Chapter 21 Seagrasses, Fish, and Fisheries
509
that compared densities of organisms, they found
that seagrass habitats had higher animal densities than various regions of a marsh (e.g. vegetated marsh edge, non-vegetated marsh, vegetated
inner marsh) (Minello et al., 2003 and references
therein). Organisms may, however, initially settle in
seagrass habitat and then move to marsh habitats
or alternatively, marsh habitats may have increased
rates of predation (e.g. Orth and van Montfrans,
1987).
Tidal inundation patterns may be important in determining habitat use (of both marsh and mangroves)
because shoreline habitats that are not flooded cannot be used by most organisms. Rozas and Minello
(1998) compared seagrass and marsh habitats where
flooding durations for each habitat type were similar.
They found that total number of fishes was not significantly different in salt marsh and seagrass. Many
Australian marshes may only be flooded during high
tides and for relatively short periods of time. Densities of fish in salt marsh in Australia are less than
1 fish per m
2 , whereas in seagrass fish are at densities greater than 1 m
2 (Connolly, 1994a; Connolly
et al., 1997; Thomas and Connolly, 2001). Thus, utilization of marsh habitats in Australia is significantly
less than seagrass habitats. In addition, differences in
submergence time between seagrass and salt marsh
habitats may also occur within the USA (Rozas and
Minello, 1998).
Survival of animals in seagrass does not appear
to be significantly different from that in salt marsh
(reviews: Heck et al., 2003; Minello et al., 2003)
(Fig. 4). For example, survival time for tethered
brown shrimp (Farfantepenaeus aztecus, Penaeidae)
was similar between seagrass and marsh habitats and
significantly greater than over non-vegetated sand
bottom (Minello, 1993).
Very few studies have compared growth of fish
or invertebrates in seagrass and marsh habitats (but
see Irlandi and Crawford, 1997; Phelan et al., 2000;
Stunz et al., 2002) and most of these studies differ in the species examined and methodology utilized. Minello et al. (2003) combined all studies in a
meta-analysis and found that growth in seagrass was
significantly greater than growth in non-vegetated
marsh (sensu Minello et al., 2003). However, their
meta-analyses also found that growth in vegetated
marsh was significantly higher than in non-vegetated
marsh. Stunz et al. (2002) using solid-walled field
mesocosms found no significant difference in growth
of Sciaenops ocellatus (Sciaenidae) between the
edge of vegetated marsh and seagrass. Irlandi and
Crawford (1997) used field enclosures containing
vegetated marsh edge with seagrass and vegetated
marsh edge without seagrass to evaluate growth
of Lagodon rhomboides (Sparidae). They found
that growth was negative (measured as biomass) in
vegetated marsh without seagrass and positive in
vegetated marsh with seagrass. Thus, seagrass appears to have an effect on growth of this species
of fish.
IV. Spatial Scales—Factors Influencing
Abundance of Fish in Seagrass Beds
Seagrass habitats represent a series of hierarchical
scales ranging from single blades to individual short
shoots to groups of short shoots composing a seagrass bed to the seagrass bed within a matrix of other
habitat types (e.g. non-vegetated habitats) to submerged vegetation within the large coastal system
(Robbins and Bell, 1994). The scale at which observations are made or experiments are conducted may
influence results and conclusions (Turner, 1989),
such that patterns and processes that are found to
be important at one scale may not be important at
another scale (Wiens, 1989).
A large amount of research has focused at the
small scale (e.g. seagrass blade length and density)
and shown that this scale is important in influencing recruitment dynamics (see below and Heck and
Orth, Chapter 22). Artificial seagrass has been used
in many of the small-scale experiments addressing a
variety of questions, as it provides a standard habitat
unit where morphological aspects of the plants can
either be held constant across a number of locations
or manipulated such that spatial variation is not a
confounding influence. Research at this small scale
has been reviewed elsewhere (e.g. Bell and Pollard,
1989) and is only briefly mentioned here. The focus
on larger-scale spatial patterns and linking seagrass
patch size, shape, or complexity of patch perimeter,
position of a patch in the seagrass bed or its position
in relation to adjacent patches and other habitats,
to the structure and function of assemblages has
only recently been investigated. At this larger spatial
scale, a range of scales (extent, sensu Wiens, 1989)
and sampling units that vary in scale (grain, sensu
Wiens, 1989) have been used making comparisons
among studies difficult.
