All raw values are based on the seagrass collected in a core with a area of 126cm².

Methods from Layman et al (2016) Ecol. Engr.:

This case study was based in the Bight of Old Robinson, Abaco, The Bahamas, a semi-enclosed bay that has a complex benthic mosaic comprised predominantly of sand, seagrass (primarily turtle grass Thalassia testudinum), and hard bottom/patch reef habitat. We sampled an artificial reef (N 26 20.549', W77 00.874') that had similar spatial patterns in aboveground seagrass traits as other patch reefs (both natural and artificial) in seagrass beds of this area (Allgeier et al., 2013; Layman et al., 2013). This reef (dimensions ~1.2 m2 at base and ~1.2 m tall) was constructed in March 2009 using 40 cinder blocks arranged in pyramid fashion (Yeager et al., 2011). Samples (December 2013) were taken in spatially-explicit fashion on 3 transects radiating from each reef; transects were oriented ~120 degrees apart in random directions. Cores were taken with a 12.7 cm diameter pvc pipe at set distances (m) from the reef on each transect: 1, 2, 3, 4, 6, 10, 15, 100. The core was driven ~16 cm into the sediment and manually excavated, placing one hand under the bottom of the core as it was pulled out. Visual inspection demonstrated cores successfully remove all seagrass tissues from the sampled area. Water was drained from the cores and sand was rinsed off with seawater and then they were placed in individual plastic bags and immediately frozen.

In the laboratory, cores of seagrass biomass were thawed and separated into aboveground biomass (all attached green leaves of shoots) and belowground biomass (rhizomes and roots). Shoots were enumerated and morphology of blades (length and width) was measured. As a proxy for grazing intensity, we measured the total number of grazing scars on all blades in the core (Valentine and Duffy, 2006). Blades were gently scraped with a razor blade to remove epiphytes; belowground material was rinsed with deionized water. Tissues were dried for 72 h at a constant temperature of 65C and dry weight (to the nearest 0.01 g) was recorded. Dried samples were ground into a fine powder using a PRECELLYS-24 grinder and subsamples of each were analyzed for percent of carbon (C), nitrogen (N), and phosphorous (P). Percent C and N content were determined using a CHN Carlo-Erba elemental analyzer (Fison NA1500). Percent P was determined by dry oxidation acid hydrolysis extraction followed by colorimetric analysis (Fourqurean and Zieman, 1992).

To explore potential thresholds in allocation of resources between different tissue structures, we first partitioned total plant nutrients into the various tissues from which these measurements were obtained (n = 48 for each tissue type, n = 144 total). Partitions included: % blade nutrients to total (whole-plant) nutrients, % total belowground (root + rhizome) to total, % root to total, % rhizome to total, % roots to total belowground, and % rhizome to total belowground. Generalized additive models (gam) were then used to describe the relationship of these data with respect to distance from the reef. If the relationship was non-linear, a changepoint analysis, using the package "changepoint" in R, was used to determine the distance from the reef at which the threshold was reached. Relationships between belowground and aboveground traits were tested using least squares regression. Data were log transformed and satisfied model assumptions.