Data for: Development of a learning community focused on sea-level rise and coastal habitat change

Rapid habitat changes are occuring in salt marshes located in the Northeastern United States, including expansion of ponded areas on the marsh platform, die off of coastal forests, and subsequent colonization of 'ghost forests' by marsh vegetation. This work focuses on two main areas: (1) environmental conditions along the marsh forest border undergoing rapid transitions; and (2) environmental conditions and plant stress on marsh platforms with extensive and expanding ponding. 

To better understand drivers of environmental change in marsh-forest borders undergoing rapid transitions, we measured shallow groundwater levels,  soil salinity, and forest health and structure along the salt marsh-upland border at three sites (in NJ, NY, MA) with varying slopes using installation of shallow groundwater wells, drone imagery and associated image processing, and geophysical methods. To better understand drivers of environmental change on the marsh platform, we again focused on three sites (in NJ, NY, MA) where we used piezometers to understand vertical gradients in marsh groundwater levels, and measured photosynthesis and plant biomass and used drone imagery to map plant stress indices. While we anticipate that this data will be published in journal articles of the next 2 years, we archive collected data to facilitate data sharing, as required by NSF.

Methods
We collected biological and environmental data to provide perspective on rapid habitat change occurring at the salt marsh-forest ecotone and on salt marsh platforms converting to open water in the Northeastern United States, and to test hypotheses about drivers of habitat change.

Marsh-Forest Ecotone

In studying the marsh forest border, we chose to work at three locations where the marsh-forest ecotone is undergoing rapid habitat transitions: the Tuckahoe Wildlife Management Area, Egg Harbor Township, NJ (39° 19.506'N;  74° 39.014'W), Pine Neck Preserve, Quogue, NY (40° 50.440'N;  72° 33.785'W), and around Sage Lot Pond, Waquoit Bay National Estuarine Research Reserve, Mashpee, MA (41° 33.438'N; 70° 30.227'W). We hypothesized that we would see steeper environmental gradients in New England than along the coastal plain. In addition, we hoped that these data sets would provide perspective on salt water intrusion vs. root zone flooding as important drivers of maritime forest dieback.

Drone Imagery - UAV surveys were conducted at the three focus sites during leaf-on and leaf-off conditions using a DJI Phantom 4 Pro (DJI, Shenzhen, CHN) flown at X m altitude at 3.5 m s−1 to record images with 80% overlap and a resolution of 3 cm pixel−1. At the Tuckahoe Wildlife Management Area, these occurred on 12 September 2022 and  30 January 2023.  At Pine Neck Preserve, these occurred on 3 April 2021 and  10 August 2022. At  Sage Lot Pond, these occurred on   27 March 2021 and 11 August 2021. Orthomosaics (5 cm px−1) were produced from UAV images following a standard photogrammetry workflow in Metashape (Agisoft, St. Petersburg, Russia).

Salinity - We used a Geonics Model EM38-MK2 Conductivity Meter (Geonics Ltd, Mississauga, Ontario, Canada) in horizontal mode, held 0.75 m over the marsh surface to record soil conductivity readings.  The readings were the result of an induced current generated by the instrument, and measures were made using the 1m setting to record measures 0.75 m, ie., at the soil surface (https://geonics.com/pdfs/downloads/catalogue.pdf). At each location occupied, the apparent electrical conductivity readings (ECa values) were recorded, in milliSeimans meter−1 (mS m−1), and the location of the measured point was recorded using a Garmin 64sx handheld GPS (accuracy of ca. 3m). We occupied 175-200 points at each of the three sites, with grid spacings on the order 10-20m. Measures were made at Waquoit Bay, MA on 7 June 2021, at Pine Neck Preserve on 19 May 2021, and at Tuckahoe Wildlife Mangement area on 21 July 2021. At each site, we also collected ca. 15-20 soil samples  for salinity calibration, to convert apparent electrical conductivity readings (ECa values) to soil salinity. Surface soil salinity was measured  on 5:1 slurries made of 5 parts water to one part dry sediment using a YSI pro30 conductivity and salinity meter. Soil salinity was not back corrected to the original water contact, as the average water:dry soil ratio was 5.13. ECa data were converted to soil salinity values using empirical calibration curves constructed from a least-squares regression of ECa values against soil salinities, forced through the origin. Regressions were y=0.0915x for Waquoit Bay (r2=0.78), y=0.1765x for Pine Neck Preserve (r2=0.90), and y=0.08x for Tuckahoe (r2=0.80),  where y is soil salinity, and x is ECa. Following field sampling and laboratory analyses, shapefiles were created in ArcGIS version 10.3 (ESRI, Redlands, CA) by creating a text file which included latitude, longitude, and estimated soil salinity. We used the kriging function in ArcGIS Spatial Analyst using a 1-m pixel size using default values of ordinary kriging using a spherical variogram model and a search radius setting of 12 to create marsh-specific salinity contour maps.

