PROJECT TITLE: Task 98-3. Mangrove Modeling of Landscape, Stand-level, and Soil-Nutrient Processes for the ATLSS Program and Everglades Restoration Project

PRINCIPAL INVESTIGATOR AND PROJECT COORDINATOR: Thomas. W. Doyle
Mail address:
USGS/BRD, National Wetlands Research Center
700 Cajundome Blvd.
Lafayette, LA. 70506
Phone: (337) 266-8647
Email: tom_doyle@usgs.gov

CO-PRINCIPAL INVESTIGATOR: Robert R. Twilley
Mail address:
Department of Biology, University of Southwestern Louisiana
Lafayette, LA 70506
Email: robert_twilley@usgs.gov

GUIDANCE AND ASSISTANCE FROM: Donald L. DeAngelis, ATLSS Coordinator

NEED, KEY QUESTIONS & RESTORATION BENEFITS: Coastal wetlands of the gulf coast and south Florida, in particular, have been and will continue to experience significant changes in their hydrologic regimes as a function of freshwater management practices and rising sea-level. Because mangroves thrive in the intertidal zone between land and sea, these systems are expected to undergo the most severe changes of both freshwater drainage effects (flushing, nutrients, pollutants) and marine influence (sea-level rise, salinity). Spatial simulation models (Doyle 1997, Doyle and Girod 1997) have been developed for south Florida mangrove communities that with additional empirical inputs and modifications can be applied to address the success of proposed freshwater restoration alternatives for maintaining and restoring the ecology of the parks and refuges of south Florida. These models will be integrated with other simulation models of hydrology and higher trophic population models developed under the ATLSS program.

The land-margin ecosystem of south Florida is subject to coastal and inland processes of hydrology largely controlled by regional climate, disturbance regimes, and water management decisions. The hydrology affecting the Everglades system is complex and variable as dictated by surface water management, sea level rise, seasonal precipitation and temperature fluctuations, and recurring hurricanes. Mangroves persist along the land-sea interface where water levels and salinity fluctuations are at extremes, daily, seasonally, and interannually. Modest changes in freshwater delivery are expected to alter the hydrologic balance and vegetative complement and boundaries of this expansive community type with adjacent wetland associations. Ecosystem models are needed that can predict vegetation response in relation to changes in hydrology that in turn can be used to test how changes in hydrology will affect vegetation health and distribution. The interaction between the natural climate extremes and surface water controls on system hydrology and vegetation will require an integrated modeling approach that describes vegetation response at the organism, population, community, and landscape-level of ecosystem organization and function.

Land managers need to know how managed and natural changes of Everglades hydrology from human management and climate change forces may alter the long-term stability of certain plant and animal assemblages in parks and refuge lands of south Florida. For example, to what degree is the altered health and dieback of some mangrove habitats of south Florida a function of altered freshwater flow, water quality, present climate conditions, hurricane, and/or sea-level rise effects. Mangrove habitats provide a valuable and critical resource for fisheries and wading bird populations. Mangrove locales vary in the degree to which they are connected to the regional hydrology as influenced by upslope drainage and tides. This proposal outlines the needs and tasks for upgrading the SELVA/MANGRO models of mangrove growth, disturbance, and succession for testing Everglades restoration alternatives and linking with ATLSS trophic simulation models. Research tasks are described as data products and model developments for accomplishing landscape and stand-level simulations of vegetation relations compatible with the ATLSS program research needs and goals. Model design and applications are specified to determine the potential benefits and consequences of active management of the Everglades freshwater delivery process and restoration program on land-margin ecosystems of south Florida.

This effort will complement field and modeling studies conducted under the NPS/FWS/NBS and EPA global change programs as well as concurrent USGS, COE, SFWMD, and NPS sponsored Everglades Restoration Initiative studies to enhance the functionality of existing landscape simulation models, ELM and ATLSS. Critical data and model products have been identified in this proposal to complement the ATLSS program needs and model applications as outlined under the Critical Ecosystem Study Initiative (CESI).

BACKGROUND AND PROGRAM HISTORY: The studies discussed here are continuations of previously peer-reviewed work for which funding has ceased and/or been inadequate. In particular, this work will continue research initiated under the USGS/BRD's South Florida Global Climate Change Program and the mangrove component of ENP's Hurricane Andrew research program. Both programs have undergone rigorous, external peer-review. Project tasks and methodologies were defined by input of agencies and scientific staff to integrate with ATLSS program goals and needs and other concurrent field projects supported under the Everglades Restoration Program.

METHODOLOGIES: The tasks discussed below will provide spatial data, critical processes, and model functions at the landscape, stand-level, and soil-nutrient interface necessary to accomplish a hierarchically integrated modeling component of vegetation relations within the land-margin ecosystem complex compatible with ATLSS. Task 1 defines a top-down approach of the pattern and process of landscape aspects that influence stand-level processes, forest growth and succession of mangrove and associated systems. Task 2 defines a bottom-up integration of water-nutrient relations and stand-level processes, above and belowground, that influence tree growth and competition, species success, and site quality.

