Area 3: Responses to Salinity and Inundation

Description Objectives Outcomes Figures Projects Publications Data Sets Show All  

Overview

GCE-II involved studies of marshes along the longitudinal gradient of the Altamaha River estuary (Q3), high marsh along the lateral gradient to upland habitat (Q4), and how these gradients affected organism distributions (Q5). These questions are integrated in Area 3, which focuses on how the dominant marsh habitats in the domain will respond to the changes in salinity and inundation that might be expected in the coming decades: sea level rise in Spartina marsh/tidal creek habitat, upstream salinity intrusion in fresh/brackish marsh, and changes in hydrologic inputs from adjacent uplands to the high marsh. The proposed research will provide data on the responses of each habitat to changes in these drivers over different time scales, allowing us to characterize their resilience and to compare their trajectories in forward and backward directions (evidence for hysteresis). As described above, the effects of these perturbations will depend on the timeframe over which the system is altered, with short-term pulses potentially affecting organismal behavior and physiology and long-term presses potentially resulting in a habitat (state) change.

To the extent that we can characterize transitions and identify state changes (through either long-term temporal patterns or spatial patterns observed along known gradients), this work also positions us to evaluate whether habitat shifts in the marsh are associated with increases in autocorrelation and variance, and slowed recovery from perturbations. Our goal is to characterize the responses of the marsh habitats in the domain (Spartina marsh, fresh/brackish marsh, high marsh) to pulses and presses in salinity and inundation.

We work in each of our key marsh habitats to assess how they will respond to changes in salinity and inundation. A) In the Spartina marsh we will assess marsh-atmosphere and marsh-creek exchange; monitor and model Spartina primary production; assess organism interactions; and evaluate ecosystem metabolism. B) In the fresh/brackish marsh our work involves long-term observations along the transect of the Altamaha River, and a large-scale field manipulation to evaluate how pulses and presses of salt water affect a tidal freshwater marsh. C) In the high marsh our work involves a field survey of high marsh areas, an experimental manipulation of runoff to the high marsh, and modeling of plant communities.

Components

Area 3A. Spartina marsh studies

Spartina alterniflora is the dominant plant species in the salt marshes in the GCE domain, and in most salt marshes in the US, and serves as the foundation species in this habitat. S. alterniflora is generally found between mean sea level and mean high water, with the tallest plants occurring at low elevations along creekbanks and shorter plants occurring at higher elevations in the midmarsh. Spartina habitat (hereafter, Spartina marsh) is flooded and drained twice daily through a network of tidal creeks, which serve as the primary hydrological link between intertidal and open water areas. These small creeks, which can drain almost completely at low tide, are important for nutrient and material exchange and also serve as conduits for aquatic organisms including nekton, epibenthic fauna (e.g. crabs), and planktonic larvae.

We expect sea level rise to be the chronic "press" driver affecting Spartina marshes. Thus, understanding how Spartina marshes respond to variation in inundation is critical for predicting the future state of coastal marshes in general. Increased inundation due to interannual variability in sea level (a pulse increase) has been found to increase NPP, but the relationship reverses at the highest sea levels, when increased flooding of tall Spartina, growing at the lower limit of their flooding tolerance, decreases productivity. At very high inundation levels plants die and no longer bind sediments with their roots, creating an unstable transition between Spartina marsh and mud flat. Understanding plant production is only part of the story, however, as the fate of this production (respired to the atmosphere, stored in the marsh, or exported to open water) will also vary with inundation patterns. Inundation also affects the amount of time that aquatic organisms can access the marsh, which may feed back in complex ways to affect plant productivity. Finally, changes in inundation may alter creek drainage patterns in the marsh.

Our goal is to characterize the response of Spartina marshes to changes in inundation with respect to C exchange, plant productivity, and organismal use, so that we can gain insight into the potential resilience of these systems to sea level rise. This research is focused on the Duplin River estuary, where 82% of the intertidal area is Spartina marsh and we have detailed information on elevation and vegetation patterns

Area 3B. Fresh/brackish Marsh Studies

Riverine estuaries such as the Altamaha have a longitudinal gradient of habitats (fresh marsh, brackish marsh, salt marsh) that vary in their delivery of ecosystem services. Our previous studies have shown that the distribution of plants along this gradient is largely defined by salinity, interacting with competition, and that changes in vegetation are accompanied by shifts in community structure, biogeochemical cycling, and other processes. We expect salinity variation, in response to changes in river inflow or salt water intrusion, to be the major driver affecting fresh and brackish marshes. Understanding how these habitats respond to salinity variation is therefore important for predicting potential state changes along the estuary.

We hypothesize that biotic and ecosystem responses to changes in salinity will be nonlinear (e.g. the system will display hysteresis), and that freshwater marshes will be less resistant than brackish habitat to increasing salinity. We will investigate these ideas using a combination of long term observations and field manipulations.

Area 3C. High Marsh Studies

The high marsh is the intertidal area closest to the upland, and is typically dominated by the plants Borrichia frutescens and Juncus roemerianus. High marsh plant and animal communities are more diverse than those at lower elevations (i.e. Spartina marsh), and the high marsh is more likely than other marsh habitats to exchange materials and organisms with upland areas. Our survey of marsh hammocks in GCE-2 found that the nature of the upland predicted multiple aspects of the high marsh habitat. High marsh habitat is potentially vulnerable to the long-term press of sea level rise, which could cause a lateral shift in its borders, to pulses in precipitation, which cause variation in salinity, and to the long-term press caused by anthropogenic alteration of the uplands, which may alter hydrologic connectivity. The vegetation in the high marsh is influenced by inundation and salinity; greenhouse experiments in which we varied both factors suggested that increasing inundation restricts Borrichia to the high marsh, whereas Juncus was more affected by increased salinity. The Juncus results support our observations that drought induces a shift in the Spartina/Juncus border towards dominance by Spartina, and with experiments that show that reducing salinity increases Juncus abundance. Once removed by a disturbance, the Juncus is slow to recover, leading to hysteresis. The high marsh is also affected by shoreline modification and upland development. A number of studies have correlated human activities in the upland to high marsh abiotic and biotic variables.

We hypothesize that the spatial extent of the high marsh is primarily determined by geomorphology and hydrologic connectivity with the upland (both of which can be altered by humans), and that temporal variation in community composition is primarily due to variation in salinity. We will investigate these ideas using a combination of long term observations, field surveys, experimental manipulations and modeling.

 

Research Objectives

  • 3A.1  Long-Term manipulations - Track recovery in the SALTEx Experiment
  • 3A.2  Long-Term manipulations - Continue the PredEx Experiment
  • 3A.3  Long-Term manipulations - Continue the High Marsh manipulation
  • 3A.4  Long-Term manipulations - Establish Disturbance manipulation
  • 3A.5  Spartina Marsh - Investigate marsh fauna interactions
  • 3A.6  Monitor headward erosion in tidal creeks (yr 1-4)
    • Description:  With external funding, Pennings and colleagues are documenting rapid (~2 m y-1) headward erosion of multiple tidal creeks in our domain and are experimentally testing the hypothesis that burrowing and herbivory by Sesarma facilitate the growth of creeks, which will affect hydrological processes as sea level rises. Data on changes in creek geomorphology will be used to inform modeling scenarios (Area 4)
    • Participants:  Steve Pennings
  • 3A.7  Develop a Spartina physiological model (yr 1-3)
    • Description:  We will measure above- and below-ground biomass monthly in short, medium and tall Spartina plots near the flux tower. These observations can be scaled up to provide an estimate of NPP for the tower footprint, and will be used as ground truth data for remote sensing of EVI and NDVI. These data and our long-term data sets (e.g. irradiance, biomass, meteorology, salinity) from GCE-I and -II will be used to produce a semi-empirical model for Spartina areal NPP and biomass that considers both above- and below-ground production. Below-ground Spartina production is affected by marsh elevation and inundation (Kirwan & Guntenspergen 2012).Existing models of Spartina photosynthesis do not consider below-ground allocation of resources (Fagherazzi et al. in press), and are unable to accurately predict long-term changes in above-ground biomass in the domain (Jung & Burd in prep). Understanding resource allocation was important for developing models in seagrasses (e.g. Burd & Dunton 2001), and we will use a similar approach here. The model will be used to evaluate how changes in salinity and inundation affect Spartina production. It will also provide the basis for a more mechanistic model that incorporates details of plant physiology (e.g. C4 photosynthesis, respiration, resource allocation and translocation).
    • Participants:  Adrian Burd
  • 3A.8  Develop a model to predict porewater salinity (yr 1-3)
    • Description:  Above- and below-ground plant biomass data near the GCE flux tower and our long-term data sets (e.g. irradiance, biomass, meteorology, salinity) from GCE-I and -II will be used to produce a semi-empirical model for Spartina areal NPP and biomass that considers both above- and below-ground production. Below-ground Spartina production is affected by marsh elevation and inundation (Kirwan & Guntenspergen 2012).Existing models of Spartina photosynthesis do not consider below-ground allocation of resources (Fagherazzi et al. in press), and are unable to accurately predict long-term changes in above-ground biomass in the domain (Jung & Burd in prep). Understanding resource allocation was important for developing models in seagrasses (e.g. Burd & Dunton 2001), and we will use a similar approach here. The model will be used to evaluate how changes in salinity and inundation affect Spartina production. It will also provide the basis for a more mechanistic model that incorporates details of plant physiology (e.g. C4 photosynthesis, respiration, resource allocation and translocation).
    • Participants:  Christof Meile
  • 3B.1  Focused Studies - Investigate controls of S. alterniflora production
  • 3B.2  Focused Studies - Investigate marsh fauna interactions
  • 3B.3  Focused Studies - Enhance our understanding of coastal carbon dynamics
  • 3C.1  High Marsh - Assess habitat dynamics at vegetation borders
  • 3C.2  High Marsh - Continue our upland manipulation
  • 3C.3  Conduct upland manipulation of water flow to high marsh areas (yr 3-6)
    • Description:  We propose to experimentally test the hypothesis that water flow from the upland to high marsh directly affects high marsh function. We will work in high marsh areas adjacent to Sapelo Island where the vegetation is a mixture of Juncus and Borrichia. Our experiment will have 3 treatments: 1) upland water reduction (to assess how upland connections, as opposed to elevation, mediate high marsh community composition); 2) impervious surface addition (to assess how increased runoff and decreased groundwater alter high marsh community composition); and 3) controls. Plots (n=8/treatment) will each extend 10 m along the upland/high marsh border and be separated by at least 10 m, with treatments fully interspersed. The water reduction treatments will have vertical plastic barriers (sheeting that extends from 60 cm deep to 30 cm above the soil surface) that will divert shallow groundwater and runoff away from the plots. The impervious surface treatment will use corrugated plastic roofing just above the soil surface to cover the upland immediately adjacent to the marsh plot (with gaskets around tree trunks to intercept stemflow). A ditch will be dug and then filled back in in the control and impervious surface treatments to control for the disturbance of installing the barrier in the water reduction treatment. The experiment will be set up in year 4 and will last a minimum of 3 y. We will use DGPS to delineate the edge of the high marsh, and establish permanent sub-plots in each plot located 2, 4 and 8 m from the upland edge, which we will survey with an RTK GPS along with the elevation of the high marsh edge. We will use standard GCE protocols to monitor pore water salinity, plant photosynthetic rates, plant biomass and composition, and benthic invertebrate density and composition in each sub-plot. Measurements will be weekly for the first 8 weeks to capture the short-term response, monthly for the next year, and quarterly thereafter.
    • Participants:  Merryl Alber, Clark Alexander, Jeb Byers, Steve Pennings, Alicia Wilson
  • 3C.4  Develop a clonal plant model to explore vegetation dynamics (yr 3-5)
    • Description:  We will use a spatially structured model of clonally growing plants (Mony et al. 2011) to evaluate interactions among the dominant high marsh plants at our study sites. The model will be parameterized with information from greenhouse experiments in GCE-II on the responses of each plant to variations in salinity and inundation, augmented by information from the Spartina physiology model. It will allow us to evaluate the dynamics of habitat shifts that might occur in response to changes in freshwater delivery or sea level rise, and to identify conditions promoting hysteresis in habitat transitions. The plant community model will also be integrated into our large-scale modeling efforts (Area 4).
    • Participants:  Marc Garbey, Steve Pennings

