|Home||Archived October 29, 2018||(i)|
projects > across trophic level system simulation (atlss) > work plan
U.S. Geological Survey Greater Everglades Science Initiative (Place-Based Studies)
Fiscal Year 2004 Project Work Plan
Project Start Date: September 1, 2003 Project End Date: June 30, 2006
Project Funding: USGS Place-Based Studies Initiative
Overview of the ATLSS Program
The ongoing goals in the ATLSS Program have been to produce models capable of projecting and comparing the effects of alternative hydrologic scenarios on various trophic components of the Everglades. The methodology involves: 1) a landscape structure; 2) a high resolution topography to estimate high resolution water depth across the landscape; 3) models to calculate spatially explicit species indices (SESI) for breeding and foraging success measures across the landscape; 4) spatially explicit individual-based (SEIB) computer simulation models of selected species populations; and 5) ability to plug into variety of visualization and evaluation tools to aid model development, validation, and comparison to field data. Included in this are numerous sub-projects for different species, vegetation succession, analysis of alternative approaches to developing high resolution, models which deal with estuarine systems, methods to allow users from a variety of agencies to access and run the models, and methods to enhance the computational efficiency of the simulations. The continuing general objective is to provide a flexible, efficient collection of methods, utilizing the best current science, to evaluate the relative impacts of alternative hydrologic plans on the biotic systems of South Florida. This is done in a spatially-explicit manner which allows different stakeholders to evaluate the impacts based upon their own criteria for the locations and biotic systems under consideration.
Currently, two important new developments are taking place. First the ATLSS models will soon migrate to a Web-based availability, so that they can be run remotely for various hydrologic scenarios and a set of different assumptions. Second, a vegetation succession model is being completed, which will allow projection of changes in vegetation types across the Everglades landscape as a function of changing hydrology, fire frequency, and nutrient loading.
These dramatic improvements in ATLSS models that will be entailed by these current developments are a strong motivation for completing important upgrading and validation testing of the existing ATLSS models and filling in several key gaps in the models. There are currently only limited funds to do this, so a carefully focused request for funding of several vital areas needed to complete the ATLSS program is proposed here. The specific projects, which are budgeted individually below, are the following.
Project 1. ATLSS Model Use in CERP Evaluations, Model Testing and Extension to Web-Based Interface This project involves completing the estuarine fish model and linkage with SICS, and a user interface for the Cape Sable Seaside Sparrow model, SIMSPAR
Project 2. Development of an Internet Based GIS to Visualize ATLSS Datasets For Resource Managers. This project involves ATLSS Data Viewer improvement, training, expansion to web.
Project 3. Spatial Decision Support for Biodiversity and Indicator Species Responses to CERP Project Activities. This project involves putting ATLSS and other models and data into a Decision Support System
Project 4. Integrating Wading Bird Empirical Data into a Model of Wading Bird Foraging Success as a Function of Hydrologic Conditions. This project involves using the most recent data on wading bird foraging parameters to complete the development of the wading bird SESI and Demographic Models.
Information Needs and Uses of ATLSS Program
It was identified early in the planning for the Everglades Restoration that models would have to play a key role in the process of choosing a restoration plan and evaluating its success. The ATLSS models have done and are doing this in the following ways, which are only a part of the overall products from ATLSS.
ATLSS Program: Key Findings from First 6 Years
The ATLSS Program has received support from the Critical Ecosystems Studies Initiative (CESI) for the past 6 years. The program has produced a set of models of spatially explicit species index, population demography, and ecosystem process models that are available for application to the Comprehensive Everglades Restoration Program (CERP). In addition, the program has supported field studies designed to produce data for model construction and validation. The main products of the modeling work are described below.