Task 1: SELVA/MANGRO Upgrades of Pattern and Process of Landscape/Stand Level Interaction: Upgrading SELVA/MANGRO Model, Map Sets, and Linkages for ATLSS (Thomas W. Doyle, USGS-BRD, National Wetlands Research Center, Lafayette, LA)

The primary goal of this task includes field and modeling activities to upgrade the functionality of an existing landscape simulation model for south Florida (SELVA-MANGRO, Doyle and Girod 1997) specifically emphasizing the physical forcings and biological qualities of the collective watersheds and estuaries of the larger Everglades landscape. Field efforts will involve stem mapping mangrove forest structure of existing permanent plots (GCC-SOFL, Doyle et al. 1995) to establish patterns of spatial ingrowth and data accuracies, including tree heights, diameter at breast height, azimuth, distance, crown class, viability status, leaf area, and light (PAR) distribution. This data will be used to initialize FORMAN (Chen and Twilley 1998a) and MANGRO (Doyle and Girod 1997) models to test model function and capability to predict post-hurricane forest recovery and tree growth. Aerial photography will be interpreted and digitized for a series of historic dates to establish rates and patterns of mangrove encroachment and retreat along ecotone boundaries. Landscape change analysis will be conducted to identify areas of varying degrees of mangrove migration and to relate ecotone dynamics to changes in freshwater drainage, land use, and sea-level rise. The SELVA-MANGRO model will be calibrated and validated to backcast/forecast ecotone migration for historic sea-level and streamflow conditions. Synoptic surveys will be conducted to determine land/water datums across the coastal zone including prairie, basin forest, riverine fringe and overwash islands. Elevational relationships will be established to refine existing digital elevation models of the coastal zone. Algorithms will be developed for SELVA-MANGRO and HYMAN models to predict tidal forcing and salinity conditions of coastal zone. Available resources including aerial photography, videography (Doyle et al. 1994), and available radar and satellite (Roberts et al. 1994) imagery will be used to classify a site quality index of mangrove habitat at the landscape level based on estimated and measured canopy height (potential). Ground surveys and overflights will be conducted to achieve a representative sampling and to establish classification accuracy and spatial interpolation. Surveys will be coordinated with associated field activities defined in Task 2 and other BRD collateral studies to characterize representative watersheds inclusive of basin, riverine, and overwash habitats by soil fertility, hydrologic and geomorphic setting, and sapling productivity measures (D/H, internode relations). Surface hydrology layers will be incorporated or digitized as necessary into a GIS layer for the study area. A landscape map of cell based units at a 500m scale will be determined relative to edge vs intact forest conditions related to juxtaposition to stream channels and open bays compatible with current ATLSS cell size resolution. HURASIM (Doyle 1994, Doyle and Girod 1997) model will be used to create a landscape map of maximum wind force and duration reconstructions of hurricane activity for the period of record 1886 to present. Aerial photography will be interpreted to incorporate the probability functions and spatial variation of lightning strike and gap creation across the study area and in SELVA-MANGRO. Freeze damage probability maps will be generated from historic climate data and interpolated across the landscape at an appropriate scale and used in SELVA-MANGRO to approximate freeze cycles and effects. Model linkages will be developed to hierarchically integrate SELVA-MANGRO with the ATLSS model. Applications will include predicting changes in ecosystem structure and function of restored waterflow alternatives for different parks and refuges, watersheds, and project locations under different climate change scenarios.

PROGRAM HISTORY: Original research for this study effort was initiated under an interagency agreement between U.S. Fish and Wildlife Service and the National Park Service under the direction and funding of DOI's Global Climate Change Program. A landscape simulation model, SELVA-MANGRO (Doyle and Girod 1997), has been developed and applied for the larger south Florida peninsula including the complex of collective state and federal parks and refuges about the Everglades System. Specific federal lands covered in this effort include Ding Darling National Wildlife Refuge (NWR), Ten Thousand Islands NWR, Florida Keys NWR, Everglades National Park (NP), Biscayne NP, and Dry Tortugas NP among other state and county reserves. The model has been used to review the impact of climate change phenomena, hurricanes and sea-level rise, on the structure and function of mangrove communities across south Florida. Field studies have been conducted to supplement model development including post-hurricane impacts (Doyle et al, 1995) describing forest structure, tree growth, and succession in established study plots in the above parks and refuges across south Florida. Experimental greenhouse and growth chamber studies have been implemented to determine species tolerance and response to growth limiting conditions of light, salinity, and hydrology. Model refinements and validation have continued with the integration of field and laboratory results as well as updated literature synthesis. Specific information on the health and status of these systems have been synthesized and reported to research, management, and interpretive staff for the above parks and refuges. Recommendations for habitat management and research priorities have also been discussed with key park and refuge officials.