Research Outcomes by Objective

  • 3A.1  Characterize temporal variability in marsh-atmosphere exchange of CO2 (yr 1-6)
    • 2 report:

      Activities:  2015: We have developed routines to check and process the data stream from the flux tower so that we can calculate fluxes. We are now testing algorithms for footprint modeling, which will be linked to vegetation characteristics. We are also interested in whether the presence of surface fog affects marsh-atmosphere interactions (see Results Fig. 1).
      2016: We have begun analyzing the data collected by the flux tower to characterize the effect of tidal inundation on CO2 exchange.
      2017: We are using these data to generate a gap­filled annual estimate of net ecosystem exchange (NEE) in the marsh.
      2018: We are evaluating the effects of environmental factors on annual and seasonal rates of NEE, R, and GPP.

      Results:  2016: Our analyses show that light response curves vary as a function of the ratio between plant height and inundation (Results Fig. 5), and suggest that flooding reduces CO2 flux by approximately 20%. A poster on this work was recognized as a runner-up in the student poster competition at the LTER All Scientist’s Meeting.
      2017: Our initial analysis of CO2 exchange at the flux tower suggests that NEE is reduced by 55% when the marsh is covered by water (Fig. 8).
      2018: Characterize temporal variability in marshatmosphere exchange of CO2 I. Forbich (PIE) is using flux tower data from PIE, GCE, and several other Atlantic coast sites to compare CO2 fluxes across a latitudinal gradient. She has found differences in terms of seasonality and magnitude, which link to patterns in NDVI and potential environmental drivers.

    • 2014 report:

      Activities:  We have developed routines to check and process the data stream from the flux tower so that we can calculate fluxes. We are now testing algorithms for footprint modeling, which will be linked to vegetation characteristics. We are also interested in whether the presence of surface fog affects marsh-atmosphere interactions (see Results Fig. 1).

    • 2015 report:

      Activities:  We have begun analyzing the data collected by the flux tower to characterize the effect of tidal inundation on CO2 exchange.

      Results:  Our analyses show that light response curves vary as a function of the ratio between plant height and inundation (Results Fig. 5), and suggest that flooding reduces CO2 flux by approximately 20%. A poster on this work was recognized as a runner-up in the student poster competition at the LTER All Scientist’s Meeting.

    • 2016 report:

      Activities:  We are using these data to generate a gap­filled annual estimate of net ecosystem exchange (NEE) in the marsh.

      Results:  Our initial analysis of CO2 exchange at the flux tower suggests that NEE is reduced by 55% when the marsh is covered by water (Fig. 8).

    • 2017 report:

      Activities:  We are evaluating the effects of environmental factors on annual and seasonal rates of NEE, R, and GPP.

      Results:  Characterize temporal variability in marshatmosphere exchange of CO2 I. Forbich (PIE) is using flux tower data from PIE, GCE, and several other Atlantic coast sites to compare CO2 fluxes across a latitudinal gradient. She has found differences in terms of seasonality and magnitude, which link to patterns in NDVI and potential environmental drivers.

    • 2019 report:

      Activities:  We deployed soil temperature sensors at the core marsh sites to evaluate its relationship to Spartina green-up. We are also planning to manipulate temperature in a greenhouse experiment.

      Results:  Hawman et al. used general additive models to evaluate the annual cycle of GPP and light use efficiency measured at the flux tower (Fig. 4), and found that the cloudiness index and daily maximum tide height are the primary factors that explain deviation in Spartina light use efficiency. This work was presented at multiple conferences and is being written up for publication.

    • 2020 report:

      Activities:  We continue to track recovery in the SALTEx experiment following cessation of experimental dosing in December 2017. We are monitoring porewater, sediment elevation, and vegetation. Although porewater nutrients returned to background levels within a few weeks, there is evidence of residual salinity in several of the press plots. Widney et al. (2019) published a synthesis of the biochemical effects of the SALTEx experiment showing that three years of saltwater intrusion increased porewater Cl-, SO4, HS and inorganic N (NH4 and NO3) and decreased plant N storage (see key findings). Solohin et al. (2020) found that declining soil surface elevation and associated carbon loss was due to reduced belowground biomass in press plots, and Mobilian et al. (2020) showed that decreased carbon inputs from plants resulted in reduced microbial diversity and decreased microbial carbon cycling (Fig. 1).

    • 2021 report:

      Activities:  We continue to track recovery in the SALTEx experiment following cessation of experimental dosing in December 2017.

      Results:  Analyses of the SALTEx experiment show differential response curves to salinity: porewater NH4 increased linearly and DOC concentrations decreased as a power function, whereas other constituents showed no significant trends (Fig. 1).

  • 3A.2  Evaluate Spartina plant phenology (yr 1-6) and above- and below-ground production (yr 1-4)
    • 2 report:

      Activities:  2014: We installed a StarDot Netcam on the GCE eddy covariance flux tower in mid-September. Images are relayed hourly to servers at UGA and then uploaded to the ecosystem phenology web camera network (phenocam.sr.unh.edu/webcam/sites/gcesapelo/). We also began monitoring above- and below-ground biomass of Spartina alterniflora in three marsh zones so that estimates of gas flux from the tower can be linked to changes in plant biomass and allocation.
      2015: The phenocam camera collects images every 30 min. Samples of S. alterniflora are collected on a monthly basis so that flux tower observations can be linked to vegetation. This year we timed plant sampling to correspond with Landsat overpasses so that our findings can ultimately be scaled up.
      2016: The phenocam camera collects images every 30 min and contributes data to the national phenocam network. We continue to sample plants in permanent plots on a monthly basis for above and below-ground production, timed to correspond with Landsat overpasses so that our findings can ultimately be scaled up.
      2017: The phenocam camera contributes data to the national phenocam network every 30 min. We continue to sample plants in permanent plots, timed to correspond with Landsat overpasses.
      2018: The phenocam camera contributes data to the national phenocam network every 30 min. We continue to sample plants in permanent plots, timed to correspond with Landsat overpasses.

      Results:  2015: Analysis of PhenoCam images using the green chromatic index showed the seasonal cycle of plant production and revealed that dates of spring green-up and fall senescence differ by marsh zone (Fig. 4).
      2016: We have developed a smart classifier for automatically extracting information from the phenocam images. This is being written up for publication.
      2017: O'Connell and Alber (2016) developed a method to filter out phenocam scenes with poor solar illumination (clouds) or non¬target objects (floods), which provides us with a subset of optimal scenes for phenology analysis (Fig 9).
      2018: Evaluate Spartina plant phenology The first 3.5 years of phenocam imagery was used to develop a phenology model for Spartina alterniflora. We found evidence for both spatial and interannual differences in phenophase length (Fig. 5).

    • 2013 report:

      Activities:  We installed a StarDot Netcam on the GCE eddy covariance flux tower in mid-September. Images are relayed hourly to servers at UGA and then uploaded to the ecosystem phenology web camera network (phenocam.sr.unh.edu/webcam/sites/gcesapelo/). We also began monitoring above- and below-ground biomass of Spartina alterniflora in three marsh zones so that estimates of gas flux from the tower can be linked to changes in plant biomass and allocation.

    • 2014 report:

      Activities:  The phenocam camera collects images every 30 min. Samples of S. alterniflora are collected on a monthly basis so that flux tower observations can be linked to vegetation. This year we timed plant sampling to correspond with Landsat overpasses so that our findings can ultimately be scaled up.

      Results:  Analysis of PhenoCam images using the green chromatic index showed the seasonal cycle of plant production and revealed that dates of spring green-up and fall senescence differ by marsh zone (Fig. 4).

    • 2015 report:

      Activities:  The phenocam camera collects images every 30 min and contributes data to the national phenocam network. We continue to sample plants in permanent plots on a monthly basis for above and below-ground production, timed to correspond with Landsat overpasses so that our findings can ultimately be scaled up.

      Results:  We have developed a smart classifier for automatically extracting information from the phenocam images. This is being written up for publication.

    • 2016 report:

      Activities:  The phenocam camera contributes data to the national phenocam network every 30 min. We continue to sample plants in permanent plots, timed to correspond with Landsat overpasses.

      Results:  O'Connell and Alber (2016) developed a method to filter out phenocam scenes with poor solar illumination (clouds) or non¬target objects (floods), which provides us with a subset of optimal scenes for phenology analysis (Fig 9).

    • 2017 report:

      Activities:  The phenocam camera contributes data to the national phenocam network every 30 min. We continue to sample plants in permanent plots, timed to correspond with Landsat overpasses.

      Results:  Evaluate Spartina plant phenology The first 3.5 years of phenocam imagery was used to develop a phenology model for Spartina alterniflora. We found evidence for both spatial and interannual differences in phenophase length (Fig. 5).

    • 2019 report:

      Activities:  We continue sampling the predator exclusion experiment initiated in summer 2016. In 2019 we began sampling pore water and decomposition, conducted tethering experiments with fiddler crabs and Littoraria, and completed a literature search to identify which species are likely being excluded by the treatment.

      Results:  The results from the PredEx manipulation indicate that nekton are exerting top-down control of marsh invertebrates, with evidence for a short-lived release that may be compensated for by mesopredators such as mud crabs (Fig. 5). Initial results suggest that pore water salinity and pH do not vary with treatment. However, decomposition in predator exclusion plots was significantly greater than controls, presumably due to greater oxygenation by crab burrows. Soil bulk density (0-5 cm) was also lower, although not significantly, while there was no difference in soil C and N.

    • 2020 report:

      Activities:  We continue the predator exclusion experiment initiated in summer 2016, with annual sampling of the invertebrate and plant communities. In 2019 we added camera traps to evaluate the presence of avian and mammalian consumers in the plots, began measuring benthic microalgae, pore water and decomposition, and conducted tethering experiments with fiddler crabs and Littoraria. Although sampling was limited in 2020 due to travel restrictions associated with COVID, we expect to resume our regular schedule of observations in 2021. Results to-date have revealed top-down control of invertebrate populations that cascades down to affect ecosystem processes such as enhanced consumption and decomposition rates. It appears that mesopredators (mud crabs) have increased in the predator exclusion treatment and have prevented other invertebrates from increasing as dramatically as we expected (Fig. 2). Camera traps showed that terrestrial predators were not major visitors to the experimental plots, and both soil pore water and snail mortality varied with elevation as opposed to predator exclusion. We did not see effects on S. alterniflora biomass, which we attribute to mesopredator release and/or compensatory facilitation by fiddler crabs. A manuscript on these results is in preparation.

