512
Bronwyn M. Gillanders
pressure between edge and center of Zostera marina
habitats was found, although these habitats had significantly less predation than non-vegetated habitats
(Peterson et al., 2001).
As habitats are fragmented, patches of different sizes may be found. It is predicted that more
individuals may be found (per unit area) as patch
size decreases (Bell et al., 2001). Naturally occurring beds of Zostera capricorni and Posidonia australis supported significantly more species of fish
and macroinvertebrates in two small beds than in
one large bed of the same total area (McNeill and
Fairweather, 1993). Despite these correlative results,
ASUs showed few significant differences between
two small (7 m
2 ) ASUs and one large (14 m
2 ) ASU
in experimental manipulations (McNeill and Fairweather, 1993). In another experimental study, only
two of the seven response variables (namely, grass
shrimp and small, mobile crustaceans) showed increased density in small patches relative to large
patches (Eggleston et al., 1999). These two species
may be responding to an edge effect since they were
two times more likely to encounter a small patch
(0.25 m
2 ) than a large (1 m
2 ) patch (Eggleston et al.,
1999). However, another study involving both a literature review and a survey of fish and epibenthic amphipods in Tampa Bay, Florida, reported that patch
size had no consistent effect on resident fauna over
spatial scales from <1 to >5,000 m
2 (Bell et al.,
2001). An organism’s response to habitats of different patch sizes may also be species-specific (Eggleston et al., 1998a). In terrestrial studies, the degree
of isolation and position of patches within the landscape is important and may account for differences
between marine studies (Bell et al., 2001) (see also
‘Landscape scale variables’, Section IV.B).
Survival of organisms may also be related to patch
size in that small patches have greater edge to area
ratios than large patches, theoretically leading to increased predation in small patches since prey should
be more available to predators. However, an inverse
relationship was found between crab (Callinectes
sapidus, Portunidae) survival and patch size, highlighting the need for empirical research on fragmentation effects (Hovel and Lipcius, 2001). Hard
clams (Mercenaria mercenaria, Veneridae) showed
increased survival in large rather than small seagrass patches and in continuous vs. patchy seagrass,
but seagrass complexity (e.g. shoot density, root
biomass) also differed among treatments making interpretation of the significance of these results difficult (Irlandi, 1994, 1997). To date, only one study
(that of Hovel and Lipcius, 2001) has demonstrated a
negative correlation between prey survival and patch
size in marine systems. This study is the only one that
has used a range of levels of fragmentation of natural
seagrass (e.g. patches <1 to >30,000 m
2 ) and tested
how these levels of fragmentation influence faunal
survival without seagrass complexity varying among
different levels of fragmentation.
Growth of clams was increased in the center
of large seagrass beds compared to small seagrass
beds, although there was no increased abundance
of organisms in the center of large beds (Irlandi,
1997). Growth of bay scallops (Argopecten irradians concentricus, Pectinidae) was slightly faster in
naturally occurring small patches of seagrass than
in large patches, although the difference was not
thought to be biologically meaningful (Irlandi et al.,
1999). Patch size did not, however, influence overall
predation rates on juvenile scallops (Irlandi et al.,
1999).
Patch shape and orientation also may influence
abundance of organisms, but there have been few
empirical studies. In one such study, fish showed decreased abundance in elongate patches perpendicular to the prevailing current compared to square
patches or elongate patches parallel to the prevailing current (Tanner, 2003). This orientation effect
was related to movements between intertidal areas
and subtidal refuges (Tanner, 2003). Although individual patch characteristics have been shown to
be important, the landscape context also needs to
be considered and may account for variation among
studies.
B. Landscape Scale Variables
1. General
Very few studies of marine systems have employed a
landscape perspective (but see Bell et al., 1988; Bell
and Hicks, 1991; Irlandi et al., 1995; Irlandi and
Crawford, 1997; and Bell et al., Chapter 26). Most
studies within a landscape perspective have focused
on either proximity to other habitats (e.g. adjacent
habitats) or on the position of seagrass beds within a
bay or estuary (e.g. position of patches within a landscape). Further research is clearly needed on landscape scale variables that operate at the scale of tens
of meters to kilometers (Table 1).
