The authors have declared that no competing interests exist.
Conceived and designed the experiments: LLG KF JB DRG. Performed the experiments: LLG KF JB DRG APB. Analyzed the data: LLG KF JB DRG APB. Wrote the paper: LLG DRG KF JB APB. Ran the SLAMM analyses: KF.
¶ These authors also contributed equally to this work.
The Sea Level Affecting Marshes Model (SLAMM) was applied at six major estuaries along Florida’s Gulf Coast (Pensacola Bay, St. Andrews/Choctawhatchee Bays, Apalachicola Bay, Southern Big Bend, Tampa Bay and Charlotte Harbor) to provide quantitative and spatial information on how coastal ecosystems may change with sea level rise (SLR) and to identify how this information can be used to inform adaption planning. High resolution LiDAR-derived elevation data was utilized under three SLR scenarios: 0.7 m, 1 m and 2 m through the year 2100 and uncertainty analyses were conducted on selected input parameters at three sites. Results indicate that the extent, spatial orientation and relative composition of coastal ecosystems at the study areas may substantially change with SLR. Under the 1 m SLR scenario, total predicted impacts for all study areas indicate that coastal forest (-69,308 ha; -18%), undeveloped dry land (-28,444 ha; -2%) and tidal flat (-25,556 ha; -47%) will likely face the greatest loss in cover by the year 2100. The largest potential gains in cover were predicted for saltmarsh (+32,922 ha; +88%), transitional saltmarsh (+23,645 ha; na) and mangrove forest (+12,583 ha; +40%). The Charlotte Harbor and Tampa Bay study areas were predicted to experience the greatest net loss in coastal wetlands The uncertainty analyses revealed low to moderate changes in results when some numerical SLAMM input parameters were varied highlighting the value of collecting long-term sedimentation, accretion and erosion data to improve SLAMM precision. The changes predicted by SLAMM will affect exposure of adjacent human communities to coastal hazards and ecosystem functions potentially resulting in impacts to property values, infrastructure investment and insurance rates. The results and process presented here can be used as a guide for communities vulnerable to SLR to identify and prioritize adaptation strategies that slow and/or accommodate the changes underway.
Coastal wetland systems and human communities will be substantially affected whether global sea level rises 0.18–0.59 m by 2100, as estimated by the IPCC (2007) [
In Florida, sustainable coastal habitats are critical drivers of both the economy and quality of life. The health and sustainability of coastal ecosystems control the future of the state’s recreational and commercial fisheries, recreational boating and diving, beach-related recreation, tourism, nature observation and other ecosystem dependent activities, collectively worth hundreds of billions of dollars a year to the state’s economy [
In many parts of the United States, human communities and infrastructure are increasingly vulnerable to storm surge effects in the face of SLR. Shepard
Among the tools developed to enhance our understanding of the effects of SLR on coastal ecosystems is the Sea Level Affecting Marshes Model (SLAMM; open source and available at:
The study objective is to illustrate the utility of SLAMM-derived qualitative, quantitative and spatial information on potential changes to coastal ecosystems due to predicted SLR for informing adaptation planning. SLR impacts on coastal ecosystems were modeled at six estuarine study areas along the State of Florida’s Gulf of Mexico coast (3 SLR scenarios at each site through the year 2100: 0.7 m, 1.0 m and 2.0 m). In addition, uncertainty analyses were conducted on selected SLAMM input parameters at three study areas to better understand how these input parameters affect model output. Model predictions have been made available through workshops, reports, and an interactive, online decision support tool (
The analysis included six study areas that encompassed wetlands and uplands surrounding the major estuarine systems along the Florida Gulf of Mexico coast. Study areas were defined to encompass lands generally up to the 2 meter elevation contour and included three in the Florida Panhandle (Pensacola Bay, St. Andrews/ Choctawhatchee Bays and Apalachicola Bay), two in Southwest Florida ((Tampa Bay and Charlotte Harbor) and one along the transitional coast between the two regions (Southern Big Bend) (
The project study areas and vegetation raster inputs are illustrated in color for all study areas. The vegetation raster inputs represent the initial condition of coastal ecosystems used in the SLAMM analyses. Subsites are identified within study areas by numbers 1–4. Subsites were created where tidal parameters vary substantially from the global site (unmarked area) or where freshwater flow is substantial as along some river floodplains. Vicinity maps for each project study area along the Florida Gulf Coast are provided next to each study area map.
