Understanding Plant Distributions Surrounding Marsh Hammocks Within the Georgia Coastal Ecosystems LTER

1
Understanding Plant Distributions Surrounding
Marsh Hammocks Within the Georgia Coastal
Ecosystems LTER
Christine Hladik1, Alana Lynes2, Chester Jackson3
, Merryl Alber1, Clark Alexander4 , Steve
Pennings2
1Marine Sciences Department, University of
Georgia, 2Department of Biology and Biochemistry,
University of Houston, 3Department of Geography,
University of Georgia, 4 Skidaway Institute of
Oceanography
Results

• Introduction
• Back barrier islands, colloquially called marsh
  hammocks, are upland areas encircled entirely or
  in part by salt marsh. There are approximately
  1,670 marsh hammocks in coastal Georgia, ranging
  in size from less than a hectare to tens of
  hectares. Most are remnants of high ground of
  either Pleistocene or Holocene age, but there are
  also man-made hammocks that have developed from
  dredge spoil or ballast stones. This project
  focused on the characteristics and extent of the
  high marsh plant community adjacent to hammocks,
  and whether/how it varies with hammock size and
  origin. Both the border plants directly adjacent
  to the hammock and the high marsh halo were
  analyzed. (The halo is defined as the area of
  marsh between the upland border and the low marsh
  region (which is dominated by Spartina
  alterniflora).)
• Predictions
• Differences in water delivery will affect the
  plant composition of the adjacent marsh. We
  hypothesize that 1) larger hammocks will have
  more freshwater runoff than smaller ones, and 2)
  hammock origin will affect sediment composition
  and slope, which will in turn affect runoff
  (i.e. with ballast stone islands being coarsest
  and steepest, Holocene islands comprised of sand
  and Pleistocene islands with a higher silt and
  clay content).
• We predict that the border vegetation adjacent to
  hammocks receiving more freshwater will be
  dominated by Juncus roemerianus, whereas those
  with less freshwater input will have border
  vegetation dominated by salt-tolerant plants such
  as Salicornia virginica and Borrichia frutescens.
  
• We predict that halo width will increase with
  hammock size, but will vary with hammock origin
  (which will likely correlate with soil
  composition and elevation). We further predict
  that halo width will vary depending on which
  plant species is dominant

Border Vegetation
Size
Origin

• B. frutescens dominated ballast, Holocene
  Pleistocene.
• S. cynosuroides dominated dredge hammocks.
• Holocene and Pleistocene had a greater number of
  dominant species.
• J. roemerianus only dominated Holocene and
  Pleistocene.

• B. frutescens dominated smaller hammocks (1-3
  ha).
• J. roemerianus dominated intermediate sized
  hammocks (3-10 ha).
• S. patens dominated on larger hammocks (gt10 ha).
• Smaller hammocks had a greater number of dominant
  species.

Overall, Borrichia and Juncus are the most
frequently dominant plants.
Marsh Halo
Size Class
Species Halo Widths
Halo-Hammock Area
The smallest halos were associated with S. patens
(SP) and B. frutescens (BF) (Group 1). These
were all significantly smaller than the halos
associated with the four plants with the largest
halos (Group 2) marsh meadow (MM), S. virginica
(SV), J. roemerianus (JR), and B. maritima (BM).
Additionally, J. roemerianus (JR) was
significantly larger than D. spicata (DIST).
When binned by size class, Wilcoxon nonparametric
t-tests showed that the halo widths for class i
hammocks were significantly smaller than those
for classes iii, v, and vii.
There was a significant (plt 0.05) linear
relationship between halo area and upland area
(R2 0.97).

• Methods
• In the summer of 2007 the GCE LTER surveyed 55
  hammocks of different origin and size.

Origin
Halo Width-Hammock Area
Data for size class I only
Border-Halo Comparison

• Hammock Border Vegetation
• At 6 sampling locations around each hammock, the
  high marsh plant community was surveyed within a
  1 m2 quadrat placed 2 m from the marsh/upland
  border. Plant species present and their percent
  cover were determined in each quadrat.
• Species relative dominance and presence was
  calculated.
• MarshHalo Characteristics
• Both the edge of the high marsh halo and the
  hammock upland line were mapped using a pair of
  Trimble GPS units.
• At 18 set points around the hammock, the
  vegetation in the halo was recorded.
  Additionally, we took detailed field notes on any
  shift in vegetation and recorded the GPS location
  where the shift occurred. Plant species used in
  the present analysis were Juncus roemerianus
  (JR), Spartina alterniflora (SA), Spartina patens
  (SP), Salicornia virginica (SV) Distichlis
  spicata (DIST) Batis maritima (BM) Borrichia
  frutescens (BF), Spartina cynosuroides (SC) and
  marsh meadow (a mix of SV, BM, and/or DIST).
• All Trimble data were processed using the GPS
  Pathfinder Office software program. Data were
  differentially corrected and exported as an
  ArcGIS shapefile format.
• The area of each marsh halo was calculated using
  the Hawths Tools ArcGIS extension and polygon
  subtraction.
• The AMBUR ArcGIS extension (C. Jackson 2008) was
  used to calculate halo width. Using the nearest
  transect method, transects were cast from the
  edge of the halo to the hammock upland at 5 m
  intervals. The mean transect length was
  calculated for each hammock and for each dominant
  plant species.

Ballast had an average width of 7 3.4 m, which
was significantly smaller than that of Holocene
hammocks (17 7.6 m). There were no
differences in the halo widths of Holocene and
Pleistocene hammocks (15 10.9 m).
There was a weak positive linear relationship
between halo width and upland area (R2 0.41).
Overall average halo width for all size classes
was 23.8 20.4 m (s.d.).
Conclusions
Ongoing Work Whole Marsh Scale
As predicted hammock size and origin affected
the border plant community composition. Larger
hammocks were dominated by J. roemerianus and S.
patens and smaller hammocks dominated by B.
frutescens. Overall, B. frutescens dominanted
ballast, Holocene and Pleistocene and J.
roemerianus was only observed to dominate on
Pleistocene and Holocene hammocks. There were
significant differences in both the mean halo
width and the dominant plant species in relation
to hammock size and origin. Larger hammocks had
larger halos, however we have limited
observations to evaluate the robustness of this
relationship. Holocene and Pleistocene hammocks
had similar average widths and Holocene halos
were significantly wider than ballast stone
halos. Halo width varied when different plants
dominated. The halo widths of S. patens and B.
frutescens were significantly smaller than those
for marsh meadow, S. virginica, J. roemerianus,
and B. maritima. Finally, dominant plants in
the halos were not always consistent with the
plants observed on the upland border of the
hammocks.
LIght Detection And Ranging (LIDAR) and the Role
of Elevation While many variables affect species
patterns, elevation is one of the most important
as it determines the frequency and duration of
tidal flooding. LIDAR has the capability to
collect elevation information for habitat mapping
at the whole marsh scale. LIDAR data were
acquired in March 2009 to enhance habitat
classifications.
Future LIDAR Work

• Accuracy assessment of LIDAR and DEM
• Map the salt marsh habitat
• Model relationship between marsh species
  distributions and controlling factors (elevation,
  proximity to creeks, proximity to uplands)
• Make predictions on how vegetation would shift in
  response to sea level rise and other potential
  changes.

http//gce-lter.marsci.uga.edu/public/file_pickup/
LiDAR/
Horizontal Accuracy 10-20 cm Vertical
Accuracy 5-10 cm 1 m DEM