540
K. L. Heck and R. J. Orth
tethering experiments (e.g. Heck and Wilson, 1987;
Heck et al., 2001) and, not surprisingly, the ability
of vegetation to provide protection from predators
was usually observed, as were substantial seasonal
and annual differences in predation rate. Less wellstudied are diurnal differences in predation rate in
vegetated habitats, and the few studies that exist
did not find identical patterns. For example, using
chronographic tethering techniques, Peterson et al.
(2001) found no significant differences in predation
rates on shrimp and crabs between light and dark periods in eelgrass beds in the Damariscotta River estuary (Maine, USA), while they found significantly
higher rates of daytime predation in St. Joseph Bay
turtlegrass beds (Florida, USA). Different numbers
of night active predators between the two areas were
suggested as an explanation for these location differences. In Newfoundland (Canada), Lineham et al.
(2001) used tethers to measure predation rates on
small cod in eelgrass and nearby unvegetated substrates during day, dusk, and night. They found that
predation rates were greatest during the day and
dusk, and lower at night, but overall rates were lower
in vegetation than on unvegetated substrates.
Because different predators forage at different
times of day, depending on the mode of foraging,
one could expect day vs. night predation rates to
fluctuate seasonally and annually as the composition
of predators changes. Overall, while the number of
studies on the effect of time of day on predation rates
in seagrass meadows is very small, the few studies
that exist are consistent with what one might expect,
given a knowledge of the predator assemblages in
any given area.
On balance, the existing literature shows that,
although there are differences in details, seagrass
presence is almost always negatively related to predation effectiveness, and that increasing seagrass
abundance is usually associated with decreasing
predator effectiveness. The only exceptions we know
are those of James and Heck (1994). In that study,
the foraging rate of the seahorse (Hippocampus erectus), an ambush predator, was not significantly negatively influenced by increasing density of artificial
seagrass. Although it seems likely that studies with
other ambush predators would produce similar results, there are no additional data to evaluate this
proposition, and more studies would be welcome.
All previous studies that have investigated the effects of vegetation density on predation success have
been done in a similar manner, and one that calls into
question much of what has become the conventional
wisdom. As far as we know, the experimental design in all previous laboratory experiments has been
to test the effects of a constant number of predators on a constant number of prey items at different
levels of vegetation density. However, as shown repeatedly in sampling studies, both the numbers of
prey and predators almost always increase with increasing vegetation density (Orth et al., 1984; Edgar,
1990; Edgar and Shaw, 1995; Bostr¨ om and Bonsdorff, 1997; Jenkins et al., 1997; Wyda et al., 2002).
Thus, the results from laboratory experiments done
to date, in which predator and prey densities were
kept constant as vegetation density was increased,
appear to have poorly mirrored reality and resulted
in an incomplete understanding of how seagrass influences predator–prey interactions in nature.
In aquarium experiments where pinfish (L. rhomboides) were used as predators and grass shrimp
(Palaemonetes pugio) as prey, Mattila and Heck (in
review) studied predation effects on an unvegetated
substrate and at three different densities (400, 2,000,
and 4,000 leaves m
−2 ) of artificial seagrass (a mimic
of turtlegrass, Thalassia testudinum). Their results
showed that seagrass provided significantly more
shelter than unvegetated substrate, but no significant
differences were found among different seagrass
densities (Fig. 2). These results conflict with those of
most previous experiments that tested the sheltering
role of seagrass, where increasing leaf density was
associated with increased prey survival. However,
Mattila and Heck’s (in review) experimental design
differed from all previous experiments in that both
the numbers of prey and predators were increased
with increasing seagrass density, thereby mimicking what is observed in nature, whereas previous experiments were conducted with constant numbers of
prey and predators at differing densities of seagrass.
These results indicate that it is not safe to assume
that increasing vegetation density will consistently
lead to proportionally higher prey survival.