PROJECT OBJECTIVES: The primary objective of this research study is to upgrade the SELVA/MANGRO model with additional spatial data sets and model algorithms that relate to issues and effects of hydrologic restoration of the Everglades drainage on the land-margin ecosystems compatible with the ATLSS program. Study initiatives coincide with hierarchical scaling of model input and functions to capture the relationship between landscape and stand-level interactions of ecosystem organization and function on the composite vegetation response to changing climate and hydrology. Hypothesis testing has been expressed in the form of model constructs of systems hierarchy and feedback mechanisms as important linkages controlling ecosystem composition, structure, and resiliency. A spatially explicit community dynamics model, MANGRO, has been constructed to simulate inter-plant relations and species tolerance to relevant environmental conditions. MANGRO encompasses those abiotic and biotic factors controlling tree and species growth, survival, and regeneration at the stand/community level. The MANGRO model is imbedded within a geospatial model, SELVA, delineating the environmental gradients and disturbance regimes exogenous to localized stand conditions but important to climatic control of species distribution and community development at the landscape scale. The following tasks describe a combination of empirical studies, GIS mapping, and model developments deemed paramount for upgrading SELVA/MANGO and for linking with ATLSS program needs.

Task 1-1: GCC Plot Remeasurement
Field efforts will involve stem mapping mangrove forest structure of existing permanent plots (GCC-SOFL, Doyle et al. 1995) to establish patterns of spatial ingrowth and data accuracies, including tree heights, diameter at breast height, azimuth, distance, crown class, viability status, leaf area, and light (PAR) distribution. This data will be used to initialize FORMAN (Chen and Twilley 1998) and MANGRO (Doyle and Girod 1997) to test their capability to predict post-hurricane forest recovery and tree growth.

The null hypothesis for Task 1 states that there is no difference in interplant spatial relations that determine tree or species success. Explicit spatial data of tree position and stature will be used to initialize model simulations to determine if there is any difference in model behavior and predictive outcome with non-spatial gap simulation of stand conditions. Specific model algorithms related to intertree competition, shade tolerance, and light distribution with forest structure will be evaluated and verified with field data. The results of this test will determine the complexity of model algorithms needed to predict stand level outcomes and the information required to pass up-scale for landscape integration with SELVA and ATLSS models.

Task 1-4. Site Quality Characterization across Mangrove Landscape
Available resources including aerial photography, videography (Doyle et al. 1994), and available radar and satellite (Roberts et al. 1994) imagery will be used to classify a site quality index of mangrove habitat at the landscape level based on estimated and measured canopy height (potential) and greenness (NDVI). Ground surveys and overflights will be conducted to achieve a representative watershed sampling and to establish classification accuracy and spatial interpolation. Measures of soil fertility, hydrologic and geomorphic setting, and sapling productivity measures (D/H, internode relations) will be gathered as part of a coordinated effort with other projects and PI's, R. Twilley and T.J. Smith. A classified site quality index map will be created in an ARCINFO coverage and converted to a raster set for the SELVA model and used to drive MANGRO growth functions for dwarf, intermediate, and tall growth forms and equations. Locations of wading bird colonies, if available, will be digitized into the same GIS coverage to alter nutrient loading and site quality for inclusive stand simulations.

The null hypothesis for Task 4 states that there is no difference in mangrove habitat structure and growth potential based on canopy heights of mature stands across the Everglades complex. Field data will be analyzed to determine whether there are distinct clusters of vegetation habits, height and growth index, with site type and fertility. An explicit map set will be generated to drive model functions of site productivity in SELVA/MANGRO in relation to site characterization.

Task 1-5. Disturbance Probability Map Sets
HURASIM (Doyle 1994, Doyle and Girod 1997) model will be used to create a landscape map of maximum wind force and duration reconstructions of hurricane activity for the period of record 1886 to present. Other parameters as identified such as wind direction and sediment deposition patterns can also be retained in SELVA for simulating other hurricane-related forces and consequences. MANGRO contains a calibrated function for predicting tree mortality related to wind force. Aerial photography will be interpreted to create an ARCINFO map set of probability estimates and spatial variation of lightning strike and gap size across the study area. Freeze damage probability maps will be generated from historic climate data and interpolated across the landscape at an appropriate scale and used in SELVA-MANGRO to approximate freeze cycles and effects.

The null hypothesis for Task 5 states that there is no difference in the probability estimates for disturbance, hurricanes, lightning, freezes, across the Everglades complex either latitudinally or longitudinally. Results of model reconstructions for historic hurricane activity from 1886 to present show that there areas of the south Florida peninsula that receive significantly greater incident of hurricane frequency and intensity (Doyle and Girod 1997). Similar analyses are needed to determine whether temperature isoclines vary spatially and whether lightning strike probabilities are lesser or greater for given locations across the Everglades. Model simulations will be conducted to determine the effects of hurricane, freeze, and lightning probabilities across the landscape are variable enough to alter forest structure and function.

Task 1-6: Coastal Hydrology and Forest Edge Map Set
SELVA already contains predicted tide and sea-level rise algorithms and amended utilities to create future sea-level rise conditions based on IPCC (1990, 1995) projections. Modifications to include additional sea-level projections will be incorporated as requested. Available USGS DLG hydrology sets will be combined into a single topology and attributed for streamflow and connectivity for the study landscape. A forest edge map set will be created at a 500m resolution compatible with ATLSS simulations to capture the spatial and functional relations of forest stands bounded or intersected by open water channels and bays. Forest type and functions will describe proximity relations for sedimentation and hurricane effects.

