Chapter 22 Predation in Seagrass Beds
539
(1990) experiments, together with the other evidence noted above, suggest that non-linear rates of
prey capture at different vegetation densities may
yet prove to be the case, although further testing is
clearly still required.
With regard to the effect of seagrass surface area
on predation rates, Stoner (1982) tested the protective abilities of seagrasses that differed in surface
area/biomass ratios. He found that for single seagrass species, predation intensity declined with plant
biomass and therefore plant surface area. However,
in multi-species trials with similar biomasses, the
seagrass species possessing the greatest amount of
total surface area (Halodule wrightii) provided the
least amount of protection. The likely reason for
this unexpected result is that the spaces between the
finely divided H. wrightii seagrass leaves and shoots
did not match the size of the amphipod prey that
were being attacked by visual fish predators. This
was also the conclusion of Ryer (1988), who studied the vulnerability of different sized amphipods to
fish predators in artificial eelgrass, and found that
only when the spaces between leaves matched the
size of prey did substantial refuge exist. And this
is also the premise underlying the index recently
proposed by Bartholomew et al. (2000) to quantify the degree of protection provided by seagrass
leaves that might be generally applied across many
types of complex habitats. The index is calculated as
the ratio of the distance between structural elements
(Sp) to predator width (Pr). The rationale is that
spaces between structural components, such as seagrass shoots, that are smaller than a predator’s width,
should restrict predator movements and provide protection in predator–prey encounters. Although the
index worked well in laboratory experiments with
regularly placed structural elements (wooden dowels), it did not work as well in field trials in an intertidal salt marsh (Spartina alterniflora) habitat, nor
did it give consistent results in mangrove prop roots
when distances between roots were calculated for
randomly, clumped, and uniform spacing patterns
(Bell et al., 2003). This led Bell et al. (2003) to conclude that the index will only be useful for cross
habitat comparisons when spatial arrangements of
structural elements are similar, but that it is not a
generally applicable means of comparing dissimilar
structure across habitat types.
Also of interest, but rarely studied, is the behavior of prey as they use the structure provided by
seagrasses to mitigate the impact of their predators. To date, the study by Main (1985, 1987) on
the antipredatory behavior of the caridean shrimp
Tozeuma carolinense in the presence of pinfish, L.
rhomboides, remains one of the few studies to document how prey consistently avoid potential predators, in this case by moving around seagrass leaves
and thereby maintaining an opaque barrier between
themselves and pinfish.
An issue that, to our knowledge, remains unstudied, is how predators that forage from above the seagrass canopy might be influenced by seagrass density and/or biomass. For example, the height of the
leaves and the degree to which they overlap each
other could have significant impacts on the ability
of down-looking predators to detect prey. In addition, the effects of wave-action could be important,
as seagrass leaves sway back and forth with wave
passage and alternately cover and uncover patches
of the meadow. In such conditions, seagrass might
provide less protection from predators foraging from
above than from those foraging within the canopy.
Because predators foraging from above the canopy
will usually be larger, more mobile fishes, the logistics of conducting experiments at appropriate scales
in the laboratory remains problematic.
With regard to studying the effect of seagrass
on survival rates of infauna, Peterson and Heck
(2001) recently reported that semi-infaunal mussels
were preyed upon significantly less frequently when
present in turtlegrass meadows than when living in
unvegetated sand. Similar findings apply to studies
of sub-lethal predation (e.g. siphon nipping), where
Irlandi (1994; Irlandi and Mehlich, 1996) and others
(Coen and Heck, 1991; Nakaoka, 2000) found reduced incidences of sub-lethal predation in seagrass
meadows than on unvegetated substrates. These conclusions matched expectations and were generally
similar to earlier investigations of the effects of
seagrasses on predation of infaunal taxa (cf. Peterson and Quammen 1982; Blundon and Kennedy,
1982). To date, there are few studies of how infaunal
species behave in the presence of predation risk, or
even if they do, with the exception of siphonous bivalves, which respond by withdrawing their siphon
in the presence of potential siphon nippers (Blundon and Kennedy, 1982; Irlandi and Peterson, 1991;
Nakaoka, 2000). How infaunal taxa detect and avoid
the different types of predators that attempt to prey
upon them remain questions for future research.