The more northerly study areas (Pensacola Bay to the Southern Big Bend) are similar in that a relatively higher percentage of their coastal wetlands are represented by coastal forest (18% to 44%) as compared to the more southerly study areas (Tampa Bay and Charlotte Harbor) with only 4% to 8% of their coastal wetlands represented by coastal forest (
Some characteristics of the 6 study areas are presented including area, mean inflow, mean depth, average salinity, great diurnal tide, extent of coastal wetlands and the type and extent of the predominant subtidal habitat (not including bare sediments).
Study Area | Area (ha) | Mean Inflow (m3/s) | Mean Depth (m) | Average Salinity (ppt) | Great Diurnal Tide (m) | Coastal Wetlands (ha) | Predominant Subtidal Habitat excluding Bare Sediment (ha) |
---|---|---|---|---|---|---|---|
Pensacola Bay | 351,679 | 328 | 4 | 23 | 0.4 | > 63,000 | seagrass; 1,800 |
St. Andrews/ Choctawhatchee Bays | 1,581,932 | 241 | 5 | 25 | 0.16 | 238,353 | seagrass; 6,500 |
Apalachicola Bay | 274,525 | 824 | 3 | 22 | 0.7 | > 117,000 | oyster reef; > 1,400 |
Southern Big Bend | 740,624 | —- |
3 |
15 | 0.6 to 1.2 | > 90,960 | seagrass; 153,470 |
Tampa Bay | 602,639 | 68 | 3 | 27 | 0.7 to 0.9 | > 94,000 | seagrass; 12,100 |
Charlotte Harbor | 761,670 | 136 | 2 | 13 | 0.4 | > 114,000 | seagrass; > 23,900 |
a Unknown.
b Southern Big Bend is not an enclosed estuary. Open ocean areas adjacent to the coast are less than 3 m in depth.
c This is likely an underestimate because seagrass has not been estimated in deeper waters due to insufficient water clarity.
Sources: [
Pensacola Bay is a relatively deep estuary along Florida’s Gulf Coast due to its location on the coast where the slope into the Gulf is steep and elevations close to shore are the highest among the study areas (
Tidal parameters and location of stations informed the creation of subsites for each study area as illustrated in
The St. Andrews/Choctawhatchee Bays study area combines two adjacent bay systems (
A key feature of Apalachicola Bay system is the large fluvial delta formed by the Apalachicola River, the largest freshwater source flowing into the Bay and the largest river by flow in Florida. Sediment has accumulated in this delta system and into the bay since the Holocene from a large drainage basin that includes much of Georgia and Alabama [
The Southern Big Bend region is not a typically shaped estuary (
Tampa Bay is a shallow estuary located in a transition zone between warm-temperate and tropical biogeographic provinces. As a consequence of its location, the bay supports a highly diverse flora and fauna [
Charlotte Harbor is a shallow, subtropical estuary made up of a number of bays and surrounded by large areas of protected coastal ecosystems. The estuary is the second largest in Florida and is separated from the Gulf of Mexico by barrier islands. Two major inlets and several smaller passes connect the estuary to the Gulf. The shoreline is mostly undisturbed except along the Caloosahatchee River where urban and residential development has proliferated [
SLAMM was applied to each study area using three SLR scenarios through the year 2100, the IPCC A1B maximum scenario (0.7m), and two additional scenarios, 1 m and 2 m to capture a range of scientifically supported scenarios. Global SLR scenarios are appropriate to use in Florida because local SLR has generally mimicked global mean SLR over the last couple of decades with some variation along Florida’s Gulf Coast (1.6 to 3.19 mm per year) as calculated from tide gauge records [
SLAMM requires two types of inputs to simulate changes in coastal wetlands due to SLR: raster (i.e., spatial) data and numeric site parameters. The raster datasets are uploaded to SLAMM via the user interface. The numeric site parameters are entered into the SLAMM user interface in a tabular format. Together the raster data and numeric parameters describe the site to be modeled. In addition to the two types of inputs, the user specifies SLR scenarios and selects optional model switches that control various aspects of the simulation via the user interface.
