Transcript One if by land, two if by sea - Natural Resource Ecology Laboratory

“One if by land, two if by sea. . .”
Documenting, Mapping, and Predicting the Invasion
of Non-native Plants, Animals, and Diseases in the United States.
Tom Stohlgren, USGS
Midcontinent Ecological Science Center, Fort Collins, CO
and Research Team
Geneva Chong and Catherine Crosier (USGS)
Mohammed Kalkhan, Robin Reich,
Dave Barnett, Sara Simonson,
and Rick Shory (CSU)
Mike Ielmini (USFWS), John Schnase and
Jim Smith (NASA), Pam Fuller, Josh Dein, John Sauer,
Carl Korcshgen, Linda Leake, Doug Posson,
Anne Frondorf, Tom Muir, Bill Gregg, Sue Haseltine,
Tom Owen, (USGS), John Kartesz (UNC), Jim Quinn
(UCD), Ken Stolte (USFS) and many more.
Invasive Species: The Top
Environmental Issue of the 21st Century
• Economic costs
($138 Billion/year).
• Environmental costs
(40% of Threatened and
Endangered Species, many
native species declines).
• Human-health costs (West
Nile Virus, Aids, malaria,
others on the way).
• Increased unintentional
spread, or threat of
ecological terrorism (hoofand-mouth, mad cow
disease, crop pathogens).
Notorious examples include Dutch
elm disease, chestnut blight, and
purple loosestrife in the northeast;
kudzu, Brazilian peppertree, water
hyacinth, nutria, and fire ants in the
southeast; zebra mussels, leafy
spurge, and Asian long-horn beetles
in the Midwest; salt cedar, Russian
olive, and Africanized bees in the
southwest; yellow star thistle,
European wild oats, oak wilt disease,
Asian clams, and white pine blister
rust in California; cheatgrass, various
knapweeds and thistles in the Great
Basin; whirling disease of salmonids
in the northwest; hundreds of invasive
species from microbes to mammals in
Hawaii; and the brown tree snake in
Guam. Hundreds new each year!
Why Us, Why Now?
• There are many data collectors, but few scientists
who specialize in data synthesis and predictive
modeling at multiple scales.
• The USGS, with the cooperation of many partners,
is uniquely qualified to lead invasive species
research that integrates species traits, vulnerability
of populations and habitats to invasion, early
detection, risk analysis, and predictive models for
“ecological forecasting.”
• There is extreme urgency at local, regional, and
national scales – the invasion is not only underway,
it is accelerating and we’re unprepared.
Why Us, Why Right Now?
• Our expanded research team was tired of writing “white
papers, budget initiatives, and progress reports on “case
studies.” We want to accomplish much more!
• Clients (USFWS, BLM, USFS, NPS, states) demanded that
we synthesis data, design new surveys and monitoring
methods, and rapidly develop predictive models for better
early detection and control of many invasive species.
• We approached several colleagues to begin a “data
cooperative” of sorts – the first nation-wide collection of
data on non-native plants, animals, and diseases integrated
with new capabilities for the predictive modeling of species,
populations, and habitats at multiple scales.
• The response has been incredible! We must seize the
moment!
On the Policy Front:
• The U.S. is beginning the
development of the
“Implementation Plan for
the National Invasive
Species Management
Plan.”
• APHIS is suggesting
strong policy changes
regarding the import of
plants and animals.
• There is increasing
awareness of the effects of
rapid biological invasions.
“Needed: A National Center
For Biological Invasions
By Don Schmitz and Dan
Simberloff” Issues in Science
And Technology Summer 2001
DEPARTMENT OF AGRICULTURE
Animal and Plant Health Inspection
Service
7 CFR Part 330
[Docket No. 95-095-2]
RIN 05789-AA80
Plant Pest Regulations; Update of
Current Provisions
AGENCY: Animal and Plant Health
Inspection Service, USDA.
ACTION: Proposed Rule.
On the Science Front:
• Better survey and
monitoring techniques
have been developed.
(multi-phase, multiscale, nested-intensity
designs).
• Better modeling
techniques have been
developed.
• More access and uses
of high performance
computing
capabilities (Beowulf
clusters,
supercomputers,
leased power).
• Fewer barriers exist to
sharing data.
40
40
Native Species
30
30
50
30
40
20
0
20
50
40
10
10
1
0.
