We sampled 86 sites within seagrass habitats throughout Back Sound, North Carolina, USA (3442\u2019 N to 3439\u2019 N, 7637\u2019 W to 7631\u2019 W during July 2013. Sampling sites were located across 21 seagrass landscapes which were defined by 200m x 80m rectangles (matching common isolated bed size and shape within our system). These landscapes were previously selected to represent independent gradients in both total seagrass cover (260-11,764 m2) and landscape patchiness (1-75 individual patches; Yeager et al. 2016). Sampling sites in the current study were haphazardly placed across all 21 landscapes, but always located within seagrass itself, and not the unvegetated matrix.\u00a0
\n
Seagrass core sampling and laboratory processing:
\nOne core sample was taken from each sampling site. Each core measured 30 cm in diameter and captured the above ground seagrass habitat as well as the top 10 cm of the sediment surface. All cores were taken within 2 h of low tide and the GPS location of each core was marked with a Garmin 72H handheld unit (Garmin International, Olathe, Kansas, USA). Low-tide depth was measured in situ at each site at the time of sampling to the nearest 10 cm.\u00a0\u00a0
\n
Seagrass tissue from the cores was separated and rinsed with clean freshwater. Seagrass was sorted by species (Zostera marina and Halodule wrightii). All shoots were enumerated to assess species-specific density and the first 20 shoots from each species were measured to assess maximum canopy height (rounded to the nearest mm). Seagrass was then sorted according to above and below-ground biomass; the above-ground biomass of each species was dried at 60 C for 48 h and weighed to the nearest 0.01 g.\u00a0
\n
The core (30-cm diameter, PVC) was gently placed by hand at each site and pushed down to a constant depth of 10 cm into the sediment.\u00a0 The core was gently rotated to break seagrass rhizomes, then dug out by hand, lifted, and placed into a resealable 1.5-gallon plastic bag. The sample was transported back to the laboratory on ice.\u00a0
\n
Shoot heights we averaged across (up to) 20 shoots measured within each core.\u00a0 Mean shoot height per core was calculated using a pivot table in Excel.\u00a0
BCO-DMO Data Manager Processing Notes:
\n* added a conventional header with dataset name, PI name, version date
\n* lat and lon rounded to five decimal places
\n* gram and mm measurement columns rounded to two decimal places.
\n* date format converted to ISO 8601 standard format yyyy-mm-dd
\n* data values that were a period, indicating no value, changed to the default missing data identifier in BCO-DMO, "nd" meaning "no data."
\n
data version 2 (2019-06-14) replaces data version 1:
\n* cells with just a period as a value replaced will no-data values. No-data values in this dataset are displayed as the missing data identifier "nd" for "no data" in the BCO-DMO system.
\n* Date for CP6\t5 changed from 2015-07-16 to 2013-07-16