SLAMM requires three raster data inputs, vegetation, elevation and slope. The source of these inputs and any required processing are described below:
The comprehensive, statewide vegetation dataset, the Cooperative Land Cover Map v1.1 (CLC) available from the Florida Natural Areas Inventory (FNAI) in vector format) [
High resolution, LiDAR-derived Digital Elevation Models (DEMs) available from a water management district or the Florida Division of Emergency Management were used to create the required elevation raster for each study area (
The third required raster input, a slope raster, was created from the final DEM using the Slope tool in the Spatial Analyst extension of ArcGIS.
Each study area was modeled using the best available site-specific numeric parameters. Twenty-one SLAMM required numeric input parameters were defined for each study area and subsite and included the dates the elevation and vegetation data were collected and site specific tidal, SLR, erosion, accretion, sedimentation and overwash parameters (
SLAMM allows for each subsite to have its own set of numeric parameters, thus allowing for spatial variation to be modeled. In addition to subsites defined due to variations in numeric input parameters, freshwater flow sites were defined in some study areas.
For each study area, the three SLR scenarios (0.7 m, 1 m and 2 m) were run with 25 year time steps through the year 2100. GIS output was obtained for each time step with every scenario. Each scenario was run twice, once with developed and undeveloped dry land allowed to become inundated, and once with developed dry land protected from inundation at all study areas except the St. Andrews/Choctawhatchee Bay study area. The latter scenario simulated protecting existing development (e.g., through shoreline hardening), a likely SLR adaptation response. All scenarios were run using the optional connectivity algorithm. This option allows dry land and freshwater wetlands to become inundated with saltwater only if there is a connection to a saltwater source.
The numerical parameters for each study area’s SLAMM simulation were developed with data available at the time. For some sites, site specific parameter information was available for all numerical input parameters (e.g., saltmarsh accretion rate, beach sedimentation rate, etc.). For other sites, site specific parameter information was not available or only available for some of the numerical input parameters. The simulated results of the scenarios presented below have all of the types of uncertainty inherent to simulation models [
SLAMM’s uncertainty analysis module was used to examine input uncertainty for selected parameters at three study areas. The uncertainty analysis module allows users to specify a distribution for any input parameter, or multiple parameters, and performs a Monte-Carlo analysis. The number of iterations was specified using a non-random seed. A limitation of SLAMM's uncertainty module is that the values drawn from the distribution are multipliers applied to a single value of the parameter in question. That is, the parameter value to which the multiplier is applied does not change even if that parameter changes across subsites.
The uncertainty analyses were limited by computer resources. With small cell sizes of 15 or 30 meters and available computer capacity, 100 iterations could take 24 hours or more. As a result, the effects of input uncertainty on 2 or 3 numeric parameters were examined at the study areas evaluated (Pensacola Bay, Southern Big Bend and Tampa Bay). The number of iterations was set to 100 for the 1 m SLR scenario. While 100 iterations can reveal trends, this number of iterations may not be enough to capture extremes. For the Pensacola Bay study area, the uncertainty module was run with a representative subset of the study area to shorten the runtime of the 100 iterations. For Tampa Bay study area, three parameter distributions were included in the 100 model iterations. All uncertainty analyses were run on the 1 m SLR scenario with developed dry land allowed to transition.
