732
south and north ends of the SE cell. Thus, the original openings have become essentially irrelevant in terms of the hydrodynamics in the restoration cells.
21.4.4 Did the Biological Communities Establish as Expected?
A number of biological communities were regularly sampled. For plants and invertebrates on the emergent marsh (area exposed to air at low tide), three permanent
transects were placed in each restoration cell and reference salt marsh perpendicular
to the old dike. These were sampled in a random stratified manner by randomly
placing three quadrats in each of five equal sections per transect (15 quadrats per
transect, 45 quadrats per site; Fig. 21.18a). All plant stems and invertebrates were
identified and counted in each quadrat. This sampling was conducted monthly from
June to August in 2010–2015 to quantify community structure 1 year pre-breach
and 5 years post-breach. These data are presented as averages over months for each
transect for a given site and year. Animals in the water column of the salt pools were
sampled using minnow traps and invertebrate activity traps (Fig. 21.18b, c), which
were deployed in the evening, left over a high tide period and recovered the following morning. This was complemented by a lift net sample (Fig. 21.18d) also done in
the morning. Animals were identified, counted and returned to their original location. This sampling was done monthly in the summer, but we present only June of
each year post-breach (2011–2015) for the large salt pool (>1000 m
2
in area) located
in each site. The large pools were sampled at five locations. The data from each
capture method was combined, and transformed to presence-absence for analysis.
The benthos of the salt pools was sampled using a benthic grab (81.7 cm
2
× 5 cm
deep) once in the summer (between mid-June and mid-July). The sediment grabs
(five per large salt pool) were sieved using a 250-μm mesh, and the contents were
preserved in 95% ethanol until they could be processed for epi- and infauna under a
dissecting microscope (i.e. identified and counted). The composition of each type of
community (on the emergent marsh and in salt pools) was analysed using nonparametric multivariate analysis of variance (not presented here) and visualized
using non-parametric multidimensional scaling (nMDS) (Clarke 1993).
The plant community is the habitat-forming community of salt marshes (Redfield
1972; Burdick et al. 1997; Bertness 2007) and so is discussed first. The plant recovery
pattern in our restorations cells showed a smooth temporal trajectory since the breaching (Fig. 21.19), and is nicely represented by the nMDS graph (Fig. 21.20a). The
reference marshes have a distinctive community composition fairly typical of a
mature salt marsh in this region (Redfield 1972; Chmura et al. 1997; Bertness 2007).
Specifically, there is a large high marsh zone dominated by salt marsh hay Spartina
patens, and a narrow low marsh zone of saltwater cordgrass S. alterniflora. The plant
community in the reference marshes is represented by a distinct grouping of data
points on the nMDS graph, and remained similar from one year to the next. The prebreach restoration cells also showed a distinct community reflecting mostly terrestrial
plants (e.g., Timothy grass Phleum pretense, Aster sp., wild morning glory Calystegia
L.K. Boone et al.
south and north ends of the SE cell. Thus, the original openings have become essentially irrelevant in terms of the hydrodynamics in the restoration cells.
21.4.4 Did the Biological Communities Establish as Expected?
A number of biological communities were regularly sampled. For plants and invertebrates on the emergent marsh (area exposed to air at low tide), three permanent
transects were placed in each restoration cell and reference salt marsh perpendicular
to the old dike. These were sampled in a random stratified manner by randomly
placing three quadrats in each of five equal sections per transect (15 quadrats per
transect, 45 quadrats per site; Fig. 21.18a). All plant stems and invertebrates were
identified and counted in each quadrat. This sampling was conducted monthly from
June to August in 2010–2015 to quantify community structure 1 year pre-breach
and 5 years post-breach. These data are presented as averages over months for each
transect for a given site and year. Animals in the water column of the salt pools were
sampled using minnow traps and invertebrate activity traps (Fig. 21.18b, c), which
were deployed in the evening, left over a high tide period and recovered the following morning. This was complemented by a lift net sample (Fig. 21.18d) also done in
the morning. Animals were identified, counted and returned to their original location. This sampling was done monthly in the summer, but we present only June of
each year post-breach (2011–2015) for the large salt pool (>1000 m
2
in area) located
in each site. The large pools were sampled at five locations. The data from each
capture method was combined, and transformed to presence-absence for analysis.
The benthos of the salt pools was sampled using a benthic grab (81.7 cm
2
× 5 cm
deep) once in the summer (between mid-June and mid-July). The sediment grabs
(five per large salt pool) were sieved using a 250-μm mesh, and the contents were
preserved in 95% ethanol until they could be processed for epi- and infauna under a
dissecting microscope (i.e. identified and counted). The composition of each type of
community (on the emergent marsh and in salt pools) was analysed using nonparametric multivariate analysis of variance (not presented here) and visualized
using non-parametric multidimensional scaling (nMDS) (Clarke 1993).
The plant community is the habitat-forming community of salt marshes (Redfield
1972; Burdick et al. 1997; Bertness 2007) and so is discussed first. The plant recovery
pattern in our restorations cells showed a smooth temporal trajectory since the breaching (Fig. 21.19), and is nicely represented by the nMDS graph (Fig. 21.20a). The
reference marshes have a distinctive community composition fairly typical of a
mature salt marsh in this region (Redfield 1972; Chmura et al. 1997; Bertness 2007).
Specifically, there is a large high marsh zone dominated by salt marsh hay Spartina
patens, and a narrow low marsh zone of saltwater cordgrass S. alterniflora. The plant
community in the reference marshes is represented by a distinct grouping of data
points on the nMDS graph, and remained similar from one year to the next. The prebreach restoration cells also showed a distinct community reflecting mostly terrestrial
plants (e.g., Timothy grass Phleum pretense, Aster sp., wild morning glory Calystegia
L.K. Boone et al.