Transcript No Slide Title

A Simulation model for Florida Panthers and white-tailed
deer in the Everglades and Big Cypress Landscape
ATLSS and the Planning
Process for Everglades
Restoration
E. JANE COMISKEY and LOUIS J. GROSS
The Institute for Environmental Modeling,
Department of Ecology and Evolutionary Biology
University of Tennessee, Knoxville, TN 37996-1610
In 1992 Congress authorized the Comprehensive Review Study
(Restudy) of the Central and Southern Florida Project to
develop modifications to the current drainage system in order
to restore the Everglades and Florida Bay ecosystems while
providing for the other water-related needs of the region.
The resulting plan was submitted to Congress in July, 1999.
The ATLSS Hierarchy of Models
Abstract
The White-tailed Deer/Florida Panther Model was developed as one
component of the Across Trophic Level System Simulation (ATLSS)
family of models, with the overall goal of providing a total ecosystem
approach to predicting the relative effects of restoration plans on the
biotic components of the natural systems of South Florida. The
Deer/Panther model was designed as a management and
evaluation tool to analyze the relative effects of alternative water
management scenarios on long-term population dynamics of the
endangered Florida panther (Felis concolor coryi) and its primary
prey species, the white-tailed deer (Odocoileus virginianus.)
The Deer/Panther model is a spatially-explicit, individual-based
model which divides the landscape into a spatial grid of cells and
tracks movements and behaviors of individual members of simulated
deer and panther populations within this spatial grid. Each individual
is described by a set of state variables which record information such
as gender, age, body weight, reproductive status and current
location on the landscape grid. Decision rules determine how
individual animals move across the landscape, interact with one
another and respond to their environment. Differences in individual
states and stochastic components in decision-making produce
variations in responses to local conditions of environmental variables
such as water depth, habitat characteristics, and proximity to other
animals. Individuals mate at different times, disperse varying
distances from their natal ranges, and gain or lose weight depending
on local forage and prey availability, individual bioenergetic status,
and stochastic factors
The four major interacting model components are hydrology,
vegetation, deer, and panthers. Daily water level inputs, provided by
scenario simulations of the ATLSS High Resolution Hydrology Model
derived from the South Florida Water Management Hydrology Model,
influence vegetation growth and restrict movement for deer and
panthers. The vegetation component provides spatial and temporal
variation in available forage for the deer herd. Deer selectively
consume forage biomass and serve as prey for panthers. Panther
predation is reflected through deer mortality. Two spatial scales,
corresponding to strategic vs. tactical decision-making, are used in
the model. The size of spatial cells for simulated animal movement is
500-m x 500-m; biomass generation and foraging behavior occur in
100-m x 100-m cells. Model events are simulated on a daily time
step. The Deer/Panther Model is coupled to PV-WAVE, a software
application for visual representation and statistical analysis of data.
Model outputs can be displayed in a variety of spatial
representations as well as using summary statistics over the entire
model area or over specified subregions.
The Plan includes:
Individual-Based
Models
Age/Size Structured
Models
Cape Sable
Seaside Sparrow
Snail Kite
White-tailed Deer
Wading Birds
Florida Panther
Fish Functional Groups
Alligators
Radio-telemetry
Tracking Tools
•Reconnecting over 80 percent of the remaining Everglades by
removing over 240 miles of internal levees and canals.
Reptiles and Amphibians
•Reducing the average of 1.7 billion gallons of fresh water discharged
every day to the ocean
Linked Cell
Models
Lower Trophic Level Components
•Additional land purchases of 47,000 acres as an addition to
Everglades National Park
Vegetation
Process Models
Spatially-Explicit
Species Index Models
Cape Sable
Seaside Sparrow
Long-legged
Wading Birds
Short-legged
Wading Birds
White-tailed Deer
Abiotic Conditions
Models
High Resolution Topography
How ATLSS was used during the Restudy:
Alligators
Snail Kite
High Resolution Hydrology
1. A hydrologic alternative plan for the next 31 years was
produced by the South Florida Water Management District (SFWMD)
using their management model, based upon historical rainfall
and assuming changes in control structures that occur all at once.