Three parameters for uncertainty that appeared influential in simulation outcomes were selected for the uncertainty analyses: marsh accretion, sedimentation and salt elevation. Generally, distributions used were plus or minus a likely spread from the input value based on values from the literature at nearby sites. Triangular distributions were used when there was a basis that a value for a parameter was more likely. In some cases, minimum and maximum values were chosen so that the distribution would encompass values from research that indicated a wider range as noted in
U = Uniform distribution (minimum, maximum); T = Triangular distribution (minimum, most likely, maximum).
Study Area | SLAMM Parameter | Input Value | Distribution and Source |
---|---|---|---|
Pensacola | Saltmarsh Accretion (mm/yr) | 2.25 | T(0.9, 3.2, 8) [ |
Pensacola | Brackish marsh Accretion (mm/yr) | 3.75 | U(3,4) [ |
Southern Big Bend | Saltmarsh Accretion (mm/yr) | 7.2 | T(0.7, 7.2, 7.6) [ |
Southern Big Bend | Beach Sedimentation Rate (mm/yr) | 0.5 | U(0.375, 0.625) |
Tampa Bay | Salt Elevation (m above MTL) | 0.55 | U (0.495, 0.605) |
Tampa Bay | Saltmarsh Accretion (mm/yr) | 2.25 | T(0.9, 2.25, 8.0) [ |
Tampa Bay | Beach Sedimentation Rate (mm/yr) | 2.7 | T (0.01, 2.5, 5) [ |
To simplify reporting of the results, focus was placed on the 1 m SLR scenario, which is in the mid-range of what are considered plausible scenarios [
In the Pensacola Bay study area (including the Florida portion of Perdido Bay), SLAMM forecast simulations estimated that 18% (-6,408 ha) of study area coastal forest would be lost under the 1 m SLR scenario (
(A) Bar graph of loss/gain of coastal ecosystems under 3 sea level rise scenarios. (B) Map of SLAMM results illustrating the change in coastal ecosystem types (from/to) under a 1 m sea level rise scenario.
Pensacola Bay | St. Andrews/ Choctawhatchee Bays | Apalachicola Bay | Southern Big Bend | Tampa Bay |
Charlotte Harbor | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Coastal Ecosystem | Change (ha) | % Change | Change (ha) | % Change | Change (ha) | % Change | Change (ha) | % Change | Change (ha) | % Change | Change (ha) | % Change |
-1,075 | -1% | -4,360 | -0.6% | -3,084 | -5% | -9,066 | -14% | -3,868 | -4% | -6,724 | -4% | |
-6,408 | -18% | -10,241 | -6% | -20,857 | -27% | -24,655 | -49% | -2,413 | -8% | -4,336 | -24% | |
1,826 | 2021% | 3,747 | 312% | 3,020 | 156% | 1,333 | 34% | -14,627 | -93% | -20,524 | -96% | |
-1,048 | -11% | -1,125 | -3% | -4,379 | -20% | -758 | -10% | -276 | -2% | -446 | -1% | |
-1,157 | -44% | na | na | -3,357 | -56% | -353 | -80% | na | na | -376 | -96% | |
-370 | -28% | -123 | -1% | -644 | -24% | -140 | -9% | -36 | -0.4% | -29 | -0.3% | |
-376 | -17% | 50 | 6% | -94 | -26% | 40 | 217% | -130 | -12% | -201 | -27% | |
17 | na | -595 | -33% | 165 | 744% | 57 | 5702% | -25 | -13% | 38 | na | |
1,536 | Na | -188 | -41% | 2,303 | 60% | -12 | -37% | -41 | -94% | na | na | |
-58 | -2% | -1,809 | -41% | 7,834 | 269% | 107 | 10668% | -180 | -92.3% | |||
na | na | na | na | na | na | -121 | -44% | 4,453 | 63% | 8,251 | 34% | |
1,357 | na | 6,547 | na | 3,081 | na | 11,592 | na | -14 | -17% | 495 | na | |
4,166 | na | 4,862 | na | 10,503 | na | 15,736 | 50% | -1,491 | -73% | -342 | -5% | |
-512 | -<1% | +1,125 | +0.5% | -2,425 | -2% | 2,826 | 3% | -14,780 | -19% | -17,507 | -15% |
1Results exclude tidal flats with no elevation data (approximately 10,000 ha).