509
that compared densities of organisms, they found
that seagrass habitats had higher animal densities than various regions of a marsh (e.g. vegetated marsh edge, non-vegetated marsh, vegetated
inner marsh) (Minello et al., 2003 and references
therein). Organisms may, however, initially settle in
seagrass habitat and then move to marsh habitats
or alternatively, marsh habitats may have increased
rates of predation (e.g. Orth and van Montfrans,
1987).
Tidal inundation patterns may be important in determining habitat use (of both marsh and mangroves)
because shoreline habitats that are not flooded cannot be used by most organisms. Rozas and Minello
(1998) compared seagrass and marsh habitats where
flooding durations for each habitat type were similar.
They found that total number of fishes was not significantly different in salt marsh and seagrass. Many
Australian marshes may only be flooded during high
tides and for relatively short periods of time. Densities of fish in salt marsh in Australia are less than
1 fish per m
2 , whereas in seagrass fish are at densities greater than 1 m
2 (Connolly, 1994a; Connolly
et al., 1997; Thomas and Connolly, 2001). Thus, utilization of marsh habitats in Australia is significantly
less than seagrass habitats. In addition, differences in
submergence time between seagrass and salt marsh
habitats may also occur within the USA (Rozas and
Minello, 1998).
Survival of animals in seagrass does not appear
to be significantly different from that in salt marsh
(reviews: Heck et al., 2003; Minello et al., 2003)
(Fig. 4). For example, survival time for tethered
brown shrimp (Farfantepenaeus aztecus, Penaeidae)
was similar between seagrass and marsh habitats and
significantly greater than over non-vegetated sand
bottom (Minello, 1993).
Very few studies have compared growth of fish
or invertebrates in seagrass and marsh habitats (but
see Irlandi and Crawford, 1997; Phelan et al., 2000;
Stunz et al., 2002) and most of these studies differ in the species examined and methodology utilized. Minello et al. (2003) combined all studies in a
meta-analysis and found that growth in seagrass was
significantly greater than growth in non-vegetated
marsh (sensu Minello et al., 2003). However, their
meta-analyses also found that growth in vegetated
marsh was significantly higher than in non-vegetated
marsh. Stunz et al. (2002) using solid-walled field
mesocosms found no significant difference in growth
of Sciaenops ocellatus (Sciaenidae) between the
edge of vegetated marsh and seagrass. Irlandi and
Crawford (1997) used field enclosures containing
vegetated marsh edge with seagrass and vegetated
marsh edge without seagrass to evaluate growth
of Lagodon rhomboides (Sparidae). They found
that growth was negative (measured as biomass) in
vegetated marsh without seagrass and positive in
vegetated marsh with seagrass. Thus, seagrass appears to have an effect on growth of this species
of fish.
IV. Spatial Scales—Factors Influencing
Abundance of Fish in Seagrass Beds
Seagrass habitats represent a series of hierarchical
scales ranging from single blades to individual short
shoots to groups of short shoots composing a seagrass bed to the seagrass bed within a matrix of other
habitat types (e.g. non-vegetated habitats) to submerged vegetation within the large coastal system
(Robbins and Bell, 1994). The scale at which observations are made or experiments are conducted may
influence results and conclusions (Turner, 1989),
such that patterns and processes that are found to
be important at one scale may not be important at
another scale (Wiens, 1989).
A large amount of research has focused at the
small scale (e.g. seagrass blade length and density)
and shown that this scale is important in influencing recruitment dynamics (see below and Heck and
Orth, Chapter 22). Artificial seagrass has been used
in many of the small-scale experiments addressing a
variety of questions, as it provides a standard habitat
unit where morphological aspects of the plants can
either be held constant across a number of locations
or manipulated such that spatial variation is not a
confounding influence. Research at this small scale
has been reviewed elsewhere (e.g. Bell and Pollard,
1989) and is only briefly mentioned here. The focus
on larger-scale spatial patterns and linking seagrass
patch size, shape, or complexity of patch perimeter,
position of a patch in the seagrass bed or its position
in relation to adjacent patches and other habitats,
to the structure and function of assemblages has
only recently been investigated. At this larger spatial
scale, a range of scales (extent, sensu Wiens, 1989)
and sampling units that vary in scale (grain, sensu
Wiens, 1989) have been used making comparisons
among studies difficult.