Toward a GIS Model of Amphibian Diversity, Distribution and Density in Amazonia: Work from Yasun Nat

1
(No Transcript)
2
Toward a GIS Model of Amphibian Diversity,
Distribution and Density in Amazonia Work from
Yasuní National Park
Paul Herbertson MSc., Kings College
London TADPOLE Board Member

• Shawn McCracken
• Texas State University
• TADPOLE
• President and Founder

Dr. Mark Mulligan Kings College London Advisor
Dr. Michael Forstner Texas State
University Advisor
3
TADPOLE An Organization Committed to the Research
and Preservation of Amphibians in Amazonia

• Our Mission
• Contribute to mapping
• diversity, distribution, and
• density of amphibians
• throughout Amazonia

4
TADPOLE An Organization Committed to the Research
and Preservation of Amphibians in Amazonia

• Our Mission
• To foster collaborative
• relationships and research
• among institutions, project
• directors and organizations

5
TADPOLE An Organization Committed to the Research
and Preservation of Amphibians in Amazonia

• Achieving Our Goal
• Innovative uses of GIS,
• remote sensing and field
• surveys to model and
• monitor amphibian
• diversity, distribution
• and density

6
Amphibians and the Amazon

• Global declines

• Potential Causes
• Habitat loss
• Contamination and
• pollution
• Climate change
• UV radiation
• Disease
• Chytridiomycosis
• Predation by
• introduced species

7
Amphibians and the Amazon

• Global declines

• Potential Causes
• Habitat loss
• Contamination and
• pollution
• Climate change
• UV radiation
• Disease
• Chytridiomycosis
• Predation by
• introduced species

8
Amphibians and the Amazon

• Global context

• Amazonia has one of the
• highest levels of diversity
• and density (Duellman 1999)
• Local (alpha) diversity is
• greatest in the upper
• Amazon Basin (Duellman 1999)
• Distributions and densities
• poorly known

9
Amphibians and the Amazon

• Yasuní National Park

• gt100 species identified at
• Tiputini Biodiversity
• Station
• gt90 species identified at
• Yasuní Research Station
• Further investigations are
• required to qualify and
• quantify potential impacts
• of direct and indirect
• anthropogenic threats to
• amphibians

10
Baseline Data Acquisition

• Field work - 2002, 2003 and 2004
• Five linear plots along existing trail system
• Quadrat and bromeliad patch sampling

11
Baseline Data Acquisition

• Over 600 specimens
• processed/photographed
• Meta data recorded
• Blood/tissue samples of
• 70 species of herps
• Audio recordings of
• 39 anuran species

• 35 quadrats in each of
• 5 plots surveyed
• 175 quadrats
• 2 trees in each of 4 plots, 5 bromeliads per tree
  40 bromeliads

Epipedobates bilinguis
12
Baseline Data Acquisition

• GIS, remote sensing and modelling

13
(No Transcript)
14
Pilot Study on Modelling Amphibian Distributions,
Density and Diversity

• Investigate environmental correlates of amphibian
  distribution

Dendrobates ventrimaculatus
15
Rationale

• CI global gap survey
• identified amphibians
• as the most deficient in
• distribution data of all
• taxa

• Lack of knowledge on amphibian distributions

Model 1 Model 2 Model 3 Mammals 6.3 7.2
4.8 Amphibians 22.6 27.7 14.7 Threaten
ed Amphibians 39.2 48.0 27.4
16
Rationale

• Lack of knowledge on amphibians
• Patterns of distribution can aid more general
• ecological understanding

• Amphibians as sensitive
• indicator species
• Understanding why
• certain areas are more
• diverse than others
• Understanding
• relationships between
• environmental variables
• and species distributions

17
Rationale

• Lack of knowledge on amphibians
• Patterns of distribution can aid more general
• ecological understanding
• Prioritization of areas for conservation

• Upscaling through
• distribution models
• can be used to move
• from the plot to the
• more policy-relevant
• regional scale
• The remoteness of some
• sites make field work
• difficult
• Reduces costs
• Reduces time

18
Rationale

• Lack of knowledge on amphibians
• Patterns of distribution can aid more general
• ecological understanding
• Prioritization of areas for conservation
• Increase awareness of threats to Amazonia

• GIS and Remote
• Sensing are very
• visual.
• Show complex
• interactions much
• more simply
• Scenario based
• modelling

19
The TOPMODEL Wetness Index
A
TOPMODln(A/tan(ß))
ß
TBS upslope area (A)
TBS slope (ß)
TBS TOPMOD
20
GIS Methodology

• Quadrats occur across the full range of wetness
  index classes and sampling is in proportion to
  the area in each class

21
Distribution Modelling Methodology

• Individual species distribution modelling

• Mahalanobis statistic uses
• association between
• landscape features and
• individual observations
• to produce a map of
• probability of occurrence
• Extension for ArcView

22
Diversity and Density Modelling

• Model amphibian
• diversity and density
• using the topographical
• wetness index

Catchment of Tiputini River
TBS
P. N. YASUNI
23
Model Development

• Empirical model development/parameterisation

R-sq for density 90.4
R-sq for diversity 92.6
24
Model Validation
Scattergram of observed vs predicted density

• Selection of 20
• of data across full range
• of WI classes. These
• data are not used in
• model parameterization
• but in validation,
• showing significant
• correlation between
• model-predicted and
• observed amphibian
• densities.

Predicted density (Individuals/100m2)
Observed density (Individuals/100m2)
25
Results

• Density model

Individuals/100m2
26
Results

• Diversity model

Species/100m2
27
An Example ApplicationDeforestation along roads
28
Results

• In a best case
• scenario road-related
• land use change
• , by 2030, 3.2 of
• the modelled
• amphibian
• population will
• be directly
• affected
• In a worse case
• scenario, by 2030
• 7.2 will have
• been directly
• affected

Best Case Scenario for 2030 (existing roads only)
Worse Case Scenario for 2030 (existing roads
only)
29
Modelling Conclusions

• Good relationship between the topographical
  wetness classes and the density of amphibians at
  TBS
• Good relationship between the wetness classes and
  the diversity of amphibians at TBS
• Validation data could be better
• Topographical wetness index is a useful approach
  to the poorly studied distribution mapping of
  amphibians
• Future model will incorporate other variables

30
Conclusion

• Use of GIS and Remote Sensed
• data is a useful tool to
• investigate the anthropogenic
• effects on amphibian status
• More work is NEEDED!

• Amphibians are suffering
• global declines
• Amazonia has one of the
• highest amphibian diversities
• Ecological status of
• amphibians in Yasuní is
• poorly documented

31
Questions?
Bufo margaritifer
32
HERB Project, Ecuador 1999-2004 http//www.kcl.ac.
uk/herb

• A long term collaborative research project,
  largely carried out through
• the efforts of PhD and MSc students.
• Investigating environmental change in neotropical
  montane and
• lowland forests.
• Using a combination of short-term intensive field
  campaigns,
• long term field environmental monitoring, GIS,
  remote sensing
• and process-based ecological modelling.
• In Yasuni (at USFQ TBS and throughout the region)
  
• Developing a climatic and environmental baseline
  dataset.
• Developing RS based methodologies for rapid
  biodiversity assessment at
• scales from the plot to the region.
• Modelling land use change and impacts.
• Modelling climate change and impacts.
• Modelling forest sensitivity to environmental
  change.
• Developing tools for minimising the impacts of
  oil-related activities,
• especially pipeline leakage.