The null hypothesis for Task 6 states that there is no difference in edge exposure of mangrove forest across the Everglades complex. This analysis will determine whether inland bays are sufficiently large for the ATLSS spatial cell size of 500m to classify categories of forest edge as either open or intact. Model simulations of MANGRO will be conducted to determine the effect on stand-level processes and estimates when simulated edge boundaries are treated as intact or open. Model comparisons will be made between cases where only 1 edge is open as compared to 2, 3, and 4 open sides. For example, islands will be simulated as 4 open sides. Open forest edges are subject to greater hurricane impact due to the lack of sheltering of winds approaching over open water. This task will determine the degree of forest edge that prevails in the mangrove zone of the Everglades and the importance to translating effects of scale and disturbance with forest edge dynamics.

Task 1-7: Develop External Model Linkages
Model linkages will be developed to hierarchically integrate SELVA-MANGRO with the FORMAN and ATLSS simulations. Direct or indirect programming developments will be devised to pass information up, down, and across scale as appropriate to establish crosslinks between models. Applications will include predicting changes in ecosystem structure and function of restored waterflow alternatives for different parks and refuges, watersheds, and project locations under different climate change scenarios.

Task 2: The Utility of Mangrove Unit Models (FORMAN and NUMAN) for the MANGRO and ATLSS Landscape Programs and Everglades Restoration Project (Robert R. Twilley, Department of Biology, University of Southwestern Louisiana, Lafayette, LA)

The goal of this proposed research program is to further develop the FORMAN and NUMAN models to link community development and biogeochemistry of mangrove ecosystems in the south Florida region. Task elements include concurrent field and experimental studies to provide parameter inputs and data for model development and testing. These models will be applied to project the response of mangrove stands to changes in quantity and quality of discharge as part of the Everglades Restoration Program under the Critical Ecosystem Study Initiative (CESI). FORMAN and NUMAN are designed to mimic ecological processes affected by water and nutrient delivery and the resultant patterns of forest development and nutrient distribution in coastal land-margin ecosystems of south Florida. One of the initial efforts will be to upgrade the existing FORMAN and NUMAN unit models by including more mechanistic control of community development. The primary objective of the FORMAN model is to simulate the mechanistic effects of spatial gradients in soil characteristics and hydroperiod on the growth and development of mangrove wetlands. The FORMAN model (Chen and Twilley 1998a) is an individual-based mangrove succession model that simulates the influence of physical factors on species distribution and productivity on yearly basis. The HYMAN model (Twilley and Chen 1998) is a hydrology model that simulates the mass balance of freshwater and tidal inputs, and calculates porewater and surface salinity. The NUMAN model was developed to simulate the availability of N and P in mangrove soil by coupling the simultaneous effect of productivity and allochthonous inputs of mineral sediment (Chen and Twilley 1998b). Resource competition theory will be utilized to model community development due to the shifts in nutrient pools and availability across the land-margin coastal gradients of the south Florida mangrove region. Task elements are designed to methodically calibrate model functions and parameter sets and to validate model behavior and results compatible with existing mangrove research programs of NBS/USGS and ATLSS to model the Everglades watershed. A large part of this effort will be to demonstrate to the various research and management groups associated with the Everglades restoration effort the utility of FORMAN in projecting the impacts of land use change on mangrove wetlands. Site-specific field studies and model applications will be conducted at NBS/USGS mangrove research sites in the coastal margins of south Florida (Tom Smith, PI).

Results from our nitrogen and phosphorous cycling studies in mangrove forests (Rivera-Monroy et al 1995a, b, Rivera-Monroy and Twilley 1996, Twilley 1995, 1997) show the regulatory nature of C:N:P ratios have on N and P transformations, particularly P and N immobilization. One of the key transformations that needs further study is documenting the role of allochthonous P inputs and N fixation in the C:N:P ratio of mangrove wetlands. Model modifications will include explicit relationships of N and P transformations such as immobilization and mineralization. By adding mechanisms that describe the partition of C, N, and P exchange (atmosphere and water) and regeneration, we expect to increase our understanding of what processes control the flux of nutrients at the mangrove:coastal water boundary of south Florida land-margin ecosystem. The NUMAN model (Chen and Twilley 1998b) used a cohort analysis of mangrove soil formation. This initial modeling effort discovered some key parameters that need to be calibrated to understand the soil organic matter and nutrient pools along the estuary gradient in southwest Florida. One of these parameters is the allocation and turnover of above- and belowground biomass of mangrove wetlands related to soil conditions. We propose field and mesocosm studies of these processes to directly test how they vary along nutrient gradients. Comparisons of organic matter production and decomposition will be evaluated among soil cohorts and compared to the input of mineral matter using Pb-210 techniques (see Nutrient Exchange below). These processes along with the NUMAN model can be compared with results of soil accretion and elevation proposed for study by Smith and Cahoon at selected sites in the Everglades. Parameterization and calibration of the NUMAN model will have important applications to understanding the impacts of specific scenarios of CESI to the long-term sustainability of mangroves under conditions of present sea-level rise.