    • 2021 report:

      Activities:  We continue sampling the predator exclusion experiment initiated in summer 2016. This past year we conducted additional mesopredator tethering trials.

      Results:  Mesopredators (mud crabs) have increased in the predator exclusion treatment after several years, and are likely the reason that other invertebrates have not increased dramatically as we expected. We are planning to conduct a mesopredator x nekton presence experiment in 2022 to tease apart their relative effects.

  • 3A.3  Quantify lateral C exchange through a small tidal creek (yr 1-3)
    • 2 report:

      Activities:  2014: We have instrumented the creek adjacent to the flux tower with a Nortek ADCP to quantify transport of water into the marsh surrounding the flux tower as a function of tidal elevation. . We have also made initial measurements of total CO2, pH and total alkalinity over a tidal cycle.
      2015: This past year we conducted quarterly field campaigns to measure total DIC in the creek that floods the flux tower. We sample water over the course of a tidal cycle for analysis of CO2, pH and total alkalinity, measure water height, and deploy an ADCP to measure water flow.
      2016: We analyzed field samples collected over the course of Year Two to evaluate DIC exchange between the small creek adjacent to the flux tower and the Duplin River. We also developed hypsometric curves (Activities Fig. 3) to characterize water flow.
      2017: We calculated metabolism over the inundated marsh based on Year 2 field observations taken in the tidal creek adjacent to the GCE flux tower.
      2018: Completed yr 4.

      Results:  2015: DIC in the ebbing creek water was higher than in entering floodwater, due to respiration by marsh flora and fauna and diffusion of DIC-rich pore waters while the marsh was flooded. Presumably CO2 flux to the atmosphere was reduced by an amount equivalent to that gained by the water column.
      2016: The creek imports water that is low in DIC concentration during flood tide, and exports water high in DIC during ebb (Results Fig. 6). This suggests that the marsh platform and creek banks are the apparent source of this C.
      2017: We found that the flooding of the marsh exports DIC to the estuary most seasons of the year and is responsible for a major portion of overall system net heterotrophy and about 1/3 of air¬water exchange of CO2.

    • 2013 report:

      Activities:  We have instrumented the creek adjacent to the flux tower with a Nortek ADCP to quantify transport of water into the marsh surrounding the flux tower as a function of tidal elevation. . We have also made initial measurements of total CO2, pH and total alkalinity over a tidal cycle.

    • 2014 report:

      Activities:  This past year we conducted quarterly field campaigns to measure total DIC in the creek that floods the flux tower. We sample water over the course of a tidal cycle for analysis of CO2, pH and total alkalinity, measure water height, and deploy an ADCP to measure water flow.

      Results:  DIC in the ebbing creek water was higher than in entering floodwater, due to respiration by marsh flora and fauna and diffusion of DIC-rich pore waters while the marsh was flooded. Presumably CO2 flux to the atmosphere was reduced by an amount equivalent to that gained by the water column.

    • 2015 report:

      Activities:  We analyzed field samples collected over the course of Year Two to evaluate DIC exchange between the small creek adjacent to the flux tower and the Duplin River. We also developed hypsometric curves (Activities Fig. 3) to characterize water flow.

      Results:  The creek imports water that is low in DIC concentration during flood tide, and exports water high in DIC during ebb (Results Fig. 6). This suggests that the marsh platform and creek banks are the apparent source of this C.

    • 2016 report:

      Activities:  We calculated metabolism over the inundated marsh based on Year 2 field observations taken in the tidal creek adjacent to the GCE flux tower.

      Results:  We found that the flooding of the marsh exports DIC to the estuary most seasons of the year and is responsible for a major portion of overall system net heterotrophy and about 1/3 of air¬water exchange of CO2.

    • 2017 report:

      Activities:  Completed yr 4.

    • 2019 report:

      Activities:  We conducted a study of marsh perturbation and recovery at headward-eroding creeks that are subject to Sesarma herbivory (Fig. 4). We also developed a protocol for field monitoring the impacts of wrack and drought disturbance.

      Results:  We have evidence that grazed creeks have become increasingly prevalent over the past few decades and are causing a significant increase in drainage density by accelerating creek incision, which has implications for invertebrate communities and predator-prey interactions on the marsh platform (Crotty et al., in review).

    • 2020 report:

      Activities:  During GCE-III we initiated a large-scale experiment wherein we attempted to manipulate surface and shallow groundwater flow from the upland border into the high marsh. Although we continue to sample the plots annually, data from groundwater wells indicates that this manipulation was largely ineffective at altering groundwater flow and porewater salinities, and we have seen no consistent changes between treatments in benthic micro-algae, invertebrates or vascular plants. However, we anticipate that data from the wells will be useful for investigating links between major perturbations (high rainfall events, very high tides), groundwater level and porewater salinity, and vegetation changes. Data from the wells are also being analyzed to calculate hydraulic gradients for groundwater flow models.

    • 2021 report:

      Activities:  The high marsh experiment was largely ineffective at altering groundwater flow, but the data are useful for understanding hydraulic gradients. We plan to decommission this experiment in 2022.

      Results:  We are using data from the wells to monitor groundwater conditions to better understand fluxes of groundwater in the high marsh (see Objective 2A4).

  • 3A.4  Evaluate net ecosystem metabolism and quantify net C exchange in the Duplin R (yr 1-4)
    • 2 report:

      Activities:  2014: We made initial measurements of diurnal changes in total CO2 along the length of the Duplin River to estimate GPP, NEP, and total system respiration. We have also begun collecting DIC, pH and TA samples at 4 stations on a quarterly basis
      2015: As part of our quarterly field campaign we conducted rapid surveys along the length of the Duplin River at dawn and dusk to evaluate diurnal changes in total CO2 and DO to estimate GPP, total system respiration, and NEP. We also collect DIC, pH and TA samples on a bimonthly basis.
      2016: We used the in situ diurnal technique to estimate metabolism from our Year 2 field measurements. Water masses were matched using a Lagrangian particle tracer model, and the rates of change of DO and DIC were used to calculate GPP, R, and NEP after correcting for air-water exchange and dispersion mixing. We also continued to collect DIC, pH and TA samples on a bimonthly basis.
      2017: We used the in situ diurnal technique to estimate metabolism from our Year 2 field measurements.
      2018: Completed yr 4.

      Results:  2015: We see substantial diurnal changes in DO and DIC in the Duplin River, signifying a high level of system metabolism. We also see deviation from saturation, which provides evidence for large air-sea exchanges as well as the overall heterotrophic nature of the Duplin system. There are also large downstream gradients in DO, DIC, chlorophyll and DOC, which suggest the export of materials to Doboy Sound and presumably the coastal ocean.
      2016: We found that the Duplin estuary is highly productive and heterotrophic, with overall levels increasing up-estuary where the ratio of marsh/water and tidal creek drainage density are highest. The dispersion and air-sea fluxes were a significant component of the overall diel changes in DO and DIC.
      2017: The metabolism of the Duplin marsh¬estuary system reflected seasonal patterns for GPP, R, and NEP (Fig 10). Annual NEP is net hetrotrophic, with maximal values during warmer months (Wang 2016).
      2018: Evaluate net ecosystem metabolism and quantify net C exchange in the Duplin R. Wang et al. (2017) published a complete budget of CO2 exchange in the Duplin River estuary. They found that the overall system was a net source of CO2 to the atmosphere and coastal ocean and a net sink for oceanic and atmospheric O2 (Fig. 6).

    • 2013 report:

      Activities:  We made initial measurements of diurnal changes in total CO2 along the length of the Duplin River to estimate GPP, NEP, and total system respiration. We have also begun collecting DIC, pH and TA samples at 4 stations on a quarterly basis

    • 2014 report:

      Activities:  As part of our quarterly field campaign we conducted rapid surveys along the length of the Duplin River at dawn and dusk to evaluate diurnal changes in total CO2 and DO to estimate GPP, total system respiration, and NEP. We also collect DIC, pH and TA samples on a bimonthly basis.

      Results:  We see substantial diurnal changes in DO and DIC in the Duplin River, signifying a high level of system metabolism. We also see deviation from saturation, which provides evidence for large air-sea exchanges as well as the overall heterotrophic nature of the Duplin system. There are also large downstream gradients in DO, DIC, chlorophyll and DOC, which suggest the export of materials to Doboy Sound and presumably the coastal ocean.

    • 2015 report:

      Activities:  We used the in situ diurnal technique to estimate metabolism from our Year 2 field measurements. Water masses were matched using a Lagrangian particle tracer model, and the rates of change of DO and DIC were used to calculate GPP, R, and NEP after correcting for air-water exchange and dispersion mixing. We also continued to collect DIC, pH and TA samples on a bimonthly basis.

      Results:  We found that the Duplin estuary is highly productive and heterotrophic, with overall levels increasing up-estuary where the ratio of marsh/water and tidal creek drainage density are highest. The dispersion and air-sea fluxes were a significant component of the overall diel changes in DO and DIC.

    • 2016 report:

      Activities:  We used the in situ diurnal technique to estimate metabolism from our Year 2 field measurements.

      Results:  The metabolism of the Duplin marsh¬estuary system reflected seasonal patterns for GPP, R, and NEP (Fig 10). Annual NEP is net hetrotrophic, with maximal values during warmer months (Wang 2016).

    • 2017 report:

      Activities:  Completed yr 4.

      Results:  Evaluate net ecosystem metabolism and quantify net C exchange in the Duplin R. Wang et al. (2017) published a complete budget of CO2 exchange in the Duplin River estuary. They found that the overall system was a net source of CO2 to the atmosphere and coastal ocean and a net sink for oceanic and atmospheric O2 (Fig. 6).

    • 2019 report:

      Activities:  We are participating in a distributed "DragNet" experiment aimed at understanding how grasslands recover from disturbances under different nutrient regimes. We also plan to initiate a standardized disturbance experiment across the GCE domain based on observations of natural marsh perturbations. We continue to monitor recovery from a wrack disturbance experiment conducted in 2011.

    • 2020 report:

      Activities:  We are planning two experiments in which we will implement a standardized 4 m2 disturbance across natural gradients of 1) salinity and 2) elevation to test the hypothesis that underlying abiotic gradients affect the response of a marsh to a disturbance. These will be initiated in the second half of the project based on our ongoing observations of natural perturbations (Area 4). We are also participating in a distributed "DragNet" experiment, set to begin in summer 2021, aimed at understanding how grasslands around the world recover from disturbances under different nutrient regimes.

    • 2021 report:

      Activities:  We started the DRAGNET distributed disturbance experiment in 2021 and are planning experiments in which we will implement a standardized disturbance across natural gradients of salinity and elevation to test the hypothesis that underlying abiotic gradients affect marsh recovery from disturbance.