B. Indirect Effects on Growth and Food Webs
A recent meta-analysis (Heck et al., 2003) found, as
expected, that growth of macrofaunal species was
significantly greater in seagrass meadows than on
unvegetated substrates, although the number of studies in the analyses was small. However, an issue addressed infrequently over the years is how seagrass
density affects the growth of animals. If prey capture
K. L. Heck and R. J. Orth
tethering experiments (e.g. Heck and Wilson, 1987;
Heck et al., 2001) and, not surprisingly, the ability
of vegetation to provide protection from predators
was usually observed, as were substantial seasonal
and annual differences in predation rate. Less wellstudied are diurnal differences in predation rate in
vegetated habitats, and the few studies that exist
did not find identical patterns. For example, using
chronographic tethering techniques, Peterson et al.
(2001) found no significant differences in predation
rates on shrimp and crabs between light and dark periods in eelgrass beds in the Damariscotta River estuary (Maine, USA), while they found significantly
higher rates of daytime predation in St. Joseph Bay
turtlegrass beds (Florida, USA). Different numbers
of night active predators between the two areas were
suggested as an explanation for these location differences. In Newfoundland (Canada), Lineham et al.
(2001) used tethers to measure predation rates on
small cod in eelgrass and nearby unvegetated substrates during day, dusk, and night. They found that
predation rates were greatest during the day and
dusk, and lower at night, but overall rates were lower
in vegetation than on unvegetated substrates.
Because different predators forage at different
times of day, depending on the mode of foraging,
one could expect day vs. night predation rates to
fluctuate seasonally and annually as the composition
of predators changes. Overall, while the number of
studies on the effect of time of day on predation rates
in seagrass meadows is very small, the few studies
that exist are consistent with what one might expect,
given a knowledge of the predator assemblages in
any given area.
On balance, the existing literature shows that,
although there are differences in details, seagrass
presence is almost always negatively related to predation effectiveness, and that increasing seagrass
abundance is usually associated with decreasing
predator effectiveness. The only exceptions we know
are those of James and Heck (1994). In that study,
the foraging rate of the seahorse (Hippocampus erectus), an ambush predator, was not significantly negatively influenced by increasing density of artificial
seagrass. Although it seems likely that studies with
other ambush predators would produce similar results, there are no additional data to evaluate this
proposition, and more studies would be welcome.
All previous studies that have investigated the effects of vegetation density on predation success have
been done in a similar manner, and one that calls into
question much of what has become the conventional
wisdom. As far as we know, the experimental design in all previous laboratory experiments has been
to test the effects of a constant number of predators on a constant number of prey items at different
levels of vegetation density. However, as shown repeatedly in sampling studies, both the numbers of
prey and predators almost always increase with increasing vegetation density (Orth et al., 1984; Edgar,
1990; Edgar and Shaw, 1995; Bostr¨ om and Bonsdorff, 1997; Jenkins et al., 1997; Wyda et al., 2002).
Thus, the results from laboratory experiments done
to date, in which predator and prey densities were
kept constant as vegetation density was increased,
appear to have poorly mirrored reality and resulted
in an incomplete understanding of how seagrass influences predator–prey interactions in nature.
In aquarium experiments where pinfish (L. rhomboides) were used as predators and grass shrimp
(Palaemonetes pugio) as prey, Mattila and Heck (in
review) studied predation effects on an unvegetated
substrate and at three different densities (400, 2,000,
and 4,000 leaves m
−2 ) of artificial seagrass (a mimic
of turtlegrass, Thalassia testudinum). Their results
showed that seagrass provided significantly more
shelter than unvegetated substrate, but no significant
differences were found among different seagrass
densities (Fig. 2). These results conflict with those of
most previous experiments that tested the sheltering
role of seagrass, where increasing leaf density was
associated with increased prey survival. However,
Mattila and Heck’s (in review) experimental design
differed from all previous experiments in that both
the numbers of prey and predators were increased
with increasing seagrass density, thereby mimicking what is observed in nature, whereas previous experiments were conducted with constant numbers of
prey and predators at differing densities of seagrass.
These results indicate that it is not safe to assume
that increasing vegetation density will consistently
lead to proportionally higher prey survival.
B. Indirect Effects on Growth and Food Webs
A recent meta-analysis (Heck et al., 2003) found, as
expected, that growth of macrofaunal species was
significantly greater in seagrass meadows than on
unvegetated substrates, although the number of studies in the analyses was small. However, an issue addressed infrequently over the years is how seagrass
density affects the growth of animals. If prey capture