Upgrading the FORMAN and NUMAN models will require a better understanding of the interaction between hydroperiod and nutrient availability on productivity of mangrove and marsh communities. This includes mesocosm experiments of N and P enrichments (three treatments, each) on growth of Rhizophora, Avicennia, Laguncularia under two hydroperiods (long vs short) at two salinity regimes. These studies will be performed in the new Environmental Greenhouse Facility (1 acre) constructed at USL and includes wetland intertidal mesocosms. The relative ecological significance of nutrient availability varies among fringe, basin, and dwarf mangroves according to tidal inundation frequency from lower to upper regions of the intertidal zone. Fertility may also vary for different ecological types of mangroves depending on the allochthonous inputs from riverine environments, compared to the lack of allochthonous inputs in lagoons and carbonate environments. Nutrient cycling in mangrove wetlands in areas with high tidal and riverine forcings are considered ëopení with high rates of material exchange across the mangrove-estuarine boundary. Whereas, less frequently inundated systems, such as inland basin mangroves, have a greater accumulation of leaf litter resulting in higher remineralization within the forest. Thus the ecological types of mangrove wetlands across different geomorphological regions with contrasting fertility may have different plant and ecosystem strategies of nutrient conservation. The FORMAN model has to capture these different strategies in nutrient conservation to accurately formulate the links between production and nutrient cycling. We hypothesize that the biogeochemical properties of mangroves such as sedimentation, mineralization, exchange, and nutrient-use efficiency will vary among sites along the nutrient gradients of the south Florida coastal landscape. We will determine the pools of carbon, nitrogen, and phosphorus in the biomass, litter, and soil of different mangrove sites in the south Florida study region as parameters of the NUMAN model (biogeochemistry component); and the nutrient use efficiency of these nutrients among these pools. Field estimates of nutrient exchange in mangrove wetlands of south Florida region are needed for the NUMAN model to simulate the role of mangrove wetlands in the distribution of nutrients in the southwesterly flow of coastal water to Florida Bay. We propose to perform quarterly measurements of nutrient fluxes (water exchange) at the boundary of mangrove sites (S1 and S4) on the Shark River slough using flume designs. We will also perform studies of nitrogen fixation quarterly at two mangrove sites to determine the significance of this input to the nitrogen exchange at the wetland boundary. In summary, this task will determine the relative significance of nutrient supply (inputs from atmosphere and water exchange) as controls of mangrove structure and function in the NUMAN model.

NEED, KEY QUESTIONS & RESTORATION BENEFITS: As ecotones of continental margins in warm temperate to tropical climates, mangrove wetlands are one of the most susceptible ecosystems to climate and land use change. The land-margin ecosystems in southwest Florida represent a combination of different mangrove ecological types in mainland carbonate environments with gradients in amount of peat accumulation. Although there exist much published literature concerning the geology and forest structure of these ecosystems, along with analyses of food webs in these environments, the biogeochemistry and productivity have not been addressed. These ecological properties are important because potential changes in nutrients and hydrology in the upland watershed may have important impacts on the structure and function of mangroves in the coastal margin of southwest Florida. Initial studies of these specific types of mangroves indicate that they are sensitive to the availability of nutrients and type of hydroperiod (Chen and Twilley 1998a). Thus changes in the water management of this coastal region could have significant ecological impacts on these mangrove ecosystems.

Restoration efforts in south Florida include issues of water management including projected impacts of fresh water diversions and nutrient enrichment on the Everglades and associated wetland ecosystems. However, the responses of mangroves to changes in water quality and the role of these land-margin ecosystems to mitigating nutrient enrichment in the coastal zone have received little attention. The biogeochemical properties of mangroves are the least understood of ecological processes along the transition from upland to coastal margin ecosystems (Twilley 1995). Thus the specific nature as to how the distribution of nutrients influences mangrove structure and productivity, and the role of mangroves in the fate of nutrients in sub-tropical estuaries, are poorly understood. Continued efforts monitoring these biogeochemical processes, together with further development of ecological models, will provide important information on the projected response of mangroves to changes in water management in this coastal watershed.

Ecological modeling is a powerful tool that can be used to describe, quantify, and forecast the response of mangroves ecosystems to different coastal management strategies. Potential impacts of coastal eutrophication and hydrological modifications on the structure and function of mangroves need to be evaluated in advance before implementing programs that modify regional land use, which can profoundly alter forcing functions controlling mangrove productivity. In this proposal, we propose the application of two mangrove unit models, FORMAN and NUMAN, to help develop and implement management plans for the restoration of the Everglades region.

The goal of this proposed research program is to further develop the FORMAN and NUMAN models to link community development and biogeochemistry of mangrove ecosystems in the south Florida region. The objective of these mangrove unit models is to understand how the restoration of these ecological processes may influence patterns of forest development and nutrient distribution in coastal land-margin ecosystems of south Florida.

PROGRAM HISTORY: We have developed three mangrove ecological unit models and propose that they can be used in the decision-making process to propose management scenarios for the Everglades Watershed. Resource competition theory will be utilized to model community development due to the shifts in nutrient pools and availability across the land-margin coastal gradients of the south Florida mangrove region. These data will be related to the development of landscape models of mangroves, MANGRO, by Tom Doyle (National Wetlands Research Center) that is to be integrated with the ATLSS modeling effort by Don DeAngelis (USGS/Miami).