YCOORD
% Soil N
30
0
0
10
20
30
40
40
1
2
20
30
40
50
% Soil Clay
3
33
2
2
20
20
5
4
10
10
20
30
XCOORD
10
10
0
0
10
30
34
20
0
40
Exotic Species
30
0
40
0
0
10
20
30
XCOORD
40
50
Current information capabilities:
analysis and outreach, primarily with USGS data
Fish +
Aquatic
plants
Birds +
Mammals
California
Wildlife
Diseases
Plants
Hawaii
Texas
SW
East/
South
Midwest
Taxonomic Approach
(grossly under-funded,
but national scale)
Geographic Approach
(slightly better funded,
but locally scaled)
Improving capabilities for synthesis, research,
and outreach, with data from all sources
Fish +
Aquatic
plants
Birds +
Mammals
California
Wildlife
Diseases
Web
Tools for
Research
+ Outreach
Plants
High
Performance
Computer
Models
Taxonomic Synergies
Hawaii
Texas
Reston/
MESC
High
Resolution
Habitat
Maps
East/
South
SW
Midwest
Thematic Approach
Geographic Synergies
Data Synergies: inputs for early detection, risk
assessment, and “ecological forecasting” models
Num be r of S pe cies
1 - 51
51 - 120
120 - 19 7
197 - 30 3
303 - 67 4
No Data
Data Synergies: Weeds in Colorado
•County Quarter-Quad
•National Parks, National
Refuges, and Military
Lands, LTER sites.
•Forest Health Monitoring
Plots, other forest data.
•County Level Data
•Natural Heritage Network
•Modified Whittaker Plots
(USGS and others)
•Nature Conservancy
INPUTS
OUTPUTS
County-level
data on vascular
plants (BONAP)
Distributed
National data on
birds, mammals, and
diseases (USGS)
Information
Management and
modeling (USGS,
NASA, CSU, UCD)
Predictive models
of habitats vulnerable
to invasion
Watershed-level
data on fishes
(USGS)
•Data gathering
• Species taxonomy
• Data formatting
• Synthesis
• Predictive modeling
• Analysis and
display tools
• Data accessibility
via the web
Predictive models
of the spread of
Invasive species
Point data on public
lands (USFWS,
NPS, USGS)
Vegetation and soils
plot data (USFS,
USGS, BLM)
CLEARING
HOUSE
Web-net
National-scale maps
of non-native
species distributions
National, regional,
and local priorities
for control efforts
Reports on the
status and trends
of non-native
species in the U.S.
Current Predictive Modeling Capabilities
1.
ArcGIS: Input satellite data,
veg., soils, topography, etc.
Field Data: Invasive species
data, veg., soils, topography, etc.
2.
S-Plus: Develop Multivariate Model,screen and normalize data, test for
tolerance/multi-colinearity, and run stepwise regression.
3. S-Plus: test residuals for auto-correlation and cross-correlation (Morans-I) and
find the best model (ordinary least squares, gausian, etc. using AICC criteria).
4.
5.
6.
S-Plus/Fortran: If spatially autocorrelated, run kriging or co-kriging models.
ArcInfo GIS: develop map of model uncertainty from S-Plus
output, Monte-Carlo simulations, observed-expected values.
ArcView: produce maps of current distributions, potential distributions,
and vulnerable habitats, with known levels of uncertainty.
Future “Ecological Forecasting” Models:
Far more automated, instantaneous, and continuous!
1.
ArcView: Input satellite
data, via new sensors or
change detection models.
2.
Web-ware:
• Develop multivariate model, screen and normalize data, test for
tolerance/multi-colinearity, and run combinatorial screening.
• Test residuals for auto-correlation and cross-correlation
(Morans-I) and find the best models.
• If spatial autocorrelation exists, run kriging or co-kriging
models.
• Develop map of models uncertainty (maps with standard
errors).
• Produce maps of current distributions, potential distributions,
and vulnerable habitats, with known levels of uncertainty.
OR
Field Data: Early detection
or monitoring data, from
many sources.
3. Repeat Step 1 – always be looking for new data
What do clients want?
• Pick and click on any
point, land management
unit, county, state, or
region and determine
The current invasion,
and vulnerability to
future invasion by many
species.
(help public and private
land managers).
Refuge:
LaCreek Wildlife
Refuge
South Dakota
Updated:10/02/02
Regional zoom? Yes
Metadata? Yes
Plot data? Yes
Control Info? Yes
Plants
P= Animals P= Diseases P=
Cheatgrass
1.0
Norway rat
1.0
Blister rust .5
Musk
thistle
.99
Fire ants
.01
plague
Leafy
spurge
.65
Brown trout .01
Water
hyacinth
.02
Cover %
Zoom?
