Transcript Pickens Longleaf Partnerships March 2016x

Identification of Restoration Sites for a Firedependent Bird in an Urbanizing Environment
Bradley A. Pickens
North Carolina Cooperative Fish and Wildlife Research Unit, NCSU
Landscape Ecology Solutions, L.L.C.
Project Coauthors:
Jeffrey F. Marcus
The Nature Conservancy
Paul J. Taillie
North Carolina State University
John P. Carpenter, Scott Anderson
North Carolina Wildlife Resources Commission
Jaime A. Collazo
U.S. Geological Survey, North Carolina Cooperative Fish and Wildlife
Research Unit, North Carolina State University
South Atlantic Landscape Conservation
Cooperative
 Indicator-driven approach
to conservation planning
 Does not explicitly
incorporate urbanization or
restoration potential...case
studies needed
Pine and Prairie Ecosystem
Bachman’s Sparrow as an indicator:
 Responsive to herbaceous ground cover, low basal
area, regular burning, habitat fragmentation
 Known to correlate with other species of concern
 Distribution provides insight into the effect of
urbanization and restoration potential
Objectives
1) Develop species distribution model for Bachman's
Sparrow based on species–habitat relationships
2) Determine effects of urban growth (2010-2090)
3) Develop scenarios to identify potential restoration
sites and determine interaction of restoration and
urban growth by 2050, including projections of patch
size and isolation
Modeling Overview
+
On-the-ground
bird data
Environmental
spatial data
Species distribution
model defines habitat
Statistics!
Validation
Restoration
scenarios
Urbanization
Final species
Calculate percent
distribution
habitat within 1, 3,
model
5, 7 km radii
Statistics!
Presence Data
Bachman’s sparrow presence locations synthesized from
North Carolina:
 North Carolina Wildlife Resources Commission
(n = 544)
 North Carolina State University (n = 99)
- Paul Taillie, Christopher Moorman, M. Nils Piels
 Virginia Tech / Camp LeJeune (n = 84)
- Jeff Walters
Modeling Methods
 4 methodologies + heterogeneous detection =
presence-only modeling
 Resource Selection Function: Use vs. Availability
- 60% of data for training / 40% for validation
 727 pseudo-absences placed within extent of survey
data. Excluded urban developed and row crops
Environmental Variables
A) LANDFIRE- fire 1 km resolution
B) Mean canopy cover- 300x300 m, 900x900 m (NLCD)
C) SD canopy cover- 300x300 m, 900x900 m (NLCD)
D) Evergreen land cover- 300x300 m, 900x900 m (NLCD)
E) Row crop agriculture (NLCD)- 3 km
F) Urbanization (SLEUTH)- 3 km
Urban Scenarios
1) Examine urban‒fire relationship
2) Urban growth effects examined for 2010 ‒ 2090
(Terando et al. 2014)– assumed Bachman’s Sparrow are
excluded by urbanization
3) Assumption! Current predicted
Bachman’s sparrow distribution is
sustained as habitat (i.e., continues
to be burned)
Restoration Scenarios
1) Canopy cover reduction by 30%
2) Prescribed fire everywhere except urban and row crop
agriculture
3) Canopy cover reduction
and prescribed fire
Fire-suppressed
pine forest
Longleaf pine
savanna
Habitat Relationships
 Localized
canopy gaps
 Moderate
canopy cover
 Moderate
heterogeneity
Habitat Relationships
Validation
 Predicted absences were 90% correct
 Predicted presences were 87% correct
Camp
Lejeune
Holly Shelter
Game Land
Urban Effects
 80,872 ha of predicted distribution on protected lands
No buffer
600 m buffer
~ fire
suppression
Current Prediction
Urban Growth by 2050
Urban Growth by 2050 and Fire
Urban Growth by 2050 and Canopy
Cover Reduction
The Future
Scenario
Mean
Mean patch
# of
patch size
isolation
patches
(ha)
(m)
Current condition
6,586
26
414
No restoration:
2050
4,097
33
537
Prescribed burn:
2050
6,508
31
423
Canopy cover
reduction: 2050
10,009
50
291
Important Conclusions
 Urban effect is projected to consolidate
Bachman’s sparrow habitat; high loss of
fragmented patches
 Canopy cover reduction scenarios resulted in 2x
the habitat
 Restoration has potential to create larger patch
sizes with less isolation compared to current
conditions
 Urban growth was problematic for fire scenario,
but not canopy cover reduction scenario
Acknowledgements
 Funding provided by the U.S. Fish and Wildlife
Service in support of the South Atlantic Landscape
Conservation Cooperative
 Data sources: North Carolina Wildlife Resources
Commission; North Carolina State University (Paul
Taillie, Christopher Moorman, M. Nils Piels); Virginia
Tech/Camp LeJeune (Jeff Walters)
QUESTIONS?
http://salcc.databasin.org
Brad Pickens,
[email protected]
Wildland‒Fire Relationship
 Based on LANDFIRE data (100 m resolution)