Modeling and Visualizing Species Movement Presented at: NASA Joint Science Workshop on Biodiversity, Terrestrial Ecology, and Applied Science College Park, Maryland, August 21-25, 2006

1
Modeling and Visualizing Species
MovementPresented at NASA Joint Science
Workshop onBiodiversity, Terrestrial Ecology,
and Applied ScienceCollege Park, Maryland,
August 21-25, 2006
Fred Watson1 Simon Cornish1 Bob Garrott2 PJ
White3 Rick Wallen3 Susan Alexander1 Wendi
Newman1 Thor Anderson1 Jon Detka1 Jason
Bruggeman2
1California State University Monterey
Bay 2Montana State University Bozeman3National
Park Service
Funding NASA NCC2-1186 NCC13-03009, NSF DEB
0074444 DEB-0413570, and NPS
2
National Parks Servicebison monitoring
management activities
3
A selection of productsin the pipeline

• Landscape visualization kiosk
• Canyon Visitor Education Center
• Real-time snowpack modeling
• Information that supports bison management at the
  boundary
• Landscape inputs to wildlife studies
• The hard science upon which policy is based
• A model for predicting species movement
• Potential contributions to any application where
  species move
• biodiversity
• invasive species
• ecological forecasting
• disease vectors

4
Landscape visualization kiosk
5
(No Transcript)
6
Real-time snowpack modeling
7
Landscape model (snowpack)helps inform decision
on when to release bison
Optimal time for release?
8
The cold face of collaboration
9
Landscape inputs to wildlife studies

• Pr( Use of location by species ) f (
  Landscape covariates,,,Temporal covariates,,, )

10
Landscape inputs to wildlife studies
Probability of bison corridor travel(Bruggeman
et al., submitted)
11
(No Transcript)
12
Some false positives
Probability of bison corridor travel(Bruggeman
et al., submitted)
13
Bison distribution landscape covariates
Vegetation
Slope
Snowpack
14
Bison utilization distributionHabitat-selection
analysisusing standard Resource Selection
Functions (RSFs)
15
Two paradigms

1. Pr( Use ) f ( Landscape covariates )
2. Pr( Use ) f (Distance to previous locations)

16
Two paradigms
17
A new approachSelective Computational Diffusion
(SCD)

• Pr( Use here )
• Pr( Use next door ) f ( Landscape here
  ) Iterated many times (computationally
  demanding)

18
Comparison of approaches
19
Taking SCD out on the road(so to speak!)---
Observed movement--- Predicted movement