Marsh terracing as a restoration technique for creating nekton habitat

1
Marsh terracing as a restoration technique for
creating nekton habitat

• USGS Louisiana Cooperative Fish and Wildlife
  Research Unit
• School of Renewable Natural Resources
• Louisiana State University Agricultural Center
• Baton Rouge, LA

2
COASTAL RESTORATION

• To restore and increase vegetated marsh and
  submerged aquatic vegetation habitat in terms of
  both quantity and quality
• To increase fishery habitat in terms of area and
  quality

3
Field of dreams hypothesis if you build it,
they will come
4
MEASURING SUCCESS

• Area of vegetated marsh created
• Establishment of marsh vegetation
• Functional equivalency
• Habitat quality (density, standing stock)
• Habitat suitability (species occurrence)
• Food chain support (diet)
• Fitness (condition or growth)

5
OBJECTIVES

• Determine the effect of marsh terraces on
  adjacent water quality and sediment
  characteristics.
• Compare nekton communities in paired terraced and
  unterraced ponds
• Density (habitat quality)
• Community assemblages (habitat suitability)
• Condition (fitness)

6
PREVIOUS TERRACE RESEARCH

• Rozas and Minello 2001
• Maximize marsh edge
• Bush Thom et al. 2004
• Differences in community composition

7
STUDY SITES

• Three sets of paired terraced and unterraced
  ponds (Rockefeller SWR (Sites 1 2) Sabine NWR
  (Site 3)).
• Sampling at three habitat types
• 1) terraced marsh edge
• 2) unterraced marsh edge
• 3) open water
• Sampled 7 times
• bi-monthly April 2004 April 2005
• 7 sample dates x 3 sites x 2 ponds x 4 sites
  168 samples

8

• Terraced Pond
• Two terraced edge (lt 1m)
• Two open water (gt 50 m edge)

9

• Unterraced Pond
• Two unterraced edge (lt 1m)
• Two open water (gt 50 m)

10
METHODS

• Nekton
• Samples were collected with a 1-m2 throw trap. A
  bar seine is used to clear all nekton from the
  trap.

11
METHODS

• Water Quality Soils
• Depth and water quality data (salinity,
  conductivity, temp., D.O., turbidity) were
  collected along with each nekton sample.
• Percent organic matter
• SAV
• All submerged aquatic vegetation was collected
  from the throw trap.

12
STATISTICAL ANALYSES

• Habitat quality
• Compare water quality, nekton density, biomass,
  richness and diversity between terraced and
  unterraced edge and open water sites (ANOVA)
• Habitat Suitability
• Compare species composition (Chi-square)
• Nekton Condition
• Compare dominant species length-weight
  relationships (ANCOVA)

13
RESULTS ENVIRONMENTAL CHARACTERISTICS

• Turbidity (P 0.23)
• lower in terraced ponds
• SAV (P lt 0.0001)
• higher biomass in terraced ponds
• Soil Organic Matter (P 0.003)
• lower at terraced edge

14
FUNCTIONAL EQUIVALENCY TRAJECTORY
Craft et al. 2003
15
45
40
35
30
25
Nekton Density (individuals/m2)
20
15
10
5
0
Edge
Open water
Edge
Open water
Terraced
Unterraced
Habitat Type
16
45
A
40
35
B
30
25
Nekton Density (individuals/m2)
20
15
10
5
0
Edge
Open water
Edge
Open water
Terraced
Unterraced
Habitat Type
17
A
45
A
40
35
30
25
Nekton Density (individuals/m2)
20
15
10
5
0
Edge
Open water
Edge
Open water
Terraced
Unterraced
Habitat Type
18
CONCLUSIONS

• Habitat characteristics differed between terraced
  and unterraced ponds (SAV, organic matter,
  turbidity).
• Habitat quality, as measured by nekton density
  and diversity, were similar between terraced and
  unterraced edges.

19
NEKTON SPECIES COMPOSITION
Daggerblade grass shrimp
Rainwater killifish
Inland silverside
Sailfin molly
Naked goby
Blue crab
Western mosquitofish
Other
n 1,623
n 1,921
20
Chi-sq P lt 0.0001
21
CONCLUSIONS

• Habitat characteristics differed between terraced
  and unterraced ponds (SAV, organic matter,
  turbidity).
• Habitat quality, as measured by nekton density
  and diversity were similar between terraced and
  unterraced edges.
• Habitat suitability, as measured by species
  abundances and community composition, differed
  significantly between terraced and unterraced
  habitats with greater proportion of benthic
  dependent species at unterraced edge, and
  greater proportions of pelagic species in
  terraced habitats.

22
FISH CONDITION

• Based on length / weight relationship
• Heavier fish better condition

Lucania parva
0.0
log10(W) a' b log10(L)
-0.5
W' aLb
Kn W / W'
log10(weight)
-1.0
-1.5
-2.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
log10(length)
23
Cyprinodon variegatus ANCOVA
terraced
unterraced
24
Condition (length-weight)
Terraced
Unterraced
NSD

• Inland silverside
• Sheepshead minnow
• Clown goby

• Rainwater killifish
• Western mosquitofish
• Naked goby

• Sailfin molly

25
CONCLUSIONS

• Habitat characteristics differed between terraced
  and unterraced ponds (SAV, organic matter,
  turbidity).
• Habitat quality, as measured by nekton density
  was similar between terraced and unterraced
  edges.
• Habitat suitability, as measured by species
  abundances and community composition, differed
  significantly between terraced and unterraced
  habitats.
• Nekton fitness or health, as measured by length
  weight relationships, was lower in terraced as
  compared to unterraced ponds for 3 species, and
  similar between terraced and unterraced ponds
  for 3 different species.

26
FUNCTIONAL EQUIVALENCY TRAJECTORY
(Hobbs and Mooney 1993)
Restoration
Alternative states
Complexity / function
Stays the same
Continued decline
Time
Plant biomass (3-5 yrs) benthic communities
(10-15 yrs) soil properties (30 years) (Craft
et al. 1999, Craft 2003, Broome et al. 1986)
27
FUNCTIONAL EQUIVALENCY

• Terraces do provide nekton habitat, largely
  through the provision of edge habitat.
• However, ecological equivalency is clearly not
  achieved within 4 years of restoration (as
  measured in this project).
• Species occurrence or abundance only provide a
  part of the picture - measures of species health
  and community assemblages need to be considered
  to fully capture the value of restored marshes.

28

• Funding provided by CREST (Coastal Restoration
  and Enhancement through Science and Technology)
• Thanks to Rockefeller SWR and Sabine NWR for
  access to sites, and logistical help.
• Thanks to Chris Cannaday, Jessica OConnell,
  Bryan Piazza, Tim Birdsong, and Seth Bordelon for
  assistance in the field.