PROJECT OBJECTIVES: The individual tasks are designed to methodically integrate the FORMAN, NUMAN, and HYMAN models into the existing mangrove research programs of NBS along with efforts to model the Everglades watershed. The goal is to develop specific guidelines as to how these mangrove unit models can be utilized in the ecosystem and landscape modeling efforts to assess the response of coastal ecosystems to changes in quantity and quality of freshwater inputs from the upland watershed. A large part of this effort will be to demonstrate to the various research and management groups associated with the Everglades restoration effort the utility of these mangrove unit models in projecting the impacts of land use change on mangrove wetlands. Along with this effort will be an opportunity of verification of FORMAN and NUMAN at NBS mangrove research sites in the coastal margins of south Florida (Tom Smith, PI). The main goal of this proposal is to determine the utility of the FORMAN and NUMAN models to project the response of mangrove stands to changes in quantity and quality of discharge as part of the Critical Ecosystem Study Initiative (CESI).

TASK 2-1: Verify the FORMAN and NUMAN models: The JABOWA-FORET individual gap models were used as a basic design to develop a model to simulate mangrove forest development (FORMAN) . The FORMAN model (Chen and Twilley 1998a) is an individual-based mangrove succession model that simulates the influence of physical factors on species distribution and productivity on yearly basis. The HYMAN model (Twilley and Chen, 1998) is a hydrology model that simulates the mass balance of freshwater and tidal inputs, and calculates porewater and surface salinity. The NUMAN model was developed to simulate the availability of N and P in mangrove soil by coupling the simultaneous effect of productivity and allochthonous inputs of mineral sediment (Chen and Twilley 1998b). In brief, the NUMAN model was modified from the SEMIDEC (Morris and Bowden 1986) and CENTURY (Parton et al. 1987) models of soil organic matter for marshes and grasslands, respectively. NUMAN was calibrated with field measurements in the Shark River estuary and estimated rates of several processes including N mineralization, P accumulation from exogenous sources, and export of organic detritus from mangroves to coastal waters (Chen and Twilley 1998b). FORMAN and NUMAN models were calibrated with sites along the Shark River estuary, but needs to be verified on other mangrove plots in the south Florida region (Tasks 1 and 2, VEGETATION DYNAMICS proposed by Tom Smith). In addition, there are several key parameters that have been identified in each of the manuscripts describing these models that will be the focus of the field and mesocosm studies proposed to continue to develop these unit models. These parameters are important in the use of FORMAN and NUMAN models to evaluate critical restoration scenarios proposed within CESI.

TASK 2-2: Linking FORMAN and NUMAN to MANGRO-SELVA. FORMAN and NUMAN are "unit" modes that do not include dynamic spatial relationships, since it only simulates structure and function of mangroves in 0.5 ha plots. During this project, efforts will be made to specifically determine how the FORMAN and NUMAN models can contribute to the MANGRO-SELVA landscape modeling efforts in the Everglades Watershed (proposal by Dr. Tom Doyle, USGS/BRD, Lafayette). Dr. Tom Doyle will use output and algorithms of the FORMAN and NUMAN models to parameterize some of the macrophyte components of the SELVA landscape model that will eventually be linked to the ATLSS model (Dr. Don DeAngelis, USGS/BRD, Miami). One approach is to use repeated simulations of the FORMAN and NUMAN models to establish parameters and input functions to the SELVA landscape models. Under certain conditions of salinity and soil nutrients, the compensatory effects of nutrient feedback from vegetation and soil nutrients can be passed to the landscape analysis of these conditions. In addition, the SELVA model can predict the landscape distribution of specific wind patterns related to hurricane disturbance that can be used as input to the FORMAN model. Thus information can be passed between both models that include relative processes at scales that are not inherent in either model. In this way, the effects of attributes from the unit to the landscape scale can be included in the evaluations of different CESI scenarios.

TASK 2-3 to 2-5: Nutrient biogeochemistry: The pool of available nutrients in mangrove soils is a product of several processes that proceed on different time scales: plant production (aboveground and belowground), decomposition of litter fall, mineralization of organic matter, input by rainfall and ground water, sedimentation by tide and runoff, and uptake by plants. The NUMAN model was developed to simulate the availability of N and P in mangrove soil by coupling the simultaneous effect of these processes. In brief, the NUMAN model was modified from the SEMIDEC (Morris and Bowden 1986) and CENTURY (Parton et al. 1987) models of soil organic matter for marshes and grasslands, respectively. NUMAN was calibrated with field measurements in the Shark River estuary and estimated rates of several processes including N mineralization, P accumulation from exogenous sources, and export of organic detritus from mangrove to coastal waters. Short term effects of processes such as the net immobilization of nitrogen into decomposition material can be crucial during periods of rapid immobilization and loss in mangrove ecosystems. Results from our nitrogen and phosphorous cycling studies in mangrove forests (Rivera-Monroy et al 1995a, b, Rivera-Monroy and Twilley 1996, Twilley 1996) show the regulatory nature C:N:P ratios have on N and P transformations, particularly P and N immobilization. One of the key transformations that needs further study is documenting the role of allochthonous P inputs and N fixation in the C:N:P ratio of mangrove wetlands. Model modifications will include explicit relationships of N and P transformations such as immobilization and mineralization. By adding mechanisms that describe the partition of C, N, and P exchange (atmosphere and water) and regeneration, we expect to increase our understanding of what processes control the flux of nutrients at the mangrove:coastal water boundary south Florida land-margin ecosystem.