Water levels - At each of three sites, we installed 10 wells (screened at 0.5 - 1.0 meters of depth) across the marsh-forest ecotone for a total of 30 wells. The wells at each site were arranged in two transects, with the well at the lower-most elevation sunk into the marsh, the well at the highest elevation in intact forest, and the other three wells tranversing the marsh-forest ecotone. The wells were constructed of SDR 21 PVC pipe of a nominal diameter of 3.175 cm (1.25"), within inner diameter of 3.82 cm  and an outer diameter of 4.22 cm.  To allow for water exchange, holes were drilled in PVC, PVC chips were removed with a file, and the outside of the piezometer was covered in well sock. An soil auger was used to bore a hole for the piezometer, which was subsequently filled with coarse sand following the insertion of the piezometer. Above the screened depth, the augered hole was backfilled with marsh peat, and near the surface  cement were used to seal the piezometer and to prevent the area around the well from acting as a macropore or or preferential flow pathway.

Marsh Platform

In studying the marsh platform, we chose to work at three locations where the marsh platform has large and expanded ponded areas and is undergoing rapid habitat transitions: the Tuckahoe Wildlife Management Area, Corbin City, NJ (39° 17.440'N; 74° 39.378'W), Bass Creek, Mashomack Preserve, Shelter Island, NY (41° 2.925'N;  72° 17.245'W), and  Sage Lot Pond, Waquoit Bay National Estuarine Research Reserve, Mashpee, MA (41° 33.222'N; 70° 30.692'W). We hypothesized that locations that were similar in elevation but that had different groundwater hydrologic regimes would display contrasts in vegetation stress and biomass. To address these hypotheses, we established well nests at four locations in three marshes (a total number of 12 well nests and 24 individual wells), where we monitored vertical hydrologic gradients for one year, measured photosynthesis and greenhouse gas exchange during two summer growing seasons, collected aboveground biomass, and collected multispectral drone imagery during peak biomass (e.g., July or August).

Water levels - At each of 12 locations, we installed two piezometers to better understand vertical water movements in healthier vs. ponded areas of the salt marsh platform. The piezometers were constructed of SDR 21 PVC pipe of a nominal diameter of 3.175 cm (1.25"), within inner diameter of 3.82 cm  and an outer diameter of 4.22 cm. The two piezometers were screened at depths of 0.25-0.50 cm and 1.0-1.25m below the soil surface. To allow for water exchange, holes were drilled in PVC, PVC chips were removed with a file, and the outside of the piezometer was covered in well sock. An soil auger was used to bore a hole for the piezometer, which was subsequently filled with coarse sand following the insertion of the piezometer. Above the screened depth, the augered hole was backfilled with marsh peat, and near the surface bentonite and cement were used to seal the piezometer and to prevent the area around the well from acting as a macropore or or preferential flow pathway.

Drone Imagery - UAV surveys were conducted at the three focus sites during leaf-on conditions using a DJI Phantom 4 Pro (DJI, Shenzhen, CHN) flown at X m altitude at 3.5 m s−1 to record images with 80% overlap and a resolution of 3 cm pixel−1. At the Tuckahoe Wildlife Management Area, these occurred on 2 September 2022.  At Mashomack Preserve, this occurred on 10 August 2022. At Waquoit Bay, this occurred on 21 July 2022. Orthomosaics (2 cm px−1) were produced from UAV images following a standard photogrammetry workflow in Metashape (Agisoft, St. Petersburg, Russia).

Greenhouse gas exchange - Photosynthesis, community respiration, net ecosystem exchange, and methane emissions were measured once in the summers for 2021 and 2022 to provide perspectives on plant stress. An LGR ultraportable greenhouse gas analyzer (ABB, San José, CA) was used to measure the greenhouse gas concentration of gases circulated through the headspace of a 20L chamber attached to 30-cm diameter)(12") PVC collars installed in the marsh soil. We conducted twelve at each marsh in each year, three  in the proximity (<3m) each well nest. Measurements of net ecosystem exchange were collected during three minute incubations in a transparent chamber, and respiration was determined by similar incubations with the chamber covered with black-out material.  Photosynthesis was calculated as the sum of respiration and net ecosystem. The Ideal Gas Law (PV = nRT) was used to convert linear changes in CO2 and CH4 concentrations within the chamber during each incubation period to fluxes standardized to the surface area of the plant pots.  Measures were made at Tuckahoe Wildlife Management Area on 17 August 2021 and 8 July 2022, at Bass Creek on 13 August 2021 and 29 June 2022, and at Sage Lot Pond on 11 August 2021 and 20 July 2022.

Leaf-based photosynthesis measures - At the same locations were greenhouse gas exchange measures were conducted, in 2021, we used a CI-340 handheld photosynthesis system (CID Bioscience, Camas, WA, USA) to measure leaf-based photosynthetic uptake and ambient fluorescence using the CI-510CF chlorophyll fluorescence module. Measures were recorded for three minute incubations. Measures were made at Tuckahoe Wildlife Management Area on 17 August 2021, at Bass Creek on 13 August 2021, and at Sage Lot Pond on 11 August 2021.

Biomass - At each location where greenhouse gas exchange measures were made, all aboveground biomass in the footprint of the (30-cm diameter)(12") PVC collars was harvested. After returning from the field, biomass material was rinsed over a 2-mm sieve to remove sediment and salt, placed in a pre-dried brown paper bag and dried in a drying oven at 60°C to constant weight.