Under the SLAMM scenarios with developed dry land allowed to transition (
SLAMM forecast simulations for St. Andrews/Choctawhatchee Bays study area estimated that approximately 6% of study area coastal forest was lost under the 1 m SLR scenario (
(A) Bar graph of loss/gain of coastal ecosystems under 3 sea level rise scenarios. (B) Map of SLAMM results illustrating the change in coastal ecosystem types (from/to) under a 1 m sea level rise scenario.
As with the previous two study areas discussed, the Apalachicola Bay study area was predicted by SLAMM to lose a large amount of coastal forest under 1 m of SLR (-20,857 ha or -27%;
(A) Bar graph of loss/gain of coastal ecosystems under 3 sea level rise scenarios. (B) Map of SLAMM results illustrating the change in coastal ecosystem types (from/to) under a 1 m sea level rise scenario.
The loss of coastal forest to marsh noted above resulted in large predicted increases in saltmarsh (+10,503 ha; percent not available), brackish marsh (+7,834 ha; +269%) and transitional saltmarsh (+3,081 ha; percent not available) and tidal freshwater marsh (+2,303 ha; +60%;
A comparison of the scenarios run with developed dry land allowed to transition revealed few differences with the developed land protected scenarios (
Similar to the Florida Panhandle study areas, SLAMM predicted a large loss of coastal forest extent in the Southern Big Bend study area (-24,655 ha; -49%;
(A) Bar graph of loss/gain of coastal ecosystems under 3 sea level rise scenarios. (B) Map of SLAMM results illustrating the change in coastal ecosystem types (from/to) under a 1 m sea level rise scenario.
SLAMM predicted an increase in saltmarsh (+15,732 ha; +50%), transitional saltmarsh (+11,592 ha) and tidal flat (+1,333 ha; +34%). Percentage changes are not reported for some coastal habitats due to their very low representation currently in the study area (approximately 1 ha or less) making their percent gain information less meaningful. A net gain of study area coastal wetlands was predicted by SLAMM in the Southern Big Bend study area (+2,826 ha; +3%) and is primarily a consequence of coastal forest loss.
As with the St Andrews/Choctawhatchee Bays study area, there appeared to be a threshold reached somewhere between 1 m and 2 m of SLR by 2100 at this study area (
Under the SLAMM scenarios with developed dry land allowed to transition (
SLAMM predicted modest to substantial changes in coastal ecosystems due to SLR in the Tampa Bay study area under a 1 m SLR scenario (
(A) Bar graph of loss/gain of coastal ecosystems under 3 sea level rise scenarios. (B) Map of SLAMM results illustrating the change in coastal ecosystem types (from/to) under a 1 m sea level rise scenario.
As with other study areas, there appeared to be a threshold between 1 m and 2 m of SLR by 2100: with 1 m SLR, mangrove forest gained 4,453 ha or 63% in extent. However, at the rate of 2 m SLR by 2100, mangrove forest was predicted by SLAMM to lose 2,589 ha in extent or 37%. In addition, loss of undeveloped dry land nearly doubled from the 1 m SLR scenario to the 2 m SLR scenario (-3868 ha versus -7263 ha). Under the 1 m scenario, several of the lowest elevation wetland types transitioned into tidal flat. Coastal forest and undeveloped dry land transitioned to mangrove forest and some mangrove forest was lost to shallow subtidal habitat.