      Results:  None to date

  • 3A.5  Conduct a predator removal manipulation (yr 4-6)
    • 2 report:

      Activities:  2015: Begins yr 4
      2016: This begins yr. 4. However, we did some initial tests and found that our proposed predator exclusion design was not effective. Our current plan is to use replicated 10 m diameter cages with different mesh sizes to tease out the effects of blue crabs, mud minnows, and larger nekton.
      2017: We initiated a predator exclusion experiment in summer 2016 by setting up 8 6­m diameter, 1.5 m high predator exclusion cages along with cage and uncaged controls in the medium Spartina zone along Dean Creek. We assessed initial conditions and will be monitoring primary production, decomposition, species diversity and abundance, biogeochemistry, and geomorphology.
      2018: We have begun regular sampling of the predator exclusion experiment initiated in summer 2016 in the medium Spartina zone along Dean Creek.

      Results:  2018: Conduct a predator removal manipulation Our initial 6 months of data show significant increases in both periwinkle snails and fiddler crabs in treatments where nekton is excluded (Fig. 7).

      Plans:  2016: We will have a planning meeting at the GCE annual meeting in January 2016 and then deploy cages in the spring.

    • 2014 report:

      Activities:  Begins yr 4

    • 2015 report:

      Activities:  This begins yr. 4. However, we did some initial tests and found that our proposed predator exclusion design was not effective. Our current plan is to use replicated 10 m diameter cages with different mesh sizes to tease out the effects of blue crabs, mud minnows, and larger nekton.

      Plans:  We will have a planning meeting at the GCE annual meeting in January 2016 and then deploy cages in the spring.

    • 2016 report:

      Activities:  We initiated a predator exclusion experiment in summer 2016 by setting up 8 6­m diameter, 1.5 m high predator exclusion cages along with cage and uncaged controls in the medium Spartina zone along Dean Creek. We assessed initial conditions and will be monitoring primary production, decomposition, species diversity and abundance, biogeochemistry, and geomorphology.

    • 2017 report:

      Activities:  We have begun regular sampling of the predator exclusion experiment initiated in summer 2016 in the medium Spartina zone along Dean Creek.

      Results:  Conduct a predator removal manipulation Our initial 6 months of data show significant increases in both periwinkle snails and fiddler crabs in treatments where nekton is excluded (Fig. 7).

    • 2019 report:

      Activities:  We began field measurements to evaluate how interactions between cordgrass and ribbed mussels influence tide water chemistry and the net import/export of materials from marshes.

      Results:  Tethering experiments with Littoraria show that the probability of survival increases with body size: for every millimeter increase in shell height, the log odds of survival increases by 0.16048.

  • 3A.6  Monitor headward erosion in tidal creeks (yr 1-4)
    • 2 report:

      Activities:  2014: We began monitoring growth of 16 headward-eroding creeks in 2011. These creeks are distinguished by an unvegetated basin at their head that supports high populations of burrowing crabs.
      2015: We continued monitoring the growth of 16 headward-eroding creeks, distinguished by an unvegetated basin at their head that supports high populations of burrowing crabs.
      2016: We continued this activity in 2014-5.
      2017: Completed yr 3.
      2018: Completed yr 3.

      Results:  2015: Annual monitoring of creeks in the GCE domain for 4 y has revealed that growth rates vary substantially among years (Fig. 5). We will continue to monitor creeks annually in order to determine the drivers of annual variation in growth rate.
      2017: Vu (2016) found that ground measurements of headward erosion correlate well with those based on aerial photographs.
      2018: Monitor headwater erosion in tidal creeks Vu (2016) found that ground measurements of headward erosion correlate well with those based on aerial photographs.

    • 2013 report:

      Activities:  We began monitoring growth of 16 headward-eroding creeks in 2011. These creeks are distinguished by an unvegetated basin at their head that supports high populations of burrowing crabs.

    • 2014 report:

      Activities:  We continued monitoring the growth of 16 headward-eroding creeks, distinguished by an unvegetated basin at their head that supports high populations of burrowing crabs.

      Results:  Annual monitoring of creeks in the GCE domain for 4 y has revealed that growth rates vary substantially among years (Fig. 5). We will continue to monitor creeks annually in order to determine the drivers of annual variation in growth rate.

    • 2015 report:

      Activities:  We continued this activity in 2014-5.

    • 2016 report:

      Activities:  Completed yr 3.

      Results:  Vu (2016) found that ground measurements of headward erosion correlate well with those based on aerial photographs.

    • 2017 report:

      Activities:  Completed yr 3.

      Results:  Monitor headwater erosion in tidal creeks Vu (2016) found that ground measurements of headward erosion correlate well with those based on aerial photographs.

  • 3A.7  Develop a Spartina physiological model (yr 1-3)
    • 2 report:

      Activities:  2014: Tests of our initial model of Spartina production suggested that a more detailed below ground component was required. We have now begun collecting samples to measure plant soluble carbohydrates in order to track plant allocation of resources.
      2015: Tests of our Spartina model suggested that a more detailed below-ground component is required. We are therefore measuring soluble carbohydrates in above- and below-ground tissue in order to track allocation of resources.
      2016: We have completed data analysis of a year's worth of above and below ground biomass and non-structural carbohydrate data, which is being used to develop model formulations for the translocation of material between above and below ground plant components.
      2017: We are using the results of our field sampling of carbohydrates to develop models for the production of above and below ground biomass.
      2018: We have developed a model of Spartina production and biomass that includes a simple phenology with regard to translocation of resources between above and below ground biomass. We are in the process of parameterizing the model for the Flux Tower data.

      Results:  2016: Although there is a significant seasonality in above and below-ground nonstructural carbohydrates, there appears to be no obvious environmental cue for the allocation of this material.
      2017: Jung & Burd (submitted) found that although Spartina non¬structural carbohydrates do not seem to correlate with any environmental variable, there is a consistent lag between above and below ground tissues indicative of transport.
      2018: Develop a Spartina physiological model Jung & Burd (2017) describes seasonal translocation of several different nonstructural carbohydrates between aboveand belowground tissues of Spartina alterniflora (Fig. 8).

      Plans:  2015: Data on soluble carbohydrates will be used to refine the existing model of Spartina growth.

    • 2013 report:

      Activities:  Tests of our initial model of Spartina production suggested that a more detailed below ground component was required. We have now begun collecting samples to measure plant soluble carbohydrates in order to track plant allocation of resources.

    • 2014 report:

      Activities:  Tests of our Spartina model suggested that a more detailed below-ground component is required. We are therefore measuring soluble carbohydrates in above- and below-ground tissue in order to track allocation of resources.

      Plans:  Data on soluble carbohydrates will be used to refine the existing model of Spartina growth.

    • 2015 report:

      Activities:  We have completed data analysis of a year's worth of above and below ground biomass and non-structural carbohydrate data, which is being used to develop model formulations for the translocation of material between above and below ground plant components.

      Results:  Although there is a significant seasonality in above and below-ground nonstructural carbohydrates, there appears to be no obvious environmental cue for the allocation of this material.

    • 2016 report:

      Activities:  We are using the results of our field sampling of carbohydrates to develop models for the production of above and below ground biomass.

      Results:  Jung & Burd (submitted) found that although Spartina non¬structural carbohydrates do not seem to correlate with any environmental variable, there is a consistent lag between above and below ground tissues indicative of transport.

    • 2017 report:

      Activities:  We have developed a model of Spartina production and biomass that includes a simple phenology with regard to translocation of resources between above and below ground biomass. We are in the process of parameterizing the model for the Flux Tower data.

      Results:  Develop a Spartina physiological model Jung & Burd (2017) describes seasonal translocation of several different nonstructural carbohydrates between aboveand belowground tissues of Spartina alterniflora (Fig. 8).

  • 3A.8  Develop a model to predict porewater salinity (yr 1-3)
    • 2 report:

      Activities:  2014: We have developed the conceptual framework for the porewater model, aimed to interface with the physical flow model.
      2015: We have incorporated precipitation, evapotranspiration, salt exchange, drainage, groundwater, tidal inundation, and surface runoff into our porewater model. We also developed an algorithm to determine the water level at which the marsh floods.
      2016: We have finalized the soil model, and begun to compare model results to measured porewater salinities. The independent estimates of freshwater inputs agree well, suggesting that the flow modeling and ET estimates are consistent.
      2017: We have validated the soil model and continue to refine it.
      2018: We have further refined the soil model, expanding the model for data comparison across different vegetation zones.

      Results:  2015: Using the critical-flooding algorithm, analysis of marsh topography revealed that the marsh platform has many depressions (on the order of 10% of the domain area) that do not flood until the tidal elevation is well above the depression elevation.
      2016: One outcome of this effort is to constrain freshwater input to the upper Duplin. This work was presented at the 2014 AGU meeting and the 2015 LTER ASM.
      2018: Develop a model to predict porewater salinity The model revealed that porewater salinity is sensitive to the composition of the creek water in the low marsh and to evapotranspiration in the high marsh.

    • 2013 report:

      Activities:  We have developed the conceptual framework for the porewater model, aimed to interface with the physical flow model.

    • 2014 report:

      Activities:  We have incorporated precipitation, evapotranspiration, salt exchange, drainage, groundwater, tidal inundation, and surface runoff into our porewater model. We also developed an algorithm to determine the water level at which the marsh floods.

      Results:  Using the critical-flooding algorithm, analysis of marsh topography revealed that the marsh platform has many depressions (on the order of 10% of the domain area) that do not flood until the tidal elevation is well above the depression elevation.

    • 2015 report:

      Activities:  We have finalized the soil model, and begun to compare model results to measured porewater salinities. The independent estimates of freshwater inputs agree well, suggesting that the flow modeling and ET estimates are consistent.

      Results:  One outcome of this effort is to constrain freshwater input to the upper Duplin. This work was presented at the 2014 AGU meeting and the 2015 LTER ASM.

    • 2016 report:

      Activities:  We have validated the soil model and continue to refine it.

    • 2017 report:

      Activities:  We have further refined the soil model, expanding the model for data comparison across different vegetation zones.

      Results:  Develop a model to predict porewater salinity The model revealed that porewater salinity is sensitive to the composition of the creek water in the low marsh and to evapotranspiration in the high marsh.

  • 3B.1  Assess changes in community composition along the salinity gradient of the Altamaha (yr 1-6)
    • 2 report:

      Activities:  2014: Epiphytic and benthic diatom samples were collected at 25 sites to characterization of diatom communities along the salinity gradient of the Altamaha.
      2015: In 2012 we began an annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha.
      2016: We continued our annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha.
      2017: We continued our annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha.
      2018: We continued our annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha. We have also developed a Random Forest classifier to extract vegetation patterns from Landsat5 imagery in areas with a fluctuating mix of oligohaline and mesohaline vegetation.

      Results:  2015: Segarra et al. (2013 a,b) studied methane fluxes and methane oxidation in freshwater and brackish sediments along the Altamaha. Rate measurements of sulfate reduction and the anaerobic oxidation of methane (AOM), two processes not typically considered relevant in low salinity habitats, revealed their importance in freshwater settings.
      2018: Assess changes in community composition along the salinity gradient of the Altamaha We have found evidence for a longterm shift (19912011) from oligohaline to mesohaline vegetation in the brackish area of the Altamaha river, with the majority of the change occurring in drought years.