Temporal differences in predation rates on seasonal and annual scales have been quantified by field
539
(1990) experiments, together with the other evidence noted above, suggest that non-linear rates of
prey capture at different vegetation densities may
yet prove to be the case, although further testing is
clearly still required.
With regard to the effect of seagrass surface area
on predation rates, Stoner (1982) tested the protective abilities of seagrasses that differed in surface
area/biomass ratios. He found that for single seagrass species, predation intensity declined with plant
biomass and therefore plant surface area. However,
in multi-species trials with similar biomasses, the
seagrass species possessing the greatest amount of
total surface area (Halodule wrightii) provided the
least amount of protection. The likely reason for
this unexpected result is that the spaces between the
finely divided H. wrightii seagrass leaves and shoots
did not match the size of the amphipod prey that
were being attacked by visual fish predators. This
was also the conclusion of Ryer (1988), who studied the vulnerability of different sized amphipods to
fish predators in artificial eelgrass, and found that
only when the spaces between leaves matched the
size of prey did substantial refuge exist. And this
is also the premise underlying the index recently
proposed by Bartholomew et al. (2000) to quantify the degree of protection provided by seagrass
leaves that might be generally applied across many
types of complex habitats. The index is calculated as
the ratio of the distance between structural elements
(Sp) to predator width (Pr). The rationale is that
spaces between structural components, such as seagrass shoots, that are smaller than a predator’s width,
should restrict predator movements and provide protection in predator–prey encounters. Although the
index worked well in laboratory experiments with
regularly placed structural elements (wooden dowels), it did not work as well in field trials in an intertidal salt marsh (Spartina alterniflora) habitat, nor
did it give consistent results in mangrove prop roots
when distances between roots were calculated for
randomly, clumped, and uniform spacing patterns
(Bell et al., 2003). This led Bell et al. (2003) to conclude that the index will only be useful for cross
habitat comparisons when spatial arrangements of
structural elements are similar, but that it is not a
generally applicable means of comparing dissimilar
structure across habitat types.
Also of interest, but rarely studied, is the behavior of prey as they use the structure provided by
seagrasses to mitigate the impact of their predators. To date, the study by Main (1985, 1987) on
the antipredatory behavior of the caridean shrimp
Tozeuma carolinense in the presence of pinfish, L.
rhomboides, remains one of the few studies to document how prey consistently avoid potential predators, in this case by moving around seagrass leaves
and thereby maintaining an opaque barrier between
themselves and pinfish.
An issue that, to our knowledge, remains unstudied, is how predators that forage from above the seagrass canopy might be influenced by seagrass density and/or biomass. For example, the height of the
leaves and the degree to which they overlap each
other could have significant impacts on the ability
of down-looking predators to detect prey. In addition, the effects of wave-action could be important,
as seagrass leaves sway back and forth with wave
passage and alternately cover and uncover patches
of the meadow. In such conditions, seagrass might
provide less protection from predators foraging from
above than from those foraging within the canopy.
Because predators foraging from above the canopy
will usually be larger, more mobile fishes, the logistics of conducting experiments at appropriate scales
in the laboratory remains problematic.
With regard to studying the effect of seagrass
on survival rates of infauna, Peterson and Heck
(2001) recently reported that semi-infaunal mussels
were preyed upon significantly less frequently when
present in turtlegrass meadows than when living in
unvegetated sand. Similar findings apply to studies
of sub-lethal predation (e.g. siphon nipping), where
Irlandi (1994; Irlandi and Mehlich, 1996) and others
(Coen and Heck, 1991; Nakaoka, 2000) found reduced incidences of sub-lethal predation in seagrass
meadows than on unvegetated substrates. These conclusions matched expectations and were generally
similar to earlier investigations of the effects of
seagrasses on predation of infaunal taxa (cf. Peterson and Quammen 1982; Blundon and Kennedy,
1982). To date, there are few studies of how infaunal
species behave in the presence of predation risk, or
even if they do, with the exception of siphonous bivalves, which respond by withdrawing their siphon
in the presence of potential siphon nippers (Blundon and Kennedy, 1982; Irlandi and Peterson, 1991;
Nakaoka, 2000). How infaunal taxa detect and avoid
the different types of predators that attempt to prey
upon them remain questions for future research.
Temporal differences in predation rates on seasonal and annual scales have been quantified by field