Under the SLAMM scenarios with developed dry land allowed to transition (
As with the Tampa Bay study area, sub-tropical Charlotte Harbor was predicted by SLAMM to lose the majority of the tidal flat ecosystem now present in this system under a 1 m SLR scenario (-29,524 ha; -96%;
(A) Bar graph of loss/gain of coastal ecosystems under 3 sea level rise scenarios. (B) Map of SLAMM results illustrating the change in coastal ecosystem types (from/to) under a 1 m sea level rise scenario.
Again, a threshold was reached between the 1 m and 2 m SLR scenarios for several of the coastal ecosystems (
Under the SLAMM scenarios with developed dry land allowed to transition (
The spatial results for all sites and SLR scenarios modeled have been posted for viewing on Coastal Resilience 2.0, a suite of map based, decision-making tools for coastal risk assessment available on the web (
Overall, coastal wetland ecosystems at all six study areas along Florida’s Gulf Coast are likely to change substantially. Under the 1 m SLR scenario, SLAMM predicted that coastal forest, tidal flat, inland freshwater marsh and tidal swamp will lose the most spatial extent, -69,309 ha, -25,552 ha, -7,733 ha, and -5,069 ha, respectively (
The SLAMM results illustrated in this bar graph are for the following 3 sea level rise scenarios: 0.7m, 1m, 2m. All scenarios were run with developed dry land protected from change in the SLAMM user interface.
All Study Areas | 0.7m | 1 m | 2 m | ||||
---|---|---|---|---|---|---|---|
Coastal Ecosystem | Initial Condition | Change (ha) | Percent Change | Change (ha) | Percent Change | Change (ha) | Percent Change |
1,172,423 | -18,071 | -2% | -28,445 | -2% | -66,523 | -6% | |
392,433 | -56,447 | -14% | -69,309 | -18% | -98,924 | -25% | |
54,069 | -27,016 | -50% | -25,552 | -47% | 2,724 | 5% | |
124,373 | -4,837 | -4% | -7,733 | -6% | -15,062 | -12% | |
9,471 | -3,490 | -37% | -5,069 | -54% | -4,143 | -44% | |
33,496 | -887 | -3% | -1,303 | -4% | -2,484 | -7% | |
5,195 | -851 | -16% | -692 | -13% | -1,832 | -35% | |
2,010 | -395 | -20% | -341 | -17% | -729 | -36% | |
4,418 | 6,563 | 149% | 3,594 | 81% | 1,692 | 38% | |
10,652 | 6,944 | 65% | 6,365 | 60% | 3,224 | 30% | |
31,230 | 11,014 | 35% | 12,583 | 40% | -18,272 | -59% | |
114 | 29,855 | na |
23,645 | na |
32,672 | na |
|
37,379 | 20,100 | 54% | 32,923 | 88% | 4,689 | 13% | |
704,839 | -19,447 | -3% | -30,890 | -4% | -96,447 | -14% |
1No transitional saltmarsh was present in the initial condition so percent change cannot be calculated.
Under the SLAMM scenarios with developed dry land allowed to transition, approximately 2%, 4% and 15% of developed dry land were predicted to be lost under the 0.7 m, 1 m and 2 m scenarios, respectively (
Uncertainty results were characterized by subtracting the minimum output from the maximum output values (
Study Area | PENS | PENS | SBB | SBB | TAM |
---|---|---|---|---|---|
Distribution Type | Uniform | Triangular | Triangular | Uniform | Multi |
Distribution Parameter | Brackish Marsh Accr | Saltmarsh Accr | Saltmarsh Accr | Sedimentation | Multi |
Variable Name | Max-Min | Max-Min | Max-Min | Max-Min | Max-Min |
0.1 | 0.8 | 0.8 | 2035.1 | ||
0 | 0.6 | 0.3 | 483.7 | ||
0 | 2.4 | 0.6 | 195.2 | ||
0 | 0 | 0 | 1.7 | ||
1.2 | 0 | 0 | 0 | 51.8 | |
0 | 0 | 0 | 0 | 0 | |
103.2 | 0.7 | 9.1 | 0.7 | 28.4 | |
144.5 | 297.1 | 26462.5 | 1.4 | 1554.6 | |
0 | 0 | 0.1 | 0 | 2591.5 | |
1.3 | 0 | 32.2 | 2.4 | 53.8 | |
186.3 | 158 | 16775 | 126.6 | 1829.5 | |
2.2 | 0 | 1.2 | 2.4 | 222 | |
333.7 | 0 | 0 | 0 | 0 | |
0 | 0 | 0 | 0 | 0 |
PENS = Pensacola Bay; SBB = Southern Big Bend; TAM = Tampa Bay; Accr = accretion.