Yes
Musk thistle Cardus iforgotus
Plant ID Help? Yes
Map Uncertainty? Yes
.4
Or . . .
Pick and click on any
species or group of
species, and get current
distributions, potential
distributions, potential
rates of change,
and levels of uncertainty.
(We have much to learn
here! HPCC example
on West Nile Virus).
Obstacles
1. Lack of data sharing.
Solutions
1. Incentives, support,
rewards for sharing.
2. Uncoordinated budget
2. Joint budget committee,
process (DOI, USDA, Commerce). share “line items” ideas.
3. Computing power,
leaving the “PC stage.”
3. Beowulf clusters,
supercomputer use.
4. Modeling spatial and temporal
variation simultaneously.
4. “Frontiers of science”
challenge.
5. Urgency combined with
inadequate funding.
5. Dedication combined with
enthusiasm and perseverance.
A growing list of partners:
U.S. Geological Survey: T. Stohlgren, G. Chong, and C. Crosier (Midcontinent Ecological Science
Center, plants, data management), J. Sauer (Patuxent Wildlife Research Center, birds, mammals), P.
Fuller (Florida Caribbean Science Center, fish), J. Dein (National Wildlife Health Center, diseases), C.
Korschgen (Columbia Environmental Research Center, web-tools), L. Leake ( Upper Midwest
Environmental Science Center, data management), T. Owen (Center for Biological Informatics,
information mapping), A. Frondorf (Reston Office, high-performance computing), M. Ruggiero
(Integrated Taxonomic Information System, taxonomy, synonyms), W. Gregg (Invasive Species
Coordinator, Reston Office). T. Muir, R. Westbrooks (early detection), S. Haseltine (HQ).
U.S. Fish and Wildlife Service: M. Ielmini (Washington Office, Wildlife Refuges), W. King (Region 6
Wildlife Refuges).
Biota of North America Program, University of North Carolina: J. Kartesz and M. Nishiko (plants).
National Park Service: G. Williams (Inventory and Monitoring), C. Axtell (Biological Resource
Management Division); M. Wotawa (Biological Inventories and NPSpecies).
U.S. Forest Service: K. Stolte (Forest Health Monitoring Program).
National Aeronautics and Space Administration: J. Schnase and J. Smith and several others (ecological
forecasting, high-performance computing, remote sensing and modeling).
Colorado State University/Natural Resources Ecology Laboratory: M. Kalkhan and R. Reich (Spatial
modeling), D. Barnett, S. Simonson, R. Shory (data management, outreach).
University of California, Davis: J. Quinn (information management, modeling).
Long Term Ecological Research (LTER): J. Gosz
Colorado Natural Heritage Program: B. Strom (director), A. Black (GIS specialist)
Center for Environmental Management of Military Lands: B. Shaw
The Nature Conservancy: A. Bartuska, J. Randall
State of Colorado: E. Lane (State Weed Coordinator), B. Cheatum (GIS)
Agriculture Experiment Station: L. Sommers
Leveraging Funds, Data, and Expertise
Funds:
USFWS ($166K, $266K),
NASA ($250K, $250K, $250K),
BRD ($50K), USGS Venture Capital
($35K), MESC ($5K), State of
Colorado Agriculture Experiment
Station ($28K, $25K, $25K).
Few USGS funds
Data:
USGS (6 centers), USFWS,
BONAP, NPS, TNC, BLM,
NPS, CEMML, UCD, UWY,
USFS (FHM), State of Colorado,
LTER, APHIS, and CNHP.
data =$multiple millions
Expertise:
USGS (6 centers), NASA, EDC,
CSU, USFWS, BONAP, NPS,
TNC, BLM, NPS, CEMML,
UCD, UWY, State of Colorado,
LTER, ITIS, APHIS, CNHP,
Students and post-docs.
No new USGS FTEs
“One if by land, two if by sea. . .”
Documenting, Mapping, and Predicting the Invasion
of Non-native Plants, Animals, and Diseases in the United States.
The Future:
It’s What We Make It!
• Many more partnerships (DoD, CDC, APHIS, many
universities, more states, and more agency offices).
• Joint budget initiatives and proposals
(coordinating at higher levels in each agency).
• More shared expertise (data base design, web tools,
metadata, parallel processing and programming,
HPCC staff, additional modeling approaches).
• More staff (discussing Ph.D. and post-doc options
with several students, EDC, NBII, and APHIS).
• “Status and Trends of the Nation’s Invasive Species”
• A much larger “Invasive Species” program in the
USGS – one of our 8 future science activities
(We can’t do it alone, but we can do it together!)