Spatially Explicit Species Index (SESI) Models. These models quantify relative effects of hydrologic conditions on the habitat suitability of species. There are several currently available ATLSS SESI models:
The ATLSS SESI models accomplish the following. They produce values for habitat suitability ranging from 0.0 to 1.0 for all 111,000 cells of the 500 x 500 m array. These can be calculated for every individual year in the historical 31-year sequence (1975-1995), or averages can be taken over any set of years (e.g., wet years, dry years, all 31 years). The SESI models are intended to be used to make relative comparisons between scenarios, not to produce absolute evaluations of habitat quality. The output can be viewed using the ATLSS Data Viewer, which allows viewing at any scale and performing of statistics. The ATLSS Data Viewer is available to all agencies that are interested. Training sessions can be scheduled when requested
Spatially Explicit Demographic Models. There are currently three available ATLSS Demographic Models. The ATLSS demographic models are spatially explicit individual-based (SEIB) models of the dynamics of the populations:
Spatially Explicit Functional Group Models. The ATLSS Structured Functional Group models simulate the size-structured and biomass dynamics of the population
There is currently one available ATLSS Structured Functional Group model:
GIS Animal Movement Tracking Tool. As a component of the development of population models, a GIS tracking tool has been developed to analyze radio-monitoring data of animals.
Landscape Models: There are background models that provide landscape information for other ATLSS models:
Landscape structure and High Resolution Hydrology:
Use of ATLSS Models. ATLSS model runs for scenario evaluations can be made in the following ways for particular models. The Snail kite demographic model (EVERKITE) is available in PC form for use (can be downloaded from Web) and user support for those wanting to use this model. Alternatively, this model will be run at the University of Miami and results posted. Currently, the remaining ATLSS models can be run at the University of Tennessee, which is funded to carry out several such runs. Results will be posted. Also, these models can be installed at agencies with Unix workstations. A NSF-funded project at the University of Tennessee is currently underway to allow dispersed resource managers to access remotely the capabilities of the SInRG (Scalable Intracampus Research Grid) at the University of Tennessee. This will allow users at resource agencies in South Florida, with relatively little computer expertise, to initiate ATLSS simulations on the computers at the University of Tennessee.
Documentation of ATLSS Models. Technical documentation of ATLSS models is available on the ATLSS web site (ATLSS.ORG) and listed in ATLSS Program Publications (available) - but will be improved. Open literature publications exist for the available ATLSS models. Nearly all models have appeared in open-literature, peer-reviewed papers (see ATLSS Publications). An internal USGS panel reviewed the ATLSS Program in May 2002. Additional review by the Model Refinement Team of CERP is planned.
Validation of ATLSS Models. Validation of models is an important issue. Some degree of model validation has been performed on some models (SIMSPAR, ALFISH). Model validation on other models will be performed depending on availability of data sets. Now data sets are becoming available for several species, and new data for others is being collected. A "validation tool", which can be applied along with the ATLSS Data Viewer, allows empirical data (e.g., nest success rate, fish biomass) to be compared spatially with SESI index values at any spatial scale. Statistical testing can be done via Excel spreadsheets. Validation is being done, or will be done soon on Cape Sable seaside sparrow, snail kite, and American alligator SESI models.
Key Current Progress
New Models. ATLSS models nearly completed or under development include:
Improvements in Use of Models. User interfaces for running models and analyses have been developed.
Progress is being made in making ATLSS runs available on the Web through new software (NetSolve and IBP). That is a huge step and is real state-of- the-art work. Users will be able to runfew different versions of the models, with different assumptions, and ill be able to look at the various individual 'layers' of the model output, as well as the whole index. Four projects needed to continue application of the ATLSS Program to CERP are described below. These projects are all deeply integrated and each will involve close interaction between the collaborating research groups at The University of Tennessee, University of Florida, Florida Atlantic University, and National Wetland Research Center.
Summary of ATLSS Program Needs from USGS Place-Based Funding
1. Project Title: ATLSS Model Use in CERP Evaluations, Model Testing and Extension to Web-Based Interface
2. Project Title: Development of an Internet Based GIS to Visualize ATLSS Datasets For Resource Managers
3. Project Title: Spatial Decision Support for Biodiversity and Indicator Species Responses to CERP Project Activities
4. Project Title: Integrating Wading Bird Empirical Data into a Model of Wading Bird Foraging Success as a Function of Hydrologic Conditions
Project Title: ATLSS Model Use in CERP Evaluations, Model Testing and Extension to Web-Based Interface
Duration of Proposed Work: July 1, 2004 - June 30, 2006
Proposed Scope of Work
Task 1. Use of ATLSS Models in CERP, Testing Models Against Empirical Data, and Extension to Web-Based User Interface. Much of the empirical data needed to test the output of the SESI models has been collected over the period 1996-2000. This could not be used to test the SESI model output against because the South Florida Water Management Model (SFWMM) output was available only up to 1995. By the middle of 2003 output of the SFWMM-2000 (Version 5.5) will be available. This will make it possible to test the SESI models. This testing can be done by the University of Tennessee ATLSS group. However, this will be an intensive process that requires careful organization of model and empirical data, and development of statistical techniques for comparing these data. The second part of this task is the extension of models to interactive use through the web. A first version of this ATLSS Network Approach is now undergoing beta-testing and will be completed in September 2003. However continual improvements will be needed to make this approach optimally useful to CERP.