The uncertainty analysis for the Pensacola Bay Study Area was conducted on only a portion of the study area (87,796 ha versus 351,679 ha) due to the extremely long runtime of the analysis on the entire site. At the Pensacola Bay study area, the uncertainty analyses on brackish marsh accretion and saltmarsh accretion revealed only small to modest changes in the 2100 results for the various wetlands types with changes to the parameters as illustrated in
At the Southern Big Bend study area, the uncertainty analyses run on the two selected input parameters, saltmarsh accretion and sedimentation rate, revealed small to modest changes in the 2100 results for the various wetlands types with changes to the sedimentation rate (
The uncertainty analysis for the Tampa Bay study area was run on three input parameters: salt elevation, saltmarsh accretion rate and beach sedimentation rate. This analysis revealed low to moderate changes in the 2100 results for the various wetland and dry land types with changes to the selected input parameters (
Overall, the uncertainty results identified small to moderate concerns over the precision of the SLAMM results. The greatest uncertainty arose in parameters directly tied to the response of coastal ecosystems to SLR such as with marsh accretion rates. Precision uncertainty is greatly diminished at sites where long-term studies of coastal ecosystem accretion, sedimentation and erosion have been conducted.
These results illustrate the utility of SLAMM for quantitatively, qualitatively and spatially describing how coastal wetland ecosystems and adjacent dry land areas may change as a result of predicted changes in SLR and how this information can be used to develop and prioritize adaptation strategies. The spatial extent of the change varies with the coastal habitat type, elevation of upland areas, the rate of SLR and other wetland processes. The uncertainty analyses highlighted the need for collecting long-term site specific data on accretion, erosion and sedimentation to improve the predictive capabilities of SLAMM. While SLAMM predicts that one coastal ecosystem type will replace another ecosystem type under a particular SLR scenario, it gives no indication of how long the predicted transition may take. For example, SLAMM may report that a tidal swamp transitions into marsh. Field studies in the Waccasassa Bay area [
In addition to changes in spatial extent, the spatial orientation and relative percent coverages of coastal wetland ecosystems are predicted to change with SLR. Dependent species and adjacent human communities will be substantially affected if the predicted large-scale changes to coastal wetland ecosystems occur. Human communities located adjacent to these changed coastal wetland ecosystems may become more or less vulnerable to coastal storm impacts depending on the changes to coastal wetland ecosystems in their vicinity [
In addition to providing protection from coastal storms, coastal wetlands provide a number of other services, including erosion control, fish and wildlife habitat, recreation, and carbon sequestration. Changes in wetland ecosystem spatial extent and composition can affect all of these services. With respect to wildlife habitat, studies of Florida species likely to be vulnerable to the effects of SLR predicted that high percentages (21 to 39%) of the plant and animal species evaluated would likely suffer adverse effects or extinction from SLR as a result of habitat contraction [
Spatial results can be used to identify promising locations for both protecting areas that will allow migration of coastal wetlands inland as sea level rises and restoring critical habitats that have been lost. The results can also be used to identify conservation priorities such as resilient wetland areas that are likely to persist despite sea level rise. Restoration can be based on where coastal wetlands are likely to become inundated, where coastal forests are likely to become marshes, or where undeveloped dry land is likely to become wetlands. Communities can use these results to assist with the development and siting of adaptation projects such as living shorelines, oyster reefs, wetland and upland habitat restoration, water flow preservation and sediment management to support vulnerable marsh, mangrove forest and dry land areas. A nature-based solution may be more cost effective and confer more benefits than an engineered solution to support shorelines [
