The Alaska Gap Analysis Project
"keeping common species common"
Our Projects
 |
|---|
 |
 |
Vertebrate Distribution Modeling
Our objectives for this project include creating consistent and current data products that are repeatable and standardized, and of high utility to resource managers. To achieve this goal, we intend to produce spatially explicit models that predict the range and distribution of terrestrial vertebrate species in Alaska to support analysis of conservation status.
Previous state and regional Gap projects have used a deductive, or “expert systems”, modeling approach in which information about species habitat associations are synthesized from experts and literature reviews. Landcover data is then used to predict a species distribution based on specific habitat associations.
In recent years, inductive modeling techniques (statistical models of climatic and physical limits at known points of occurrence) have gained in popularity as GIS technologies have advanced. Advantages to using inductive modeling techniques include predictions that are precise and repeatable and calculation methods that are explicit and transparent.
In order to improve the quality and utility of our species distribution maps, we plan to use a combination of deductive modeling and inductive modeling, similar to methods being explored by the Northwest Regional Gap Program, to predict the distribution of terrestrial vertebrate species across their range in Alaska. By combining the strengths of these two modeling techniques, we aim to produce more robust distribution models.
We are currently preparing for GAP modeling efforts by:
• Preparing a list of target species to model
• Preparing preliminary range maps at the watershed scale
• Gathering occurrence data for distribution modeling efforts
• Assembling teams of experts to help with the review process
More detailed information about our project objectives and methods are provided in the Alaska Gap Analysis Work Plan.
This map shows the results of an inductive model generated for the Solitary Sandpiper: 48 records were used to develop the model (black circles), 14 (about 22% of the records) were used to test the model (open circles). Predicted distribution is shown in green. Range maps (blue line) will be used to delimit the extent of the modeled distribution.