Chapter 21 Seagrasses, Fish, and Fisheries
511
seagrass shows an asymptotic response. For example, the abundance of fish increased greatly between
0 and 25 seagrass blades per m
2 , but abundances of
fish were similar when density of seagrass was between 25 and 400 blades per m
2 (Worthington et al.,
1991).
Although abundances of fish and invertebrates
within single beds of seagrass (e.g. plots of 25 m
2 )
showed noticeable effects when leaf height and density were manipulated, correlations between abundances of organisms and complexity of seagrass at
the larger spatial scale were rare (Bell and Westoby, 1986c). Environmental features (e.g. distance
from ocean, grain size of sediment, area of seagrass
bed) also failed to account for differences in abundances and Bell and Westoby (1986c) concluded that
stochastic settlement from the plankton may cause
variation in density of organisms. A model of hierarchical habitat selection, in which juveniles select
the first seagrass bed they encounter and then after
settlement select microhabitats within the seagrass
bed that provide food and a high level of protection from predation, was proposed to explain the
distribution of fish in seagrass beds (Bell and Westoby, 1986a; Bell et al., 1987). Settlers of blue groper
(Achoerodus viridis, Labridae) did not discriminate
between artificial seagrass units (ASUs) with dense
and sparse covers of leaves, which supports the above
model (Bell et al., 1987). This ‘settle and stay’ hypothesis was not supported by other studies in which
the turnover rate appeared to be high (Jenkins and
Sutherland, 1997). Although the study of Jenkins
and Sutherland (1997) did not distinguish between
movement and mortality, the appearance of postsettlement juveniles within 24 h of clearances suggests rapid migration to seagrass. Thus, correlations
between physical complexity of seagrass and abundances of fish and invertebrates would be unlikely at
broad spatial scales, but if individuals move around
within seagrass beds to select habitats (e.g. density
and height of seagrass) that favor survival, then correlations would be common at the small spatial scale
(Bell and Westoby, 1986c).
2. Among Patches
Although abundances of fish and invertebrates
are often correlated with seagrass structure (e.g.
biomass, shoot density, shoot length) within seagrass
beds, these correlations tend to break down at larger
spatial scales. At the individual patch level (i.e. contiguous area of seagrass habitat that differs in appearance from surrounding habitat), which generally
covers scales of meters to tens of meters, the size,
shape, and orientation of patches, as well as percent
cover of seagrass has been shown to influence not
only the abundance and species composition of organisms within patches, but also rates of predation
on patch inhabitants, and growth (Table 1). Most
of this research has focused on the effect of habitat fragmentation (e.g. edge vs. center of habitats;
patch size) and on responses by invertebrates. The
individual habitat patch also may be considered in a
landscape mosaic (see ‘Landscape scale variables’,
Section IV.B, and Bell et al., Chapter 26).
Habitat destruction resulting in habitat loss per se
and habitat fragmentation (which creates a greater
number of habitat patches that are smaller in size
than the original contiguous tract of habitat, Bender
et al., 1998) may influence abundance of species.
Although the effect of habitat loss is obvious in
that when habitat is removed, species within them
are subsequently displaced and may also be lost,
the effects of habitat fragmentation are less obvious
(Bender et al., 1998). Habitat fragmentation may differentially affect the abundance of species depending on whether individual species preferentially utilize interior sections of habitat, areas along edges
of patches, or show no preference. One would expect that as habitats are fragmented, interior species
would decrease in abundance, edge species would
increase in abundance, and generalist species that
show no preference for either the edge or the interior
of habitats would show no difference in abundance
(Bender et al., 1998). Bell et al. (2001) found little evidence identifying any taxonomic group to be
edge sensitive, although infaunal bivalves appeared
to act as interior species in the majority of studies
surveyed. Thus, infaunal bivalves would be expected
to show a disproportionate decrease in abundance as
patch size decreased due to the increase in edge habitat relative to interior habitat (Bell et al., 2001).
Higher densities of bivalves (including bay scallops) have been found at edges of seagrass beds than
in the interior or in nearby non-vegetated sediments
(Bologna and Heck, 1999, 2000). Since predation
rates were increased along edges of Thalassia testudinum beds relative to the other habitats, and
growth rates also were significantly higher, individual scallops appeared to trade off an increased predation rate for increased growth rates (Bologna and
Heck, 1999). No significant difference in predation
511
seagrass shows an asymptotic response. For example, the abundance of fish increased greatly between
0 and 25 seagrass blades per m
2 , but abundances of
fish were similar when density of seagrass was between 25 and 400 blades per m
2 (Worthington et al.,
1991).