    • 2013 report:

      Activities:  Epiphytic and benthic diatom samples were collected at 25 sites to characterization of diatom communities along the salinity gradient of the Altamaha.

    • 2014 report:

      Activities:  In 2012 we began an annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha.

      Results:  Segarra et al. (2013 a,b) studied methane fluxes and methane oxidation in freshwater and brackish sediments along the Altamaha. Rate measurements of sulfate reduction and the anaerobic oxidation of methane (AOM), two processes not typically considered relevant in low salinity habitats, revealed their importance in freshwater settings.

    • 2015 report:

      Activities:  We continued our annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha.

    • 2016 report:

      Activities:  We continued our annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha.

    • 2017 report:

      Activities:  We continued our annual survey to document the transition in bankside vegetation from S. cynosuroides to S. alterniflora along the salinity gradient of the Altamaha. We have also developed a Random Forest classifier to extract vegetation patterns from Landsat5 imagery in areas with a fluctuating mix of oligohaline and mesohaline vegetation.

      Results:  Assess changes in community composition along the salinity gradient of the Altamaha We have found evidence for a longterm shift (19912011) from oligohaline to mesohaline vegetation in the brackish area of the Altamaha river, with the majority of the change occurring in drought years.

    • 2019 report:

      Activities:  We conduct an annual survey of bankside vegetation and sample plots with mixed vegetation to document transitions along the Altamaha River salinity gradient. In 2018 we added an annual photo survey of trees in the tidal fresh forest and observations on Broughton Island, which has a dynamic mix of oligohaline and mesohaline species. We have also sampled bald cypress deposits for a longer dendrochronology analysis (Fig. 5, Napora et al. 2019).

      Results:  Herbert et al. (in press) found that long-term addition of N and P to a tidal freshwater marsh in the Altamaha increased above-ground biomass, microbial biomass and N cycling, and N, P, and C assimilation and burial more than either nutrient alone (Fig. 6), and suggest that the ability of these habitats to mitigate eutrophication will depend on the quantity and relative proportion of N versus P entering the system.

    • 2020 report:

      Activities:  Temperature and flooding are two key variables known to affect S. alterniflora production, and both are experiencing long-term change (winter temperatures are increasing; sea-level rise is expected to increase flooding of low-lying areas). We have multiple efforts underway to collect information on these variables and how they affect plant production (see Area 2 Objective 4). We are planning a greenhouse experiment to evaluate how winter soil temperature interacts with salinity and nutrients to affect belowground processes and S. alterniflora phenology. (Our planned trials for this work were postponed due to COVID, but we developed experimental apparatus in 2020-1.) Hawman et al. (2021) evaluated the annual cycle of GPP and light use efficiency measured at the flux tower (Fig. 3) and found that the cloudiness index and daily maximum tide height are the primary factors that explain deviation in S. alterniflora light use efficiency. Nahrawi et al. (2020) reported that tidal flooding depressed NEE and that this effect varied seasonally as a function of plant phenology.

    • 2021 report:

      Activities:  We have multiple efforts underway to collect information on how temperature and flooding affect plant production. In June 2021 we began monthly sampling of vertical profiles of leaf area index in S. alterniflora canopies to quantify the effects of tidal flooding; in December we set up a hydroponic experiment to evaluate how winter soil temperature interacts with salinity and nutrients to affect belowground processes and S. alterniflora phenology (Fig. 2).

      Results:  Hawman and Mishra (in prep) found that NEE decreased sharply under flooded conditions, and that accounting for changes in emergent leaf area index due to tidal flooding improved relationships with the Sentinel-2 near-infrared reflectance index.

  • 3B.2  Conduct field manipulation of salt water intrusion in a low-salinity tidal marsh (yr 1-6)
    • 2 report:

      Activities:  2014: We have begun collecting baseline data (aboveground biomass, species composition, photosynthesis-respiration, greenhouse gas emissions, SET measurements). In a companion experiment conducted in the greenhouse, we exposed plants from tidal fresh and brackish marshes to 7 different salinity regimes over a period of 3 months.
      2015: In the GCE SALTex (Seawater Addition Long Term) experiment we began dosing experimental plots with saltwater (press and pulse plots) and are monitoring porewater, vegetation, herbivores, soil surface elevation, and gas exchange. In a companion experiment, we exposed mixtures of plants from tidal fresh and brackish marshes to water of differing salinity and for different lengths of time in the greenhouse.
      2016: We have been dosing the SALTEx plots for almost 2 years and are continuing monitoring of porewater, soil surface elevation, and gas exchange. We have begun monitoring surface algae and extracellular enzyme activity in the plots also. We are collecting soils for future analysis of microbial communities. This past year we also harvested the greenhouse mesocosm experiment in which we exposed mixtures of plants from tidal fresh and brackish marshes to water of differing salinity and for different lengths of time.
      2017: We dosed the SALTEx plots for a 3rd year and continue to monitor porewater, soil surface elevation, gas exchange, vegetation and invertebrates. Soil samples from SALTEx plots were all extracted and analyzed for molecular biomarkers using GC­MS. We have also begun monitoring surface algae and extracellular enzyme activity. In May 2016 we initiated an experiment to compare the rates of decomposition of Zizaniopsis miliacea roots between treatments.
      2018: We dosed the SALTEx plots for a 4th year and continue to monitor porewater, soil characteristics, plants and animals. This past summer we used the 15N in situ pushpull method to determine denitrification rates in the plots (samples are currently being processed). Recruitment of brackish marsh plants into SALTEx plots has been slow, so in May 2017 we transplanted 3 species to see how well they would perform if they did arrive.

      Results:  2014: Initial results from the greenhouse experiment indicated that sensitivity to saline conditions varied markedly among plant species. For example, Polygonum hydropiperoides aboveground biomass decreased sharply when exposed to saline conditions for 16 or 31 days per month, while Zizaniopsis miliacea biomass showed no response to salinity treatments (Fig. 6).
      2015: Porewater salinity in the press plots has increased from 0 to about 4. Some plant species, especially forbs, are visibly stressed, and hydrogen sulfide emissions are evident. In the greenhouse, plant community composition was highly sensitive to salinity regime. We will monitor mixtures again in the spring to assess overwintering survival.
      2016: Seawater additions resulted in increased sulfate and chloride immediately, and porewater nitrogen and phosphorus increased 2-4 months after initial seawater application. Most of the freshwater plants have died in the press plots, which are continually exposed to brackish water (Results Fig. 7). Ecosystem respiration and methane emissions are also significantly lower and we are seeing a measurable decline in elevation and an increase in soil surface temperature. In the pulse treatments, the most sensitive plant species died back temporarily but recovered in the following year. In the mesocosms, plant community responses were a function of both salinity and the duration of exposure to saline water.
      2017: Craft et al. (2016) provided an overview of the SALTEX project and summarized initial results (Fig. 11). Plots subject to press additions of seawater have increased porewater Cl, SO4, H2S, N, and DRP, increased temperature, vegetation loss, and up to 2 cm of soil elevation loss. Preliminary results suggest a strong decrease of plant wax biomarker concentrations.
      2018: Conduct field manipulation of salt water intrusion in a lowsalinity tidal marsh Li and Pennings described the vegetation response in field as part of the SALTEx experiment (Li and Pennings in review). In an accompanying mesocosm experiment, they found that both species richness and plant biomass decreased with increasing pulse duration and salinity (Fig. 9).

    • 2013 report:

      Activities:  We have begun collecting baseline data (aboveground biomass, species composition, photosynthesis-respiration, greenhouse gas emissions, SET measurements). In a companion experiment conducted in the greenhouse, we exposed plants from tidal fresh and brackish marshes to 7 different salinity regimes over a period of 3 months.

      Results:  Initial results from the greenhouse experiment indicated that sensitivity to saline conditions varied markedly among plant species. For example, Polygonum hydropiperoides aboveground biomass decreased sharply when exposed to saline conditions for 16 or 31 days per month, while Zizaniopsis miliacea biomass showed no response to salinity treatments (Fig. 6).

    • 2014 report:

      Activities:  In the GCE SALTex (Seawater Addition Long Term) experiment we began dosing experimental plots with saltwater (press and pulse plots) and are monitoring porewater, vegetation, herbivores, soil surface elevation, and gas exchange. In a companion experiment, we exposed mixtures of plants from tidal fresh and brackish marshes to water of differing salinity and for different lengths of time in the greenhouse.

      Results:  Porewater salinity in the press plots has increased from 0 to about 4. Some plant species, especially forbs, are visibly stressed, and hydrogen sulfide emissions are evident. In the greenhouse, plant community composition was highly sensitive to salinity regime. We will monitor mixtures again in the spring to assess overwintering survival.

    • 2015 report:

      Activities:  We have been dosing the SALTEx plots for almost 2 years and are continuing monitoring of porewater, soil surface elevation, and gas exchange. We have begun monitoring surface algae and extracellular enzyme activity in the plots also. We are collecting soils for future analysis of microbial communities. This past year we also harvested the greenhouse mesocosm experiment in which we exposed mixtures of plants from tidal fresh and brackish marshes to water of differing salinity and for different lengths of time.

      Results:  Seawater additions resulted in increased sulfate and chloride immediately, and porewater nitrogen and phosphorus increased 2-4 months after initial seawater application. Most of the freshwater plants have died in the press plots, which are continually exposed to brackish water (Results Fig. 7). Ecosystem respiration and methane emissions are also significantly lower and we are seeing a measurable decline in elevation and an increase in soil surface temperature. In the pulse treatments, the most sensitive plant species died back temporarily but recovered in the following year. In the mesocosms, plant community responses were a function of both salinity and the duration of exposure to saline water.

    • 2016 report:

      Activities:  We dosed the SALTEx plots for a 3rd year and continue to monitor porewater, soil surface elevation, gas exchange, vegetation and invertebrates. Soil samples from SALTEx plots were all extracted and analyzed for molecular biomarkers using GC­MS. We have also begun monitoring surface algae and extracellular enzyme activity. In May 2016 we initiated an experiment to compare the rates of decomposition of Zizaniopsis miliacea roots between treatments.

      Results:  Craft et al. (2016) provided an overview of the SALTEX project and summarized initial results (Fig. 11). Plots subject to press additions of seawater have increased porewater Cl, SO4, H2S, N, and DRP, increased temperature, vegetation loss, and up to 2 cm of soil elevation loss. Preliminary results suggest a strong decrease of plant wax biomarker concentrations.

    • 2017 report:

      Activities:  We dosed the SALTEx plots for a 4th year and continue to monitor porewater, soil characteristics, plants and animals. This past summer we used the 15N in situ pushpull method to determine denitrification rates in the plots (samples are currently being processed). Recruitment of brackish marsh plants into SALTEx plots has been slow, so in May 2017 we transplanted 3 species to see how well they would perform if they did arrive.

      Results:  Conduct field manipulation of salt water intrusion in a lowsalinity tidal marsh Li and Pennings described the vegetation response in field as part of the SALTEx experiment (Li and Pennings in review). In an accompanying mesocosm experiment, they found that both species richness and plant biomass decreased with increasing pulse duration and salinity (Fig. 9).

    • 2019 report:

      Activities:  We sampled porewater, greenhouse gases, vegetation, invertebrates, and soil biomarkers in the SALTEx experiment to track recovery following cessation of dosing in December 2017.