It is essential that the ATLSS models be run for all or a substantial number of the proposed CERP scenarios in 2003 and 2004. The University of Tennessee currently has $20K available for such model runs. This will probably not be sufficient for all of the demands on scenario evaluations. Some of the proposed $90K will support manpower to produce and post a large amount of SESI and other model output over the period 2003/2004.
ATLSS SESI model output is currently being used within the ATLSS DataViewer (a product of the NWRC staff of USGS) to allow ease of visualization of the basic yearly spatially-explicit indices produced by these models. Modifications to allow the output of other ATLSS models within this are underway, though the only effort to date has been with the ALFISH freshwater fish model. Yet to be developed are standards for visualizing the outputs of SEIB models, and that task will be ongoing during year one of the project. An additional task will be to include outputs from the ALFISHES estuarine fish model. A separate task concerns the development of a database structure for all ATLSS model output. Although a formal metadata structure has been established and utilized within the ATLSS models, there has not been any formal database constructed to maintain the results of ATLSS runs. With the decreasing cost of storage, it is now feasible to construct such a database (using Oracle), and allow web access to it from within the DataViewer to allow users to visualize alternative ATLSS model runs without having the files shipped to them on a CD. This database establishment effort will be ongoing throughout the two years of the project.
Task 2. Integrating Estuarine Fish Model (ALFISHES) with Southern Inland Coastal System (SICS) Model
The technical development of the estuarine fish model has been completed. However, testing, upgrading, and interfacing with the Southern Inland Coastal System (SICS) model and with users (e.g., Jerry Lorenz, Rob Bennetts) is still necessary. ALFISHES can be run remotely or on a PC, and is available to users.
Work has been done to try to get ALFISHES working correctly on the short-term without the SICS model. Hydrologic and salinity data from three of Jerry Lorenz's sites have been input to the ALFISHES, to generate fish densities in only these three sites (three pixels) over a multi-year period. This is being compared with empirical data on fish densities to see if the model is making prediction in the right ballpark. (Some modifications of ALFISHES will be needed to make it work for selected sites. In particular, consideration is needed of what assumptions to use regarding reintroduction of fish to a re-wetted cell following drying, when, since there will be no adjacent cells in the model.
It is anticipated that SICS will be providing 5-year outputs within a couple of months. It is essential that Jon Cline be able to interface SICS with ALFISHES and upgrade the model such that it can be used for scenario evaluations.
Task 3. Developing User Interface for Cape Sable Seaside Sparrow Model (SIMSPAR). The Cape Sable sparrow demographic model (SIMSPAR) is an excellent model that has undergone a great deal of testing. However, the model is not currently usable to a broad range of users. The effective use of SIMSPAR requires a good user interface that will adapt it to use on a PC. It is almost certain that this could be done within two to four months with an appropriate postdoctoral assistant.