Bronwyn M. Gillanders
pressure between edge and center of Zostera marina
habitats was found, although these habitats had significantly less predation than non-vegetated habitats
(Peterson et al., 2001).
As habitats are fragmented, patches of different sizes may be found. It is predicted that more
individuals may be found (per unit area) as patch
size decreases (Bell et al., 2001). Naturally occurring beds of Zostera capricorni and Posidonia australis supported significantly more species of fish
and macroinvertebrates in two small beds than in
one large bed of the same total area (McNeill and
Fairweather, 1993). Despite these correlative results,
ASUs showed few significant differences between
two small (7 m
2 ) ASUs and one large (14 m
2 ) ASU
in experimental manipulations (McNeill and Fairweather, 1993). In another experimental study, only
two of the seven response variables (namely, grass
shrimp and small, mobile crustaceans) showed increased density in small patches relative to large
patches (Eggleston et al., 1999). These two species
may be responding to an edge effect since they were
two times more likely to encounter a small patch
(0.25 m
2 ) than a large (1 m
2 ) patch (Eggleston et al.,
1999). However, another study involving both a literature review and a survey of fish and epibenthic amphipods in Tampa Bay, Florida, reported that patch
size had no consistent effect on resident fauna over
spatial scales from <1 to >5,000 m
2 (Bell et al.,
2001). An organism’s response to habitats of different patch sizes may also be species-specific (Eggleston et al., 1998a). In terrestrial studies, the degree
of isolation and position of patches within the landscape is important and may account for differences
between marine studies (Bell et al., 2001) (see also
‘Landscape scale variables’, Section IV.B).
Survival of organisms may also be related to patch
size in that small patches have greater edge to area
ratios than large patches, theoretically leading to increased predation in small patches since prey should
be more available to predators. However, an inverse
relationship was found between crab (Callinectes
sapidus, Portunidae) survival and patch size, highlighting the need for empirical research on fragmentation effects (Hovel and Lipcius, 2001). Hard
clams (Mercenaria mercenaria, Veneridae) showed
increased survival in large rather than small seagrass patches and in continuous vs. patchy seagrass,
but seagrass complexity (e.g. shoot density, root
biomass) also differed among treatments making interpretation of the significance of these results difficult (Irlandi, 1994, 1997). To date, only one study
(that of Hovel and Lipcius, 2001) has demonstrated a
negative correlation between prey survival and patch
size in marine systems. This study is the only one that
has used a range of levels of fragmentation of natural
seagrass (e.g. patches <1 to >30,000 m
2 ) and tested
how these levels of fragmentation influence faunal
survival without seagrass complexity varying among
different levels of fragmentation.
Growth of clams was increased in the center
of large seagrass beds compared to small seagrass
beds, although there was no increased abundance
of organisms in the center of large beds (Irlandi,
1997). Growth of bay scallops (Argopecten irradians concentricus, Pectinidae) was slightly faster in
naturally occurring small patches of seagrass than
in large patches, although the difference was not
thought to be biologically meaningful (Irlandi et al.,
1999). Patch size did not, however, influence overall
predation rates on juvenile scallops (Irlandi et al.,
1999).
Patch shape and orientation also may influence
abundance of organisms, but there have been few
empirical studies. In one such study, fish showed decreased abundance in elongate patches perpendicular to the prevailing current compared to square
patches or elongate patches parallel to the prevailing current (Tanner, 2003). This orientation effect
was related to movements between intertidal areas
and subtidal refuges (Tanner, 2003). Although individual patch characteristics have been shown to
be important, the landscape context also needs to
be considered and may account for variation among
studies.
B. Landscape Scale Variables
1. General
Very few studies of marine systems have employed a
landscape perspective (but see Bell et al., 1988; Bell
and Hicks, 1991; Irlandi et al., 1995; Irlandi and
Crawford, 1997; and Bell et al., Chapter 26). Most
studies within a landscape perspective have focused
on either proximity to other habitats (e.g. adjacent
habitats) or on the position of seagrass beds within a
bay or estuary (e.g. position of patches within a landscape). Further research is clearly needed on landscape scale variables that operate at the scale of tens
of meters to kilometers (Table 1).