Although abundances of fish and invertebrates
within single beds of seagrass (e.g. plots of 25 m
2 )
showed noticeable effects when leaf height and density were manipulated, correlations between abundances of organisms and complexity of seagrass at
the larger spatial scale were rare (Bell and Westoby, 1986c). Environmental features (e.g. distance
from ocean, grain size of sediment, area of seagrass
bed) also failed to account for differences in abundances and Bell and Westoby (1986c) concluded that
stochastic settlement from the plankton may cause
variation in density of organisms. A model of hierarchical habitat selection, in which juveniles select
the first seagrass bed they encounter and then after
settlement select microhabitats within the seagrass
bed that provide food and a high level of protection from predation, was proposed to explain the
distribution of fish in seagrass beds (Bell and Westoby, 1986a; Bell et al., 1987). Settlers of blue groper
(Achoerodus viridis, Labridae) did not discriminate
between artificial seagrass units (ASUs) with dense
and sparse covers of leaves, which supports the above
model (Bell et al., 1987). This ‘settle and stay’ hypothesis was not supported by other studies in which
the turnover rate appeared to be high (Jenkins and
Sutherland, 1997). Although the study of Jenkins
and Sutherland (1997) did not distinguish between
movement and mortality, the appearance of postsettlement juveniles within 24 h of clearances suggests rapid migration to seagrass. Thus, correlations
between physical complexity of seagrass and abundances of fish and invertebrates would be unlikely at
broad spatial scales, but if individuals move around
within seagrass beds to select habitats (e.g. density
and height of seagrass) that favor survival, then correlations would be common at the small spatial scale
(Bell and Westoby, 1986c).
2. Among Patches
Although abundances of fish and invertebrates
are often correlated with seagrass structure (e.g.
biomass, shoot density, shoot length) within seagrass
beds, these correlations tend to break down at larger
spatial scales. At the individual patch level (i.e. contiguous area of seagrass habitat that differs in appearance from surrounding habitat), which generally
covers scales of meters to tens of meters, the size,
shape, and orientation of patches, as well as percent
cover of seagrass has been shown to influence not
only the abundance and species composition of organisms within patches, but also rates of predation
on patch inhabitants, and growth (Table 1). Most
of this research has focused on the effect of habitat fragmentation (e.g. edge vs. center of habitats;
patch size) and on responses by invertebrates. The
individual habitat patch also may be considered in a
landscape mosaic (see ‘Landscape scale variables’,
Section IV.B, and Bell et al., Chapter 26).
Habitat destruction resulting in habitat loss per se
and habitat fragmentation (which creates a greater
number of habitat patches that are smaller in size
than the original contiguous tract of habitat, Bender
et al., 1998) may influence abundance of species.
Although the effect of habitat loss is obvious in
that when habitat is removed, species within them
are subsequently displaced and may also be lost,
the effects of habitat fragmentation are less obvious
(Bender et al., 1998). Habitat fragmentation may differentially affect the abundance of species depending on whether individual species preferentially utilize interior sections of habitat, areas along edges
of patches, or show no preference. One would expect that as habitats are fragmented, interior species
would decrease in abundance, edge species would
increase in abundance, and generalist species that
show no preference for either the edge or the interior
of habitats would show no difference in abundance
(Bender et al., 1998). Bell et al. (2001) found little evidence identifying any taxonomic group to be
edge sensitive, although infaunal bivalves appeared
to act as interior species in the majority of studies
surveyed. Thus, infaunal bivalves would be expected
to show a disproportionate decrease in abundance as
patch size decreased due to the increase in edge habitat relative to interior habitat (Bell et al., 2001).
Higher densities of bivalves (including bay scallops) have been found at edges of seagrass beds than
in the interior or in nearby non-vegetated sediments
(Bologna and Heck, 1999, 2000). Since predation
rates were increased along edges of Thalassia testudinum beds relative to the other habitats, and
growth rates also were significantly higher, individual scallops appeared to trade off an increased predation rate for increased growth rates (Bologna and
Heck, 1999). No significant difference in predation