If we find a species and document its presence at a location, then we can say with certainty it occurs there. If we do not find that species, however, it’s a little more complicated: does the species occur there but we failed to cross paths with it on our search, or is this location truly outside of its range? NatureServe uses advanced ecological modeling techniques to develop predictions of where imperiled species are most likely to occur.
Species Habitat Modeling
It’s impossible for our scientists to inventory everywhere. Using the mapped locations of the species we have documented, and data characterizing environmental conditions at those locations, NatureServe develops models that predict where the habitat for at-risk species is likely to be found. The models produce maps of species distributions, which helps rule out places where at-risk species probably don’t occur while directing new field surveys and conservation efforts to places that matter most. Better species distribution maps save time, minimize conflict, and make conservation more efficient.
Building off our habitat models, NatureServe has released a portfolio of maps that identify areas critical to sustaining our nation's rich biodiversity.
More generalized information on species distributions also helps land managers, scientists, and the interested public understand which species might be encountered in a given area. NatureServe generates range maps, which depict the broad geographic areas in which species are found, using our documented occurrence data and the collective knowledge of our network of expert scientists.
Project Highlight: OneRange
Refining Habitat Maps for Imperiled Species Via Human-AI Collaboration
A lack of precise information on the habitat distribution of endangered plants and animals inhibits efficient management actions and creates unnecessary regulatory burdens. Using broad-range maps at the county or watershed level results in inconsistent data and could be more efficient for directing management and conservation efforts. Without robust, consistent, transparent, and scale-appropriate maps to guide managers and regulators, potential impacts on imperiled species may hold up land use decisions.
NatureServe recently partnered with the Fish and Wildlife Service and NatureServe Network member programs in Florida, Georgia, and South Carolina to demonstrate the power of our human-AI collaborative approach for modeling species habitats. This effort, called the 'OneRange project,' created new, refined habitat maps for four endangered species of high conservation conern: Britton bear's grass, Apalachicola false rosemary, fringed campion, and the frosted flatwoods salamander.
The maps below show the power of fine-scale predictive habitat maps. The map on the left shows the range map for Apalachicola false rosemary (Conradina glabra) from the Environmental Conservation Online System (ECOs), with the new map developed by the One Range project on top of it. The map on the right zooms in to just the One Range map layer, which highlights where the conservation efforts would be most effective.
This project demonstrates how NatureServe's species habitat modeling process combines state-of-the-art AI and expert input. This collaborative approach could result in better refined, robust, and transparent species habitat maps for threatened and endangered species to support regulatory activities and improve species recovery outcomes.