Question 4: Lateral Gradients
|
||||||||
|---|---|---|---|---|---|---|---|---|
Research QuestionQ4: What are the underlying mechanisms by which proximity of marshes to upland habitat drives ecosystem change along lateral gradients in the intertidal zone? OverviewThe GCE LTER is investigating the underlying mechanisms by which proximity of marshes to upland habitat drives ecosystem change along lateral gradients in the intertidal zone. Our approach to this question involves taking advantage of marsh hammocks as a natural laboratory for evaluating the influence of landscape structure and freshwater input on marsh processes. The hammock research includes basic characterization of groundwater flow as well as physical and biological characteristics at selected sites, experiments designed to understand the effects of manipulating water flow on marsh processes, and modeling. Research Components
|
||||||||
Research Question BackgroundThe GCE domain includes a diverse array of marsh hammocks, which are upland areas nested between the mainland and larger barrier islands. Marsh hammocks range in size from less than a hectare to tens of hectares. Most are remnants of high ground of either Pleistocene or Holocene age, but there are also man-made hammocks that have developed from dredge spoil or ballast stones. Although many of the largest hammocks are developed and have paved roads and houses associated with them, the majority are currently uninhabited. Most, however, were utilized to varying degrees by humans (primarily Native American) in the past, as evidenced by discarded shell deposits and signs of agricultural activity. The large number and diversity of hammocks in terms of size, development, and origin provide a natural laboratory for evaluating the influence of landscape structure and freshwater input on marsh processes. The GCE research is focused primarily on high-marsh communities, as the underlying question is whether (and how) freshwater from an adjacent upland controls marsh plant and invertebrate distributions. In particular, we are interested in the distribution and extent of plants and animals at the marsh-upland interface. At elevations above those where Spartina alterniflora is found the marsh is generally dominated by Juncus romerianus and/or Borrichia frutescens. We hypothesize that uplands of different size (ranging from small to large hammocks to mainlands) will support a different extent of upland marsh, with different associated fauna because we expect that 1) groundwater input to adjacent marshes will increase with increasing upland size and 2) only larger uplands will support populations of vertebrates (birds, raccoons, deer) that forage in the marsh. We further hypothesize that hammocks of different elevation will have different associated marsh plant and invertebrate communities because we expect that elevation will affect the hydraulic gradient and hence groundwater inflow to the adjacent marsh. We anticipate correlations among some of our independent variables (i.e. hammock origin will likely correlate with soil composition, with ballast stone islands being coarsest, Holocene islands comprised of sand and Pleistocene islands with a higher silt and clay content). Our predictions are summarized in the table below.  |
||||||||
Hammock SurveyIn summer 2007 a team of technicians, graduate and undergraduate students led by M. Alber (UGA) and C. Alexander (SKIO) conducted a broad survey of hammocks representing a range of sizes (Table 1) and origin (20 each of Pleistocene and Holocene origin; 9 dredge spoil islands; 6 ballast stone islands; 4 mainland transects; see Figure 7).   They used GIS and field methods to characterize each site in terms of its geomorphology, stratigraphy, water table characteristics, flora and fauna (Table 2). Field surveys at each hammock involved circumnavigating the hammock with a real-time sub-meter GPS to delineate the extent of the upper marsh. Transects were conducted at six locations around the perimeter of each hammock to determine the slope and profile (e.g., relative elevation) using traditional surveying techniques (i.e., rod and level). Surficial sediment was sampled along each transect to characterize trends in grain size (i.e., for permeability) and carbon content of soils.  