Disturbance
© TIEM / University of Tennessee 1999
Shown in red are the ATLSS components which are used for Deer-Panther Model
2. Major hydrologic results from the plan were posted on the Web
and detailed results were provided to ATLSS staff.
3. The ATLSS staff ran the collection of ecological models for
the alternative, compared to the base plan (F2050) which
assumed only control structure changes currently planned to
be implemented by 2050. Within a week, results from these were
bound and shipped overnight to 14 agency representatives, as
well as being placed on the Web.
4. A few days later the Alternative Evaluation Team, consisting
of representatives from interested agencies, met to evaluate
the alternative and make suggestions for additional changes,
based upon ATLSS results, hydrologic performance measures,
and their agencies mandate.
5. Suggestions from the AET were sent to the Alternative
Design Team, consisting mostly of hydrologists, who then
revised the plan and ran a new hydrologic scenario.
6. Go back to step 2 and repeat many times.
There were over 15 major iterations of the above, each
iteration taking approximately 3 weeks, from September 1997
to January 1999. Additional plans taking account of phasing of
structure changes are still ongoing.
ATLSS Models are ALWAYS used in a relative
assessment framework. We do not claim these
are predictive models. Rather, they allow a
relative ranking of the effects of alternative
scenarios on the various species modeled.
Inputs to Deer/Panther Model
•Government agencies and programs provide spatial representations of the
landscape including vegetation, land use, and elevation
Expanded ATLSS
Model Area
Florida Gap Analysis
Habitat Map v. 6.6
•Field researchers provide historical observations of breeding, foraging behavior,
and abundances as affected by hydrology
•Hydrology planners provide scenarios for future ponding over 31 years at a 2-mile
by 2-mile resolution
Work in Progress
Using empirical panther data to define panther behavior rules:
habitat use and habitat/homerange associations
Deer/Panther model development efforts, deferred while the extensive Restudy
scenario evaluations were being produced during 1998 and 1999, are ongoing to
expand the model area and enhance the realism of the model.
The Deer/Panther model was designed to be used as a management and evaluation
tool to analyze the relative effects of alternative water management scenarios on
long-term population dynamics of the Florida panthers and their primary prey species,
white-tailed deer. Originally conceived as one component of a group of
interconnected ATLSS models and driven by an improved and expanded SFWMD
Hydrology Model, the model's projected functionality is somewhat compromised by
the necessary modifications of original plans which resulted in an indefinite delay in
provision of expanded hydrology inputs from the District and postponement of
development of a vegetation succession model.
The following efforts have been undertaken to make Deer/Panther simulations more
realistic and computationally efficient:
Panther telemetry: 1999
Recent panther locations
• Analyses of radio telemetry data and review of published literature about panthers in
South Florida led to the realization that population dynamics could not be realistically
modeled within the the original ATLSS model area, restricted by availability of
hydrologic inputs from the SFWMD Hydrology Model. An expanded model area is
being provided by methods described elsewhere in this document.
• A stable version of the FGAP vegetation map (Version 6.6) has recently become
available, and the extensive work required to adapt this new and substantially
modified version for use in Deer/Panther is near completion. The vegetation
component of Deer/Panther, which provides spatial and temporal variation in
available forage for the deer herd, is being modified to provide realistic levels of Type
I, II, and III forage based on the FGAP Version 6.6 vegetation types, and deer and
panther behaviors which are influenced by habitat type are likewise being modified to
account for the habitat map change.
• The most recent version of the evolving ATLSS landscape classes has been
incorporated into Deer/Panther, dramatically reducing the run-time required for
simulations. Final timings and comparisons to observational data have been deferred
until incorporation of the new habitat map has been completed.