Project Title. Development of an Internet Based GIS to Visualize ATLSS Datasets For Resource Managers
Duration of Proposed Work: May 1, 2004 - April 30, 2006
Proposed Scope of Work
The Across Trophic Level System Simulation (ATLSS) Program attempts to predict the responses of a suite of higher trophic level species to different alterations in the Everglades/Big Cypress region of South Florida to represent the biotic community and various factors that affect this community. A tremendous amount of digital data have resulted from running these scenarios. To make these data available to resource managers and scientists, the USGS-National Wetlands Research Center has developed the ATLSS Data Viewer System (ADV). It is a spatial query and visualization GIS tool that provides the capability of retrieving, displaying, and analyzing ATLSS model data by using a user-friendly graphical interface and project-oriented procedures: The project has
The above goals have largely been completed. In addition, DVS has been upgraded to ATLSS Data Visualization System (DVS) 2.0. The upgrades include
The ATLSS spatially-explicit species index (SESI) models were used in the Restudy evaluations and have continued to be used for subsequent evaluations. They are available for use in the many evaluations that must be made in the future. The ATLSS DVS developed by this project serves the following purposes
ATLSS Viewer has been developed by using ESRI-ArcView 3.2 GIS. Most procedures implemented in this project use raster-based functions provided by the ArcView extension Spatial Analyst 1.1. ATLSS Viewer provides the capability of converting binary raster files produced by running ATLSS models to ESRI ArcInfo grids. Annual data can be retrieved and displayed to show alternative water management changes and their effects on numerous species used in ATLSS, and compare numerous scenarios for one species. The viewer also allows the user to calculate average data based on specified time intervals or time points during the 31-years simulation period. Statistics and/or graph representations of the whole study area, or user-defined spatial subdivisions can be performed, displayed, and exported to other applications. Tables, histograms, maps, and metadata can be generated to report the comparison between a basic and an alternative scenario for the whole Restudy area or user-defined areas. A specific section of the project is devoted in providing detailed information about each ATLSS model included into the system. Several other procedures have also been developed to simplify project tasks and documents management.
The ATLSS Data Viewer system needs to be upgraded to allow a larger amount of ATLSS model output to be viewed and analyzed. More training sessions will be needed. Continued collaboration is needed between NWRC and the University of Tennessee group, Wolf Mooij, Ken Rice, and others to make sure that output from various ATLSS models is input to the Data Viewer. The Data Viewer must be extended to a Web-based system.
Although the ADV has been upgraded to display output of the ATLSS model ALFISH, it must be further upgraded to allow the display of other models, including both EVERKITE and the Alligator Demographic model.
Project Title. Spatial Decision Support for Biodiversity and Indicator Species Responses to CERP Project Activities
Duration of Proposed Work: January 1, 2004 - December 31, 2005
Proposed Scope of Work
The objective of this project is to provide a working prototype spatial decision support system (SDSS) for impacts to biodiversity and indicator species in the landscape within and surrounding CERP projects. The proposed SDSS tools will be designed to assist managers and other users in understanding wildlife habitat response to hydropattern and land use changes. An integral task in the SDSS is the provision for assistance in defining the problem given uncertainty in the data and then providing problem specific models. Successful implementation of the SDSS presumes cooperative development with the users to meet resource managers needs.
Decision support systems (DSS) are broadly defined as computer-based systems used to aid decision makers using data and models to solve unstructured problems. These model and knowledge-based system tools become necessary when complex geographic or domain interrelationships are considered. The development of a computerized DSS makes economic sense in integrated Everglades restoration efforts because of the large amount of data that must be collected and processed to produce and analyze decision alternatives, decision-making procedures that are applied to many cases within a domain or periodically repeated, many potential users, short time frames for making critical decisions, the expense of accessing top-level expertise, and the possibility of a large number of alternative decisions with significant and different implications. This project is developing a spatial decision support system (SDSS) is proposed in which the decision models are tightly integrated with, or directly generated from, geographic information systems (GIS) analyses and display. This is integrating ATLSS and other models, as well as empirical data. The SDSS will incorporate analyses, but the scope and range of scales considered in the SDSS models will be carefully restrained to a specific subset of landscape problems. Part 1 of this work developed a framework for the SDSS. This will next be extended into concrete development of the SDSS.
This project addresses 2 high priority CESI science objectives: 3007-19, Monitor the status of indicator species, their communities, and species of special concern for evaluation of Everglades restoration success; and 3070-8, Develop and implement methodologies and decision support tools that will permit effective and timely assessment of CERP projects on DOI natural resources.
This project will develop a case study functioning spatial decision support system for a CERP project area using a modular architecture that allows rapid transfer of the prototype system to other project areas. The SDSS will assist managers in assessing issues and alternatives for wildlife habitat response to CERP project activities. Specific objectives in support of the goal include:
The project area for prototyping the SDSS will be chosen as representative of diverse unstructured issues. The area will be selected with consultation and agreement from the RECOVER Leadership Group (RLG) and/or the RECOVER Regional Evaluation Team (RET). The study area will be defined for this project as the selected project area plus additional contiguous lands necessary to adequately model adjacent impacts of project activities.