At each of the six transect locations, flora and fauna were surveyed using standardized GCE protocols in 0.5 m2 quadrats set up 2.5 m from the upland edge. Plant species, shoot height (to the nearest cm) and flowering status were recorded; epifauna were counted and measured for size frequency information. Stratigraphy and water table height were determined at the top of one of the transects (n = 1 per hammock) using a 25-cm throw hand auger. Water was sampled for salinity and nutrient analysis (DOC, TDN). The data will be used to characterize hammocks and explore relationships among a series of independent (i.e. upland physical characteristics) and dependent (i.e. marsh biodiversity, plant and animal distributions) variables. Intensive CharacterizationIn 2008 we selected two hammocks for more detailed study: HN_I_1 is of Holocene origin and is located adjacent to Blackbeard Island to the north of Sapelo; PC_I_29 is of Pleistocene origin and is located adjacent to the south end of Sapelo Island. These are of similar size, with similar vegetation zones in the high marsh. We set up transects in each hammock that run from the nearby upland (Blackbeard and Sapelo Islands, respectively), through the marsh, and up and over each hammock to the marsh adjacent to the Sound (Sapelo and Doboy, respectively).  Sediment characteristicsC. Alexander (SKIO) led a field team that took vibracore samples along the transects on each hammock to get detailed information about stratigraphy. These cores show that the sedimentology of these features argues for distinctly different physical and stratigraphic processes dominating their formation. AMS C-14 and Optically Stimulated Luminescence dating are being used to constrain the times over which these features formed, and grain size and x-radiography illustrate the differences in sedimentary processes important in the genesis of each. Groundwater wellsOn each hammock, groundwater wells were installed along the transects in the locations sampled by vibracore. Wells were placed in sand layers in order to prevent clogging and to allow sampling of a continuously connected groundwater source. In addition to the vibracore stations, two additional lateral wells were installed at HN_I_1 approximately 5 meters to either side of the main transect near well #4 in order to get data on lateral inflow of groundwater. An additional upland well was also installed at this site in order to obtain a true zero salinity end member. The wells range in depth from 41 to 110 cm below a constant zero datum (99 to 232 cm below the surface). At PC_I_29, the two lateral wells are located near well #6. Well depths at this site range between 75 and 168 cm below a zero datum (130 to 247 cm below the surface). Pressure, salinity, and temperature loggers were placed into the select wells at both sites to collect data at 10-minute intervals; these data will be used in the groundwater modeling efforts. Groundwater biogeochemistry and radium samples were collected in August and September 2008. Sample collection will continue on monthly intervals, with additional samples planned to capture spring/neap tide and storm event signals. ModelingGroundwater modelThe observations collected on groundwater nutrients, flow, and radium isotopes as part of the intensive hammock characterization will form the core data set for calibration and validation of water balance models for each hammock. Large scale shallow subsurface flow patterns between hammock and creeks will be established by building upon available finite element codes. Measured hydraulic head and permeability measurements will be spatially interpolated. Together with infiltration/evapotranspiration rate estimates they will be employed to assess water movement using a Darcy approximation and water continuity. Given the uncertainty in driving forces and observed heterogeneity in marsh soil hydraulic conductivities, model simulations will be validated by the radium based water balance. Water fluxes to and from marsh soils will be assessed following Gardner and Reeves (2002). Vertically resolved fluid transport estimates will be validated using temperature profiles.  Plant modelTo investigate how changes in the relative amounts of groundwater and sub-surface flow affect plant species growth and competition, we will develop a plant model linked to the water model described above. The plant model will involve explicit descriptions of plant below- and above ground biomass of Spartina, Juncus and Borrichia, with particular attention to rooting depth (since that will affect pore water availability and quality). The biomass models will be driven primarily by irradiance within the canopy, salinity and sediment nutrient availability. Competition between the species will be based upon salt tolerance and light competition. The plant model will be coupled with nutrient distributions computed from a simplified soil model in order to predict changes in the distribution of Spartina, Juncus and Borrichia with changes in surficial groundwater. Results from the plant manipulations will be compared with model predictions and used to help refine the model assumptions. Plant ObservationsGreenhouse experimentsIn order to help parameterize the plant model, we are conducting a greenhouse experiment evaluating the sensitivity of three common salt marsh plants, Spartina alterniflora, Juncus roemerianus, and Borrichia frutescens, to varying levels of shade (0 to 80% of light blocked) and