The modeling results point to several potential adaptation strategies (see
Project Area | SLR Result, 1 m by 2100 | Potential Adaptation Action |
---|---|---|
Large amounts of brackish marsh transitions to saltmarsh, tidal flat and open water in several areas. | Consider areas adjacent to open water for living shorelines to minimize or reduce the rate of transition. As a result, brackish marsh will persist longer for the benefit of dependent species. | |
Ocean beaches are at risk of diminishing in extent as SLR rises. | Prioritize beach renourishment activities to maximize social and ecological benefits (e.g., tourism, sea turtle and shorebird nesting habitat). | |
Coastal forest transitions to marsh in several areas. One of these areas is adjacent to a developed area (Santa Rosa Beach). | Install/construct living shorelines to support brackish marsh and coastal forest in proximity to developed areas to maintain to their wind and wave energy reduction properties as long as possible. | |
A large area of freshwater marsh at the mouth of the Apalachicola River transitions to saltmarsh. | This area may be a good candidate for the installation of living shoreline habitat along the edge bordering open water to support the freshwater marsh and its floodwater absorption values in this area as long as possible. | |
This entire area is likely to experience a large transition of coastal forest to saltmarsh. | Maintain freshwater flows to this coast to the greatest extent possible. Doing so will slow the transition of coastal wetland ecosystems. | |
There is a large area around the bay where wetlands and tidal flats transition to open water. | Install living shorelines where wetlands are transitioning to open water. Doing so will support these coastal wetland ecosystems as long as possible and will extend the life of the storm surge reduction benefits that they provide. | |
There are areas where undeveloped dry land transitions to mangrove forest. | Protect the land and shoreline where mangroves are predicted to expand and facilitate their expansion where needed to take advantage of wave attenuation and reduced flooding where mangroves are present (see text). | |
There are a few areas where mangrove forest transitions to open water. | These areas are good candidate sites for installing living shorelines to maintain the protective qualities of mangroves as long as possible. |
Communities along Florida’s Gulf Coast and in other vulnerable areas that take action in the near term to implement adaptation strategies will have a longer planning horizon across which to spread adaptation costs and reap benefits. These communities will have a greater ability to mainstream adaptation into existing activities with the result of greatly increasing the affordability of these actions [
(PDF)
Study areas are listed from west to east.
(PDF)
(PDF)
(PDF)
(PDF)
The St. Andrews/Choctawhatchee Bay Results were provided by the Texas Chapter of The Nature Conservancy.
(PDF)
(PDF)
We thank staff of Warren Pinnacle Consulting, Inc. for the generosity of their time helping troubleshoot various applications of SLAMM. We also thank staff of the Florida Natural Areas Inventory for providing their expertise on the various vulnerable species analyses. Staff from the following organizations assisted us by providing data and advice on our project site work and with helping us to organize the study area workshops, namely, the US Environmental Protection Agency in the Pensacola Bay Area; the Choctawhatchee Basin Alliance and the South Walton Center of the Northwest Florida State College in the St. Andrews/Choctawhatchee Bays area; Apalachicola National Estuarine Research Reserve in the Apalachicola Bay area; Florida Department of Environmental Protection in the Southern Big Bend Area; Tampa Bay Estuary Program in the Tampa Bay Area; and Sanibel-Captiva Conservation Foundation and the Charlotte Harbor National Estuary Program in the Charlotte Harbor Area. Thank you also to staff of The Nature Conservancy: Lara Rainbolt for assisting with literature searches, Chris Bergh and Christine Shepard for editing assistance, and Linda Finch for helping with workshop logistics.