Conference PaperPDF Available

Distribution and connectivity of Sierra Nevada red fox in the Oregon Cascades

  • Wildlife Ecology Institute
  • Cascades Carnivore Project
Distribution and Connectivity of Sierra Nevada Red Fox in the Oregon Cascades
Cate Quinn1, Tim Hiller2, Jaime McFadden-Hiller3, Jocelyn Akins4, Ben Sacks1
Sierra Nevada Red Fox
Native red fox in the western US are primarily montane.
Since the early 1900s montane red fox populations have suffered
declines. The Sierra Nevada red fox (Vulpes vulpes necator; SNRF) is
the most imperiled subspecies, and was recently evaluated for listing
under the Endangered Species Act.
Historically the SNRF designation was limited to California. More
current genetic data showed that the Columbia River Gorge has been a
long-time barrier to gene flow between the Washington and Oregon
Cascades, and as a result Oregon red fox were reclassified with
California populations as SNRF.
Little is known about the status of the SNRF in Oregon. How many are
there? How connected are populations? Where do we find them?
Obtaining this basic information on red fox in the
Oregon Cascades is vital to development of a
comprehensive conservation strategy for SNRF
Use occurrence data to build predictive distribution model to guide
future surveys.
Use genetic data to qualitatively assess population size and
connectivity among known sighting areas.
Step 1 – Build Model Key Findings
Contemporary connectivity does not appear to be a limiting factor in
the Oregon Cascades. Both Maxent model and genetic structure
support capacity for gene flow.
Genetic data show two distinct populations: a southern population
with low genetic diversity, and a northern population with markedly
higher genetic diversity. A possible contact zone occurs at the
latitude of Bend, where admixed individuals are concentrated.
Implications & Future Work
This pattern suggests the following tentative conclusions:
The southern population may represent a historically isolated remnant of
the montane lineage, similarly bottlenecked as populations in California. In
this case, the low genetic diversity and effective population size make it
vulnerable to extirpation, and a high priority for protection.
The higher genetic diversity of northern population indicates a larger
and/or more connected population. Two scenarios could be responsible:
“Ancient Ties”: The northern population could be part of a larger,
more connected population. For example, it may have shared more
recent ancestry with the Wallowa and Rocky Mountain than Sierra
Nevada lineages. Even absent of contemporary gene flow, the larger
genetic effective population would cause diversity to decline at a slower
“Traveling Wave”: Contemporary gene flow following recent
secondary contact could have resulted in a southward-moving wave of
introgression. Potential source populations include native montane red
fox from the Wallowa and Blue Mountain ranges, or introduced lowland
red fox from human-dominated valleys to the east and west.
These findings are tentative and require a larger sample for confirmation.
To resolve uncertainties, next steps are to a) gather more individual
genotypes from the Oregon Cascades, with particular focus on sampling
gaps, b) compare the Oregon Cascade genotypes to reference samples
from eastern Oregon, Idaho, and known nonnative lowland populations,
and c) compare modern Oregon mitogenome sequences to those from
historical specimen from the Sierra Nevada and Rocky Mountains.
Collaborators who provided red fox observations and
Willamette National Forest, J. Doerr , C. Ferland, & R. Seitz
Deschutes National Forest, L. Turner & M. Gregg
Rogue River-Siskiyou National Forest, Shelia Colyer
Mt. Hood National Forest, A. Dyck
Region 6 US Forest Service, J. Chapman
High Desert Museum, J. Nelson
Crater Lake National Park, S. Mohren & M. Immel
Oregon Department of Fish and Wildlife, C. Heath
Cascade Carnivore Project, Cascadia Wild
Oregon State University, D. Gumtow-Farrior
Contact me
We correlated red fox observations with environmental variables
to build a presence-only distribution model in Maxent. A total of
109 genetic and photographic detections were rarified to 33
observations >4 km apart and background points were sampled
within a 20-km radius of fox detections. A stepwise variable
removal process was used to select the best model from the 9
least correlated of 38 possible variables. The model with the
lowest AICc, shown here, used 4 variables: vegetation, minimum
temperature in January, ruggedness, and isothermality.
Structure plot with individuals (vertical bars) ordered from south to north. The
number of clusters with the most support was K=2. Nineteen of 22 individuals
had q-values >0.75, suggesting moderate to strong population differentiation
along a north-south axis.
The Groundwork
Step 3 – Relate Genetics To Space
We split the global population by the
proportion of ancestry (q) estimated in
Structure. The southern population had
significantly lower expected and observed
heterozygosity relative to the northern
population. Genetic effective population sizes
(Ne) were estimated from linkage
disequilibrium and showed a similar
increasing trend with latitude (above).
Structure results are represented on map as
pie charts of q-values (right). The genetic
break indicated by Structure does not
coincide with a break in predicted suitable
habitat. The lack of correlation suggests
dispersal is not the limiting factoring causing
genetic differentiation.
population size (Ne)
(Ho, He)
Over the last five
years, a collaborative
effort by Federal,
State, Tribal, UC
Davis, and nonprofit
organizations have
advanced field
surveys for SNRF
over the extent of the
Cascade range.
Red fox detections
from 2011 include
photographs from
systematic and
opportunistic camera
surveys, and fecal
DNA collected through
carnivore scat
Figure shows the Oregon study area in relation to the
historical California range and known SNRF populations
at Lassen Peak and the central Sierra Nevada.
Photographic ( ) and DNA ( ) red fox detections in
Oregon occurred 2011 2016. Scats collected from other
carnivores ( ) indicate search effort.
Step 2 – Genotype Individuals
Ancestry Proportion (q)
(1) Mammalian Ecology and Conservation Unit, Veterinary Genetics Laboratory, UC Davis, Davis, CA; (2) Wildlife Ecology Institute, Starkville, MS;
(3) Department of Geosciences, Mississippi State University, MS; (4) Cascade Carnivore, Hood River, OR
Robert Campbell, Mt. Bachelor
We genotyped 66 red fox
samples (fecal, hair, and
tissue DNA from roadkills)
at 33 nuclear microsatellites
and a sex marker, yielding
37 usable genotypes that
matched to 22 unique
We estimated standard
genetic diversity indices,
effective population size
from linkage disequilibrium
(below), and evaluated
genetic differentiation using
a Bayesian clustering
approach implemented in
the program Structure
Oregon Department of Fish and Wildlife
Willamette & Deschutes National Forests
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.