Panther locations over
forested habitat
Habitat edge metric
Some References on ATLSS
ATLSS Home Page http://atlss.org/
Comiskey, E.J., L.J. Gross, D.M. Fleming, M.A. Huston, O.L. Bass, H.K. Luh,
and Y. Wu. 1994. A Spatially-explicit Individual-based Simulation Model
for Florida Panther and White-tailed Deer in the Everglades and Big
Cypress Landscapes. In D. Jordan, ed. Panther Conference Proceedings,
Ft. Myers, Florida, U.S. Fish and Wildlife Service.
Curnutt, J. L., E. J. Comiskey, M. P. Nott and L. J. Gross. 2000. Landscape-based
spatially-explicit species index models for Everglades restoration. Ecological
Applications 10: 1849-1860.
DeAngelis, D. L., L. J. Gross, M. A. Huston, W. F. Wolff, D. M. Fleming,
E. J. Comiskey, S. M. Sylvester. 1998. Landscape Modeling for Everglades
Ecosystem Restoration. Ecosystems 1:64-75.
DeAngelis, D. L., L. J. Gross, W. F. Wolff, D. M. Fleming, M. P. Nott and E. J.
Comiskey. 2000. Individual-based models on the landscape: applications
to the Everglades. P. 199-211 in J. Sanderson and L. D. Harris (eds.),
Landscape Ecology: A Top-Down Approach. Lewis Publishers, Boca Raton,
FL.
Duke-Sylvester, S. and L. J. Gross. 2000. Integrating spatial data into an
agent-based modeling system: ideas and lessons from the development
of the Across Trophic Level System Simulation (ATLSS). To appear in:
Agent-based models and GIS, R. Gimblett (ed.), Oxford University Press.
Gaff, H., D. L. DeAngelis, L. J. Gross, R. Salinas and M. Shorrosh.
2000. A dynamic landscape model for fish in the Everglades
and its application to restoration. Ecological Modelling 127:33-52.
• The predictive capabilities of an individual-based model is closely tied to the realism
of the decision rules which determine how individual animals move across the
landscape, interact with one another and respond to their environment. The definition
of these rules is in turn tied to the availability and interpretation of empirical
observations about these behaviors and movement patterns. Recent developments in
panther recovery have contradicted and confounded published theories and opinions
about panther ecology in South Florida, including ideas about habitat use, home
range establishment, female fecundity, kitten survival, and dispersal patterns. For
example, panther literature has reflected the commonly held perception that Big
Cypress south of I-75 could not sustain a breeding population of panthers because of
thinning forest cover and poor soils. However, the largest population expansion over
the past two years has taken place in this area. With this in mind, the telemetry data
set of panther location observations (1981-1995) on which preliminary panther
behavior rules for the Deer/Panther model were based has been expanded to include
observations through December 1999 so that recent trends can be reflected in revised
behavior rules. Observations from January-July 2000 will be incorporated when data
are made available.
Work will begin soon on an ATLSS model of vegetation succession in South Florida,
which will enhance the predictive capabilities of Deer/Panther as well as many of the
other ATLSS models.
ATLSS Panther Modeling
Dr. Louis J. Gross
Jane Comiskey
Scott M. Duke-Sylvester
The Institute for Environmental Modeling
University of Tennessee
Dr. Donald DeAngelis
University of Miami
Oron L. Bass
South Florida Natural Resources Center
Everglades National Park
ATLSS Collaborators and Data Providers
Big Cypress National Preserve:
Deborah Jansen, Wildlife Biologist, National Park Service
The Deer/Panther modeling effort ongoing at the University of Tennessee has been
supported by the U. S. Geological Survey, Biological Resources Division, through
Cooperative Agreement #1445-CA09-95-0094.
Florida Fish and Wildlife Conservation Commission:
Darrell Land, Florida Panther Recovery Project Leader
Roy McBride, Florida Panther Capture Team