      Results:  Widney et al. (2019) published a synthesis of the biochemical effects of the SALTEx experiment showing that three years of saltwater intrusion increased porewater Cl-, SO4, HS and inorganic N (NH4 and NO3) and decreased plant N storage (see Accomplishments). Manuscripts regarding the changes in soil elevation and the vegetation response are both in review.

    • 2020 report:

      Activities:  Marsh fauna can often drive marsh response to perturbations, and we are conducting several investigations to understand their effects on marsh structure and function. For example, Angelini et al. (2016) demonstrated that there is a mutualism between S. alterniflora and mussels in which S. alterniflora patches associated with mussels are more resilient to drought because mussels enhance water storage and reduce soil stress. In 2019 we began field measurements to evaluate how interactions between cordgrass and ribbed mussels influence tide water chemistry and the net import/export of materials from marshes. Sharp and Angelini (in press) found that birds and nekton can enhance S. alterniflora resilience to drought by increasing soil aeration via probing as well as their suppression of snail grazers and transmission of disease to snails. GCE researchers have also conducted several studies examining top-down effects of megafauna such as feral hogs on marsh ecosystems (see key findings).

    • 2021 report:

      Activities:  We continue to conduct focused studies to understand the relationships between marsh fauna and environmental variables. Activities this past year included studies of detritivores, variation in metabolic rate of snails, and the effects of herbivory by megafauna on salt marsh invertebrates. We also continue to refine designs for biomimic sensors.

      Results:  Seer et al. (2021) found that the colonization of litter by infauna shifts during decomposition as litter becomes less labile (Fig. 3).

  • 3B.3  Apply SLAMM to the GCE domain (yr 1-3)
    • 2 report:

      Activities:  2014: We are collecting field measurements of elevation of different vegetation communities, including tidal forest, to improve SLAMM’s modeling capabilities, and are also working to link it to both the point-based MEM model of marsh accretion developed by Jim Morris (USC) and to Squeezebox, our estuarine salinity model.
      2015: New inputs were acquired for SLAMM, including updated LiDAR, bathymetry and ground-truthed RTK data in the Altamaha River. Modeling runs are underway using various scenarios of sea level rise and river discharge.
      2016: In the Altamaha River we have run the SLAMM model with the new inputs and are also working with Jim Morris (Univ. South Carolina) to run his point-based model (Marsh Equilibrium Model). In the larger domain, we have mapped the distribution of salt, brackish and tidal fresh marsh in the GCE domain using both the GCE new high-resolution imagery, LiDAR and regression tree models. These distributions will provide better initial conditions to better inform future SLAMM modeling.
      2017: E. Herbert worked with the SLAMM research group to include ecogeomorphic feedbacks between flooding depth, biomass, and accretion. Updated elevation and habitat datasets were also used (See Objective 4B3).
      2018: Completed yr 4.

      Results:  2016: Hauer et al. (2015) ran the SLAMM model as part of a coast-wide study of GA designed to improve population projections at sub-county scales. In McIntosh County, where the GCE is located, the projected population at risk to the threat of sea-level rise varied from 291 to 2,887, depending on the increase in sea level (1 or 2 m) and the year considered (2050 or 2100).
      2017: When the improved SLAMM model was run using a variable accretion rate, the predicted outcome for wetlands along the Altamaha River in the face of SLR changed from one of marsh loss to one of marsh gain.

    • 2013 report:

      Activities:  We are collecting field measurements of elevation of different vegetation communities, including tidal forest, to improve SLAMM’s modeling capabilities, and are also working to link it to both the point-based MEM model of marsh accretion developed by Jim Morris (USC) and to Squeezebox, our estuarine salinity model.

    • 2014 report:

      Activities:  New inputs were acquired for SLAMM, including updated LiDAR, bathymetry and ground-truthed RTK data in the Altamaha River. Modeling runs are underway using various scenarios of sea level rise and river discharge.

    • 2015 report:

      Activities:  In the Altamaha River we have run the SLAMM model with the new inputs and are also working with Jim Morris (Univ. South Carolina) to run his point-based model (Marsh Equilibrium Model). In the larger domain, we have mapped the distribution of salt, brackish and tidal fresh marsh in the GCE domain using both the GCE new high-resolution imagery, LiDAR and regression tree models. These distributions will provide better initial conditions to better inform future SLAMM modeling.

      Results:  Hauer et al. (2015) ran the SLAMM model as part of a coast-wide study of GA designed to improve population projections at sub-county scales. In McIntosh County, where the GCE is located, the projected population at risk to the threat of sea-level rise varied from 291 to 2,887, depending on the increase in sea level (1 or 2 m) and the year considered (2050 or 2100).

    • 2016 report:

      Activities:  E. Herbert worked with the SLAMM research group to include ecogeomorphic feedbacks between flooding depth, biomass, and accretion. Updated elevation and habitat datasets were also used (See Objective 4B3).

      Results:  When the improved SLAMM model was run using a variable accretion rate, the predicted outcome for wetlands along the Altamaha River in the face of SLR changed from one of marsh loss to one of marsh gain.

    • 2017 report:

      Activities:  Completed yr 4.

    • 2020 report:

      Activities:  The sources and sinks of carbon in coastal ecosystems are an important component of the global carbon budget. In GCE-III Wang et al. (2017) published a complete budget of CO2 exchange in the Duplin River estuary (based on DIC exchange in a creek in combination with estuary metabolism measurements and data from the flux tower), which showed that although the marsh is a net sink for CO2 it is also a source of C to estuarine and coastal water.. High-frequency monitoring conducted with funds provided through an ROA supplement showed evidence for DIC export during spring tides, which is consistent with the Duplin River budget. We are following up on these observations with additional sampling this summer. In the marsh, our eddy covariance data provide estimates of vertical flux for the marsh carbon budget. We have evidence that S. alterniflora photosynthesizes when fully submerged, suggesting that part of the fixed C is likely exchanged with the water rather than the atmosphere. Spivak et al. (2019) wrote a synthesis paper highlighting the importance of understanding the key biogeochemical mechanisms within the marsh that control decomposition of soil organic matter when evaluating the effects of changing environmental conditions on coastal wetland C storage (Fig. 4), and we are using this to guide sampling and analysis of cores collected along elevation, salinity, and disturbance gradients. (See also cross-site research.)

    • 2021 report:

      Activities:  We are collaborating with the Univ. of Wisconsin to collect pCO2 data at the GCE flux tower to understand CO2 export by tidal waters. We also completed a lab experiment on decomposition controls under varying oxygen conditions and bioavailable C inputs.

      Results:  GCE was part of a cross-site study that highlighted the significant role of transport in organic matter dynamics (Harms et al. 2021).

  • 3C.1  Continue to monitor groundwater salinity, temperature and pressure on instrumented hammocks (yr 1-2)
    • 2 report:

      Activities:  2014: We are analyzing pressure time series from groundwater well transects connecting the marsh, a hammock and the main upland at Blackbeard Island to delineate the relative importance of tidal pumping, precipitation and density variations in driving groundwater flow across a marsh transect.
      2015: We continued our analysis of the pressure data collected at groundwater wells installed along an upland-to-marsh transect. However, the sensors in these wells continue to fail under harsh coastal conditions.
      2016: Sensors have been removed from hammock wells because of high failure rate. We have worked with Schlumberger (the sensor company) to redesign the sensors to be more resistant to the harsh conditions in the wells.
      2017: We have outlined time series analyses to interpret the groundwater observational data, building on Ledoux (2015).
      2018: Completed yr 3.

      Results:  2015: We have separated out tidal influences on the pressure signals by fitting the signal using constituents with known tidal frequencies. We also corrected erroneous ground elevations. Although the identification of other driving forces is ongoing, we have determined that marsh topography has a critical impact on the pressure signal.
      2016: Ledoux (2015) used the sensor data to quantify the role of upland / precipitation driven groundwater flow. He found that it exceeds the impact of tidal forcing on net groundwater flow from the upland to the marsh.

      Plans:  2015: We have halted data collection due to sensor failure. The manufacturer is working to design more robust sensors that can be field-tested at the GCE site.
      2016: New sensors have been developed and will be deployed for testing at GCE sites.

    • 2013 report:

      Activities:  We are analyzing pressure time series from groundwater well transects connecting the marsh, a hammock and the main upland at Blackbeard Island to delineate the relative importance of tidal pumping, precipitation and density variations in driving groundwater flow across a marsh transect.

    • 2014 report:

      Activities:  We continued our analysis of the pressure data collected at groundwater wells installed along an upland-to-marsh transect. However, the sensors in these wells continue to fail under harsh coastal conditions.

      Results:  We have separated out tidal influences on the pressure signals by fitting the signal using constituents with known tidal frequencies. We also corrected erroneous ground elevations. Although the identification of other driving forces is ongoing, we have determined that marsh topography has a critical impact on the pressure signal.

      Plans:  We have halted data collection due to sensor failure. The manufacturer is working to design more robust sensors that can be field-tested at the GCE site.

    • 2015 report:

      Activities:  Sensors have been removed from hammock wells because of high failure rate. We have worked with Schlumberger (the sensor company) to redesign the sensors to be more resistant to the harsh conditions in the wells.

      Results:  Ledoux (2015) used the sensor data to quantify the role of upland / precipitation driven groundwater flow. He found that it exceeds the impact of tidal forcing on net groundwater flow from the upland to the marsh.

      Plans:  New sensors have been developed and will be deployed for testing at GCE sites.

    • 2016 report:

      Activities:  We have outlined time series analyses to interpret the groundwater observational data, building on Ledoux (2015).

    • 2017 report:

      Activities:  Completed yr 3.

    • 2019 report:

      Activities:  We monitor vegetation dynamics in 9 high marsh mixtures and have begun annual drone flights to scale up to the surrounding landscape. We also operate two web applications where citizen scientists align and extract data from photographs taken along transects that begin in the high marsh.

  • 3C.2  Survey high marsh characteristics in sites with different land-use categories (yr 1-2)
    • 2 report:

      Activities:  2014: In summer 2013 we conducted field surveys of the high marsh at 60 sites. The survey included residential sites both with and without bulkhead structures as well as forested areas. Plant and animal distributions, sediment characteristics, porewater nutrient concentrations and microbial community composition are currently being analyzed to assess patterns and processes associated with upland class. GCE investigators also participated in a cross- site working group to evaluate the ecological impacts of coastal armoring on soft sediment coastal ecosystems.
      2015: We analyzed plant and animal distributions, sediment, porewater nutrients, parasite loads, and microbial community from samples collected during the yr 1 high marsh survey. We also continued working with a cross- LTER group to evaluate the ecological impacts of coastal armoring.
      2016: We conducted further analysis on the plant and animal distributions, sediment, porewater nutrients, parasite loads, and microbial community from samples collected during the yr 1 high marsh survey. We also continued working with a cross- LTER group to evaluate the ecological impacts of coastal armoring.
      2017: We have completed all analyses on this project.
      2018: Completed yr 4.