TASK 2-3: Nutrient biogeochemistry-Biomass Allocation and Soil Formation. The NUMAN model (Chen and Twilley 1998b) used a cohort analysis of mangrove soil formation. This initial modeling effort discovered some key parameters that need to be calibrated to understand the soil organic matter and nutrient pools along the estuary gradient in southwest Florida. One of these parameters is the allocation and turnover of above- and below-ground biomass of mangrove wetlands related to soil conditions. We propose field and mesocosm studies of these processes to directly test how they vary along nutrient gradients. Comparisons of organic matter production and decomposition will be evaluated among soil cohorts and compared to the input of mineral matter using Pb-210 techniques (see Nutrient Exchange below). These processes along with the NUMAN model can be compared with results of soil accretion and elevation proposed for study by Smith and Cahoon at selected sites in the Everglades. Parameterization and calibration of the NUMAN model will have important applications to understanding the impacts of specific scenarios of CESI to the long-term sustainability of mangroves under conditions of present sea-level rise.

TASK 2-4: Nutrient biogeochemistry-Nutrients and Hydroperiod. Calibrating the FORMAN and NUMAN models will require a better understanding of the interaction between hydroperiod and nutrient availability on productivity of mangrove and marsh communities. This includes mesocosm experiments of N and P enrichments (three treatments, each) on growth of Rhizophora, Avicennia, Laguncularia under two hydroperiods (long vs short) at two salinity regimes. This is a major effort to provide parameters for the FORMAN and MANGRO-SELVA models proposed by Tom Doyle, USGS/BRD, Lafayette. Decrease in P concentrations alone could not explain 20% of the decrease in forest structure along the Shark River estuary; and there is evidence that this may be due to the additional effects of increased hydroperiod. Thus accurate projections of marsh and mangrove distributions and community structure will depend on model formulations that capture effects of fresh water and nutrient inputs from upland watersheds to the coastal vegetation. These studies will be performed in the new Environmental Greenhouse Facility (1 acre) constructed at USL and includes wetland intertidal mesocosms.

METHODOLOGIES

Estuary: Sampling of inorganic (NH₄, NO₃, NO₂, PO₄, SiO₄) and total nitrogen and phosphorus will be sampled in the Shark River estuary adjacent to the six mangrove sites proposed in this study. This water quality survey will include measurements of water salinity, temperature, and dissolved oxygen at each of the sites. Forest Soil: Three 40 cm deep cores will be obtained seasonally from each site using a 10 cm diameter aluminum coring apparatus. Soil total carbon (C), nitrogen (N), and phosphorus (P), and extractable N and P, along with soil bulk density will be measured at 5 cm intervals. Forest Pore water analysis: Pore water samples will be collected at each site seasonally (three times during the study) using a plastic siphon with syringe from depth 0, 10, 20 and 40 cm. On aliquot of pore water is filtered and assayed for inorganic nutrient (NH4, NO3 and PO4) and salinity. Another aliquot of the porewater sample will be added to an antioxidant buffer in the field, then brought to the laboratory to be analyzed for sulfide concentrations. Modeling: Results of these field studies will be used to parameterize the NUMAN model and focus on estimates of nutrient exchange between mangroves and estuarine waters of the Shark River estuary. These estimates will be linked to crude box model of the estuary using information from USGS to estimate the loading of nutrients from Shark River estuary to coastal boundary current.

SCHEDULE OF PROJECT MILESTONES: This schedule assumes a start-up during the first quarter of FY99 and initial funding of three years with expected 5 year planning table and integration with ATLSS program elements.

	YR1	YR2	YR3
TASK 1: SELVA-MANGRO
1-1. Plot Mapping	X	X
1-4. Site Quality	X	X	X
1-5. Disturbance Maps		X	X
1-6. Hydrology/Edge Maps		X	X
1-7. Model Linkages		X	X
TASK 2: FORMAN-NUMAN
2-1. Verification of FORMAN and NUMAN	X	X
2-2. Model Interface with MANGRO-SELVA	X	X	X
2-3. Biomass Turnover	X	X	X
2-4. Nutrients/Hydroperiod	X	X	X

SCHEDULE OF ACTIVITIES AND DELIVERABLES – 2000 & 2001:

