2.1. THE PHYSICAL ENVIRONMENT
regimes normally found on most coral reefs. Both theoretical and empirical approaches have been undertaken to address this issue. Models of coral
feeding under simplified, laminar flow conditions (Abelson et al. 1993) have
suggested that the size and type of particle captured is dependent on the
height of the organism above the substrate relative to the width of the organism in the direction of flow. Corals which extend above the substrate are
expected to feed on finer particles which are resuspended from the bottom,
whereas corals which lie close to the bottom were predicted to feed primarily
on heavier, bed-load transported particles (Abelson et al. 1993). While this
model was supported by measurements conducted using physical models, it
has yet to be tested using corals living under more realistic flow conditions.
Empirical measurements of particle capture by corals and other suspension feeding invertebrates have also been employed under both artificial
(laboratory flume) and semi-natural (field flume) conditions, and have shown
that patterns vary significantly with coral morphology (Heidelberg et al.
1997). Sebens and Johnson (1991) measured rates of feeding by two species
of scleractinian corals using brine shrimp cysts as food particles. They
found that the branching, cylindrical species of coral, Madracis decactis
(see Fig. 2.5), showed an increase in particle capture with increasing flow
speed. In contrast, a flat species of coral, Meandrina meandrites, showed no
effect of flow speed on particle capture, due to the tendency of the coral's
tentacles to flatten under ever higher flow speeds (Johnson and Sebens 1993).
Helmuth and Sebens (1993) examined particle capture by several morphotypes of Agariciaagaricites, and found that particle capture in unidirectional
flow increased with flow speed up to a velocity of approximately 30 ctn ]s,
but then decreased at flows above this level. Similarly, Sebens et al. (1997)
found that feeding by solitary branches of the coral Madracis mirabilis (see
Fig. 2.5, a cylindrical coral that exists almost exclusively in aggregations of
clonemates) experienced maximum rates of food capture at a flow speed of
10 -15 cm/s . These studies show that while increasing flow speeds do increase
the rates of particle delivery to the coral surface, particle capture efficiency
often decreases with increasing flow due to the tendency for a coral's tentacles to flatten under high flows, rendering them unable to capture the
particles moving across their surfaces (Patterson 1984, Lasker 1981, Johnson
and Sebens 1993, Sebens et al. 1997). Thus, particle capture rates are often
highest at intermediate flow speeds, where particle delivery rate is high, but
tentacles are still capable of retaining particles.
Some species have apparently been able to at least partially circumvent
this limitation through the formation of aggregations in which the spacing
between ramets varies as a function of ambient flow (McFadden 1986, Sebens
et al. 1997). For example, McFadden (1986) found that under low flow conditions, the presence of neighboring colonies reduced the rate of particle
capture by the soft coral Alcyonium. However, at higher flow speeds , particle capture rates were enhanced by the presence of neighbors. Sebens et
al. (1997) found that branches within aggregations of the cylindrical coral
Madracis mirabilis were more widely spaced in slower moving water than
in areas with higher average water velocities. Feeding trials in a laboratory
flume confirmed that particle capture increased with branch spacing in low
flows, but decreased with branch spacing in higher flows, suggesting that
plasticity in branch spacing represents a means of acclimatizing to the characteristics of the local flow environment. Thus, living in aggregations may
29
~ Fig. z.sa-e. The stony coral Madracis
decactis collected at different depths,
sample (e) originates from a depth of
6 m, (d) wascollected at a depth of IS m.
ThestonycoralMadracis mirabilis, samples (c), (b), and (a), were collected at
depths of respectively 6, 8, and 20 m.
regimes normally found on most coral reefs. Both theoretical and empirical approaches have been undertaken to address this issue. Models of coral
feeding under simplified, laminar flow conditions (Abelson et al. 1993) have
suggested that the size and type of particle captured is dependent on the
height of the organism above the substrate relative to the width of the organism in the direction of flow. Corals which extend above the substrate are
expected to feed on finer particles which are resuspended from the bottom,
whereas corals which lie close to the bottom were predicted to feed primarily
on heavier, bed-load transported particles (Abelson et al. 1993). While this
model was supported by measurements conducted using physical models, it
has yet to be tested using corals living under more realistic flow conditions.
Empirical measurements of particle capture by corals and other suspension feeding invertebrates have also been employed under both artificial
(laboratory flume) and semi-natural (field flume) conditions, and have shown
that patterns vary significantly with coral morphology (Heidelberg et al.
1997). Sebens and Johnson (1991) measured rates of feeding by two species
of scleractinian corals using brine shrimp cysts as food particles. They
found that the branching, cylindrical species of coral, Madracis decactis
(see Fig. 2.5), showed an increase in particle capture with increasing flow
speed. In contrast, a flat species of coral, Meandrina meandrites, showed no
effect of flow speed on particle capture, due to the tendency of the coral's
tentacles to flatten under ever higher flow speeds (Johnson and Sebens 1993).
Helmuth and Sebens (1993) examined particle capture by several morphotypes of Agariciaagaricites, and found that particle capture in unidirectional
flow increased with flow speed up to a velocity of approximately 30 ctn ]s,
but then decreased at flows above this level. Similarly, Sebens et al. (1997)
found that feeding by solitary branches of the coral Madracis mirabilis (see
Fig. 2.5, a cylindrical coral that exists almost exclusively in aggregations of
clonemates) experienced maximum rates of food capture at a flow speed of
10 -15 cm/s . These studies show that while increasing flow speeds do increase
the rates of particle delivery to the coral surface, particle capture efficiency
often decreases with increasing flow due to the tendency for a coral's tentacles to flatten under high flows, rendering them unable to capture the
particles moving across their surfaces (Patterson 1984, Lasker 1981, Johnson
and Sebens 1993, Sebens et al. 1997). Thus, particle capture rates are often
highest at intermediate flow speeds, where particle delivery rate is high, but
tentacles are still capable of retaining particles.
Some species have apparently been able to at least partially circumvent
this limitation through the formation of aggregations in which the spacing
between ramets varies as a function of ambient flow (McFadden 1986, Sebens
et al. 1997). For example, McFadden (1986) found that under low flow conditions, the presence of neighboring colonies reduced the rate of particle
capture by the soft coral Alcyonium. However, at higher flow speeds , particle capture rates were enhanced by the presence of neighbors. Sebens et
al. (1997) found that branches within aggregations of the cylindrical coral
Madracis mirabilis were more widely spaced in slower moving water than
in areas with higher average water velocities. Feeding trials in a laboratory
flume confirmed that particle capture increased with branch spacing in low
flows, but decreased with branch spacing in higher flows, suggesting that
plasticity in branch spacing represents a means of acclimatizing to the characteristics of the local flow environment. Thus, living in aggregations may
29
~ Fig. z.sa-e. The stony coral Madracis
decactis collected at different depths,
sample (e) originates from a depth of
6 m, (d) wascollected at a depth of IS m.
ThestonycoralMadracis mirabilis, samples (c), (b), and (a), were collected at
depths of respectively 6, 8, and 20 m.