      Results:  2014: We used geographic information systems to investigate the spatial distribution of armored structures in the state of Georgia with respect to land use and land cover. We found that upland immediately adjacent to hardened shorelines was highly developed at the parcel scale (Fig. 7), and the extent of armoring was tightly linked with indicators of urbanization at the county scale (impervious surface coverage; r = 0.98). GCE investigators participated in cross-site soft-sediment working group that conducted to a literature synthesis and the development of a matrix of ecological responses to shoreline armoring, both of which will be included in a manuscript that is currently being developed.
      2015: Elevations at bulkheaded, developed sites are lower than those at either unarmored developed sites or forested areas, and have a higher % cover of S. alterniflora (Fig. 6). The sediments at these areas are finer-grained, suggesting that transfer of sandy upland materials has been cut off by the installation of armoring.
      2016: Freshwater input from the upland structures the high marsh community in our study sites. Spartina dominates near developed uplands with bulkheads; Juncus dominates near developed, but unarmored shorelines. We also found that movement across the ecotone by the land crab Armases cinerum varied with upland structure. This latter observation is especially interesting given that Hubner et al. (2015) found evidence for differences in both dietary and habitat preferences of Armases that can move freely between high marsh and coastal forests.
      2017: Gehman et al. (submitted) found that the effects of shoreline armoring were subtle in the high marsh. Armoring tends to make a site more like the low marsh (higher Spartina density and siltier sediment).

    • 2013 report:

      Activities:  In summer 2013 we conducted field surveys of the high marsh at 60 sites. The survey included residential sites both with and without bulkhead structures as well as forested areas. Plant and animal distributions, sediment characteristics, porewater nutrient concentrations and microbial community composition are currently being analyzed to assess patterns and processes associated with upland class. GCE investigators also participated in a cross- site working group to evaluate the ecological impacts of coastal armoring on soft sediment coastal ecosystems.

      Results:  We used geographic information systems to investigate the spatial distribution of armored structures in the state of Georgia with respect to land use and land cover. We found that upland immediately adjacent to hardened shorelines was highly developed at the parcel scale (Fig. 7), and the extent of armoring was tightly linked with indicators of urbanization at the county scale (impervious surface coverage; r = 0.98). GCE investigators participated in cross-site soft-sediment working group that conducted to a literature synthesis and the development of a matrix of ecological responses to shoreline armoring, both of which will be included in a manuscript that is currently being developed.

    • 2014 report:

      Activities:  We analyzed plant and animal distributions, sediment, porewater nutrients, parasite loads, and microbial community from samples collected during the yr 1 high marsh survey. We also continued working with a cross- LTER group to evaluate the ecological impacts of coastal armoring.

      Results:  Elevations at bulkheaded, developed sites are lower than those at either unarmored developed sites or forested areas, and have a higher % cover of S. alterniflora (Fig. 6). The sediments at these areas are finer-grained, suggesting that transfer of sandy upland materials has been cut off by the installation of armoring.

    • 2015 report:

      Activities:  We conducted further analysis on the plant and animal distributions, sediment, porewater nutrients, parasite loads, and microbial community from samples collected during the yr 1 high marsh survey. We also continued working with a cross- LTER group to evaluate the ecological impacts of coastal armoring.

      Results:  Freshwater input from the upland structures the high marsh community in our study sites. Spartina dominates near developed uplands with bulkheads; Juncus dominates near developed, but unarmored shorelines. We also found that movement across the ecotone by the land crab Armases cinerum varied with upland structure. This latter observation is especially interesting given that Hubner et al. (2015) found evidence for differences in both dietary and habitat preferences of Armases that can move freely between high marsh and coastal forests.

    • 2016 report:

      Activities:  We have completed all analyses on this project.

      Results:  Gehman et al. (submitted) found that the effects of shoreline armoring were subtle in the high marsh. Armoring tends to make a site more like the low marsh (higher Spartina density and siltier sediment).

    • 2017 report:

      Activities:  Completed yr 4.

    • 2019 report:

      Activities:  The high marsh experiment has had little effect on water flow, and hence there have been no effects on plants or invertebrates. However, data from the wells are being analyzed to calculate hydraulic gradients for groundwater flow models (Fig. 3).

  • 3C.3  Conduct upland manipulation of water flow to high marsh areas (yr 3-6)
    • 2 report:

      Activities:  2015: Begins yr 3
      2016: We have identified a potential site for this experiment. We have surveyed and taken cores along three transect across the marsh-upland boundary to characterize the stratigraphy at the site and have requested permission from DNR to conduct the experiment.
      2017: We received a permit from DNR for this experiment and collected pre­treatment data on plants, benthic algae, invertebrates, and groundwater. Treatments are set to begin later in 2016.
      2018: We have two years of "before" data in the plots (2015 and 2016) and began treatments in 2017. We are tracking groundwater pressure continuously and sampled plants and invertebrates three times in the summer of 2017.

      Results:  2016: The site proposed for the upland manipulation has a well-developed, clean, sandy, porous layer at ~1.0 m depth that serves as a conduit for freshwater from the uplands to influence the marsh.
      2018: Conduct upland manipulation of water flow to high marsh areas Groundwater data are being used to characterize the horizontal hydraulic gradient at the upland manipulation site (Fig. 10).

    • 2014 report:

      Activities:  Begins yr 3

    • 2015 report:

      Activities:  We have identified a potential site for this experiment. We have surveyed and taken cores along three transect across the marsh-upland boundary to characterize the stratigraphy at the site and have requested permission from DNR to conduct the experiment.

      Results:  The site proposed for the upland manipulation has a well-developed, clean, sandy, porous layer at ~1.0 m depth that serves as a conduit for freshwater from the uplands to influence the marsh.

    • 2016 report:

      Activities:  We received a permit from DNR for this experiment and collected pre­treatment data on plants, benthic algae, invertebrates, and groundwater. Treatments are set to begin later in 2016.

    • 2017 report:

      Activities:  We have two years of "before" data in the plots (2015 and 2016) and began treatments in 2017. We are tracking groundwater pressure continuously and sampled plants and invertebrates three times in the summer of 2017.

      Results:  Conduct upland manipulation of water flow to high marsh areas Groundwater data are being used to characterize the horizontal hydraulic gradient at the upland manipulation site (Fig. 10).

  • 3C.4  Develop a clonal plant model to explore vegetation dynamics (yr 3-5)
    • 2 report:

      Activities:  2015: Begins yr 3
      2016: We measured the clonal architecture of 4 species of plants to parameterize the model.
      2017: The graduate student assigned to this project dropped out, and we have shifted our emphasis to documenting vegetation patterns on photographs collected as part of intensive annual surveys of a high marsh. These data will inform the upland manipulation and future work on a clonal plant model.
      2018: We have shifted our emphasis to documenting vegetation patterns in photographs collected in annual surveys of a high marsh to inform the upland manipulation and future work on a clonal plant model. The web sites that allow citizen scientists to contribute data towards this project have been migrated to servers at UGA.

      Results:  2016: We have established a web site (ScalingUpMarshScience.cs.uh.edu) that allows volunteers to help us align thousands of photographs into a mosaic of the marsh. A second web site under development will use volunteers to score plant and animal abundance in the images. This information will be used to compare with model predictions.
      2017: We are using citizen scientists to align and extract data from over 70,000 photographs that we have collected from a high marsh site over 7 years.

    • 2014 report:

      Activities:  Begins yr 3

    • 2015 report:

      Activities:  We measured the clonal architecture of 4 species of plants to parameterize the model.

      Results:  We have established a web site (ScalingUpMarshScience.cs.uh.edu) that allows volunteers to help us align thousands of photographs into a mosaic of the marsh. A second web site under development will use volunteers to score plant and animal abundance in the images. This information will be used to compare with model predictions.

    • 2016 report:

      Activities:  The graduate student assigned to this project dropped out, and we have shifted our emphasis to documenting vegetation patterns on photographs collected as part of intensive annual surveys of a high marsh. These data will inform the upland manipulation and future work on a clonal plant model.

      Results:  We are using citizen scientists to align and extract data from over 70,000 photographs that we have collected from a high marsh site over 7 years.

    • 2017 report:

      Activities:  We have shifted our emphasis to documenting vegetation patterns in photographs collected in annual surveys of a high marsh to inform the upland manipulation and future work on a clonal plant model. The web sites that allow citizen scientists to contribute data towards this project have been migrated to servers at UGA.

Research Projects by Objective

3A. Spartina Marsh Studies

2. Evaluate Spartina plant phenology and above- and below-ground production

Flux tower monthly vegetation monitoring
description: GCE web page, plain web page
date range: ongoing (since 2013)
principal investigator(s): Steven C. Pennings

6. Monitor headward erosion in tidal creeks

Do hydrological conditions at creek heads stimulate Sesarma reticulatum recruitment?
description: GCE web page, plain web page
date range: 2012 to 2015
principal investigator(s): Steven C. Pennings

Effects of Sesarma reticulatum on tidal creek growth.
description: GCE web page, plain web page
date range: 2012 to 2017
principal investigator(s): Steven C. Pennings

7. Develop a Spartina physiological model

Models of salt marsh plant species
description: GCE web page, plain web page
date range: ongoing (since 2009)
principal investigator(s): Adrian B. Burd

3A. Spartina marsh studies

7. Develop a Spartina physiological model

Marsh plant studies in support of hammock plant model
description: GCE web page, plain web page
date range: 2009 to 2012
principal investigator(s): Steven C. Pennings

3A. Spartina Marsh Studies

8. Develop a model to predict porewater salinity

Hammock groundwater modeling
description: GCE web page, plain web page
date range: ongoing (since 2010)
principal investigator(s): Samantha B. Joye, Christof Meile

3B. Fresh/brackish Marsh Studies

1. Assess changes in community composition along the salinity gradient of the Altamaha

Algal survey and light availability experiment along salinity gradient
description: GCE web page, plain web page
date range: 2013 to 2013
principal investigator(s): Christopher B. Craft

Nutrient Limitation of Benthic Biofilms Along the Estuarine Salinity Gradient
description: GCE web page, plain web page
date range: ongoing (since 2012)
principal investigator(s): Christopher B. Craft

2. Conduct field manipulation of salt water intrusion in a low-salinity tidal marsh

Freshwater plant responses to saltwater pulses
description: GCE web page, plain web page
date range: ongoing (since 2014)
principal investigator(s): Steven C. Pennings

Long-term Nutrient (N, P) Additions to a Tidal Freshwater Marsh
description: GCE web page, plain web page
date range: ongoing (since 2004)
principal investigator(s): Christopher B. Craft

Seawater Addition Long Term Experiment (SALTEx), a long-term field manipulation experiment in a Zizaniopsis marsh in the Altamaha River
description: GCE web page, plain web page
date range: ongoing (since 2011)
principal investigator(s): Christopher B. Craft

2. Conduct field manipulation of salt water intrusion in a low-salinity tidal marsh

Short-term responses of tidal low-salinity marsh vegetation to saltwater intrusion
description: GCE web page, plain web page
date range: 2013 to 2013
principal investigator(s): Steven C. Pennings

3C. High Marsh Studies

1.  Continue to monitor groundwater salinity, temperature and pressure on instrumented hammocks

Intensive hammock characterization
description: GCE web page, plain web page
date range: ongoing (since 2008)
principal investigator(s): Merryl Alber, Clark R. Alexander Jr.