salinity (0 to 60 PSU). Monthly measurements of plant size and photosynthetic rate over the course of a single growing season (2009) will help inform the model about how the three plant species respond to competition for light (as simulated by shade cloth) and salinity. Field observationsIn order to help parameterize the plant model, we are collecting data on high marsh plant communities and associated soils. In 2009 we collected light profiles in stands of Spartina alterniflora, Juncus roemerianus, and Borrichia frutescens, with associated data on plant heights and densities. These data will provide information on how stands of different plant species at different densities affect the light environment for other plants (one mechanism of competition among plants). In 2009 we also collected data on soil porewater content and porewater salinity from stands of the same three plant species. These data will provide information on how the physical environment experienced by each plant species varies over space and time. To determine if plants are in fact rooting in the soils that we are sampling, we collected pre-dawn xylem pressure potential readings on all three species of plants (in theory, pre-dawn XPP readings are in equilibrium with soil water potentials). ManipulationWe are planning to use both experimental results and modeling predictions to guide us in initiating a long-term experiment designed to alter the water balance and flow patterns within designated portions of hammocks monitored during our intensive studies. We anticipate that this experiment will involve “paving” an appropriate area of the upland consistent with statewide building setbacks from the marsh (25’) with rainout shelters. This should increase overland runoff at the expense of infiltration, which should influence pore water dynamics and hence the marsh community. This manipulation will be ongoing to allow us to evaluate the long-term consequences of flow alteration, which we expect will affect a variety of ecosystem processes. Archeological studiesWe are particularly interested in the distribution of shell deposits left by the Native American occupants, because such deposits affect soil chemistry and landform exposure, and therefore may mediate high-marsh ecosystem processes by affecting the quality and quantity of water reaching the high marsh. Other legacy effects on current ecosystems might be mediated by changes in soil nutrient structure and permeability associated with fire, latrines, and seafood processing. V. Thompson and his students and colleagues have performed archeological surveys of four hammocks to date. Surveying these areas involves shovel test probes at 20 meter intervals. All sites are mapped and drawn according to standard archaeological procedures using GPS and a total station. All observable human modification to the landscape such as existing historic structures, canals, fences indicating livestock, etc. are also included. Thompson is also currently working on a database that includes reconnaissance information on over 100 islands.  |
||||||||
ArcheologyArcheological studies (more info) Hammock surveyCorrection for estimated upland areas of marsh hammocks (more info) Plant and soil characteristics of hammock uplands (more info) Intensive characterizationIntensive hammock characterization (more info) ModelingHammock groundwater modeling (more info) Plant observationsMarsh plant studies in support of hammock plant model (more info) |
||||||||
Theses and DissertationsConference Proceedings (Published Papers and Abstracts)Conference Posters and PresentationsAlbers, G. and Alber, M. 2003. Presentation: Vascular Plant Composition and Soil Characteristics of Undeveloped Back-barrier Islands near Sapelo Island, Georgia. 2003 Ecological Society of America meeting. Aug. 2003, Savannah, GA. Albers, G. and Alber, M. 2002. Presentation: The Impact of Residential Development on Georgia's Marsh Hammocks: Exploring the link between land and water in the coastal zone. The Coastal Society Conference. The Coastal Society Conference, May, 2002, Galveston, Texas. |
||||||||
Research Project Principal InvestigatorsMerryl Alber, University of Georgia Clark R. Alexander Jr., Skidaway Institute of Oceanography Samantha B. Joye, University of Georgia Christof Meile, University of Georgia Steven C. Pennings, University of Houston Victor D. Thompson, The Ohio State University Other Associated PersonnelGayle Albers, University of Georgia Kristen Anstead, University of Georgia Marine Institute Adrian B. Burd, University of Georgia Christine M. Hladik, University of Georgia Alana R. Lynes, University of Houston Willard S. Moore, University of South Carolina Jacob Shalack, University of Georgia Marine Institute John A. Turck, University of Georgia Alicia M. Wilson, University of South Carolina Caitlin Yeager, University of Georgia |
||||||||
