Modeling and
Mapping Reptile Distributions on the Idaho National Laboratory
Site
Background
This study was designed for the purpose of understanding
factors affecting reptile distribution and to make predictive
distribution maps for individual species across the development
zone (central 259 km2) of the INL Site. This
information will be used to help develop the conservation
management plan for the INL Site, which will help the Department
of Energy make decisions about future facility locations.
Objectives
The main objective of this project is to assist in
development of a conservation management plan that will be used
for future facility siting decisions: Specific objectives for the
2007 season included:
Develop and apply a habitat-based sampling design for
reptiles
Determine the occurrence, distribution, and habitat
relationships of reptiles on the development zone
Develop a habitat model for each reptile species in the
development zone
Make and test predicted distribution maps.
Accomplishments Through
2007
Reptile data collection was completed and
modeling procedures for six reptile species on the INL Site
were developed. Three distribution modeling techniques were
also completed: 1) Boolean Modeling, 2) Trapping and
observational probability modeling, and 3) Mahalanobis
Distance Modeling.
Results
Incidental observation was the sampling
technique that provided most reptile observations.
However, skinks were not detected using this method.
Visual Encounter Surveys produced the
second highest number of reptile observations. All six
species were detected.
Trapping detected all species but in few
numbers. However, the use of traps was required to
sample for night snakes, which potentially occur in the
study area.
Distribution Modeling. Boolean Model.
Uses all positive data to determine in which environmental
types each species occurs, then map all suitable
environmental type polygons
Trapping / Observational Probability
Model. Uses all positive and negative trapping and
visual encounter survey data to calculate a probability of
trapping or observing a particular species in a particular
environmental type.
Mahalanobis Distance Model. Uses all
positive data to create a habitat similarity index based on
the characteristics of pixels in the GIS layers for sites
where a species of interest is known to occur.
Model Ranking. We used our sample
sizes, statistical analyses, and knowledge of species
ecology to rank each model for each species. We also used
this information to determine our relative confidence for the
highest ranked model for each species.
Species Richness
The species richness map was made by
overlapping all of the boolean distribution model
results.
The number of species within the
Development Zone varied from 2 to 6.
The area with the highest reptile
species richest is located on the southern end of the L
shaped corridor where development is most likely to
occur.
The highest species richness areas are
characterized by big sagebrush and no recent burns.
Plans for Continuation
Additional modeling approaches will be tried
(e.g., DOMAIN and Maximum Entropy).
Publications, Theses, Reports, etc.
Thesis and publications are in progress.
Investigators and Affiliations
David P. Hilliard, Graduate
Student, Herpetology Laboratory, Department of Biological
Sciences, Idaho State University, Pocatello, Idaho
Charles R. Peterson,
Professor, Herpetology Laboratory, Department of Biological
Sciences, Idaho State University, Pocatello, Idaho
Funding Sources
U.S.
Department of Energy, Idaho Operations Office.
