Data Accompanying "Leveraging successional facilitation to improve restoration of foundational dune grasses along a frequently disturbed coastline"

Experimental Data accompanying "Leveraging successional facilitation to improve restoration of foundational dune grasses along a frequently disturbed coastline"
Authors: Hallie Fischman, Copeland Cromwell, Joe Morton, Ralph Temmink, Tjisse van der Heide, Pete Adams, Christine Angelini

Manuscript Abstract

Coastal ecosystems provide critical storm and flood protection services, but are rapidly degrading worldwide, making their restoration urgent. Here, we evaluated whether successional facilitation, where pioneers facilitate climax species, could be leveraged to accelerate coastal dune revegetation. A survey spanning 270km of Southeast US coastline revealed that Panicum amarum (bitter panicum) supported higher plant richness than Uniola paniculata (sea oats), and that sea oat cover was 230% greater on mature dunes than disturbed dunes, suggesting bitter panicum functions as a pioneer and sea oats is a climax species. A reciprocal transplant experiment confirmed this interpretation: bitter panicum stem production and height fell by 37% and >20 cm, respectively, when planted proximate to sea oats versus in isolation, whereas sea oats produced 38% more and >12 cm taller stems when planted proximate to bitter panicum versus in isolation. A second experiment evaluating the density-dependence of this facilitative interaction revealed that sea oats transplanted into low densities of bitter panicum grew >15% taller than isolated and high-density treatments. However, within seven months, wave inundation eliminated >60% propagules in both experiments. To explore foredune inundation frequency and its implications for dune revegetation, we applied empirical wave runup models at 101 locations throughout Volusia County, Florida. While disturbance frequency varied seasonally and annually, sites with low dune crests and steep beach slopes experienced frequent inundation (>50 events/year). Given the interactions between geomorphology and vegetation success, we present decision matrix to guide managers in determining optimal revegetation methods tailored to project goals and site conditions.

Methods and Data Description

Experiment 1: Uniola paniculata (sea oats) and panicum amarum (bitter panicum) were planted in one of three treatments. Controls are in bare areas, Adjacent treatments (adj) are planted 50cm away from naturally occuring patches of the other species (ie, SO-adj-NPA is sea oats planted adjacent to natural bitter panicum), and Within treatments are planted in a naturally occuring patch of the other species (ie, SO-in-NPA is sea oats planted into a patch of natural bitter panicum). Each replicate consisted of 5 transplant propagules in an X configuration. At each monitoring, all stems were counted and the maximum height of each plot was measured. Plot treatments were reassigned during monitoring, if needed, such that any plot with the heterospecific growing within the 0.25 m2 area was considered a “within plot” and any plot with the heterospecific <50 cm away was considered an “adjacent plot”. See maunscript for additional details.

Experiment 2: Sea oats were planted in the same configuration (5 propagules in an X) surrounded by varrying densities of natural bitter panicum. At each monitoring, we counted all bitter panicum stems within a 1m radius of the transplant, all propogule stems, and measured the height of the tallest transplant in each plot. Treatments were binned based on the number of surrounding bitter panicum stems at the final monitoring point: “Zero”, containing no stems within the 1 m radius (n=20 plots); “Low”, 1-9 stems in the surrounding 1 m radius (n=46 plots); “Mid” 10-34 stems in the 1 m radius (n=29 plots); and “High”, 35-70 stems in the 1m radius (n=25 plots). See manuscript for additional details.

Datasets contains stem counts and heights of planted grasses over time. Stem counts are provided for each of the 5 propagules that comprise a single experimental unit, and a single maximum height is provided per plot. Plots were monitored approximately bimonthly until grasses were eroded by waves (see manuscript). Additional information on treatments are provided in the manuscript and data methods sections.