Reports and Deliverables:
Type of Product*	No. of Copies	Due Date
1. Brief letter report with the following: [1] date of receipt of executed contract (i.e. start date), [2] this "table" with specific due dates, [3] status of progress made towards providing data/metadata/model-source-code, [4] draft text w/ graphics/slides for 2-pager Fact Sheet.	Original + 2 + electronic copy	60 days after award of contract.
2. 1st Trimester Report – a brief report updating progress/problems to date and all data and metadata file, &/or model source code to date. Plus, final of 2-pager Fact Sheet. For ENDING projects, in addition to above, provide a plan for final report in a detailed outline format.	Original + 2 + electronic copy	120 days after award of contract.
3. 2nd Trimester Report – with updated data/metadata file[s] &/or model source code. Plus, "Request for Continued Funding" for the next year funding. For ENDING projects, provide a DRAFT final report suitable for external review.	Original + 2 + electronic copy	240 days after award of contract.
4. Annual or Final Report with all data to date with metadata file[s], fully documented model source code. For FINAL modeling projects, a fully executable, fully documented model source code is required.	Original + 2 + electronic copy	1 Year after award of contract.
* Note: Manuscripts, peer reviewed publications, book chapters, graduate student thesis/dissertation, etc. are both acceptable and desirable as chapters or sections of annual/final reports. At least one 2-page fact sheet is required for each of these major types of publications.

PROJECT MANAGEMENT AND COLLABORATION: The overall coordination of the project will be directed by Dr. T. W. Doyle involving cooperation and exchange with Dr. Don DeAngelis of the ATLSS program and Dr. T. J. Smith regarding program code standards and field data for model calibration and verification. Dr. Doyle will direct Task 1 assignments and deliverables while Dr. Robert Twilley, USL and project co-investigator, will be responsible for Task 2 elements and deliverables.

LITERATURE CITED:

Chen, R.H. 1996. Ecological analysis and model simulation of mangrove forest development and soil characteristics in the Shark River estuary, Florida. Ph.D. Dissertation, University of Southwestern Louisiana, Lafayette.

Chen, R. And R. R. Twilley. 1998a. A gap dynamic model of mangrove forest development along gradients of soil salinity and nutrient resources. Journal of Ecology 86:37-52.

Chen, R. And R. R. Twilley. 1998b. A simulation model of organic matter and nutrient accumulation in mangrove wetland soils. Biogeochemistry XX:000-000.

Doyle, T. W. 1994. Field and modeling studies of hurricane disturbance on Gulf Coastal Forest Ecosystems. pp.167-173. IN: Global Strategies for Environmental Issues: NAEP 19th Annual Conference Proceedings held June 12-15, 1994, New Orleans, Louisiana. NAEP Publications, Washington D. C.

Doyle, T. W. 1997. Modeling hurricane effects on mangrove ecosystems. USGS Fact Sheet-095-97. June 1997.

Doyle, T. W., C. J. Wells, K. W. Krauss, and M. Roberts. 1994. The Use of Videography to Analyze the Spatial Impact of Hurricane Andrew on south Florida mangroves. pp. 222-227. IN: Proceedings of GIS/LIS 94 Annual Conference held October 27-30, 1994, Phoenix, AZ. Published by American Congress on Surveying and Mapping, Bethesda, Maryland.

Doyle, T. W., T. J. Smith, and M. B. Robblee. 1995. Wind damage effects of Hurricane Andrew on mangrove communities of southwest Florida. Journal of Coastal Research 18: 144-159.

Doyle, T. W. and G. F. Girod. 1997. The frequency and intensity of Atlantic hurricanes and their influence on the structure of south Florida mangrove communities. Pages 109-120 in H.F. Diaz, and R. S. Pulwarty, editors. Hurricanes: Climate and Socioeconomic Impacts. Springer-Verlag, Heidelberg, Germany.

Rivera-Monroy, V. H., Day, J. W., Twilley, R. R., Vera-Herrera, F., and Coronado-Molina, C. 1995. Flux of nitrogen and sediment in a fringe mangrove forest in Terminos Lagoon, Mexico. Estuarine, Coastal and Shelf Science 40: 139-160.

Rivera-Monroy, V. H., Twilley, R. R., Boustany, R. G., Day, J. W., Vera-Herrera, F., and Ramirez, M. C. 1995. Direct denitrification in mangrove sediments in Terminos Lagoon, Mexico. Marine Ecology Progress Series 126: 97-109.

Rivera-Monroy, V. H., and Twilley, R. R. 1996. The relative role of denitrification and immobilization on the fate of inorganic nitrogen in mangrove sediments of Terminos Lagoon, Mexico. Limnology and Oceanography 41: 284-296.

Roberts, M., C. Wells, and T. W. Doyle. 1994. Component analysis for interpretation of time series NDVI imagery. pp.538-550. IN: Volume 1, 1994 ASPRS/ACSM Annual Convention and Exposition, Technical Papers held April 25-28, 1994 in Reno, NV. Published by ASPRS/ACSM, Bethesda, MD.

Twilley, R.R. 1997. Mangrove wetlands, pp. 445-473. In, M.Messina and W. Connor (eds.), Southern Forested Wetlands: Ecology and Management, CRC Press, Boca Raton, Florida.

Twilley, R. R. 1995. Properties of mangrove ecosystems related to the energy signature of coastal environments. Pages 43-62 in C. Hall, Editors. Maximum power. University of Colorado Press, Boulder, Colorado, USA.

Twilley, R.R. and R. Chen. A water budget and hydrology model of a basin mangrove forest in Rookery Bay, Florida. 1998. Australian Journal of Freshwater and Marine Research. 49:309-323.

Back to Project Homepage