Wildlife models to locate potential habitats suitable for the species will use land cover, hydroperiod, soils, and other available GIS data layers where appropriate. The wildlife models will allow for rapid incorporation of change scenarios from other existing models of land use and hydrology. Field surveys directed at sampling target species presence and abundance will validate the wildlife models. Field surveys will also address uncertainties in target species modeling parameters and be used to re-calibrate and improve the models in an adaptive approach.
The decision model will be selected after review of existing "off-the-shelf" knowledge-engines and DSS systems. The rationale is to not reinvent already effective solutions, but rather to wrap those solutions in an interface specific to the needs of the local projects. This approach is expected to reduce the time and costs of development and take advantage of solutions that have been proven in application. Development and programming will still be a major component of the project, however, as adaptation of the decision model goes through an evolutionary process of addressing local user needs, refinement and modification. Software development will be a dynamic process that responds to continuing user feedback as well as verification and validation of the refinements. Validation of the decision models will be performed with sensitivity analyses and empirical testing following accepted published methods.
The final report will include complete documentation of the final methods and programming code for both the wildlife habitat and decision models. All source code created for this project will be included. Field monitoring results will be summarized in GIS data layers and tables. Validation, verification and uncertainty analyses will be documented and summarized. The principal product will be a spatial decision support system for the selected project area, user documentation, a final workshop to introduce users to the system operation, and an example scenario, including GIS data layer inputs and outputs, using the SDSS to address a task or problem. Final products will include a manuscript submitted to a peer-reviewed scientific publication. Interim semi-annually reports will be provided to ensure communication between PIs and contract office technical representatives (COTR's). Annual reports will include data summaries that follow the data management policies of the SFNRC.
Project Title: Integrating Wading Bird Empirical Data into a Model of Wading Bird Foraging Success as a Function of Hydrologic Conditions
Principal Investigator: Dale E. Gawlik, Florida Atlantic University, Boca Raton, Florida
Project Scope of Work
Declining wading bird populations in the last century have been considered one of the most prominent signs of the degradation of the Everglades ecosystem. Consequently, recovery of these populations will be a key indicator of successful restoration efforts. It has been hypothesized that the specific mechanism by which Everglades degradation has led to declining bird populations is related to changes in hydropatterns. These changes have most likely altered the availability of prey to wading birds.
Prey availability is determined by both the abundance of prey and the vulnerability of prey to capture. Prey abundance is affected by factors such as nutrient levels and hydroperiod whereas vulnerability to capture is affected by such things as behavior of the prey species, water depth, vegetation density, and body size.
Each component of prey availability is affected differently under various water management scenarios. For example, management for long periods without severe drydowns changes the species composition of the fish community. Different species of fishes exhibit different behavioral response to predators, thus changing their availability to capture. Moreover, these behavioral differences of the prey may be dependent on water depths. When the water column is deep, social species like the golden shiner (Notemigonus crysoleucas) may occur in schools and present wading birds with a very different capture probability than more solitary fishes like the bluegill (Lepomis macrochirus). However, as the water level recedes, capture probabilities of the two species may converge.
Ongoing modeling efforts in south Florida, such as the Federal Across Trophic Level System Simulation (ATLSS) program, integrate information on hydrology and wading bird food availability to provide predictive power for future water management decisions. Currently, the biggest information gap limiting the wading bird component of ATLSS is foraging success as a function of prey availability and water depths. We conducted a series of experiments aimed at determining the effects of water management (manifested through changes in prey availability) on the use of foraging sites by wading birds. The wading bird data collected and analyzed provides strong evidence for the relationships of water depths, fish densities, fish sizes, and fish species to availability to wading birds.
These data complete what is needed to make the wading bird bioenergetic model developed by Wilfried Wolff a useful tool for science and management. The individual wading bird in that model is described by a set of species-specific rules that govern its behavioral activities. A model wading bird does not operate on a fixed time scale, because its behavioral activities are of different duration. Instead, the wading bird model uses an event-driven approach, in which each bird sets its own time scales depending on its current activities. In its current version, the wading bird model operates on spatial grid of 500 m x 500 m grid cells. The model keeps track of colony sizes and the number of nesting adults as well as the number of successfully fledged nestlings after the breeding season is over. Because energetic constraints drive most of their activities, in particular the onset and timing of nesting, different environmental conditions will lead to varying reproductive behavior and recruitment of young wading birds into the population.