3. Conduct upland manipulation of water flow to high marsh areas

Upland manipulation of water flow to high marsh areas
description: GCE web page, plain web page
date range: 2015 to 2018
principal investigator(s): Steven C. Pennings

4. Develop a clonal plant model to explore vegatation dynamics

Photomapping of marsh vegetation
description: GCE web page, plain web page
date range: ongoing (since 2010)
principal investigator(s): Steven C. Pennings

4. Develop a clonal plant model to explore vegetation dynamics

Marsh plant studies in support of hammock plant model
description: GCE web page, plain web page
date range: 2009 to 2012
principal investigator(s): Steven C. Pennings

Outcomes Figures and Tables

2014 Results Figure 1

2014 Results Figure 1

2014 Results Figure 4

2014 Results Figure 4

2014 Results Figure 5

2014 Results Figure 5

2013 Outcomes Figure 3

2013 Outcomes Figure 3

2013 Results Figure 6

2013 Results Figure 6

2013 Results Figure 7

2013 Results Figure 7

2014 Results Figure 6

2014 Results Figure 6

Literature Cited in Outcomes

Craft, C.B. 2012. Tidal freshwater forest accretion does not keep pace with sea level rise. Global Change Biology. 18:3615-3623. (DOI: 10.1111/gcb.12009)

Segarra, K., Comerford, C., Slaughter, J.B. and Joye, S.B. 2013. Impact of electron acceptor availability on the anaerobic oxidation of methane in coastal freshwater and brackish wetland sediments. Geochimica et Cosmochimica Acta. 115:15 - 30. (DOI: 10.1016/j.gca.2013.03.029)

Segarra, K., Samarkin, V., King, E., Meile, C. and Joye, S.B. 2013. Seasonal variations of methane fluxes from an unvegetatedtidal freshwater mudflat (Hammersmith Creek, GA). Biogeochemistry. 115:349 - 361. (DOI: 10.1007/s10533-013-9840-6)

All Related Publications

Journal Articles

Li, F., Angelini, C., Byers, J., Craft, C.B. and Pennings, S.C. 2022. Responses of a tidal freshwater marsh plant community to chronic and pulsed saline intrusion. Journal of Ecology. 110:1508-1524. (DOI: 10.1111/1365-2745.13885)

Simon, J., Hopkinson, B.M. and Pennings, S.C. 2022. Insights into Salt Marsh Plant Community Distributions Through Computer Vision and Structural Equation Modeling. Estuaries and Coasts. (DOI: https://doi.org/10.1007/s12237-022-01147-w)

Li, F. and Pennings, S.C. 2019. Response and Recovery of Low-Salinity Marsh Plant Communities to Presses and Pulses of Elevated Salinity. Estuaries and Coasts. 42:708-718. (DOI: 10.1007/s12237-018-00490-1)

Herbert, E., Schubauer-Berigan, J.P. and Craft, C.B. 2018. Differential effects of chronic and acute simulated seawater intrusion on tidal freshwater marsh carbon cycling. Biogeochemistry. 138:137–154. (DOI: 10.1007/s10533-018-0436-z)

Li, F. and Pennings, S.C. 2018. Responses of tidal freshwater and brackish marsh macrophytes to pulses of saline water simulating sea level rise and reduced discharge. Wetlands. 38:885-891. (DOI: 10.1007/s13157-018-1037-2)

Alexander, C.R. Jr., Hodgson, J. and Brandes, J. 2017. Sedimentary processes and products in a mesotidal salt marsh environment: insights from Groves Creek, Georgia. Geo-Marine Letters. 37:345-359. (DOI: 10.1007/s00367-017-0499-1)

Jung, Y. and Burd, A.B. 2017. Seasonal changes in above- and below-ground non-structural carbohydrates (NSC) in Spartina alterniflora in a marsh in Georgia, USA. Aquatic Botany. 140:13-22. (DOI: https://doi.org/10.1016/j.aquabot.2017.04.003)

Craft, C.B., Herbert, E., Li, F., Smith, D., Schubauer-Berigan, J.P., Widney, S., Angelini, C., Pennings, S.C., Medeiros, P.M., Byers, J. and Alber, M. 2016. Climate change and the fate of coastal wetlands. Wetland Science and Practice. 33(3):70-73.

Hawkes, A., Kemp, A., Donnelly, J., Horton, B., Peltier, W., Cahill, N., Hill, D., Ashe, E. and Alexander, C. 2016. Relative Sea-Level Change in Northeastern Florida (USA) During the Last ~8.0 KA. Quaternary Science Reviews. (DOI: 10.1016/j.quascirev.2016.04.016)

Herbert, E., Boon, P., Burgin, A.J., Neubauer, S.C., Franklin, R.B., Ardon, M., Hopfensperger, K.N., Lamers, L. and Gell, P. 2015. A global perspective on wetland salinization: Ecological consequences of a growing threat to freshwater wetlands. Ecosphere. 6(10)(206):1-43. (DOI: 10.1890/ES14-00534.1)

Wieski, K. and Pennings, S.C. 2014. Latitudinal variation in resistance and tolerance to herbivory of a salt marsh shrub. Ecography. 37:763-769. (DOI: 10.1111/ecog.00498)

Porubsky, W.P., Joye, S.B., Moore, W.S., Tuncay, K. and Meile, C. 2011. Field measurements and modeling of groundwater flow and biogeochemistry at Moses Hammock, a backbarrier island on the Georgia coast. Biogeochemistry. 104:69-90. (DOI: 10.1007/s10533-010-9484-8)

Meile, C., Porubsky, W.P., Walker, R.L. and Payne, K. 2009. Natural Attenuation Of Nitrogen Loading From Septic Effluents: Spatial And Environmental Controls. Water Research. 44(5):1399-1408. (DOI: 10.1016/j.watres.2009.11.019)

Theses and Dissertations

Jung, Y. 2018. Modeling Growth and Production Dynamics of Spartina Alterniflora. Ph.D. Dissertation. University of Georgia, Athens, GA. 148 pages.

Ledoux, J.G. 2015. Drivers of groundwater flow at a back barrier island - marsh transect in coastal Georgia. M.S. Thesis. The University of Georgia, Athens. 104 pages.

Conference Papers (Peer Reviewed)

Porubsky, W.P. and Meile, C. 2009. Controls on groundwater nutrient mitigation: Natural attenuation of nitrogen loading from septic effluents. In: Hatcher, K.J. (editor). Proceedings of the Georgia Water Resources Conference. Athens, Georgia.

Conference Posters and Presentations

Widney, S., Smith, D., Schubauer-Berigan, J.P., Herbert, E., Desha, J. and Craft, C.B. 2017. Poster: Changes in sediment porewater chemistry in response to simulated seawater intrusion in tidal freshwater marshes, Altamaha River, GA. Society of Wetland Scientists Annual Meeting, June 5-8, San Juan, Puerto Rico.

Smith, D., Herbert, E., Li, F., Widney, S., Desha, J., Schubauer-Berigan, J.P., Pennings, S.C., Angelini, C., Medeiros, P.M., Byers, J., Alber, M. and Craft, C.B. 2016. Poster: Seawater Addition Long Term Experiment (SALTEx). Georgia Department of Natural Resources Coastal Resources Division 2016 Climate Conference, November 2-3, 2016, Jekyll Island, GA.

Ledoux, J.G., Alexander, C.R. Jr. and Meile, C. 2015. Poster: Groundwater flow at the Georgia coast: Magnitude and drivers across a back barrier island – marsh transect. LTER All Scientists Meeting, Aug 30-Sept 2, Estes Park, CO.

Miklesh, D.M., McKnight, C.J., Di Iorio, D. and Meile, C. 2015. Poster: Controls on porewater salinity distributions in a southeastern salt marsh. LTER All Scientists Meeting, Aug 30-Sept 2, Estes Park, CO.

Ledoux, J.G., Alexander, C.R. Jr. and Meile, C. 2014. Poster: Delineating groundwater flow along a marsh transect at a back barrier island on the coast of Georgia. Southeastern Estuarine Research Society Fall meeting, November 6-8, Carolina Beach, NC.

Alexander, C.R. Jr., Alber, M., Hladik, C.M. and Pennings, S.C. 2010. Presentation: Physical-Biological Interactions in Coastal Settings: The Georgia Coastal Ecosystem LTER Example. American Geophysical Union - Meeting of the Americas, 9-13 August 2010, Foz do Iguacu, Brazil.

Alexander, C.R. Jr. 2008. Presentation: Stratigraphic Development of Holocene and Pleistocene Marsh Islands. Tidalites 2008 - Seventh International Conference on Tidal Environments, 25th-27th September, 2008, Qingdao, China.

Data Sets by Research Topic

Core LTER Data Sets

Algal Productivity

Green algae, cyanobacteria and diatom concentrations from the GCE-LTER Seawater Addition Long-Term Experiment (SALTEx) Project

Aquatic Invertebrate Ecology

Summer 2007 crab population survey based on crab hole counts in fifty-four GCE LTER Hammock sites and four GCE LTER study sites

Summer 2007 mollusc population survey in fifty-five GCE LTER Hammock sites and five GCE marsh monitoring sites

Botany

Grasshopper counts and feeding damage at the GCE-LTER Seawater Addition Long-Term Experiment (SALTEx) in 2016

General Nutrient Chemistry

Baseline soil chemistry data measurements from the GCE-LTER Seawater Addition Long-Term Experiment (SALTEx)

Geology

Soil surface temperature measurements from the GCE-LTER Seawater Addition Long-Term Experiment (SALTEx) Project

Groundwater Hydrology

Continuous groundwater well temperature, salinity and water level measurements at the GCE-LTER Seawater Addition Long-Term Experiment (SALTEx) site from May 2014 to February 2018

Hydrography/Hydrology

Groundwater well pressure, temperature, conductance, salinity and oxygen measurements for GCE-LTER study hammock PC_i_29 from 15-Aug-2008 to 13-Dec-2013

Groundwater well pressure, temperature, conductance, salinity and oxygen measurements for GCE-LTER study hammock HN_i_1 from 30-Oct-2008 to 03-Dec-2013

Multi-Disciplinary Study

GCE-LTER Hammock Well Vegetation and Invertebrate Monitoring - July 2008

Plant Ecology

Monthly Vegetation and Invertebrate Population Monitoring near the Georgia Coastal Ecosystems LTER Flux Tower

Leaf area index for Spartina alterniflora near the GCE-LTER Flux Tower in 2018 and 2019

Survey of high marsh plant structure and biomass for Spartina, Juncus, Borrichia and Batis specimens on Sapelo Island, Georgia during May to June 2015

Pot experiment on fresh and brackish marsh plants responses to salinity pulses in summer 2013

Summer 2007 survey of plant presence and abundance at fifty-five GCE LTER Hammock sites and five GCE marsh monitoring sites

Pore-water Chemistry

High frequency sediment surface temperature measurements at GCE-LTER study hammock PC_i_29 from 15-Aug-2008 to 24-Sep-2012

Terrestrial Insect Ecology

Survey of grasshopper abundance in GCE LTER hammock sites and five GCE LTER study sites
 
LTER
NSF

This material is based upon work supported by the National Science Foundation under grants OCE-9982133, OCE-0620959, OCE-1237140 and OCE-1832178. Any opinions, findings, conclusions, or recommendations expressed in the material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.