and range restricted organisms
Proposed change: Add the word “resilient” to clarify that these are places that are both biodiversity hotspots now and likely to be hotspots in the future in the face of future change (Climate, Urban growth, etc).
Learn more about the idea behind this metric
Number of historic register sites without nearby development
Provides an indicator of historic sites that have maintained their historic surroundings of rural/natural land cover
Proposed change: None
Hi John and Cindy,
Thank you both for your comments. The low road density indicator is now new and improved! I’ve incorporated some changes and tested the sensitivity of the results to other possibilities. Here’s what I’ve come up with:
1) Paved vs. Unpaved Roads
I tested paved-only road density vs. ALL road density. The resulting “low road density areas” were exactly the same in both cases. While this might seem strange, there are definitely inherent spatial patterns in the data. Most unpaved roads were in areas that already had a high road density. The unpaved roads that were in low road density areas did not substantially increase the density, and they were not enough to cross our threshold of 1.5 km/km2, which is based on available scientific literature. While weighting different road types (TIGER: primary, secondary, and unpaved) is possible, the weighting would be subjective and may not be easy to interpret on-the-ground. In fact, we investigated the possibility of quantitatively ranking roads based on traffic, but there appears to be a lack of available data for the LCC. With this in mind, we are open to new possibilities if you have seen a method to do this objectively. There are also inherent spatial patterns in the data that probably make the results robust to such a ranking. e.g., 6-lane interstates tend to be in areas with higher road density than state highways, state highways tend to have more roads around them compared to unpaved roads, etc.
2) The 3-km radius, open water, and the barrier island problem:
Yes, the original GIS calculation did include large water bodies and the original mask of the data did not affect the analysis as I expected. We have a solution! The new, detailed methods are here:
Basically, we excluded water bodies mapped at a 1: 1,000,000 scale. Also, the new road density measure is now: (length of road within a 2-km radius) / (terrestrial area within a 2-km radius). The resulting measure is still road length per km2, but the areas in the density calculation are not always the same. This resulted in the narrow barrier island segments with a single road being excluded, but wider barrier islands with few roads are still included as “low road density.”
During the process of developing this indicator, I experimented with several radii. For example, the 1-km radius did not work well because the interior of most farm fields were identified. However, after the initial indicator review, we added the minimum patch size of 10,000 acres (see methods), so I think a 2-km radius is now suitable and this will help with the urbanization problem John mentioned.
3) We’ll get this revised indicator posted in the next week. Overall, I think it’s important to remember that this measurement alone is not meant to identify absolutely pristine areas or areas with no road effects. As one piece of the puzzle, using this indicator does say that we should favor conservation in low road density areas compared to high road density areas. As we have indicators for all major, southeastern ecosystems (Beaches and Dunes, Estuarine, Forested Wetlands, Upland Hardwood Forests, etc.), the “Landscapes” indicators will be combined with them to identify high priority conservation areas. As Cindy noted, Ag fields are particularly difficult to deal with; plus, they don’t directly fit into one of our southeastern ecosystems. Therefore, their treatment will continue to evolve as we work our way to inland ecosystems. The good news is that the National Landcover classifications (NLCD) do a pretty good job at identifying Ag lands.
Again, we thank you both for your thoughts and hope you continue to follow our indicator process as we make our way through each ecosystem. The more people we can get involved, the better our indicators will be.
Sincerely,
Brad
Brad,
Thanks for the update. Sounds like you’ve done a good job addressing the issues I raised. It seemed odd at first that adding unpaved roads didn’t make any difference, but when I went back and looked at the road types you DID include originally it looks like they probably did include many unpaved roads and only excluded a few types that probably aren’t that common (such as 4WD trails, bike paths, and bridle paths). I agree that weighting the roads by type would likely be problematic unless the road class information is sufficiently reliable and clearly interpretable (and somewhat subjective even in that case), so it’s probably best to stick with the unweighted metric unless more quantitative data become available. Thanks for being so responsive; I look forward to seeing the new map!
WORKS FOR ME, THANKS!
John and Cindy – Glad to hear those updates addressed the issues you raised. Thanks again for digging deep in your review and helping improve the low road density indicator. It was really helpful.
For everyone else looking at these comments – The comment period for the landscapes indicator revision is now officially closed. These proposed revisions are now official and the updated set of indicators will be used in developing the South Atlantic Blueprint 2.0